THE INFLUENCE OF GROWTH CONDITIONS ON

THE INFLUENCE OF GROWTH CONDITIONS ON THE INTRACELLULAR MICROCYSTIN QUOTA OF MICROCYSTIS AERUGINOSA

Submitted by Benedict Michael Long, B.Sc. (Hons.)

A thesis submitted in total fulfilment of the requirements for the degree of Doctor of Philosophy

Department of Botany School of Life Sciences Faculty of Science, Technology and Engineering La Trobe University Bundoora, Victoria 3083 Australia

February, 2001

Table of contents

ii

Table of Contents List of figures

vi

List of tables

ix

List of equations

x

Nomenclature and Abbreviations

xi

Summary

xiii

Statement of Authorship

xiv

Acknowledgments

xv

CHAPTER 1: Toxic cyanobacterial blooms and microcystin production 1.1

Characteristics of cyanobacteria

2

1.2

Factors influencing cyanobacterial dominance in freshwater systems

5

1.2.1 Nutrients

5

1.2.2 Light

6

1.2.3 Temperature

6

1.2.4 CO2 and pH

7

1.2.5 Grazing

7

1.3

The occurrence of toxic cyanobacterial blooms

8

1.4

The genus Microcystis

9

1.5

The Microcystins (MCYSTs)

10

1.5.1 MCYST structure and occurrence in cyanobacteria

10

1.5.2 MCYST Toxicity

18

1.5.3 MCYST biosynthesis

20

1.6

1.7

Factors influencing MCYST production and content in cyanobacteria grown in culture

21

1.6.1 Temperature and Light

21

1.6.2 Nitrogen and phosphorus

23

1.6.3 Iron

25

1.6.4 CO2 and pH

25

1.6.5 Time (culture age)

26

1.6.6 Problems associated with MCYST content studies

34

Aims of this thesis

34

Table of contents

iii

CHAPTER 2: M. aeruginosa strains : their sources, growth and MCYST production 2.1

Introduction

38

2.2

Sources of M. aeruginosa strains and their growth in culture

38

2.2.1 Collection of M. aeruginosa from lakes and reservoirs

38

2.2.2 Isolation and growth of M. aeruginosa strains

39

2.3

Investigation of MCYST production capabilities of selected strains of M. aeruginosa

45

2.3.1 Production of MCYSTs by selected M. aeruginosa strains as determined by protein phosphatase (PP) inhibition

45

2.3.2 Production of MCYSTs by selected M. aeruginosa strains as determined by reaction to polyclonal anti-MCYST-LR antibodies

52

2.3.3 Production of MCYSTs by selected M. aeruginosa strains as determined by HPLC

2.4

55

2.3.4 Characterisation of MCYSTs from MASH-01A19

59

Concluding remarks

66

CHAPTER 3: MCYST quota of M. aeruginosa MASH-01A19 in N-limited chemostat cultures 3.1

Introduction

69

3.2

Chemostat theory

70

3.3

Materials and Methods

72

3.3.1 Organism and growth conditions

72

3.3.2 Sampling and analysis

74

Results

76

3.4.1 Steady-state Growth

76

3.4.2 MCYST analysis

77

3.5

Discussion

91

3.6

Implications of results from chemostats with respect to batch cultures

95

3.4

CHAPTER 4: The influence of growth conditions on the intracellular MCYST quota of M. aeruginosa MASH-01A19 in batch culture 4.1

Introduction

4.2

Growth and intracellular MCYST quota of M. aeruginosa MASH-01A19 grown under standard conditions in batch culture

99

101

Table of contents

4.3

4.4

iv

4.2.1 Introduction

101

4.2.2 Standard growth conditions

101

4.2.3 Sampling and analysis

102

4.2.4 Results

104

4.2.5 Discussion

112

The effect of Pi supply on intracellular MCYST quota of M. aeruginosa MASH-01A19 in batch culture

118

4.3.1 Introduction

118

4.3.2 Growth conditions, sampling and analysis

118

4.3.3 Results

119

4.3.4 Discussion

129

The effect of SO42- supply on intracellular MCYST quota of M. aeruginosa MASH-01A19 in batch culture

132

4.4.1 Introduction

132


133

4.4.3 Results

133

4.4.4 Discussion 4.5

4.6

4.7

The effect of Fe

3+

140 supply on intracellular MCYST content of M. aeruginosa

MASH-01A19 in batch culture

142

4.5.1 Introduction

142


142

4.5.3 Results

143

4.5.4 Discussion

153

The effect of irradiance on intracellular MCYST quota of M. aeruginosa MASH-01A19 in batch culture

157

4.6.1 Introduction

157


158

4.6.3 Results

159

4.6.4 Discussion

167

The effect of temperature on intracellular MCYST quota of M. aeruginosa MASH-01A19 in batch culture

172

4.7.1 Introduction

172


172

4.7.3 Results

173

4.7.4 Discussion

183

CHAPTER 5: Concluding Remarks

188

Table of contents

v

Appendices Appendix 1: Procedure for the detection of MCYST-producing strains of M. aeruginosa on nitrocellulose filters using polyclonal anti-MCYSTLR-BSA.

195

Appendix 2: Efficiency of various solvents for the extraction of MCYST-LR from M. aeruginosa MASH-01A19. Appendix 3: HPLC methods

197 198

Appendix 4: Turbidostat culture system for the mass production of M. aeruginosa MASH-01A19.

200

Appendix 5: Publication pertaining to this thesis

201

References

208

vi

List of figures Page numbers refer to the location of the figure legends. Where figures occupy an entire page, figure legends appear on the facing page. Fig. 1.1 ........................................................................................................................................... 12 Fig. 2.1 ........................................................................................................................................... 39 Fig. 2.2 ........................................................................................................................................... 50 Fig. 2.3 ........................................................................................................................................... 50 Fig. 2.4 ........................................................................................................................................... 50 Fig. 2.5 ........................................................................................................................................... 54 Fig. 2.6 ........................................................................................................................................... 57 Fig. 2.7 ........................................................................................................................................... 64 Fig. 2.8 ........................................................................................................................................... 64 Fig. 2.9 ........................................................................................................................................... 64 Fig. 2.10 ......................................................................................................................................... 66 Fig. 3.1 ........................................................................................................................................... 73 Fig. 3.2 ........................................................................................................................................... 82 Fig. 3.3 ........................................................................................................................................... 82 Fig. 3.4 ........................................................................................................................................... 82 Fig. 3.5 ........................................................................................................................................... 85 Fig. 3.6 ........................................................................................................................................... 87 Fig. 3.7 ........................................................................................................................................... 89 Fig. 3.8 ........................................................................................................................................... 89 Fig. 3.9 ........................................................................................................................................... 96 Fig. 4.1 ......................................................................................................................................... 107 Fig. 4.2 ......................................................................................................................................... 107 Fig. 4.3 ......................................................................................................................................... 107 Fig. 4.4 ......................................................................................................................................... 107 Fig. 4.5 ......................................................................................................................................... 107 Fig. 4.6 ......................................................................................................................................... 107 Fig. 4.7 ......................................................................................................................................... 109 Fig. 4.8 ......................................................................................................................................... 109 Fig. 4.9 ......................................................................................................................................... 109 Fig. 4.10 ....................................................................................................................................... 109 Fig. 4.11 ....................................................................................................................................... 109 Fig. 4.12 ....................................................................................................................................... 109 Fig. 4.13 ....................................................................................................................................... 122

List of Figures

vii

Fig. 4.14 ....................................................................................................................................... 122 Fig. 4.15 ....................................................................................................................................... 122 Fig. 4.16 ....................................................................................................................................... 122 Fig. 4.17 ....................................................................................................................................... 122 Fig. 4.18 ....................................................................................................................................... 124 Fig. 4.19 ....................................................................................................................................... 124 Fig. 4.20 ....................................................................................................................................... 124 Fig. 4.21 ....................................................................................................................................... 124 Fig. 4.22 ....................................................................................................................................... 126 Fig. 4.23 ....................................................................................................................................... 135 Fig. 4.24 ....................................................................................................................................... 135 Fig. 4.25 ....................................................................................................................................... 135 Fig. 4.26 ....................................................................................................................................... 137 Fig. 4.27 ....................................................................................................................................... 137 Fig. 4.28 ....................................................................................................................................... 137 Fig. 4.29 ....................................................................................................................................... 137 Fig. 4.30 ....................................................................................................................................... 137 Fig. 4.31 ....................................................................................................................................... 137 Fig. 4.32 ....................................................................................................................................... 146 Fig. 4.33 ....................................................................................................................................... 146 Fig. 4.34 ....................................................................................................................................... 146 Fig. 4.35 ....................................................................................................................................... 146 Fig. 4.36 ....................................................................................................................................... 146 Fig. 4.37 ....................................................................................................................................... 148 Fig. 4.38 ....................................................................................................................................... 148 Fig. 4.39 ....................................................................................................................................... 148 Fig. 4.40 ....................................................................................................................................... 148 Fig. 4.41 ....................................................................................................................................... 150 Fig. 4.42 ....................................................................................................................................... 162 Fig. 4.43 ....................................................................................................................................... 162 Fig. 4.44 ....................................................................................................................................... 162 Fig. 4.45 ....................................................................................................................................... 162 Fig. 4.46 ....................................................................................................................................... 162 Fig. 4.47 ....................................................................................................................................... 164 Fig. 4.48 ....................................................................................................................................... 164 Fig. 4.49 ....................................................................................................................................... 164 Fig. 4.50 ....................................................................................................................................... 164

List of Figures

viii

Fig. 4.51 ....................................................................................................................................... 176 Fig. 4.52 ....................................................................................................................................... 176 Fig. 4.53 ....................................................................................................................................... 176 Fig. 4.54 ....................................................................................................................................... 176 Fig. 4.55 ....................................................................................................................................... 176 Fig. 4.56 ....................................................................................................................................... 178 Fig. 4.57 ....................................................................................................................................... 178 Fig. 4.58 ....................................................................................................................................... 178 Fig. 4.59 ....................................................................................................................................... 178 Fig. 4.60 ....................................................................................................................................... 180 Fig. A.1 ........................................................................................................................................ 200

ix

List of tables TABLE 1.1

14

TABLE 1.2

27

TABLE 2.1

42

TABLE 2.2

44

TABLE 2.3

66

TABLE 3.1

80

TABLE 3.2

81

TABLE 3.3

84

TABLE 4.1

111

TABLE 4.2

128

TABLE 4.3

139

TABLE 4.4

152

TABLE 4.5

166

TABLE 4.6

182

TABLE 5.1

193

TABLE A1

197

x

List of equations Equation 3.1.................................................................................................................................. 71 Equation 3.2.................................................................................................................................. 71 Equation 3.3.................................................................................................................................. 71 Equation 3.4 ....................................................................................................................... 72 Equation 3.5.................................................................................................................................. 72 Equation 3.6.................................................................................................................................. 78 Equation 3.7.................................................................................................................................. 79 Equation 4.1................................................................................................................................ 102 Equation 4.2................................................................................................................................ 161 Equation 4.3................................................................................................................................ 166

xi

Nomenclature and Abbreviations The nomenclature and abbreviations used in this thesis comply with the conventions of Phycologia.

(6Z)-Adda

steUHRLVRPHU RI $GGD DW WKH û6 double bond

(E)-Dhb

stereoisomer of Dhb

(H4)Y

1,2,3,4-tetrahydrotyrosine

(Z)-Dhb

stereoisomer of Dhb

Aba

aminoisobutyric acid

Adda

3-amino-9-methoxy-2,6,8-trimethyl-10-phenyldeca-4,6-dienoic acid

ADMAdda

O-acetyl-O-demethylAdda

ANOVA

analysis of variance

BSA

bovine serum albumin

Dha

dehydroalanine

Dhb

dehydrobutyrine

DMAdda

O-demethylAdda

DMAdda

O-demethylAdda

E(OMe)

glutamic acid methyl ester

EDTA

ethylenediaminetetraacetic acid

EGTA

ethylene glycol-bis[-aminoethyl ether]N,N,N',N'-tetraacetic acid

FMOC

9-fluorenmethyl chloroformate

g

relative centrifugal force

GC

gas chromatography

Har

homoarginine

Hil

homoisoleucine

Hph

homophenylalanine

HPLC

high performance liquid chromatography

HRP

horseradish peroxidase

Hty

homotyrosine

i.p.

intraperitoneally

LC-MS

liquid chromatography- mass spectrometry

LD50

median lethal dose

LDmin

minimum lethal dose; minimum dose required to kill a 25 g mouse

M(O)

methionine-S-oxide

M-amine

methylamine (acid hydrolysis product of Mdha)

MCYST

microcystin

MeAsp

D-erythro--methylaspartate

Nomenclature and Abbreviations

xii

Mdha

N-methyldehydroalanine

MeLan

methyllanthionine

MeSer

N-methylserine

MS

mass spectrometry

MU

4-methylumbelliferone

MUP

4-methylumbelliferyl phosphate

PAR

photosynthetically active radiation

PMSF

phenylmethylsulfonyl fluoride

p-NOH

p-nitrophenol

p-NPP

p-nitrophenyl phosphate

33 33

protein phosphatase type 1, catalytic subunit of a human PP1

SAM

S-adenosylmethionine

SD

standard deviation

SEM

standard error of the mean

Other terms requiring definition Term

Units

Definition

µ

d-1

Specific growth rate

µc

d-1

Specific cell division rate

µ c/µ max

unitless

Relative growth rate.

µ crit

d-1

Critical specific cell division rate at which QMCYST deviates from a linear relationship with µ c.

-1

µ max

d

Maximum specific growth / cell division rate

µ MCYST

d-1

Specific MCYST production rate

d

day

1 d = 24 h = 86,400 s

h

hour

1 h = 60 min = 3,600 s

irradiance

µmol photons m-2 s-1

Photon flux density or rate of light supply per unit area per unit time.

min

minute

1 min = 60 s -1

-1

QMCYST

MCYST cell d

Intracellular MCYST quota

QMCYSTmax

MCYST cell-1 d-1

Maximum intracellular MCYST quota

QMCYSTmin

MCYST cell-1 d-1

Minimum intracellular MCYST quota

RMCYST

fmol MCYST cell-1 d-1

Net intracellular MCYST production rate

xiii

Summary Microcystins (MCYSTs) are cyclic heptapeptides that occur in several species of cyanobacteria. They potently inhibit protein phosphatases and are powerful hepatotoxins. This thesis examines the production of MCYST by Microcystis aeruginosa by determining changes in the intracellular MCYST quota (QMCYST) under various growth conditions. A novel protein phosphatase inhibition assay and an antibody detection method were developed for the detection and analysis of MCYST-producing strains of M. aeruginosa. However, for the routine analysis of MCYST an HPLC method was adopted as it permitted the quantification of specific MCYSTs. In N-limited chemostat-grown cells of M. aeruginosa MASH-01A19, QMCYST is strongly correlated with the growth rate (µ). For a specified growth rate, QMCYST can be predicted from the maximum (QMCYSTmax) and minimum (QMCYSTmin) intracellular MCYST quotas and the maximum growth rate (µ max).

Since the cell volume was inversely related to µ, the intracellular concentration of

MCYST was >10-fold greater in rapidly growing cells (µ = 0.96 d-1) than in slow growing cells (µ = 0.10 d-1). These data have important implications with respect to the toxicity of M. aeruginosa under field conditions.

QMCYST was examined in batch cultures of M. aeruginosa MASH-01A19 to determine whether the relationship between QMCYST and its determinants µ, µ max, QMCYSTmin and QMCYSTmax, as they apply in N-limited chemostat cultures, also apply under other growth conditions. In batch cultures, µ max increased with temperature and irradiance but was not significantly affected by the level of P, S, N or Fe nutrition. QMCYST decreased as the level of S and Fe nutrition decreased but low P enhanced QMCYSTmin. QMCYSTmin decreased as the temperature and the irradiance increased but QMCYSTmin was not significantly affected by N nutrition. QMCYSTmax decreased as the level of S and N nutrition decreased but was not significantly affected by P and Fe nutrition.

However,

QMCYSTmax showed evidence of optima at intermediate temperature (26°C) and irradiance (40 µmol photons m-2 s-1). The strong effect of S and Fe nutrition on QMCYSTmin suggests that these elements play an important role in MCYST production. Due to the relationship between µ and QMCYST, environmental factors that affect µ will therefore affect QMCYST. The data in the thesis emphasise the importance of µ in determining QMCYST under a particular set of growth conditions and indicate that QMCYST can be predicted provided the parameters µ max, QMCYSTmax and QMCYSTmin are known for those conditions.

The thesis also highlights the

physiological inadequacies of determining MCYST relative to biomass indicators other than cell number.

xiv Statement of Authorship Except where reference is made in the text of the thesis, this thesis contains no material published elsewhere or extracted in whole or in part from a thesis by which I have qualified or been awarded another degree or diploma.

No other person’s work has been used without due acknowledgment in the main text of the thesis.

This thesis has not been submitted for the award of any degree or diploma in any other tertiary institution.

A publication based on the material reported in Chapter 3 of this thesis is listed below.

BENEDICT MICHAEL LONG Melbourne, 28 February 2001

Publication:

LONG B.M., JONES G.J. & ORR P.T. 2001. Cellular microcystin content in N-limited Microcystis aeruginosa can be predicted from growth rate. Applied and Environmental Microbiology 67: 278-283.

xv Acknowledgments I would like to express my sincere thanks to my supervisor, John Anderson, for his guidance friendship throughout the project and for his infectious enthusiasm for all things scientific. It has been an enormous privilege to learn from John, both as an undergraduate and as a post-graduate student. I also extend my thanks to all those in the Botany Department who have helped in one way or another throughout my stay and made life in the department a great joy. I would especially like to thank my fellow students for sharing the experience of PhD candidature, especially Olga Nicolaou, Nicole Morcom and Joshua Heazlewood, and to the new breed of students making their way through post-graduate studies in the department. I would also like to thank Peter Crouch for the daily chance to chat over coffee and discuss how to fix the HPLC.

I must acknowledge the unselfish assistance of Phil Orr (CSIRO Land & Water, Brisbane) who provided invaluable discussion and great ideas. Also to Gary Jones (CSIRO Land & Water, Brisbane/ CRC for Freshwater Ecology, Canberra University), whose faith in my ability has developed my interest cyanotoxins.

Together Gary and Phil provided the impetus for the

direction I decided to take in my project, and ultimately what I would discover. For that I am extremely grateful. I also acknowledge the assistance of John Beardall (Monash University) in teaching me about the intricacies of chemostats, and how mind-bogglingly difficult they can be to understand. I am also indebted to the CSIRO Microalgae Research Centre (Hobart) for supplying the axenic strain of Microcystis aeruginosa used in the project. I am very grateful for the mathematical assistance of Dr Rouben Rostamian (University of Maryland) in determining Equations 4.2 and 4.3. Thanks also to Edgar Sakers, for creating the ingenious little circuit used to detect culture density in my turbidostat system, and to Trevor Philips for his imaging expertise. Thanks must also go to Margaret Lingam for the final proof reading of the thesis.

Peter Mascini must take some blame in leading me up the garden path to the Botany Department, and for that I am extremely grateful. He not only taught me as an undergraduate in Botany, but taught me how to teach undergraduate Botany. Thanks for everything Pete. I would also like to thank his mother Lorraine for her kindness and friendship, and especially the great food.

Last but not least I want to extend my thanks to my family and close friends for their love and support throughout my time as a student. I am indebted to Mum and Dad, not only for their love but for the opportunities they have given me. I am glad they finally ‘retired’ and planted a vineyard for us to play in. Most of all I want to thank my loving wife and best friend, Jenny. She has endured this thesis as much as I have and she has kept me going all the way through. Thank you for everything.

1

CHAPTER 1 Toxic cyanobacterial blooms and microcystin production

A toxic bloom of Microcystis aeruginosa superimposed by microcystin-LR

Chapter 1 Introduction

2

This thesis is concerned with the intracellular content and production of the cyclic peptide toxins, microcystins (MCYSTs), in the organism from which they take their name, the cyanobacterium Microcystis aeruginosa Kütz. emend. Elenkin. The importance of the MCYSTs is highlighted by their potent toxicity in a wide range of organisms, including humans, and their relative abundance in the environment. MCYST production by M. aeruginosa has been studied at length by a number of authors yet the cellular physiology of MCYST production remains unknown. This chapter summarises some of the important physiological features of cyanobacteria and introduces the current knowledge of MCYSTs.

This information is used to identify several problems

associated with the current understanding of MCYST production by these organisms. These problems form the basis of the investigations described in the ensuing chapters.

1.1

Characteristics of cyanobacteria

The cyanobacteria are a unique group of prokaryotic organisms considered amongst the most ancient of all life. Fossilised remains of unicellular and filamentous forms of cyanobacteria have been found in sedimentary rocks formed 3.5 billion years ago (Wilmotte 1994; van den Hoek et al. 1995). Comparison of the fossil record with present day cyanobacteria suggests that this group of organisms has undergone very little change since it first evolved (Schopf 1994). Indeed, even those current day cyanobacteria that live in disparate climates such as hot springs (Ward et al. 1994) and the Antarctic (Hitzfeld et al. 2000) have counterparts in the fossil record, suggesting that the physiological diversity of cyanobacteria has not arisen from recent evolutionary change (Schopf 1994). Most contemporary cyanobacteria are found in freshwater, marine and terrestrial environments. Some cyanobacteria are often found in symbiotic associations with organisms as diverse as fungi (i.e. lichens, Honegger 1993), cycads (Ahern & Staff 1994), liverworts (Knight & Adams 1996), and marine sponges (Hinde et al. 1994) amongst others (Fay 1983). However, free-living, aquatic cyanobacteria are abundant and generally the most conspicuous, forming blooms in both freshwater and brackish/marine environments (Mur et al. 1999).

The predominance of cyanobacteria in earth's early history has been a factor contributing to an oxygen rich atmosphere (Schopf 1994) since the majority carry out oxygenic photosynthesis, utilising H2O as an electron donor for CO2 reduction (Mur et al. 1999). Like eukaryotic algae, cyanobacterial cells contain photosynthetic pigments located in thylakoids and the light reactions of photosynthesis involve photosystems I and II as in higher plants (van den Hoek et al. 1995). It was for such reasons that the cyanobacteria were originally thought to be members of the algae and hence given the name 'blue-green algae'.

However, the cellular arrangement of

photosynthetic pigments in thylakoids occurring freely in the cytoplasm, and the lack of


3

membrane-bound organelles, clearly define the cyanobacteria as prokaryotic organisms (van den Hoek et al. 1995). In addition, the cyanobacteria contain only chlorophyll a (not chlorophylls b and c), utilise phycobilins (giving them their blue colour) as accessory pigments in photosynthesis (van den Hoek et al. 1995), and carry out protein synthesis on 70S ribosomes exclusively (Fay & van Baalen 1987). Thus, the name cyanobacteria indicates their position in relation to the eubacteria, and defines their uniqueness. The cyanobacteria are therefore assigned to their own division (Cyanophyta) in the kingdom Eubacteria (van den Hoek et al. 1995). Despite this, there is compelling evidence to suggest that the cyanobacteria have much in common with the chloroplasts of green plants (Giovannoni et al. 1988), suggesting a symbiotic event in evolutionary history which led to the emergence of the eukaryotic photosynthetic cells of eukaryotic algae and higher plants.

The cyanobacteria exist in a variety of forms including unicellular (e.g. Synechococcus), colonial (e.g. Microcystis), filamentous (e.g. Planktothrix) and pseudoparenchymatous (e.g. Pleurocapsa). In most cases, cyanobacterial cells are embedded in a mucilaginous sheath of hydrated polysaccharide. In some species, including Microcystis spp., the sheath creates an environment suitable for bacterial associations (Parker 1982) and, hence, colonial forms. The cells themselves are enclosed by a plasma membrane surrounded by a peptidoglycan wall and enveloped by a lipopolysaccharide layer (van den Hoek et al. 1995). Cyanobacterial cells contain a variety of electron dense inclusions including carboxysomes (or polyhedral bodies) which contain the primary enzyme for CO2 assimilation, ribulose-1,5-bisphosphate carboxylase-oxygenase (RuBisCO) (Allen 1984).

Phycobilisomes, which play a major role in energy transfer in

photosystem II, are the most abundant structures within cyanobacterial cells (Allen 1984). Most cyanobacterial cells also contain cyanophycin granules (multi-L-arginyl-poly-[L-aspartic acid]), polyphosphate and polyglucose bodies, all of which are utilised as nutrient storage compounds (Allen 1984). These storage compounds are accumulated when particular nutrients occur in excess in the environment. Conversely, cyanobacteria make use of these stores when nutrient limitation occurs. Under conditions of P stress, for example, polyphosphate stores in some cyanobacteria are sufficient to support cell division for four to five cell doublings when the external source of P has become limiting (Kromkamp et al. 1989).

Many cyanobacterial cells contain protein-bound spaces filled with gas, known as gas vesicles, (Allen 1984) which help aquatic species to control their depth in the water column by adjusting their buoyancy (Walsby 1994). This allows cyanobacteria to not only utilise the euphotic zone for photosynthesis but also access the nutrient-rich waters of the hypolimnion (Fogg & Walsby 1971).

During photosynthesis in the euphotic zone, carbohydrate accumulation leads to an


4

increase in the osmotic movement of water into the cells. Since the cells are protected from rupture by their peptidoglycan wall, the turgor pressure increases, causing gas vesicles to collapse. This results in an increase in cell density so the cells sink in the water column where they can access soluble nutrients. Here, there is a net consumption of carbohydrate with the result that the osmotic uptake of water into the cells decreases and gas vesicles re-form, allowing the cells to rise to the euphotic zone to photosynthesise (Oliver & Walsby 1984; Utkilen et al. 1985). There is now evidence to suggest that this process is light-dependent and driven by carbohydrate accumulation (Klemer et al. 1996; Bormans et al. 1999), and that gas vesicles play a minor role in buoyancy (Rijn & Shilo 1985). The isolation of both vacuolate and non-vacuolate strains of some cyanobacteria is further evidence for this (Bolch & Blackburn 1996).

Some species of cyanobacteria are capable of fixing atmospheric nitrogen (N2), allowing them to grow in waters where soluble inorganic forms of N are limited. Predominantly, dinitrogen fixation occurs in the heterocystous filamentous species of cyanobacteria although there are reports of this occurring in several non-heterocystous species (Bergman et al. 1997). Where reduced forms of N (e.g. NH4+) are limited, blooms of N2-fixing genera (e.g. Anabaena, Aphanizomenon, Cylindrospermopsis, Nodularia and Nostoc) are common (Mur et al. 1999).

In temperate regions, as the weather becomes colder, respiration slows relative to photosynthesis in aquatic cyanobacteria. As a result, there is a net accumulation of carbohydrate in cells and they remain at depth for longer periods. Over winter, cells may survive at the bottom of a water body, utilising their carbohydrate stores as respiration continues at a very slow rate (Mur et al. 1999). Eventually as their density decreases, surviving cells return to the euphotic zone and rapid growth can resume with the subsequent onset of warmer temperatures and increased irradiance. Aquatic cyanobacteria are often found as dense populations (blooms) in temperate lakes and slow-moving streams or rivers, especially when temperatures rise and the water column becomes stratified. Under these conditions, growth can become so rapid that the concentration of dissolved CO2 in the water column becomes limiting and photosynthesis slows (Walsby 1977).

As a result,

cyanobacterial cells undergo a net loss of carbohydrate, they become more buoyant (Walsby 1977), and a visible surface scum of cyanobacteria appears. In a surface scum, the cells utilise any remaining soluble nutrients. Sometimes the dense packing of the cells in a surface scum may result in anoxia (Revsbech et al. 1983) and anaerobic respiration (Moezelaar & Stal 1994). Cells that become confined to the water surface are exposed to extended periods of high irradiance which commonly leads to photooxidative death (Abeliovich & Shilo 1972) and cell lysis. This series of events creates conditions that are favourable for the production of toxic water supplies, as intracellular toxins are released directly into the water (Sivonen & Jones 1999).


1.2

5

Factors influencing cyanobacterial dominance in freshwater systems

In many freshwater systems, cyanobacterial growth dominates the phytoplankton, causing scumforming blooms. Due to the physiological differences between cyanobacterial species and the complex nature of the interaction of environmental factors on cyanobacterial growth, it is difficult to identify individual factors allowing a single species to dominate (Mur et al. 1999). This section summarises the growth responses of cyanobacteria to individual environmental factors contributing to the dominance of cyanobacteria in aquatic environments.

1.2.1

Nutrients

Cyanobacterial growth, and for that matter the growth of most phytoplankton, is often limited by the nutrients N and P in freshwater systems (Lange 1971). Data suggest that many cyanobacterial species have high affinities for the uptake of these nutrients compared with other phytoplankton (Mur et al. 1999), giving cyanobacteria a distinct advantage under conditions where these nutrients are available in low concentrations. Moreover, cyanobacteria require N and P in smaller quantities than many other phytoplankton and have lower optimum N:P ratios (Rhee & Gotham 1980). In addition, many cyanobacteria are able to fix N2, and most store cellular reserves of N and P as cyanophycin and polyphosphate respectively (Section 1.1). This allows cell growth to proceed for a short time in the complete absence of extracellular sources of these nutrients as the reserve materials are utilised (Allen 1984; Kromkamp et al. 1989). The form in which these nutrients are available may also play an important role in cyanobacterial dominance since ammonia promotes cyanobacterial growth under some conditions while nitrate does not (Blomqvist et al. 1994). The availability of nutrients is also affected by factors such as climate (e.g. rainfall), shape and depth of the water body (which affect the recirculation of nutrients from bottom sediments), drainage basin characteristics (i.e. geology and morphology affect catchment size and the rate of nutrient input into a water body), and flow rate (slow moving water bodies maintain a higher nutrient loading) (Heeps 1996). Although all phytoplankton are capable of taking advantage of eutrophic waters, characteristics such as buoyancy control, production of nutrient storage compounds and their persistence at relatively slow growth rates, regardless of the prevailing growth conditions, often mean that cyanobacteria dominate the phytoplankton in the long term (Mur et al. 1999).

Despite the advantages cyanobacteria have over other phytoplankton when nutrients are limiting, high nutrient loads also contribute to cyanobacterial blooms. Human activity (especially intensive farming practices and domestic use of water resources) has resulted in a rise in the eutrophication of water supplies and a marked increase in the occurrence of cyanobacterial blooms (Skulberg et


6

al. 1984). An increase in the concentration of dissolved nutrients in water bodies often results in a dramatic increase in phytoplankton biomass, and cyanobacteria often dominate. A contributing factor here is that domestic and farming activity often results in the production of forms of nutrients, such as reduced forms of N (e.g. NH4+), which cyanobacteria utilise preferentially (Blomqvist et al. 1994). Polluted water systems also have increased concentrations of other nutrients essential for cyanobacterial growth (e.g. P, S, Ca, Mg and Fe) so that blooms can easily form once a cyanobacterial species has become dominant. Under these conditions growth may be limited by low CO2 or O2 availability as scum formation occurs.

1.2.2

Light

Cyanobacteria have the ability to use a wide range of wavelengths in the visible spectrum due to the presence of an array of photosynthetic pigments including the phycobilins (phycoerythrin (red), phycocyanin (blue) and allophycocyanin (blue)) chlorophyll a and a number of carotenoids (Fay 1983). In combination, these compounds allow the absorption of light between 400 and 700 nm (Fay 1983), providing a unique advantage over other phytoplankton. Also, cyanobacteria are able to adapt to changes in light quality (i.e. spectral composition) through a process known as complementary chromatic adaption (Tandeau de Marsac & Houmard 1988), thus permitting them to compete successfully in environments with a restricted spectral range. In addition, aquatic cyanobacteria can grow at very low irradiance compared with their eukaryotic counterparts, and are capable of continued growth despite overshading by other phytoplankton (van Liere & Walsby 1982).

In the long term, persistent but slow growth under low irradiance can lead to

cyanobacterial dominance (Mur et al. 1978). This, coupled with the ability of cyanobacteria to regulate their buoyancy, also permits cyanobacteria to compete successfully with other phytoplankton in highly turbid environments.

1.2.3

Temperature

Robarts & Zohary (1987) found that maximum growth rates of many bloom-forming cyanobacteria are achieved at temperatures greater than 25 °C. This is higher than for green algae and diatoms (Mur et al. 1999). In temperate freshwater environments in summer, the water column becomes thermally stratified and the temperature in the warm upper level (the epilimnion) often reaches 20 to 25 °C (Welch 1952). This relatively high temperature, which benefits cyanobacterial growth, is a likely reason for the predominance of cyanobacterial blooms in temperate and boreal water bodies during summer months (Mur et al. 1999). As noted above, in turbid water, the buoyancy control offered by carbohydrate and the production of gas vesicles


7

enables cyanobacteria to make use of the warmer conditions as well as the higher irradiance in the epilimnion (Mur et al. 1999).

1.2.4

CO2 and pH

Aquatic cyanobacteria are expert scavengers of CO2, utilising both free CO2 and bicarbonate ions in the water (Tabita 1987). Green algae, on the other hand, preferentially take up bicarbonate and dehydrate it to CO2 for use in photosynthesis. The assimilation of CO2 by cyanobacteria is catalysed by RuBisCO, which can account for up to 30 % of total cellular protein in some species (Fay 1983).

Thus, cyanobacteria devote considerable cellular resources to carbon fixation

enabling them to compete successfully in waters that contain relatively low concentrations of dissolved CO2.

In addition, several species of cyanobacteria are capable of anoxygenic

photosynthesis, fixing CO2 using H2S as an electron donor, while others can utilise phosphoenol pyruvate carboxylase to fix CO2 in dim light (Fay 1983). During periods of rapid phytoplankton growth, the rate of uptake of dissolved CO2 into phytoplankton can often exceed the rate at which it is replaced from the atmosphere. However, cyanobacteria generally have lower half-saturation constants for CO2 uptake compared with other phytoplankton (Shapiro 1990) and therefore an advantage over green algae in these situations.

Freshwater aquatic cyanobacteria are often found in alkaline waters (e.g. Dwivedi et al. 1994) and grow well at relatively high pH (Gerloff et al. 1950). For example, the photosynthetic rate of Anabaena flos-aquae increases with pH up to pH 10 (Olofsson & Woodward 1977: cited by Shapiro 1990). Despite living in a highly alkaline external environment, several species of cyanobacteria can maintain an almost neutral cytosolic pH (Belkin & Boussiba 1991; Dwivedi et al. 1994). Shapiro (1973) suggested that the dominance of cyanobacteria at high pH/low CO2 is due to their superior CO2 uptake kinetics compared with other phytoplankton. On the other hand, green algae tend to dominate when bicarbonate concentrations are high and the pH is relatively low. Thus, in water bodies where pH does not increase significantly in summer, cyanobacteria tend not to dominate the phytoplankton (Shapiro 1990). A review of the important role of high pH and low CO2 availability in contributing to cyanobacterial dominance in freshwater bodies is provided by Shapiro (1990).

1.2.5

Grazing

In the freshwater environment, phytoplankton represent a large biomass of primary production utilised directly by zooplankton and planktivorous fish as a source of nutrition.

However,


8

cyanobacteria tend to have relatively low nutritional value either because they are difficult to digest or they lack essential nutrients for zooplankton growth and reproduction (DeMott & Moxter 1991). Moreover, zooplankton have some difficulty ingesting certain cyanobacterial species due to their large colonial or filamentous morphologies (Ganf 1983; DeMott & Moxter 1991; Kirk & Gilbert 1992). In addition, chemical defences, including the cyanobacterial toxins (MCYSTs) examined in this thesis, can inhibit the ingestion of cyanobacteria by particular zooplankton, adversely affecting their survival (DeMott & Moxter 1991; DeMott et al. 1991; DeMott 1999; Rohrlack et al. 1999). There is also evidence that cyanobacterial toxins adversely affect planktivorous fish (Gaete et al. 1994; Carbis et al. 1996; Carbis et al. 1997) perhaps acting as a deterrent towards further grazing. Together, these factors contribute to a relatively stable population of many cyanobacterial species within a water body, often at the expense of eukaryotic green algae and other phytoplankton.

As a result, cyanobacteria tend to dominate the

phytoplankton when environmental conditions favour their growth.

1.3

The occurrence of toxic cyanobacterial blooms

The first scholarly report of a toxic cyanobacterial bloom concerned the death of sheep after drinking water contaminated with Nodularia spumigena in Lake Alexandrina, South Australia, at the mouth of the Murray River (Francis 1878; Codd et al. 1994). However, there are also records that toxic cyanobacterial blooms were responsible for stock deaths from 1833 in Denmark (Moestrup 1996). There is also anecdotal evidence of toxic blooms from twelfth century Scotland (Codd 1996) and even a report of a cyanobacterial or algal scum in the Roman era from AD 77 (Codd et al. 1994). It is believed that, for centuries Australian aborigines have been aware of the toxicity of water bodies containing blooms as it was general practice to gather water from waterholes dug in the shores of lakes and rivers to allow filtration through the sand before drinking. Thus, despite an apparent increase in their occurrence (Skulberg et al. 1984; Bartram et al. 1999), the formation of toxic cyanobacterial blooms is not a recent phenomenon.

Geographically, toxic cyanobacteria have a wide distribution (Hoffmann 1996).

Toxicity

attributable to cyanobacteria has now been described on all continents (e.g. Skulberg et al. 1984; Galey et al. 1987; Jones 1994; van Halderen et al. 1995; Mahakhant et al. 1998; Yunes et al. 1998; Hitzfeld et al. 2000), making public awareness and management of toxic blooms a matter of widespread concern (Chorus & Bartram 1999). Comprehensive lists of the locations of published toxic cyanobacterial blooms and the toxins they contained are provided by Rapala (1998) and Sivonen & Jones (1999), although reports of toxic blooms in previously uninvestigated locations are becoming more frequent (e.g. Bouvy et al. 2000; Mohamed & Carmichael 2000). Generally,


9

toxic cyanobacterial blooms occur during late summer and into autumn in temperate and boreal regions (Mur et al. 1999; Sivonen & Jones 1999) when the demand for water for both agricultural and urban use is greatest. In regions with Mediterranean or sub-tropical climates, blooms may occur earlier in summer and persist for longer periods (Sivonen & Jones 1999).

Many cyanobacterial blooms tend to involve a single cyanobacterial species. However, some blooms involve several genera. More than half of the reported cyanobacterial blooms are toxic (Rapala 1998; Sivonen & Jones 1999) while, within a toxic bloom, both toxic and non-toxic strains of a single cyanobacterial species can often be found (Carmichael & Gorham 1981). This distribution makes the prediction of bloom toxicity difficult. Thus, until appropriate examination of the water is carried out, water managers must consider all cyanobacterial blooms toxic. As noted above, mass occurrences of several cyanobacterial genera can often be found in a single water body, producing different toxins and contributing to extremely poisonous conditions.

Generally, most of the toxic cyanobacterial blooms analysed contain the liver damaging hepatotoxins, MCYSTs and/or nodularins (Sivonen & Jones 1999). Blooms containing these toxins most commonly contain the toxic genera Microcystis, Anabaena, Planktothrix (syn. Oscillatoria) and Nodularia, which have been reported almost worldwide (Sivonen & Jones 1999). Interestingly, hepatotoxic blooms of Anabaena have not yet been reported in Australia. Neurotoxic blooms (affecting the nervous system) are also common worldwide and are most commonly associated with the genus Anabaena, although they are produced by several other genera (Sivonen & Jones 1999; Section 1.4).

1.4

The genus Microcystis

Microcystis spp. are widespread in their distribution. Blooms of this genus have been reported throughout Europe, the Americas, Africa, Australia and Asia (Rinehart et al. 1994; Sivonen & Jones 1999). Microcystis accounts for a very large proportion of toxic freshwater blooms so far reported (Sivonen & Jones 1999), thereby making it by far the most important cyanobacterium in this regard. The toxicity of the genus is attributable predominantly to the potently hepatotoxic MCYSTs (Table 1.1; Sivonen & Jones 1999).

Microcystis is a single-celled, planktonic cyanobacterium that forms clathrate (net-like) colonies, ranging in size from macroscopic to microscopic, in which the cells are densely aggregated in a mucilaginous envelope (Baker 1992). Cell size and shape as well as the presence or absence of gas vacuoles, are distinguishing features amongst different species of Microcystis (Baker 1992),


10

while colony morphology is a characteristic feature of several formae within species (Reynolds et al. 1981; Baker 1992). The most commonly reported bloom-forming species of Microcystis is M. aeruginosa, which is delineated from other species by large cells (> 2 µm in diameter) containing gas vacuoles and a colonial mucilage whose boundary is not clearly defined by light microscopy (Baker 1992; Watanabe 1996). Under some conditions in culture, M. aeruginosa often loses its colonial morphology after extended periods, and grows as isolated cells (e.g. Hughes et al. 1958). M. aeruginosa is further divided into two formae (M. aeruginosa f. aeruginosa (Kütz.) Elenkin and M. aeruginosa f. flos-aquae (Wittr.) Elenkin). Forma aeruginosa has large cells (4 to 7 µm in diameter) which aggregate predominantly in irregularly shaped colonies, while forma flos-aquae has smaller cells (2.5 to 4.5 µm in diameter) and more compact, regular (e.g. spherical) shaped colonies (Baker 1992). Both formae are reportedly toxic (Baker 1992).

Several other species of Microcystis (viz. M. viridis (A. Br.) Lemm., M. wesenbergii (Kom.) Kom. in Kondr., M. novacekii (Kom.) Compère and M. ichtyoblabe Kütz.) also form hepatotoxic blooms (Namikoshi et al. 1992d; Watanabe 1996). However, M. aeruginosa appears to be the most prevalent and widespread toxic species of this genus (Sivonen & Jones 1999) and is therefore regarded as the most important. In addition to producing the potently hepatotoxic MCYSTs, M. aeruginosa produces a suite of novel toxins and metabolic inhibitors. Many of these compounds (e.g. microviridins, micropeptins, cyanopeptolins and aeruginosins; Ishida et al. 1995; Jakobi et al. 1995; Okino et al. 1995; Ishida et al. 1999) are bioactive peptides which can occur in cells in greater quantities than MCYSTs (e.g. microginin FR1, Neumann et al. 1997). However, the considerably higher molar toxicity of MCYSTs and the widespread distribution of freshwater M. aeruginosa blooms mean that the production of MCYSTs by blooms of this species poses a particular threat to public health.

1.5 1.5.1

The Microcystins (MCYSTs) MCYST structure and occurrence in cyanobacteria

Konst (1965) first used the term microcystin (MCYST) to describe the toxic factor produced by Microcystis aeruginosa NRC-1. However, the MCYSTs have been known by various other names including fast-death factor (Bishop et al. 1959), cyanoginosin (Botes et al. 1984) and cyanoviridin (Kusumi et al. 1987). A standard nomenclature for these compounds was suggested by Carmichael et al. (1988). A family of more than 60 MCYSTs has now been described from a number of fresh and brackish water, bloom-forming cyanobacterial genera (Table 1.1) and several other sources (e.g. marine mussels, Chen et al. 1993).


11

MCYSTs are cyclic heptapeptides (consisting of seven amino acids). The general structure can be represented in abbreviated form as follows: 1

2

3

4

5

6

7

Cyclo(-|D-Ala|-|L-X|-|D-erythro--Methylasp|-|L-Z|-|Adda|-|D-Glu|-|N-Methyldehydroala|). The amino acids are numbered sequentially from the D-alanine residue (D-Ala, Fig. 1.1) enabling variations in this structure to be easily identified. Adda, at position 5, is the unusual C20 amino acid (2S,3S,8S,9S)3-amino-9-methoxy-2,6,8-trimethyl-10-penyldeca-4,6-dienoic acid (Fig. 1.1). N-Methyldehydroala (N-Methyldehydroalanine, Mdha), at position 7, is possibly a dehydrated, methylated L-serine residue (Pearce & Rinehart 1979; Namikoshi et al. 1992a). The amino acids D-erythro--Methylasp (D-erythro--methylaspartate),

at position 3, and D-Glu (D-glutamate), at

position 6, are iso-linked in the peptide via their and carboxyl groups respectively (Fig. 1.1). The amino acids at positions 2 and 4, identified as X and Z in the general structure, are L amino acids, the identity of which varies between the various specific MCYSTs. According to the nomenclature of Carmichael et al. (1988), this feature is used for naming each of the microcystins (MCYSTs). This involves representing the abbreviated forms of the amino acids (using the L amino acid single letter code nomenclature) at positions 2 and 4 respectively, as a suffix for each MCYST.

Thus, MCYST-LR contains the amino acids L-leucine (L) and L-arginine (R) at

positions 2 and 4 (positions X and Z in the general structure), respectively.

Using this

terminology a growing family of MCYSTs with variable L amino acids has been described, including MCYSTs with structural modifications in positions 1, 3, 5, 6 and 7 (Table 1.1; Fig. 1.1). From the currently characterised MCYSTs the molecular masses of these peptides ranges from 909 to 1115 (Table 1.1) due to variation in L amino acids or other modifications in their structure.


12

Fig. 1.1. The chemical structure of MCYST-LR. The constituent amino acids are as follows: 1, D-alanine; 2, L-leucine; 3, D-erythro--methylaspartate (iso-linked to amino acid 4); 4, L-arginine; 5, (2S,3S,8S,9S)3amino-9-methoxy-2,6,8-trimethyl-10-penyldeca-4,6-dienoic acid (Adda); 6, D-glutamate (iso-linked to amino acid 7); 7, N-Methyldehydroalanine (Mdha). The single letter code nomenclature for the constituent L amino acids at positions 2 and 4 (in this case leucine (L) and arginine (R)) is utilised in defining specific MCYSTs. Other MCYSTs are named according to the nomenclature of Carmichael et al. (1988) using the position numbers given in the general structure (see the text). For example, an MCYST demethylated in positions 3 and 7, and containing L-arginine in positions 2 and 4 is given the name [D-Asp3,Dha7]MCYSTRR.

The MCYSTs are very stable peptides. They are resistant to chemical hydrolysis and oxidation at neutral pH, and even remain toxic after boiling (Sivonen & Jones 1999). In addition, the cyclic structure of MCYSTs make them resistant to enzymic hydrolysis by some proteases (e.g. trypsin). Under low light conditions MCYSTs can remain stable for long periods (perhaps years, Sivonen & Jones 1999). Jones & Orr (1994) found that the MCYSTs released into the water, after treating a bloom of M. aeruginosa with an algicide, remained stable for several days before degradation began at a slow rate. Crusts of M. aeruginosa surface scums found on a lake shore still contained MCYSTs after five to six months (Jones et al. 1995). However, under high irradiance (especially in the presence of photosynthetic pigments or metal catalysts) MCYSTs can be degraded rapidly (Takenaka & Tanaka 1995; Tsuji et al. 1995; Twist & Codd 1997; Shephard et al. 1998). Also, using UV irradiation the potently toxic MCYST-LR (Table 1.1) can be rendered almost non-toxic by the formation of [(6Z)Adda5]MCYST-LR (Table 1.1) (Tsuji et al. 1995).

Bacterial degradation of MCYSTs is likely once they are released into the water (Jones et al. 1994). Bourne et al. (1996) discovered a Sphingomonas species capable of degrading MCYST-


13

LR, and determined the catabolic pathway involved. In animals, detoxification appears to involve the formation of a glutathione-MCYST adduct (Pflugmacher et al. 1998).

First reported in the genus Microcystis, MCYSTs have been found in the planktonic cyanophytes Anabaena, Anabaenopsis, Nostoc, Phormidium (syn. Oscillatoria) and Planktothrix (syn. Oscillatoria) (Sivonen & Jones 1999) and also in the terrestrial genus Hapalosiphon (Prinsep et al. 1992). With the exception of the discovery of MCYST-LA in Hapalosiphon, MCYSTs appear to be confined to the genera Chroococcales, Nostocales and Oscillatoriales (Table 1.1). The glutamic acid methyl-esterified (E(OMe)) MCYSTs are thus far found only in members of the Nostocales while MCYSTs containing L-homo-amino acids in the Z position (position 2 of the MCYST structure, Fig. 1.1) are found almost exclusively in the Nostocales and Oscillatoriales (Table 1.1). In addition, the acetylated variant of the amino acid Adda (ADMAdda, O-acetyl-OdemethylAdda) has so far been described in the only Nostocales (Table 1.1; Rinehart et al. 1994). The specificity of some MCYST variants to certain orders of cyanobacteria is probably due to differences in the biochemistry of MCYST biosynthesis, suggesting some form of divergent evolution of MCYST production processes. However, further investigation may reveal a more widespread distribution of MCYST variants amongst cyanobacteria. In summary, the data in Table 1.1 show that more MCYSTs have been reported in Microcystis than in any other genus, consistent with the well known toxic blooms of this organism.

TABLE 1.1 MCYSTs, their toxicity and the cyanobacterial genera in which they have been characterised to date. 1, 2 Molecular

LD50

weight

(µg kg-1) 4

MCYST-LA

909

50

MCYST-LAba

923

NR

MCYST-LL

951

MCYST-AR MCYST-YA

Microcystin 3

Genus/Genera (Order)

Reference(s)

Hapalosiphon (Stigonematales), Microcystis

(Botes et al. 1984; Kaya & Watanabe 1990; Prinsep et al.

(Chroococcales)

1992)

Microcystis (Chroococcales)

(Gathercole & Thiel 1987)

+


(Craig et al. 1993)

952

250


(Namikoshi et al. 1992d)


(Botes et al. 1985)

959

NR

3

7

966

+

Microcystis (Chroococcales), Anabaena (Nostocales)

(Harada et al. 1991b; Sivonen et al. 1992b)

3

7

[D-Asp ,Dha ]MCYST-EE(OMe)

969

+

Anabaena (Nostocales)

(Namikoshi et al. 1998)

MCYST-VF

971

NR


(Bateman et al. 1995)

980

160-300

[D-Asp ,Dha ]MCYST-LR

3

[D-Asp ]MCYST-LR 7

[Dha ]MCYST-LR

[DMAdda]MCYST-LR

980

250

Microcystis (Chroococcales), Anabaena (Nostocales),

(Krishnamurthy et al. 1989; Harada et al. 1990b; Cremer &

Planktothrix [syn. Oscillatoria] (Oscillatoriales)

Henning 1991; Luukkainen et al. 1993)


(Harada et al. 1991b; Sivonen et al. 1992b; Luukkainen et

Planktothrix (Oscillatoriales)

al. 1993)

Microcystis (Chroococcales), Nostoc (Nostocales)

(Namikoshi et al. 1992d; Sivonen et al. 1992d)

980

90-100

3

7

981

70


(Sano et al. 1998; Sano & Kaya 1998)

3

7

981

NR


(Sano et al. 1998)

983

+



[D-Asp ,Dha ]MCYST-E(OMe)E(OMe)

983

+



MCYST-LF

985

+


(Azevedo et al. 1994)

MCYST-LR

994

50


(Botes et al. 1985; Rinehart et al. 1988; Watanabe et al.

Aphanocapsa (Chroococcales), Phormidium [syn.

1988; Krishnamurthy et al. 1989; Domingos et al. 1999;

Oscillatoria] (Oscillatoriales)

Brittain et al. 2000)

[D-Asp ,(E)-Dhb ]MCYST-LR [D-Asp ,(Z)-Dhb ]MCYST-LR 7

[Dha ]MCYST-EE(OMe) 3

7

continued

14

Table 1.1 continued Molecular

LD50

weight

(µg kg-1)

994

NR

[(6Z)-Adda ]MCYST-LR

994

>1,200

[Dha7]MCYST-E(OMe)E(OMe)

997

+

7

998

+


(Namikoshi et al. 1992a)

MCYST-LY

1,001

90


(Stoner et al. 1989)

1,001

+



[D-Asp ,L-Ser ]MCYST-E(OMe)E(OMe)

1,001

+



MCYST-HilR

1,008

100



1,008

160

Nostoc (Nostocales)

(Namikoshi et al. 1990; Sivonen et al. 1990)

1,008

>1,000


(Sivonen et al. 1992a; Rinehart et al. 1994; Bateman et al.

Microcystin [D-Asp3,D-Glu(OCH3)6]MCYST-LR 5

[L-Ser ]MCYST-LR 7

[L-Ser ]MCYST-EE(OMe) 3

7

3

5

[D-Asp ,ADMAdda ]MCYST-LR 6

[D-Glu(OCH3) ]MCYST-LR


Reference(s)


(Sivonen et al. 1992a)


(Harada et al. 1990b; Harada et al. 1990a)



1995) MCYST-LHar

1,009

NR

3

1,009

+

7

[D-Asp ,Dha ]MCYST-RR 3

5

7

[D-Asp ,ADMAdda ,Dhb ]MCYST-LR

Phormidium [syn. Oscillatoria] (Oscillatoriales)

(Brittain et al. 2000)


(Krishnamurthy et al. 1989; Sivonen et al. 1992b;


Luukkainen et al. 1994)

Nostoc (Nostocales)

(Beattie et al. 1998)

1,009

+

7

1,012

150


(Namikoshi et al. 1992d; Namikoshi et al. 1995)

[Dha ]MCYST-FR

1,014

NR


(Luukkainen et al. 1994)

1,015

+



1,022

60

Nostoc (Nostocales)


1,022

+

Nostoc (Nostocales)

(Sivonen et al. 1992d)

1,023

250


(Meriluoto et al. 1989; Sivonen et al. 1992b; Luukkainen et


al. 1994)

[L-MeSer ]MCYST-LR 7

7

[L-Ser ]MCYST-E(OMe)E(OMe) 5

[ADMAdda ]MCYST-LR 3

5

[D-Asp ,ADMAdda ]MCYST-LHar 3

[D-Asp ]MCYST-RR

continued

15

Table 1.1 continued Molecular

LD50

weight

(µg kg-1)

[Dha7]MCYST-RR

1,023

180

MCYST-LW

1,024

NR

[D-Asp ,(E)-Dhb ]MCYST-RR

1,024

MCYST-FR MCYST-M(O)R

Microcystin

3

7

7

[Dha ]MCYST-HphR 3

7

[D-Asp ,Dha ]MCYST-HtyR 3

[D-Asp ]MCYST-YR 7


Reference(s)


(Kiviranta et al. 1992; Sivonen et al. 1992b; Luukkainen et


al. 1993)


(Bateman et al. 1995)

250


(Sano & Kaya 1998)

1,028

250



1,028

700-800



1,028

+


(Namikoshi et al. 1992b)

1,030

+



1,030

+


(Namikoshi et al. 1992c)

[Dha ]MCYST-YR

1,030

+


(Sivonen et al. 1992c)

MCYST-YM(O)

1,035

56


(Elleman et al. 1978; Botes et al. 1985)

[ADMAdda5]MCYST-LHar

1,036

60

Nostoc (Nostocales)


MCYST-RR

1,037

600


(Kusumi et al. 1987; Painuly et al. 1988; Watanabe et al. 1988; Sivonen et al. 1992b)

5

[(6Z)-Adda ]MCYST-RR 1

[D-Leu ]MCYST-LR 1

5

[D-Ser ,ADMAdda ]MCYST-LR 5

7

[ADMAdda ,MeSer ]MCYST-LR 7

1,037

>1,200 5


(Harada et al. 1990a)


(Matthiensen et al. 2000; Park et al. 2001)

1,038

100

1,038

+

Nostoc (Nostocales)

(Sivonen et al. 1992d)

1,040

+

Nostoc (Nostocales)

(Sivonen et al. 1992d) (Namikoshi et al. 1992a; Luukkainen et al. 1994)

]MCYST-RR

1,041

+


3

[D-Asp ,MeSer ]MCYST-RR

1,041

+


(Luukkainen et al. 1993)

MCYST-YR

1,044

70


(Botes et al. 1985; Watanabe et al. 1988)

1,044

160-300



[L-Ser

3

7

[D-Asp ]MCYST-HtyR

continued

16

Table 1.1 continued Microcystin [Dha7]MCYST-HtyR 3

7

Molecular

LD50

weight

(µg kg-1)

1,044

+

Genus/Genera (Order) Anabaena (Nostocales)

Reference(s) (Namikoshi et al. 1992b)

,(E)-Dhb ]MCYST-HtyR

1045

70


(Sano et al. 1998; Sano & Kaya 1998)

[D-Asp3,(Z)-Dhb7]MCYST-HtyR

1045

NR


(Sano et al. 1998)

MCYST-(H4)YR

1,048

NR



1,052

>1,000



[D-Asp ,ADMAdda ,Dhb ]MCYST-RR

1,052

200

Nostoc (Nostocales)


MCYST-HtyR

1,058

80-100



1,062

+



1,067

150-200



[D-Asp

6

[D-Glu-OC2H3(CH3)OH ]MCYST-LR 3

5

7

7

[L-Ser ]MCYST-HtyR MCYST-WR 3

5

7

[D-Asp ,ADMAdda ,Dhb ]MCYST-HtyR

1,073

100

[L-MeLan7]MCYST-LR

1,115

1,000

1 2

3

4

5

Nostoc (Nostocales)




Adapted and updated from Sivonen & Jones (1999). Not listed are several MCYSTs whose structures have not been completely characterised (see Boland et al. 1993; Craig et al. 1993; Jones et al. 1995; Sivonen et al. 1995, and Chapter 2 of this thesis). All MCYSTs are listed according to the nomenclature of Carmichael et al. (1988). Single letter code nomenclature for L amino acids is used. Other abbreviations are as follows: Aba aminoisobutyric acid Har homoarginine O-acetyl-O-demethylAdda Hil homoisoleucine ADMAdda Dha dehydroalanine Hph homophenylalanine Dhb dehydrobutyrine Hty homotyrosine (E)Dhb geometrical stereoisomer of Dhb MeLan methyllanthionine (Z)Dhb geometrical stereoisomer of Dhb M(O) methionine-S-oxide O-demethylAdda N-methylserine DMAdda MeSer JOXWDPLF DFLG PHWK\O HVWHU û (6Z)-Adda 6WHUHRLVRPHU RI $GGD DW WKH û6 double bond E(OMe) 1,2,3,4-tetrahydrotyrosine (H4)Y LD50 refers to the amount of each MCYST (in µg) required to kill 50 % of test animals (in kg) exposed to the toxin by intraperitoneal injection (i.p.). A '+' indicates that the MCYST has been found to be toxic in qualitative mouse bioassays or it inhibits protein phosphatase (PP) activity. 'NR' denotes that toxicity has not been reported. The toxicity value determined for [D-Leu1]MCYST-LR is the minimum lethal dose, LDmin (Matthiensen et al. 2000).

17


1.5.2

18

MCYST Toxicity

The importance of MCYST toxicity toward humans was made apparent in 1996 when patients at two haemodialysis centres in Caruaru (Pernambuco, Brazil) were poisoned when water taken from a reservoir containing a Microcystis bloom was used during dialysis (Azevedo 1997). A number of patients presented symptoms similar to those observed in animals suffering MCYSTinduced hepatotoxicosis, including diarrhoea, vomiting, headaches, muscle weakness, visual disturbances and painful hepatomegaly (Azevedo 1997; Jochimsen et al. 1998). Within just over a month 26 patients had died from liver failure (Jochimsen et al. 1998). By October of 1997, 49 deaths were attributed to what is now known as 'Caruaru Syndrome' (Jochimsen et al. 1998). Analysis of water, dialysis centre filtration systems, liver biopsies and post-mortem samples provided the evidence that MCYSTs from M. aeruginosa were responsible (Jochimsen et al. 1998; Pouria et al. 1998).

Also in Brazil, 88 deaths (mostly children) were reported from some 2,000 cases of gastroenteritis that occurred after drinking water from a newly flooded reservoir that developed an immense Anabaena/Microcystis bloom. Cases were confined to areas supplied with drinking water from the reservoir, and no other infectious agents or toxins could be found. This, and the fact that illness resulted even after drinking water was boiled, suggested that a cyanobacterial toxin (possibly MCYST) was responsible (Teixera et al. 1993: cited by Kuiper-Goodman et al. 1999). There has been a considerable number of cases of human illness which can be attributed to cyanobacterial toxins (Kuiper-Goodman et al. 1999) and it is quite possible that illness associated with drinking water may have been caused by cyanobacterial toxins more often than has been reported in the past.

MCYSTs cause liver damage in animals (including humans), primarily through potent and specific inhibition of protein serine/threonine phosphatases (PP) 1 and 2A (MacKintosh et al. 1990; Runnegar et al. 1993). In hepatocytes, PP inhibition leads to hyperphosphorylation of intermediate cytoskeletal filaments and subsequent cell disruption (Falconer & Yeung 1992). Orally ingested MCYST-LR is rapidly taken up and localised in the liver (Runnegar et al. 1986) through a multi-specific bile acid transport system present in the small intestine and in hepatocytes (Runnegar et al. 1991; Kuiper-Goodman et al. 1999). Once cell disruption occurs, hepatocyte disaggregation leads to hepatic haemorrhage and rapid death (Carmichael 1994). Hooser (1990) found that rats injected intraperitoneally (i.p.) with MCYST-LR (160 µg kg-1; a lethal dose) showed signs of hepatic damage (lesions) within 10 min after dosing and that liver damage progressed from centilobar degeneration and necrosis within 20 min to almost complete


19

lobule degeneration and necrosis within 60 min. Thus, MCYST toxicosis can be rapid and destructive at high doses, hence the term 'fast-death factor' given to the MCYSTs by Bishop et al. (1959) before their structure was determined.

The MCYSTs are not only potent hepatotoxins. There is some evidence of carcinogenic and tumour promoting effects of MCYSTs in mice (Kuiper-Goodman et al. 1999). The incidence of hepatocellular carcinoma is relatively high in Chinese villages where the primary water source is subject to cyanobacterial blooms, raising the possibility that MCYSTs contribute to this disease if exposure is prolonged (Kuiper-Goodman et al. 1999). Due to the potent toxicity of the MCYSTs and their potential to cause other illness, a provisional tolerable daily intake (TDI) of MCYSTs has been set by the World Health Organisation at 0.04 µg MCYST kg-1 body weight per day (Falconer et al. 1999; Kuiper-Goodman et al. 1999). This corresponds to a TDI of 1.0 µg MCYST l-1 in water containing cyanobacterial cells (both intracellular and extracellular MCYST) (Falconer et al. 1999).

The high specificity of MCYST for serine/threonine PP binding and the degree of conservation amongst PPs from diverse groups of organisms (from plants to animals; Cohen & Cohen 1989) means that MCYSTs are also toxic to a very broad range of organisms. For example, Mackintosh et al. (1990) discovered that MCYST-LR potently inhibits PPs from the protozoan Paramecium tetraurelia, several plant species (including Brassica napus) and rabbit skeletal muscle. The toxicity of MCYSTs can now be extended to include several species of fish (e.g. Carbis et al. 1996; Zimba et al. 2001), birds (e.g. Matsunaga et al. 1999), insects (e.g. Saario et al. 1994) and zooplankton (e.g. DeMott & Dhawale 1995), and is likely to extend to most eukaryotic organisms. The MCYSTs vary in their toxicity from the relatively non-toxic [(6Z)-Adda5]MCYST-LR (Harada et al. 1990b; Harada et al. 1990a) to the potently toxic MCYST-LR and MCYST-LA (Table 1.1). Harada (1996) classified MCYSTs which vary only in their L amino acids, into three groups according to their toxicity. MCYSTs -LR, -LA, -YR, were listed as 'strong' toxins (Harada 1996), to which MCYST-YM(O), -LY, -HilR and -HtyR could be added. MCYST-WR was considered a 'medium' toxin (Harada 1996) and shows similar toxicity to MCYSTs-FR and -AR. Harada listed the MCYSTs -RR and -M(O)R as 'weak' (Stotts et al. 1993; Harada 1996), and no other MCYSTs with similar toxicities and only variable L amino acids have since been described.

While some structural modifications result in reduced or increased toxicity of particular MCYSTs, changes to the Adda moiety of MCYST can result in complete loss of toxicity. The structural isomer of MCYST-LR, [6(Z)Adda5]MCYST-LR, is practically non-toxic (An & Carmichael


20

1994; Harada 1996) while removal of Adda or saturation of the double bonds in the Adda moiety of MCYST-LR greatly reduces its toxicity (Dahlem 1989: cited by Harada 1996). However, hydroxy or acetoxy groups instead of the methoxy group on the C9 of Adda (Fig. 1.1) do not substantially affect toxicity (Namikoshi et al. 1990). considerably

lower

toxicity

of

Compared with MCYST-LR, the

[D-Glu-(OCH3)6]MCYST-LR

and

[D-Glu-

6

OC2H3(CH3)OH ]MCYST-LR (Table 1.1) indicate the importance of the free carboxyl group of the D-Glu residue in toxicity (Harada 1996). The importance of the Mdha residue at position 7, with respect to toxicity of MCYSTs, is unclear. The Mdha residue reportedly forms a covalent bond with a Cys residue in PPs (Goldberg et al. 1995), indicating its importance in PP inhibition and hence an essential role in MCYST toxicity. However, synthetic addition of glutathione to this residue (Kondo et al. 1992) or its reduction with sodium borohydride (Meriluoto et al. 1990) had little effect on toxicity. In addition, the presence of dehydrobutyrine (Dhb) at position 7 of the MCYST structure does not significantly alter its toxicity (Table 1.1; Sano & Kaya 1998).

1.5.3

MCYST biosynthesis

Studies involving the application of

13

C-labelled substrates to cultures of M. aeruginosa have

provided some details on the origin of carbon atoms in the structure of MCYSTs. Moore et al. (1991) showed that the 2, 6, 8 methyl and 9 methoxy carbon groups of Adda, and the N-methyl group of Mdha, are derived from the methyl group of methionine. The same authors also showed that the terminal phenyl group of Adda is derived from the phenyl, C-2 and C-3 groups of Lphenylalanine, while acetate and pyruvate residues contribute to a number of carbon atoms within MCYST-LR (Moore et al. 1991). The cyclic peptide berinamycin contains dehydroalanine, an analogue of the Mdha residue present in MCYSTs, which is derived from L-serine (Pearce & Rinehart 1979). Rinehart et al. (1994) proposed that the Mdha residue of MCYSTs is also derived from L-serine in processes involving dehydration and N-methylation. The occurrence of variants of MCYST containing L-serine or L-N-methylserine in position 7 (Fig. 1.1, Table 1.1) supports this proposal. The Dhb residue in position 7 of some MCYSTs (Fig. 1.1, Table 1.1) is possibly derived from L-threonine (Rinehart et al. 1994; Sano & Kaya 1998). The discovery of a linear form of MCYST-LR in a M. aeruginosa bloom containing MCYSTs (Choi et al. 1993) provided evidence for Rinehart et al. (1994) to propose that MCYST-LR is synthesised by the condensation of amino acids (beginning with Adda), a process involving peptide bond formation and subsequent cyclisation. Rinehart et al. (1994) also proposed that Adda is synthesised from Lphenylalanine and malonyl-CoA subunits followed by methylation (via L-methionine) and that the first amino acid added to the growing peptide is D-glutamate.


21

The biosynthesis of many cyclic peptides in other organisms is often carried out non-ribosomally, using a number of enzymes or large multi-functional proteins (Kleinkauf & von Döhren 1996). Arment & Carmichael (1996) showed that the synthesis of MCYSTs involves a thio-template mechanism (unlike the system utilised on ribosomes) and that MCYST synthesis continues in the presence of the 70S ribosomal inhibitor chloramphenicol.

Thus, a peptide synthetase (or

synthetases) is likely to be responsible for the biosynthesis of MCYSTs. Dittmann et al. (1997) successfully disrupted MCYST synthesis by insertional mutagenesis of a peptide synthetase gene from M. aeruginosa PCC 7806, providing conclusive evidence that MCYSTs are synthesised by a particular enzymic process. The genes involved in this process have since been described in further detail (Nishizawa et al. 1999; Nishizawa et al. 2000) and the procedure for the synthesis of MCYSTs proposed by Rinehart et al. (1994) has been confirmed by the identification of specific domains within the MCYST biosynthesis gene cluster (Tillett et al. 2000).

1.6

Factors influencing MCYST production and content in cyanobacteria grown in culture

The potent toxicity of MCYSTs and their prevalence in freshwater environments makes them a considerable threat to human and domestic animal health. As a result, there has been considerable interest in determining the factors that control or influence the production of these compounds in the species of cyanobacteria that produce them. Understandably, much of the initial work in this area concerned the factors that affect toxicity in MCYST-producing cyanobacteria, so that environmental conditions leading to a rise in toxin content could be predicted. Thus, early studies determined parameters such as LD50s associated with cultured cyanobacterial biomass (e.g. van der Westhuizen et al. 1986). Studies more deliberately related to the determination of the physiology of MCYST production by cyanobacteria were carried out predominantly in the 1990s, though to date the physiology of MCYST production is still somewhat confused (see below). Those factors primarily affecting cyanobacterial growth (e.g. light, temperature, pH, N, P and micronutrients) have been studied most aggressively with respect to MCYST production (see Table 1.2).

1.6.1

Temperature and Light

Since both temperature and light are important factors in cyanobacterial mass bloom events (Section 1.2), many studies have examined their influence on MCYST production in M. aeruginosa and several other cyanobacteria (Table 1.2). Considered in isolation, the effects of either temperature or light on MCYST content show some trends although the outcomes of


several studies are somewhat contradictory.

22

For example, Gorham (1964) found that high

irradiance had no significant effect on toxicity (measured as minimum lethal dose, LDmin) whereas van der Westhuizen & Eloff (1985) found that toxicity (LD50) increased with increasing irradiance (Table 1.2). In addition, Sivonen (1990) showed that MCYST measured as a proportion of dry biomass (MCYST:dry weight) was greatest at low irradiance, while Utkilen & Gjølme (1992) showed that MCYST measured as a proportion of either protein or dry biomass increased with increasing irradiance up to 40 µmol photons m-2 s-1, thereafter decreasing or remaining constant, depending upon the unit of measurement. With respect to the effect of temperature, most studies suggest that the greatest toxicity or MCYST:dry weight occurs at an optimum temperature (Table 1.2), but this is contradicted by findings such as those by Song et al. (1998) and Ohtake et al. (1989) which indicate that temperature has no effect on MCYST, either on a dry weight basis or toxicity (Table 1.2). Thus, while in general there appears to be optimum temperatures for maximum MCYST content, reports on the effects of light on MCYST content are highly variable.

Several studies have addressed the apparent inconsistencies of the effects of light and temperature on MCYST content by analysing their combined effects. Rapala & Sivonen (1998) analysed the MCYST:dry weight content of two toxic strains of Anabaena in turbidostat cultures grown under varied conditions of light and temperature, and found that while the MCYST-LR content appeared to be regulated by irradiance, MCYST-RR was regulated by temperature (Table 1.2). Song et al. (1998) found no significant effect of irradiance on MCYST:dry weight content at 15 °C but MCYST:dry weight content decreased with increasing irradiance at 25 °C. Several other studies suggest that the MCYST composition of toxic strains may vary depending upon the conditions of light and temperature (Table 1.2). Thus, both temperature and light may play a part in the production of different MCYSTs from cyanobacteria. Due to the variation in toxicities of various MCYSTs (Table 1.1), it is possible that the contradictory results from studies examining changes in toxicity (e.g. LD50) could arise as a result of changes in the content of specific MCYSTs in response to changes in light and/or temperature.

In contrast to previous studies, recent studies by Hesse & Kohl (2001) and Böttcher et al. (2001) have examined the influence of irradiance on MCYST production and content by determining MCYST content on a cell/trichome quota (e.g. fg MCYST cell-1 or trichome-1) basis or as an amount per unit biovolume (µg MCYST mm-3) (Table 1.2). Böttcher et al. (2001) found that intracellular MCYST quota and MCYST biovolume-1 in turbidostat cultures varied little as irradiance increased, whereas MCYST expressed as a proportion of dry weight or chlorophyll a showed different responses to irradiance (Table 1.2). The estimates of MCYST quota and MCYST biovolume are based on the assumption that the cell volume does not vary during growth


23

(Böttcher et al. 2001). Hesse & Kohl (2001) showed that total MCYST content biovolume-1 increased as irradiance increased in two strains of M. aeruginosa and decreased in a third strain in semi-continuous cultures. These experiments provide a novel approach to MCYST analysis and allow some insight into the cellular production of MCYSTs. However, although Böttcher et al. (2001) showed there was no significant difference in cell volumes based upon microscopic examination of cells, Hesse & Kohl (2001) presented no evidence that cell volume remains constant when growth conditions vary. Of particular importance is that the work of Böttcher et al. (2001) and Utkilen & Gjølme (1992) (Table 1.2) highlights the fact that different units of MCYST measurement can result in different conclusions. However, the determination of the amount of MCYST per cell, hereafter referred to as the intracellular MCYST quota (QMCYST), provides particular information about cellular MCYST physiology. Thus, the findings of Böttcher et al. (2001) are the first to show that irradiance has little affect on the intracellular MCYST quota of cyanobacteria. In the absence of further studies, the variation seen in previous work is possibly due to the effects of irradiance on cell dry weight or protein content, or, contrary to the assumption made above, cell volume.

The influence of light quality on MCYST content has also been addressed by several studies. Measured both as a proportion of dry weight and protein, MCYST content is greatest in cultures exposed to red light, compared with green light and white light (Utkilen & Gjølme 1992). Kaebernick et al. (2000) recently showed that the levels of mcyB and mcyD transcripts (mRNAs of MCYST synthetase genes) are greatest in cells exposed to red light compared with those exposed to blue light or dark treatment, indicating that MCYST production is enhanced under dim or red light. Unfortunately, they were not able to show any significant change in the intracellular MCYST content (pg MCYST cell-1) (Kaebernick et al. 2000). These data suggest that both light quality and quantity could regulate MCYST production.

1.6.2

Nitrogen and phosphorus

Since N and P are often considered to be the limiting nutrients for phytoplankton, these nutrients possibly playing a major role in cyanobacterial bloom formation (Section 1.2.1), their influence on MCYST production by M. aeruginosa has been of particular interest. Like the data available for light and temperature, the influence of P on MCYST production is difficult to interpret due to inconsistencies in the experimental outcomes reported to date (Table 1.2).

For example,

Watanabe & Oishi (1985) found that P-sufficient batch cultures of M. aeruginosa were more toxic (i.e. MCYST measured as LD50s) than P-limited cultures, whereas the opposite was the case in a study by Codd & Poon (1988). On the other hand, there appears to be a consensus with respect to


24

the effect of N on MCYST content since, for the most part, high N results in greater toxicity or MCYST:dry weight (Table 1.2).

Orr & Jones (1998) provided a unique insight into the effect of N supply on the MCYST content of cells by measuring intracellular MCYST quotas (QMCYST, fg MCYST cell-1). They found that the concentration of N supplied to batch cultures of M. aeruginosa determined the total amount of MCYST produced within a batch culture (on a volumetric basis). In experiments using the nonaxenic M. aeruginosa strain MASH01, the N supply had a significant effect on QMCYST determined during the exponential growth phase (Orr & Jones 1998). However, at the same phase of growth in batch cultures of the axenic M. aeruginosa strain MASH-01A19, QMCYST was not significantly affected by the N supply (Orr & Jones 1998). The latter finding conflicts with the generalisation that high N leads to greater MCYST content but is a more accurate interpretation of the effects of N supply on the intracellular MCYST content of cyanobacteria. Despite this, the conflicting results obtained for axenic and non-axenic strains of M. aeruginosa leave the effect of N on QMCYST unclear. Orr & Jones (1998) also showed that the specific rate of MCYST production (µ MCYST, d-1) was equal to the specific cell division rate (µ c), indicating that QMCYST should remain constant over a range of specific cell division rates. Application of this idea to previously published studies indicated that their proposed relationship held true for a wide variety of growth conditions (Orr & Jones 1998), including the N and P limitation studies carried out by Sivonen (1990) (Table 1.2). Thus, application of Orr & Jones' (1998) concept to other N and P studies suggests that, in theory, the intracellular MCYST quota of cultured cyanobacteria is not affected by N and P supply.

Although studies on the effect of N and P on MCYST content have been carried out since the work of Orr & Jones (1998) (Table 1.2), none has determined intracellular MCYST quotas. Oh et al. (2000) recently showed that in P-limited chemostats, MCYSTs -LR and RR decreased as proportions of both dry weight and protein as the growth rate increased (Table 1.2). This contradicts the relationship proposed by Orr & Jones (1998). However, the effect of P limitation on cell dry weight and cellular protein content was not determined by Oh et al. (2000), so it is not possible to establish if the intracellular MCYST quota remains constant under P limitation, as proposed by Orr & Jones (1998). In a recent batch culture study by Lee et al. (2000) MCYST:dry weight correlated positively with the amount of N supplied to cultures (Table 1.2), but again, no data on intracellular MCYST quotas are available.


25

The interpretation of results from previous studies on the effects of N and P on MCYST content of toxic cyanobacteria are confounded by the variety of forms in which data are expressed. The approach of Orr & Jones (1998) provides a more precise method for understanding the cellular physiology of MCYST production, providing some understanding of the effect of N supply on the intracellular MCYST quota of M. aeruginosa. However, this approach has not yet been applied to the study of P limitation, and the role of N in determining QMCYST has not been clearly established (Orr & Jones 1998). Thus, while the existing pool of evidence provides an unclear explanation of the effect of N and P supply on MCYST content (expressed as toxicity or as a proportion of various biomass indicators; Table 1.2) the effects of N and P supply on the intracellular MCYST quota of hepatotoxic cyanobacteria are yet to be clearly established.

1.6.3

Iron

/XNDþ $HJHUWHU (1993) proposed that, since the production of several cyclic peptide antibiotics (e.g. gramicidin S and mycobacillin) is regulated by heavy metals (Vining & Wright 1977), possibly particular metals (including micronutrients) could regulate MCYST production in M. aeruginosa.

8VLQJ EDWFK FXOWXUHV /XNDþ $HJHUWHU (1993) found that MCYST:dry weight

increased when the concentration of Fe3+ was low and that Zn was required for optimum growth and MCYST production (Table 1.2). They proposed that this was indicative of MCYST being produced by M. aeruginosa in response to an environmental stress (i.e. low Fe3+ availability). In contrast, Utkilen & Gjølme (1995) found that high Fe3+ availability stimulated MCYST production (expressed as a proportion of dry weight and protein) in M. aeruginosa (Table 1.2). They suggested that MCYST production increased in cells in response to a build-up of intracellular Fe2+ ions that could be sequestered within the cyclic structure of MCYSTs. Lyck et al. (1996) found that Fe3+ starvation led to an increase in MCYST measured as a proportion of dry weight, protein and chlorophyll a (Table 1.2), thereby casting doubt on whether Fe has a role in MCYST production. Further studies on Fe3+ supply have been carried out by Bickel et al. (2000) but the role of Fe3+ in MCYST production remains unclear. Thus, although the published data are conflicting, there appears to be evidence that Fe3+ plays a role in MCYST production.

1.6.4

CO2 and pH

Several studies have examined the influence of pH on the MCYST content of Microcystis spp. As noted in Section 1.2.4, Microcystis, like most cyanobacteria, grows best at alkaline pH. Recently, Song et al. (1998) found that growth at each end of the alkaline range (pH 7.0 and 9.2) resulted in high MCYST:dry weight content, whereas growth at intermediate pH (7.8 and 8.6), which were


26

the most conducive to growth, resulted in lower MCYST:dry weight values (Table 1.2). A similar result was obtained by van der Westhuizen & Eloff (1983) who showed that the toxicity of M. aeruginosa (LD50) was greatest when it was grown at extreme pH values (6.5 and 10.5) but considerably lower in cultures grown at intermediate pH (7.5 to 8.0). It is not clear whether this is related to the uptake and assimilation of CO2 (Section 1.2.4). Codd & Poon (1988) found that the toxicity of M. aeruginosa was far greater in cultures supplied with CO2 than in cultures without CO2 (Table 1.2). Low CO2 availability is usually associated with extremely high pH (Section 1.2.4) but this would not explain the high toxicity at extreme pH values as observed by Song et al (1998) and van der Westhuizen & Eloff (1983). Since the intracellular MCYST quota has not been determined in relation to the effect of pH or CO2, it is difficult to know if the observations described above are a result of changes in cell dry weight which occur in response to pH or CO2 concentration changes.

1.6.5

Time (culture age)

In batch cultures the physiology of cells changes continuously as they proceed from nutrientreplete to nutrient-deplete conditions over time.

As a result, the physiology of MCYST

production and the MCYST content of cyanobacteria might also be expected to change over time. Several workers have investigated the MCYST content of cyanobacteria with time in batch cultures. They found that MCYST content (either LD50 or MCYST:dry weight) is generally highest when batch cultures are in exponential growth phase, and that the MCYST content decreases as cultures approach decelerating, stationary and decline phases (see Fig. 3.9) (Gorham 1964; van der Westhuizen & Eloff 1983; Watanabe et al. 1989; Song et al. 1998).

This

information suggests that the production of MCYSTs is affected in a complex manner as the growth conditions change with time and as nutrients are consumed. The temporal change in MCYST content can only be observed in batch cultures, as time is not a relevant factor in steadystate continuous cultures. However, changes in cellular physiology resulting from temporal environmental change, are manifested in a decrease in specific growth rate as batch cultures age. In contrast to batch cultures, the specific growth rate of continuous cultures can be manipulated. In theory, any variation in MCYST content observed over time in batch cultures in response to nutritional stress can also be achieved in steady-state continuous cultures by varying the replacement rate of the limiting nutrient. This provides an experimentally convenient way of investigating the effects of various parameters on MCYST production, and other aspects of cell physiology, in response to the nutrient controlled manipulation of the growth rate.

TABLE 1.2 Summary of the outcomes of laboratory studies on the effects of culture growth conditions on the toxicity/MCYST content of MCYST producing cyanobacteria. 1 Parameter Temperature

Test range 20 - 35 °C

Cyanobacterium M. aeruginosa NRC-1

Culture type Batch/semi-

Toxicity / MCYST

Outcome

content 2 LDmin

continuous

Optimum growth at 32.5 °C. Five-fold greater toxicity at 25 °C

Reference (Gorham 1964)

than at 20 or 30 °C

18 - 34 °C

M. aeruginosa

Batch

LD50

Four-fold greater toxicity at 18 °C than at 29 °C

(Runnegar et al. 1983)

16 - 36 °C

M. aeruginosa UV-006

Batch

LD50

Growth rate increased with increasing temperature. Toxicity

(van der Westhuizen &

greatest at 20 °C but decreased with increasing temperature.

Eloff 1985)

Greatest toxicity at 20 °C. MCYST complement varied with

(van der Westhuizen et al.

temperature.

1986)

Optimum growth at 32 °C. Greatest toxicity at 18 °C. Toxicity

(Watanabe & Oishi 1985)

16 - 36 °C


Batch

LD50 -1

SAU g 18 - 32 °C

M. aeruginosa M228

Batch

LD50

decreased as temperature increased. 10 - 34 °C

M. aeruginosa 7813

Batch

LD50

Optimum growth at 25 °C. Greatest toxicity at 25 °C. Lower

(Codd & Poon 1988)

toxicity at 10 and 34 °C. 22, 30 °C

M. aeruginosa K-139

Batch

LD50

Greater growth at 30 °C. No effect on toxicity.

(Ohtake et al. 1989)

15 - 30 °C

Planktothrix (syn.

Batch

MCYST:dry wt

Optimum growth at 25 °C. Greatest MCYST content at 25 °C.

(Sivonen 1990)

15 - 30 °C

P. agardhii CYA 128

Batch

MCYST:dry wt

Optimum growth at 25 - 30 °C. MCYST content the same at 15 -

Oscillatoria) agardhii 97

MCYST content decreased as temperature increased. (Sivonen 1990)

25 °C. MCYST content lower at 30 °C. 12.5 - 30 °C

Anabaena 90, 202A1

Batch

MCYST:dry wt

Greatest total MCYSTs at 25 °C. Different MCYST variants

(Rapala et al. 1997)

predominant at different temperatures. 10 - 28 °C

3

Anabaena 90, 202A1

Turbidostat

MCYST:dry wt

Growth rate varied with temperature. MCYST content correlated

(Rapala & Sivonen 1998)

positively with growth rate. Production of MCYST-RR regulated by temperature. continued

27

Table 1.2 continued Parameter

Test range

Cyanobacterium

Culture type

Temperature

15 - 30 °C 3

M. viridis

Batch

Light 4

220, 1,600 ft.-c.

M. aeruginosa NRC-1

Batch/semi-

Toxicity / MCYST

Outcome

content MCYST:dry wt

Reference

No significant affect of temperature on MCYST content.

(Song et al. 1998)

LDmin

Optimum growth at 1,600 ft-c. No significant effect on toxicity.

(Gorham 1964)

LD50

Growth rate increased with increasing irradiance. Toxicity


increased with increasing irradiance.

Eloff 1985)

continuous 21 - 205 µmol

21 - 205 µmol



Batch

Batch

LD50, -1

SAU g 7.5 - 75 µmol

M. aeruginosa M228

Batch

LD50

-2 -1

Greatest toxicity at 145 µmol m s . Lower toxicity at low

(van der Westhuizen et al.

irradiance

1986)

Growth rate increased with increasing irradiance. Toxicity greatest


at high irradiance and lowest at low irradiance. 5 - 50 µmol

M. aeruginosa 7813

Batch

LD50

12 - 95 µmol

P. agardhii 97

Batch

MCYST:dry wt

No significant effect of irradiance on toxicity

(Codd & Poon 1988)

Optimum growth at 50 µmol m-2 s-1. MCYST content decreased

(Sivonen 1990)

with increasing irradiance. 12 - 95 µmol

P. agardhii CYA 128

Batch

MCYST:dry wt

Optimum growth at 50 µmol m-2 s-1. MCYST content greatest at

(Sivonen 1990)

-2 -1

12 and 24 µmol m s but lower at high irradiance. 20 - 75 µmol

M. aeruginosa CYA

red, white, green.

228/1

Chemostat

MCYST:dry wt

MCYST:dry wt increased with irradiance up to 40 µmol m-2 s-1

MCYST:protein

thereafter decreasing. MCYST:protein increased with irradiance

(Utkilen & Gjølme 1992)

up to 40 µmol m-2 s-1 thereafter remaining constant. MCYST:dry wt and MCYST:protein greatest in red > green > white light. continued

28

Table 1.2 continued Parameter Light

Test range

Cyanobacterium

Culture type

Toxicity / MCYST

Outcome

content

2 - 100 µmol

Anabaena 90, 202A1

Batch

MCYST:dry wt

Greatest total MCYST at 25 µmol m-2 s-1. Different MCYST

7 - 42 µmol 3

Anabaena 90, 202A1

Turbidostat

MCYST:dry wt

Growth rate increased with increasing irradiance. MCYST content

Reference (Rapala et al. 1997)

variants predominant at different irradiances. (Rapala & Sivonen 1998)

correlated positively with growth rate. Production of MCYST-LR regulated by irradiance. 15 - 100 µmol (at both 15 / 25 °C) 16 - 68 µmol

M. viridis

Batch

MCYST:dry wt

3

No significant effect of irradiance on MCYST content at 15 °C.

(Song et al. 1998)

MCYST content decreased with increasing irradiance at 25 °C. M. aeruginosa PCC 7806

Batch

dark, red, blue

QMCYST

Greatest levels of mcy D and B transcripts at high irradiance and

mcy transcripts

red light. Low transcript levels at low irradiance and blue light.

(Kaebernick et al. 2000)

No significant changes in QMCYST. 10 - 150 µmol

M. aeruginosa HUB 5-2-

Semi-

4, W334, W368

continuous

MCYST µm-3

MCYST content increased with increasing irradiance in two strains

(Hesse & Kohl 2001)

but decreased with increasing irradiance in W334. Total MCYST complement varied with irradiance.

2.4 - 73 µmol

M. aeruginosa HUB 5-24, P. agardhii HUB 076

Turbidostat

QMCYST/MCYST µm-1 essentially unresponsive to irradiance.

QMCYST -1

MCYST µm

MCYST:dry wt increased with increasing irradiance. MCYST:chl

MCYST:dry wt

a increased with increasing irradiance. Culture MCYST

MCYST:chl a

concentration [MCYST] greatest at 10 - 20 µmol photons m-2 s-1

[MCYST]

(Böttcher et al. 2001)

and tended to decrease at higher and lower irradiances.

continued

29

Table 1.2 continued Culture type

Toxicity / MCYST

Parameter

Test range

Cyanobacterium

Phosphorus

23 - 460 µM P

M. aeruginosa M228

Batch

LD50

Slightly higher toxicity at 460 µM P than at lower P.


0, 225 µM P

M. aeruginosa 7813

Batch

LD50

Greater toxicity without added P.

(Codd & Poon 1988)

14.4, 143.5 µM P

M. aeruginosa CYA

Chemostat

MCYST:dry wt

MCYST:dry wt greater at high P than at low P. MCYST:protein


MCYST:protein

slightly higher at low P than at high P.

228/1

Outcome

content

Reference

3.2 - 177 µM P

P. agardhii 97

Batch

MCYST:dry wt

No growth at 3.2 µM P. MCYST:dry wt greatest at high P.

(Sivonen 1990)

3.2 - 177 µM P

P. agardhii CYA 128

Batch

MCYST:dry wt

No growth at 3.2 µM P. MCYST:dry wt greatest at high P.

(Sivonen 1990)

1.6 - 177 µM P

Anabaena 90, 202A1

Batch

MCYST:dry wt

Greatest total MCYST at 177 µM P. At low P, MCYST:dry wt


decreased over time. At high P, MCYST:dry wt increased over time. 0, 460 µM P

M. viridis

Batch

MCYST:dry wt

Higher growth rate and MCYST:dry wt without P

6 µM P

M. aeruginosa UTEX

Chemostat

MCYST:dry wt

As growth rate increased, MCYST:dry wt and MCYST:protein

2388

MCYST:protein

decreased. Net MCYST production rate (µg MCYST(mg dry wt)

(Song et al. 1998) (Oh et al. 2000) -1

d-1) increased with increasing growth rate. 0.7 - 71 µM P, fixed N

5

5 - 150 µM P

M. aeruginosa UTEX

Batch

MCYST:dry wt

2388 M. aeruginosa NIVACYA 228/1

MCYST:dry wt was constant at all P concentrations when N fixed

(Lee et al. 2000)

at 71 µM. Chemostat

MCYST:protein

MCYST:protein increased as P decreased. MCYST correlated

(Bickel et al. 2000)

positively with energy charge over all P concentrations. continued

30

Table 1.2 continued Parameter Nitrogen

Test range

Cyanobacterium

Culture type

0.1 to 2.0 mM N

M. aeruginosa M228

Batch

0, 0.9 mM N

M. aeruginosa 7813

Batch

0.35 - 5.8 mM N

M. aeruginosa CYA

Chemostat

Toxicity / MCYST

Outcome

Reference

Greatest toxicity at intermediate N (0.2 mM). Lower toxicity at 0.1


content LD50

and 2.0 mM N.

228/1 0.03 - 6.0 mM N

P. agardhii 97

Batch

Greater toxicity with added N.

(Codd & Poon 1988)

MCYST:dry wt

MCYST:dry wt increased with increasing N. MCYST:protein


MCYST:protein

unaffected by N supply

MCYST:dry wt

Best growth at high N. MCYST:dry wt increased with increasing

LD50

(Sivonen 1990)

N. 0.03 - 6.0 mM N

P. agardhii CYA 128

Batch

MCYST:dry wt

Best growth at high N. MCYST:dry wt increased with increasing

(Sivonen 1990)

0 - 3.6 mM N

Anabaena 90, 202A1

Batch

MCYST:dry wt

Total MCYST:dry wt greatest in N-deplete cultures.


0, 2 mM N

M. viridis

Batch

MCYST:dry wt

Cultures without N showed little growth but had a higher

(Song et al. 1998)

N.

MCYST:dry wt. 0.009 - 1.2 mM N

M. aeruginosa MASH01A19 (axenic), MASH01 (non-axenic)

Batch

MCYST:dry wt

N supply had no significant effect on MCYST:dry wt. N supply

QMCYST

had a significant effect on QMCYST in the non-axenic strain but not

[MCYST]

in the axenic strain. Culture MCYST concentration increased with

(Orr & Jones 1998)

increasing N. Specific rate of MCYST production = specific cell division rate. continued

31

Table 1.2 continued Parameter Nitrogen

Test range 6.5 - 650 µM N 5

Cyanobacterium M. aeruginosa UTEX

Culture type Batch

Toxicity / MCYST

MCYST:dry wt

2388 Micronutrients

Al, Cd, Cr, Cu,

M aeruginosa PCC 7806

Outcome

content

Reference

MCYST:dry wt greatest at intermediate N (104 and 325 µM) when

(Lee et al. 2000)

P fixed at 6.5 µM. Batch

MCYST:dry wt

Zn required for optimum growth. MCYST:dry wt decreased with

/XNDþ $HJHUWHU

3+

Fe, Mn, Ni, Sn

increasing Fe .

and Zn. 0.3 - 10 µM Fe3+ and +/- EDTA 3+

0, 10 µM Fe

M. aeruginosa CYA

Chemostat

228/1 M. aeruginosa CYA

Chemostat

228/1

MCYST:dry wt

MCYST:dry wt and MCYST:protein increased with increasing

MCYST:protein

Fe and decreased with increasing EDTA concentration.

MCYST:dry wt

MCYST:dry wt, MCYST:protein and MCYST:chl a greatest under

MCYST:protein


3+

(Lyck et al. 1996)

3+

Fe starvation.

MCYST:chl a 0.5 - 10 µM Fe3+

M. aeruginosa NIVACYA 228/1

Chemostat

MCYST:protein

MCYST:protein decreased as Fe3+ decreased. MCYST correlated

(Bickel et al. 2000)

3+

positively with energy charge over all Fe concentrations. continued

32

Table 1.2 continued Parameter pH & CO2

Test range

Cyanobacterium

Culture type

Toxicity / MCYST

Outcome

content

pH 6.5 - 10.5


Batch

LD50

+/- CO2

M. aeruginosa 7813

Batch

LD50

pH 7.0 - 9.2

M. viridis

Batch

MCYST:dry wt

Reference

Optimum growth at pH 9. Lowest toxicity at pH 7.5-8. Toxicity


1.5-fold greater at pH 6.5-7 and at pH 10-10.5.

Eloff 1983)

Toxicity greatest in cultures supplied with CO2.

(Codd & Poon 1988)

MCYST greatest at pH 7.0 and 9.2, lowest at intermediate pH (7.8

(Song et al. 1998)

and 8.6). 1

Adapted, revised and updated from Rapala (1998). Units of measurement of toxicity or MCYST content are as follows; LDmin; minimum lethal dose of MCYST required to kill a 25 g mouse by intraperitoneal injection (i.p.). LD50; amount of MCYST (in µg) required to kill 50 % of test animals by i.p. injection. -1 SAU g-1; Standard absorbance units (A 1.0cm 240nm ml ) per gram of dry culture material (Eloff 1982). MCYST:dry wt; MCYST expressed as a proportion of culture dry weight (e.g. µg MCYST (mg dry weight)-1). MCYST:protein; MCYST expressed as a proportion of protein (e.g. µg MCYST (mg protein)-1). mcy transcripts; Northern blot analysis of MCYST synthetase gene (mcy) transcripts (mRNA). MCYST µm-3; Amount of MCYST per unit biovolume (e.g. fg MCYST µm-3). QMCYST; Intracellular MCYST quota (e.g. fg MCYST cell-1). MCYST µm-1; Amount of MCYST per trichome length in filamentous cyanobacteria (e.g. fg MCYST µm-1). MCYST:chl a; MCYST expressed as a proportion of chlorophyll a (e.g. µg MCYST (mg chl a)-1). [MCYST]; Volumetric MCYST concentration or amount of MCYST per unit culture volume (e.g. µmol MCYST ml-1). This can include both intracellular and extracellular MCYST. 3 Combination studies looking at the effects of both temperature and light. 4 For light experiments, irradiance is measured either as foot candles (ft.-c.) or µmol photons m-2 s-1. 5 The study by Lee et al. (2000) examined the effect of N:P ratios on MCYST production. Thus, the range of P concentrations are examined at a fixed N concentration and the range of N concentrations are examined at a fixed P concentration. 2

33


1.6.6

34

Problems associated with MCYST content studies

It is apparent from Table 1.2 that the MCYST content of M. aeruginosa and other cyanobacteria has been expressed in various ways by different authors, creating some difficulty in unifying such disparate studies. Almost without exception, the reports published to date express MCYST content as a proportion of a biomass indicator (e.g. culture dry weight, protein, and chlorophyll a). However, this provides no information on the physiological response of the organism, as the expression of MCYST content with respect to a biomass indicator merely provides a ratio of MCYST to a parameter that could also be affected by the growth conditions (Sivonen & Jones 1999). Similarly the determination of toxicity (e.g. LD50) provides only a measure of lethality of cultured cyanobacterial biomass since it is usually expressed on a dry weight basis (e.g. µg culture biomass required to kill 1 kg of mice). Moreover toxicity data provide no indication of MCYST identity or relative quantity.

Thus, the majority of studies regarding the effects of growth

conditions on MCYST production by cyanobacteria provide little information about the cellular physiology of MCYST production. In addition, without full knowledge of the effects of growth conditions on the denominators of MCYST content (e.g. dry weight, protein, and chlorophyll a) it is impossible to compare one study with another with any degree of surety.

1.7

Aims of this thesis

The cellular physiology of MCYST production remains unclear despite a considerable pool of data on the changes in MCYST content in relation to biomass indicators (Table 1.2), simply because the intracellular MCYST content has rarely been determined in studies examining the effect of growth conditions on MCYST production. However, the study of MCYST production in relation to biomass indicators (such as dry weight or biovolume) would appear to have serious deficiencies since it is known that cell size can vary in response to growth conditions (Krüger & Eloff 1981) which in turn has a major outcome on biomass. To some degree, the absence of intracellular MCYST content data results from the difficulty associated with determining individual cell numbers in filamentous cyanobacteria (e.g. Planktothrix and Anabaena). However, application of some lateral thought allowed Blackburn et al. (1996) to determine intracellular quotas of the pentapeptide MCYST analogue, nodularin, in the filamentous Nodularia spumigena, while Böttcher et al. (2001) determined MCYST content per trichome of Planktothrix. Sivonen & Jones (1999) made the point that there has been some confusion regarding the use of terms to describe the content of MCYSTs within cyanobacteria (see Sivonen & Jones 1999, Box 3.1, p. 64). The problems outlined in Section 1.6.6 provide further evidence on how this practice confounds the collation of many of the studies made with cyanobacteria.


35

Sivonen & Jones (1999) also refer to toxin quota: the amount (in mass or moles) of toxin per cyanobacterial cell. This would appear to be the most appropriate form in which to express MCYST content in the study of the cellular physiology of MCYST production. Determination of MCYST as a proportion of biomass (or other biomass indicators; e.g. protein or chlorophyll a) or the determination of toxicity cannot provide the required information to allow any conclusions to be made regarding cellular MCYST production. Thus, this thesis aims to investigate the influence of growth conditions on the intracellular MCYST quota (QMCYST) of M. aeruginosa. In order to examine the effects of various growth conditions on QMCYST in M. aeruginosa it is imperative that an appropriate strain of this organism be utilised for this study. Chapter 2 describes the isolation and screening of strains of M. aeruginosa for their MCYST production capabilities. Several novel techniques for the detection and analysis of MCYSTs are described and the MCYST production capabilities of M. aeruginosa MASH-01A19 are examined in detail.

The dearth of intracellular MCYST data in the existing literature raises the question whether the observed variation in MCYST content expressed in relation to other biomass indicators reported for cultured cyanobacteria under varied growth conditions (Table 1.2) is also manifested at an intracellular level. Anderson et al. (1990) showed that the intracellular quota of saxitoxin in Alexandrium spp. varied in a different manner over time compared with the ratios of saxitoxin to various biomass indicators such as N, P, C, chlorophyll a, protein and carbohydrate. Thus, the effect of growth conditions upon cellular components to which MCYST may be proportioned must be considered. This issue is addressed in Chapter 3 by determining not only variations in QMCYST, but changes which occur in cell dry weight, chlorophyll a content, protein and amino acid content in N-limited chemostat cells of M. aeruginosa MASH-01A19. The independent variation in these cellular components on a cell quota basis suggests that they are inappropriate for addressing the cellular physiology of MCYST production.

Orr & Jones (1998) found that the intracellular MCYST content (QMCYST) of M. aeruginosa MASH-01A19 varied by no more than two-fold over a complete batch culture growth cycle under N limitation. However, they also found that the specific rate of MCYST production is equal to the specific cell division rate. According to these data, QMCYST, in theory, should remain constant. Chapter 3 also seeks to address this apparent anomaly by determining QMCYST over a range of growth rates in N-limited chemostats. Significant variation in Q MCYST was observed over a range of growth rates. The data not only indicate that QMCYST does not remain constant, but also that QMCYST increases with the growth rate in a predictable manner.


36

Collectively the results from Chapter 3 indicate that the cellular physiology of MCYST production can only be examined using QMCYST, which follows a predictable pattern of change with respect to growth rate. Thus, the ensuing chapter (Chapter 4) seeks to bring together the apparent disparities which exist in the literature regarding MCYST production (Table 1.2) by determining QMCYST in M. aeruginosa MASH-01A19 grown under a variety of growth conditions in batch culture. In addition, the apparent relationship between QMCYST and growth rate is examined to determine if it applies to different growth conditions. The results from this study indicate that, under most conditions, QMCYST does vary with growth rate in a predictable manner, but that the parameters defining the relationship between QMCYST and growth rate (namely the minimum and maximum intracellular MCYST quotas) depend to some extent upon the growth conditions. The influence of particular growth conditions on QMCYST and their relevance to the production of MCYSTs is discussed in Chapter 5.

37

CHAPTER 2 M. aeruginosa strains : their sources, growth and MCYST production

M. aeruginosa bloom. Lake Mokoan, Victoria

Chapter 2 MCYST analysis

2.1

38

Introduction

The production of MCYSTs by cyanobacteria can be studied from a number of perspectives to determine the MCYST content of these organisms under a variety of environmental conditions. For example, a number of studies have examined cyanobacterial growth and MCYST production in freshwater lakes and reservoirs in situ in relation to the prevailing environmental parameters (Kotak et al. 1995; Yunes et al. 1998).

Another approach involves routine monitoring of

phytoplankton populations and both intracellular and extracellular MCYST over extended periods (e.g. Park et al. 1998). Such studies have been invaluable in predicting the occurrence of toxic Microcystis blooms and in developing strategies to cope with them. Alternatively, MCYST production by cyanobacteria can be studied in a more controlled way by removing them from their natural habitat and growing them in culture, enabling the influence of specific growth conditions on MCYST production to be studied. This approach is particularly suitable for determining the effect of environmental factors on the intracellular MCYST content of M. aeruginosa cells. However, the studies of Bolch et al. (1997) and Watanabe et al. (1991) indicate that strains of M. aeruginosa can differ greatly in their ability to produce MCYSTs. Since this thesis is concerned with the study of MCYST production by cells of M. aeruginosa, it is essential to employ strains with an intracellular MCYST content sufficient to allow quantification of any significant changes in content that might occur in response to varied growth conditions. Furthermore, techniques for the detection and quantification of MCYSTs must be employed and, if necessary, developed. This chapter describes the collection of strains of M. aeruginosa and an analysis of their MCYST content, using several procedures, with a view to selecting an appropriate technique for measuring MCYSTs and a strain suitable for further study.

The

MCYST production capabilities of four strains are examined using a novel protein phosphatase (PP) inhibition assay, antibody techniques and high performance liquid chromatography (HPLC). The MCYST complement of one particular strain (MASH-01A19) is examined in further detail using amino acid analysis.

2.2 2.2.1

Sources of M. aeruginosa strains and their growth in culture Collection of M. aeruginosa from lakes and reservoirs

Field samples containing Microcystis spp. and other cyanobacteria were collected from the surface waters of various freshwater bodies in Victoria and South Australia, which contained cyanobacterial blooms at the time of sampling. The location of collection sites is indicated in Fig. 2.1. Surface water samples were collected in sterile 50 ml culture containers and returned to the laboratory within 48 h. For growth, the samples were diluted approximately 1:1 with sterile


39

modified MLA medium (Table 2.1). The growth of eukaryotic phytoplankton was minimised by the addition of filter sterilised cycloheximide (final concentration approximately 50 µg ml-1). The nutrient-enriched samples were grown at 26 °C under constant illumination at 15 µmol photons m-2 s-1 until further processing (Section 2.2.2). These samples are subsequently termed field collections.

Fig. 2.1 The location of freshwater bodies in Victoria and South Australia from which strains of M. aeruginosa were isolated. 1 Shepparton. 2 Lake Mokoan. 3 Bendigo. 4 Drouin. 5 Lake Boga. 6 Mount Bold Reservoir.

2.2.2

Isolation and growth of M. aeruginosa strains

Isolation methods Under field conditions, M. aeruginosa occurs as free-floating colonies of densely aggregated cells (Section 1.4). Isolation of individual strains was carried out using several methods. Primarily, isolation allowed the removal of contaminating green algae or other cyanobacteria from M. aeruginosa cells to ensure 'uni-algal' strains of M. aeruginosa. Isolated strains were achieved either through spread, streak or pour plating of field collections onto solid growth media or by micromanipulation as described below.


40

Individual colonies of M. aeruginosa in field collections were isolated by micromanipulation, using micropipettes under a high power dissecting microscope. Isolated colonies were then washed sequentially in a series of drops of sterile modified MLA medium (Table 2.1) to remove adhering contaminating cyanobacteria and other microalgae. Individual washed colonies were then transferred to 2 ml of sterile modified MLA medium, containing 50 µg ml-1 cycloheximide, in glass culture tubes (Kimax) and incubated at 20 or 25 °C under constant illumination (15 µmol photons m-2 s-1). After three to four weeks, successful colonies were sub-cultured into larger volumes of growth medium in sterile conical flasks, stoppered with sterile cotton wool plugs, and incubated under identical conditions.

Several strains of M. aeruginosa were isolated from field collections, using solid growth media made by combining either agar or low melting temperature agarose with double-strength liquid growth medium after autoclaving. Before inoculation of agar or agarose plates, Microcystis colonies were disaggregated in twice-distilled de-ionised water (Milli-Q H 2O) according to the method of Parker (1982) in order to promote dispersion of cells and the removal of contaminating organisms from within colonies. Inoculation of solid media was achieved by spreading or streaking inoculum over plates, or by incorporating inoculum in the medium before gelling (i.e. pour plating). Plates inoculated with field collections heavily contaminated with bacteria were treated with Na2S to select for cyanobacterial growth (Parker 1982). Inoculated plates were incubated, inverted, at 20 °C under constant illumination (15 µmol photons m-2 s-1). Successful colonies were transferred to sterile growth medium as for micromanipulated strains (above).

Field collections containing high concentrations of Microcystis sp. were sub-cultured into fresh growth medium containing 50 µg ml-1 cycloheximide in an attempt to selectively enrich the collections with the dominant species.

Attempts were made to make strains axenic by eliminating contaminating bacteria. This was done by repeated streak plating of strains and growing them in the presence of Na2S according to the method of Parker (1982). Colonies showing no associated bacterial growth on solid media were re-streaked onto agar or agarose plates and the resulting individual colonies subsequently used to inoculate liquid modified MLA medium. Although no attempt was made to show that the strains prepared using these procedures were axenic, by demonstrating that bacteria could not be isolated from them, these strains showed no visible signs of contamination. All strains were 'unialgal'.

Cyanobacterial species were identified by microscopic examination using the keys provided by Baker (1991; 1992).


41

Growth media A modified version of MLA medium (Bolch & Blackburn 1996) (Table 2.1) was used for the isolation and growth of M. aeruginosa strains in this study. This medium differs from that described by Bolch & Blackburn (1996) in that it contains the organic buffer CHES (2-(Ncyclohexylamino) ethanesulfonic acid), 0.02 mM K2HPO4 (compared with 0.2 mM in the original medium) and no H2SeO3. The growth medium was prepared from filter-sterilised stock solutions of each constituent (Table 2.1) which were stored in the dark at 4 °C until required. Aliquots of the stock solutions of each constituent, except for the CaCl2, NaHCO3 and vitamin stocks, were combined in a 1 l Schott bottle and the volume made up to 1 l with Milli-Q H2O. The solution was then autoclaved (20 min, 121 °C) and allowed to cool before the sterile addition of appropriate volumes of CaCl2, NaHCO3 and vitamin stocks. For the growth of M. aeruginosa cultures in liquid modified MLA medium, fresh medium in sterile 100 to 500 ml conical flasks, stoppered with cotton wool plugs and covered with aluminium foil caps, was inoculated with cells under sterile conditions in a laminar flow cabinet. Flasks were then transferred to a constant temperature growth room (26 °C) and placed under lights with a constant irradiance of 40 µmol photons m-2 s-1 (PAR). Cultures were shaken gently every two to three days and sub-cultured into fresh medium at approximately 30 day intervals.

Solid media were made using agar or low melting point agarose. Oxoid No.3 agar (15 g) or agarose (Type 1-A: low EEO, Sigma) (7 g) was suspended in 500 ml of Milli-Q H2O and sterilised by autoclaving. Upon cooling, but while the agar or agarose was still molten, 500 ml of double-strength MLA medium was added. Requisite volumes of filter-sterilised stock solutions of vitamins, CaCl2 and NaHCO3 were also added under sterile conditions in a laminar flow cabinet to give final concentrations of agar or agarose of 1.5 % (w/v) or 0.7 % (w/v) respectively. The agar/agarose - modified MLA mixture was then poured into petri dishes and allowed to solidify. Petri dishes were either used immediately or stored in a sterile container at 4 °C until required. When pour plating was used to isolate strains of M. aeruginosa, cells disaggregated using Milli-Q H2O were added to approximately 15 ml of cooling agar/agarose modified MLA mixture, mixed by inversion and then poured into petri dishes.


42

TABLE 2.1 Composition of modified MLA medium used for growth of M. aeruginosa under standard growth conditions Component

Concentration

Volume (ml) of stock

Final concentration in complete

of stock solution

solution required for

medium

(mM)

1.0 litre of medium (mg l-1)

(mM)

Macronutrient stock solutions MgSO4· 7H2O

200

1.0

49.3

0.2

NaNO3

1000

2.0

170

2.0

200

1.0

29.4

0.2

1000

2.0

168

2.0

H3BO3

40

1.0

2.47

0.04

K2HPO4· 3H2O

200

0.1

4.56

0.02

CaCl2· 2H2O NaHCO3

1

1

Combined micronutrient

1.0

stock solution Na2EDTA

12

4.47

0.012

FeCl3· 6H2O

6.0

1.62

0.006

NaHCO3

14

1.18

0.014

MnCl2· 4H2O

1.8

0.36

1.8 × 10-3

CuSO4· 5H2O

0.04

0.01

40 × 10-6

ZnSO4· 7H2O

0.08

0.02

80 × 10-6

CoCl2· 6H2O

0.04

0.01

40 × 10-6

Na2MoO4· 2H2O

0.025

0.006

25 × 10-6

Thiamine HCl

1.2 × 10

-3

0.1 × 10-3

0.3 × 10-6

Biotin

8.2 × 10-6

0.5 × 10-6

2.05 × 10-9

Cyanocobalamin

1.48 × 10-6

0.5 × 10-6

0.37 × 10-9

621.9

3.0

Vitamin stock solution 1

0.25

Organic Buffer CHES/NaOH, pH 8.0 1

300

10

These solutions were filter sterilised and added to the autoclaved medium after cooling.

Strains Collected Using the methods described above, a total of seven strains of M. aeruginosa were grown, with a number of sub-lines from each strain (Table 2.2). Sub-lines were given the strain name and numbered according to the order of isolation (e.g. LB-A2 was the second sub-line of strain LBA). An additional axenic strain, M. aeruginosa MASH-01A19 (Table 2.2) was kindly supplied by


the CSIRO Microalgae Research Centre (CMARC) in Hobart, Tasmania.

43

The strain M.

aeruginosa MB-M was collected from a sample taken from a Microcystis sp. bloom in Mount Bold Reservoir, South Australia (Fig. 2.1), which was kindly supplied by Peter Baker (Australian Centre for Water Quality Research, Adelaide, South Australia). Bacterial contamination remained visible in several strains after sub-culturing. The growth of these strains was discontinued. The remaining strains, DTW-3, LB-A1, LM-B3 and MASH-01A19, were kept for further analysis of MCYST production (Section 2.3).

TABLE 2.2 Summary of the strains of M. aeruginosa isolated from freshwater blooms in Victoria and South Australia, used in this study. Sampling site 1

Goulburn River, Shepparton, Vic.

Number of

Axenic sub-

sub-lines

lines 4

C. Bolch (1992)

24

Yes

MASH-01A19

B. Long (1993)

5

No

LM-M 3

D. Fraser/B. Long (1993)

3

No

LW-M 3

Selective enrichment

B. Long (1993)

6

NC

DTW

Spread plating/Na2S

B. Long (1993)

2

No

STW-M 3

Method of isolation Centrifugation and spread plating 2

Lake Mokoan, Vic.

Selective enrichment

Lake Weroona, Bendigo, Vic.

Pour plating

Drouin Sewage Treatment Ponds, Drouin, Vic. Shepparton Sewage Treatment Ponds, Shepparton, Vic.

Collector/Isolator (Year)

Strain designation

Lake Boga, Vic.

Streak plating

D. Fraser/B. Long & D. Fraser (1993)

10

NC

LB-A

Mt. Bold Reservoir, S.A.

Micromanipulation

P. Baker/B. Long & D. Fraser (1993)

34

No

MB-M 3

Lake Mokoan, Vic.

Streak plating

B. Long (1996)

3

NC

LMB

1

Sampling sites are shown in Fig. 2.1.

2

The isolation and purification of this strain is described by Bolch & Blackburn (1996).

3

These strains were discontinued due to excessive bacterial contamination.

4

NC, Axenicity not confirmed.

44


2.3

45

Investigation of MCYST production capabilities of selected strains of M. aeruginosa

The four strains of M. aeruginosa, DTW-3, LB-A1, LM-B3 and MASH-01A19 were selected as candidates for the analysis of MCYST production due to their low (or absent) bacterial contamination (Table 2.2). Each strain was grown in modified MLA medium (Table 2.1) under standard growth conditions (26 °C, constant illumination at 40 µmol photons m-2 s-1) and the ability of each strain to produce MCYSTs was determined according to the methods described below (Sections 2.3.1 to 2.3.3).

2.3.1

Production of MCYSTs by selected M. aeruginosa strains as determined by protein phosphatase (PP) inhibition

Background The potent inhibition of PPs by MCYSTs has been well established (MacKintosh et al. 1990). As a result, the inhibition of PP activity by MCYSTs has now become a routine method for the determination of MCYST content in both field and cultured cyanobacterial material (e.g. Jones & Orr 1994; Kaebernick et al. 2000). A variety of PP substrates, in conjunction with a source of PP, have been applied to MCYST analysis (Honkanen et al. 1990; An & Carmichael 1994; Fontal et al. 1999). In this section, the potential of four strains of M. aeruginosa to produce MCYSTs is determined by inhibition of PP activity using a novel fluorescent PP inhibition assay and a purified, recombinant catalytic subunit of a type 1 PP.

3UHSDUDWLRQ SXULILFDWLRQ DQG DVVD\ RI 33

33 ZDV SUHSDUHG IURP WKH KXPDQ 33 cDNA clone (Barker et al. 1993) which was a kind gift from Dr P.T.W. Cohen (MRC Protein Phosphorylation Unit, Department of Biochemistry,

8QLYHUVLW\ 'XQGHH 'XQGHH 6FRWODQG 8. 33 LV DQ LVRIRUP RI WKH FDWDO\WLF VXEXQLW RI human PP1, a protein serine/threonine phosphatase (Barker et al. 1993). Recombinant 33 was supplied as a sub-clone in the bacterial expression vector plasmid pCW under an isopropyl-thio-D-galactopyranoside (IPTG)-inducible duplicated trp-lac hybrid promoter (Alessi et al. 1993).

The plasmid, also containing an ampicillin resistance marker gene, was introduced into Escherichia coli MV 1190 cells using the method of Inoue et al. (1990). Transformants were selected by growing MV 1190 cells overnight at 37 °C on agar plates containing Luria-Bertani (LB) medium and ampicillin (50 µg ml-1). Successful colonies were further streaked onto LB agar plates (50 µg ml-1 ampicillin) to ensure clonal separation.


46

Transformed cells were prepare

G IRU H[SUHVVLRQ RI 33 E\ UHPRYLQJ FRORQLHV WR OLTXLG /% EURWK

containing ampicillin (100 µg ml-1) and growing the cells overnight at 37 °C with constant shaking. Successive sub-culturing allowed for the development of large cell densities required for

SXULILFDWLRQ RI 33 ,QGXFWLRQ RI 33 V\QWKHVLV ZDV DFKLHYHG ZLWK O FXOWXUHV RI /% EURWK containing 100 µg ml-1 ampicillin, using the method of Alessi et al. (1993). When the cultures reached an optical density at 600 nm (OD600) of 0.3, expression of the PP1γ gene was induced by the addition of IPTG to a final concentration of 1 mM. The cultures were then grown at 28 °C for 16 h to maximise PP1γ as a proportion of cellular protein. Cells were harvested by centrifugation (5,000 × g, 15 min, 4 °C) and the pellet was snap frozen in liquid N2, then stored at -70 °C until required.

33 ZDV SXULILHG LQ D UDSLG VLQJOH-step

procedure.

This involved chromatography on a

MCYST-LR-sepharose affinity column which was made using purified MCYST-LR (Section

)UR]HQ FHOOV FRQWDLQLQJ 33 ZHUH WKDZHG DQG UHVXVSHQGHG LQ DSSUR[LPDWHO\ PO RI 33 H[WUDFWLRQ EXIIHU P0 +(3(6 S+ 2.3.4) according to the method of Moorhead et al. (1994)

7.5; 100 mM KCl; 5 % (v/v) glycerol; 1 mM EDTA; 2 mM dithiothreitol; 0.5 mg ml-1 lysozyme; 5 µg ml-1 DNase-1; 0.1 mM phenylmethylsulfonyl fluoride (PMSF); 1 mM benzamidine). Cells were lysed at 4 °C for 2 min using a Vibrinogen Cell Mill (Edmund Bühler, Germany) containing approximately 5 ml of dry 100 µm diameter glass milling beads. The resulting bacterial lysate was collected by centrifugation (10,000 × g, 5 min, 4 °C) and the pellet was returned to the cell mill and re-extracted in a small volume of buffer for an additional two min. The second bacterial lysate was centrifuged as before and the resulting supernatants were pooled and passed through a 45 µm sterile filter before storing at -20 °C until required. Protein in the filtered lysate was determined according to the method of Bradford (1976) using bovine serum albumin as a

VWDQGDUG 33 IURP DSSUR[LPDWHO\ PJ RI SURWHLQ LQ WKH ILOWHUHG O\VDWH ZDV SXULILHG XVLQJ WKH MCYST-LR-Sepharose column according to the method of Moorhead et al. (1994).

RQWDLQLQJ 33 ZHUH FROOHFWHG IURP WKH FROXPQ DQG 33 DFWLYLW\ LQ

Approximately 1 ml fractions c

the fractions was determined using the p-NPP (p-nitrophenyl phosphate) assay described below.

7KH PRVW DFWLYH IUDFWLRQV ZHUH SRROHG DSSUR[LPDWHO\ PO DQG GLDO\VHG WZLFH DJDLQVW 33 storing buffer (50 mM triethanolamine/HCl, pH 7.5; 0.1 mM ethylene glycol-bis[-aminoethyl

ether]N,N,N',N'-tetraacetic acid (EGTA); 5 % (v/v) glycerol; 0.1 % (v/v) -mercaptoethanol; 1

P0 EHQ]DPLGLQH IROORZHG E\ D VLQJOH GLDO\VLV DJDLQVW 33 VWRULQJ EXIIHU FRntaining 55 % (v/v) JO\FHURO 7KH SXULW\ RI WKH 33 ZDV GHWHUPLQHG E\ 6'6-PAGE using the discontinuous method of Laemmli (1970) )LJ 3XULILHG 33 DSSUR[LPDWHO\ PO ZDV VWRUHG LQ µl aliquots at -20 °C with no loss in activity over five months. Although Mn LV UHTXLUHG IRU 33 DFWLYLW\ 2+

(Alessi et al. 1993), it was found that MnCl2 in the dialysis buffer caused precipitation of PP1γ


47

ZLWKLQ WKH GLDO\VLV EDJ 7KXV WKH SURFHGXUH IRU WKH SXULILFDWLRQ RI 33 differs from that of Moorhead et al. (1994) in that MnCl2 was omitted from both the growth medium for transformed cells and all extraction/dialysis buffers. This had no adverse affect on PP1γ yield, purity or activity.

The PP activity in fractions eluting from the MCYST-LR-Sepharose column was determined using p-NPP (p-nitrophenyl phosphate) as substrate. Reactions were carried out at 37 °C in 1.5 ml reaction tubes in a volume of 850 µl consisting of 550 µ

O RI 33 DVVD\ EXIIHU P0

imidazole/HCl, pH 7.6; 0.1 mM EDTA; 2 % (v/v) -mercaptoethanol; 1 mg ml-1 bovine serum albumin; 4 mM MnCl2) and 50 µl samples from the MCYST-LR-Sepharose column. Reactions were initiated by the addition of 250 µl of 20 mM p-NPP and terminated after 15 min by the addition of 150 µl of ice-cold 90 % saturated (NH4)2SO4. Terminated reaction mixtures were stored on ice for 5 min at 0 °C and then clarified by centrifugation (10,000 × g, 1 min, 4 °C). The amount of product (p-nitrophenol, p-NOH) in the supernatant was determined by measuring the absorbance at 405 nm using a Pye-Unicam SP-8 UV/Vis spectrophotometer.

,QKLELWLRQ RI 33

DFWLYLW\ E\ H[WUDFWV IURP

M. aeruginosa strains

,QKLELWLRQ RI 33 E\ H[WUDFWV RI M. aeruginosa strains DTW-3, LB-A1, LM-B3 and MASH01A19 was determined by the production of the fluorochrome 4-methylumbelliferone (MU) using 4-methylumbelliferyl phosphate (MUP) as substrate. Exponential phase batch cultures of the M. aeruginosa strains DTW-3, LB-A1, LM-B3 and MASH-01A19 were harvested by centrifugation (5,000 × g, 30 min) and freeze-dried. Extracts were prepared from each strain by suspending dried cell material in Milli-Q H2O (3.3 mg ml-1) in a 1.5 ml tube which was then placed in boiling water for 10 min according to the method of Metcalf & Codd (2000). Cell debris was removed by centrifugation (10,000 × g, 5 min) and supernantants were serially diluted (ranging from undiluted

WR GLOXWLRQ LQ 33 DVVD\ EXIIHU VHH GHWDLOV IRU p-NPP assays above) prior to analysis XVLQJ WKH 33-083 LQKLELWLRQ DVVD\ 33-MUP inhibition assays were carried out in microtitre plates using 5 µ O SXULILHG 33 LQ D ILQDO UHDFWLRQ YROXPH RI µl. Incubation mixtures also contained 20 µl of diluted extract of M. aeruginosa strains or 20 µl of standard MCYST-LR

&DOELRFKHP GLOXWHG LQ 33 DVVD\ EXIIHU DQG µ O RI 33 DVVD\ EXIfer. Reactions were initiated by the addition of 20 µ O RI P0 083 $SSURSULDWH FRQWUROV ODFNLQJ 33 083 RU

MCYST/diluted extract were also included. Reactions were carried out at 37 °C in a Fluoroskan II 96 well microtitre plate fluorescence detector (Labsystems).

Activity in each well was

PHDVXUHG E\ WKH LQFUHDVH LQ IOXRUHVFHQFH QP = 460nm) over a period of 45 PLQ XVLQJ *HQHVLV VRIWZDUH /DEV\VWHPV $ VWDQGDUG FXUYH RI 33 DFWLYLW\ YHUVXV 0& D, biomass will accumulate within the culture vessel. Once a flow rate has been set in a chemostat system, fluctuation in biomass concentration may occur initially as cells adjust to the rate of supply of the limiting nutrient. Eventually, cells achieve a physiological state in which they are able to utilise the limiting nutrient at the same rate that it is supplied. Under these conditions µ = D, and the flow rate therefore determines the specific growth rate. This is known as steady-state, the physiological state of cells under these conditions being constant. Thus, biomass (measured by either cell concentration or dry weight concentration) also remains constant, cellular processes generally being controlled by the concentration of the limiting nutrient. When steady-state prevails in a chemostat, Monod (1950) proposed that the specific growth rate is dependent upon the concentration of the limiting nutrient (S) according to Michaelis-Menten type kinetics.

Chapter 3 MCYST quota in N-limited chemostats

µ = µ max

S Ks + S

72

(Equation 3.4)

The term Ks is the concentration of the limiting nutrient at which half maximal specific growth rate is achieved. The concept that nutritional requirements determine specific growth rate was further developed by Droop (1973), who showed more specifically that the internal quota of the limiting nutrient (Q) determines the specific growth rate:

µ = µ ′ max

Q − Q min Q

(Equation 3.5)

where µ'max is the apparent maximum specific growth rate at infinite Q, and Qmin is the minimum intracellular quota of the limiting nutrient. The interaction of these two concepts is made clear in a review by Morel (1987).

Thus, cells grown in a chemostat are controlled by both the dilution rate and the internal concentration of the limiting nutrient. As dilution rates are changed in such a system, there is a period of fluctuation in biomass as cells readjust to changes in the rate of supply of the limiting nutrient. Therefore, whenever dilution rate is altered, time is required to allow cells to reach steady-state once again before subsequent analysis at the new dilution rate. Proceeding in this way for a range of dilution rates allows physiological analysis of cellular processes over a range of specific growth rates above µ = 0 and approaching µ max (Pirt 1975; Pritchard & Tempest 1982).

3.3 3.3.1

Materials and Methods Organism and growth conditions

Microcystis aeruginosa MASH-01A19 (Bolch & Blackburn 1996; Bolch et al. 1996; Bolch et al. 1997) was provided by the CSIRO Marine Laboratories culture collection.

The MCYST

production capabilities of this strain are described in Chapter 2. For the analysis of intracellular MCYST quota under N-limited growth conditions, cells were grown in three replicate 500 ml continuous culture vessels under constant illumination (40 µmol photons m-2 s-1), using cool white fluorescent lights, at 26 ± 1 °C in a constant temperature room. Cultures were supplied with a continuous flow of modified MLA (Bolch & Blackburn 1996) medium (Table 2.1) containing 0.2 mM NaNO3 (1/10 original concentration) and 3.0 mM 2-(N-cyclohexylamino) ethanesulfonic acid (CHES)/NaOH, pH 8.0 buffer via a Gilson multiplus 2 peristaltic pump. A single air pump provided constant airflow to all cultures throughout the experiment. Using the air pump to


73

maintain a positive pressure in each culture vessel ensured that a constant volume was maintained via a vertical overflow tube (Fig. 3.1). Flow rates were set using the peristaltic pump, and accurate dilution rates were determined by measuring overflow volumes from each culture at regular intervals (one to two days). The configuration of the chemostat system is shown in Fig. 3.1. Using this system, dilution rates of between 0.1 and 1.08 d-1 were achieved through control of the peristaltic pump flow rate. Fresh, sterile growth medium was made up routinely from stock solutions kept in the dark at 4 °C and autoclaved before use. The medium reservoir was filled, using an additional peristaltic pump, via sterile 0.22 µm filters, and the formation of precipitates within the medium was minimised by constant stirring of the medium (Fig. 3.1).

Hello!

Fig. 3.1. Schematic diagram of the chemostat system used in this experiment. 1 Reservoir of sterile MLA medium. 2 Peristaltic pump. 3 Replicate 500 ml culture vessels. 4 Air pump. 5 Overflow vessels from each culture vessel. 6 Sterile sampling port and sampling syringe. 7 Tube clamp. 8 Sterile, stainless-steel airstones. 9 Vertical overflow device to maintain constant volume. 10 Two sterile 0.22 µm filters. 11 Magnetic stirring plates. 12 Sterile cotton wool air filters. The arrows mark the direction of air/liquid flow. The sterile growth medium reservoir can be refilled via the two sterile inlet filters.

The growth medium reservoir, culture vessels, overflow vessels and connecting tubing were sterilised by autoclaving the entire system with connections in place. Attachment of sterile filters, addition of initial growth medium and inoculation of cultures was carried out under sterile conditions in a laminar flow cabinet. Cultures were grown initially under batch conditions until approximately eight days after inoculation. After this time the pump was turned on and steadystate conditions allowed to develop. The procedures used to ascertain the attainment of steadystate conditions are given in Section 3.4.1. Batch cultures were analysed to determine that the concentration of NO3- supplied to chemostats (0.2 mM) ensured N limitation and low biomass concentration. Thus, biomass concentration in the cultures does not result in light limitation through mutual shading. Conical flasks (500ml)


74

containing 250 ml of either MLA medium, which contains 2.0 mM NO3- (Table 2.1), or MLA medium containing 0.2 mM NO3-, were inoculated with stationary phase (N-depleted) cells of M. aeruginosa MASH-01A19. Each NO3- treatment was conducted in triplicate. Culture flasks were grown under conditions of light and temperature identical to those for the chemostats. For approximately 21 days, cultures were sampled at regular intervals (two to three days) for cell concentrations (cells ml-1).

In addition, triplicate batch cultures were used to estimate the maximum specific growth rate (µ max) of M. aeruginosa MASH-01A19 grown under the same NO3-, temperature and light conditions as the chemostats. In this instance, exponentially growing cells were used to inoculate 500 ml culture flasks containing 250 ml of MLA medium with 0.2 mM NO3-. The specific cell division rate (µ c) and the specific rate of dry weight accumulation (µ g, Orr & Jones 1998) were determined using the method described for batch cultures in Section 4.2.3 (Equation 4.1) after frequent sampling post-inoculation.

3.3.2

Sampling and analysis

To ensure steady-state conditions at each specific growth rate, the stability of culture populations was determined by cell counting and dry weight analysis. Once populations had stabilised at each specific growth rate (not more than 3% variation in cell concentration or dry weight between four successive samplings), cultures were allowed to grow at steady-state for at least five residence times before sampling. Approximately 75 ml of each culture was removed for analysis of cell concentrations (cells ml-1), cell dry weight (pg cell-1), MCYST content, chlorophyll a, cell volume, total cellular protein and amino acid content. Cell counting was carried out using a haemocytometer (Neubauer) after disrupting colonies at 80 °C for 20 min, using the method of Humphries & Widjaja (1979). For the determination of cell volume, the cells treated for cell counting were concentrated by centrifugation (13, 000 × g, 5 min) and photomicrographs of the cells and a standard scale (10 µm) were taken using an Olympus compound microscope fitted with a camera. Once developed, photomicrographs were digitally scanned and cross-sectional cell area determined using NIH Image software (National Institutes of Health, Bethesda, Md, USA. rsb.info.nih.gov/nih-image/). The cross-sectional area of between 40 and 100 cells from each culture at each specific growth rate was determined, and was then used to calculate mean cell volume assuming spherical cells. Both dividing and non-dividing cells were treated equally. Additional cell samples taken from cultures grown at µ = 0.96 d-1 were used to examine the effect of heat treatment on cell volume by carrying out the procedure above on cells which had not been heated to 80 °C for 20 min.


75

MCYST extraction and analysis Dry weight was determined by collecting specific volumes of culture material on pre-weighed 47 mm diameter Whatman GF/C filters and allowing them to dry overnight in a vacuum desiccator before reweighing. MCYSTs were recovered from the filters by extracting four times in 2.0 ml of 80% (v/v) methanol. The extracts were pooled and dried in vacuo using a Speedie-Vac vacuum centrifuge. The resulting residue was redissolved in 2.0 ml of 80% (v/v) methanol and analysed for MCYSTs by HPLC using the method of Jones & Orr (1994). MCYSTs were separated on a Alltima C18 column (4.6 mm × 250 mm, Alltech), using a linear gradient of 20 to 35 % acetonitrile (v/v) in 8 mM ammonium acetate, and detected by measuring the absorbance of the eluant at 238 nm (see Appendix 3).

Quantification of MCYSTs was made possible by

comparison with a MCYST-LR standard (Calbiochem) and all results are expressed as MCYSTLR molar equivalents.

Amino acid analysis For analysis of total amino acid content, whole cells (containing free amino acids and the amino acids in proteins and other cellular peptides including MCYSTs) from approximately 10 ml of culture were collected by centrifugation and hydrolysed in 6.0 N HCl / 0.2 % (w/v) phenol for 24 hours. Hydrolysates were dried under a stream of N2 gas and the residues were redissolved in 100 µl 0.25 M borate buffer (pH 8.5) before derivatisation, using 9-fluorenylmethyl chloroformate (FMOC), and HPLC analysis according to the procedure of Ou et al. (1996). A chromatogram of a derivatised standard amino acid mixture separated by HPLC using this method is shown in Appendix 3. All protein amino acids except for glutamine and asparagine were analysed using this method. The amides glutamine and asparagine were measured as glutamate and aspartate respectively due to the deamidation of these amino acids during hydrolysis. For this reason total glutamate + glutamine and total aspartate + asparagine are represented as Glx and Asx respectively. A standard mixture of amino acids was hydrolysed with samples to determine the rate of loss of each amino acid under the hydrolysis conditions. Subsequent analysis allowed for any losses by using a correction factor for each amino acid. Amino acids are presented as total protein amino acid content of biomass (i.e. fmol cell-1 or µg (mg dry weight)-1 - see Table 3.1), or individually on a cell quota basis (i.e. fmol cell-1 - see Table 3.2). In addition to total cellular amino acids, free amino acids, soluble in the 80 % methanol fraction obtained during MCYST extraction, were also measured by FMOC derivatisation. However, amino acid concentrations in the 80 % methanol extracts were insufficient to allow quantification of free amino acids in these samples.


76

Protein, chlorophyll a and dry weight analysis Total cellular protein was estimated using the method of Lowry et al. (1951) as adapted by Walsh et al. (1997), using bovine serum albumin (BSA) as a standard. Cells from 5 to 10 ml of culture were collected by centrifugation (13, 000 × g, 5 min) and dried in vacuo. Cells were resuspended in 250 µl of 0.5 M NaOH and heated to 70 °C for 20 minutes before analysis.

Chlorophyll a was estimated in 80% (v/v) acetone extracts using the method of Arnon et al. (1974) in all cultures grown at dilution rates above 0.3 day-1. An absorption coefficient of 82.04 (mg chlorophyll a ml-1)-1 cm-1 at 663 nm was employed. In cultures grown at dilution rates of 0.3 d-1 and below, chlorophyll a in cells was below the detection limit and could not be determined using this method.

3.4 3.4.1

Results Steady-state Growth

As determined by consistent dry weight and cell concentrations, the chemostat cultures were at steady-state at each specific growth rate prior to sampling. The steady-state cell concentration significantly increased with increasing specific growth rate (P < 0.001, Table 3.1) but cell dry weight decreased from 43.4 to 17.7 pg cell-1 (Table 3.1).

Hence, steady-state biomass

concentrations increased only slightly with increasing specific growth rates (P < 0.03) ranging from 50.1 to 61.9 mg ml-1 between the lowest and highest specific growth rates (0.10 – 1.08 d-1). The reduction in cell weight with increasing specific growth rate was associated with a decrease in cell volume of approximately five-fold from 111 to 19.2 µm3 (Table 3.1). The effect of the heat treatment used to disrupt colonies for cell counting had a minimal effect on cell volume; heat-treated cells were slightly smaller in size (approximately 4 µm3) compared with untreated cells (Fig. 3.2). This is consistent with the minor volume changes noted by Porter & Jost (1976) after collapsing gas vacuoles. An example of the difference in cell volume at different specific growth rates is given in Fig. 3.5. Cell protein quota (pg protein cell-1) showed no specific correlation with µ (Table 3.1). In contrast, protein expressed per unit dry weight (µg protein (mg dry weight)-1) increased significantly from lowest to highest specific growth rates (P < 0.001, Table 3.1).

Where

chlorophyll a was quantifiable, there was a significant increase in chlorophyll a content with specific growth rate (P < 0.001, Table 3.1) expressed both as cell quota (pg cell-1) and per unit dry weight (µg (mg dry weight-1)). Chlorophyll a content was not determined at low specific growth


77

rates (below detection limits), but the cultures were visibly more yellow while cell concentrations did not differ significantly (Table 3.1), indicating low chlorophyll a quotas in slower growing cells.

The total cellular protein amino acid content decreased significantly as µ increased (P < 0.05) varying approximately three-fold both on a cell quota basis (fmol cell-1) and as a proportion of dry weight (µg (mg dry weight)-1) over the range of specific growth rates examined (Table 3.1). Interestingly, the total amino acid content (µg (mg dry weight)-1) does not quantitatively account for the protein content as measured by the Lowry method, except for cultures growing at the lowest specific growth rate (Table 3.1), indicating possible problems associated with protein determination using one or both methods. Potential reasons for this discrepancy are discussed in Section 3.5.

The intracellular quota of individual protein amino acids also decreased with increasing µ (Table 3.2). Notably, those amino acids which are constituents of MCYST-LR (viz. alanine, arginine, leucine, and perhaps proportions of Asx and Glx) are present in sufficient quantities to account for the MCYST present in the cells (Table 3.2 and Section 3.4.2). The cellular amount of the Scontaining amino acids cysteine and methionine and the amino acid tryptophan were relatively low compared with other amino acids (Table 3.2). A preliminary batch culture study revealed that 0.2 mM NO3- resulted in a decrease in biomass yield compared with normal MLA medium which contains 2 mM NO3- (Fig. 3.3), establishing that at 0.2 mM NO3-, N is the limiting nutrient.

3.4.2

MCYST analysis

As described in Chapter 2, two MCYST peaks were determined in M. aeruginosa MASH-01A19 by HPLC (Fig. 2.6). The first of these is MCYST-LR which, although it probably contains desmethyl isomers of MCYST-LR (Orr & Jones 1998), is expressed in MCYST-LR molar equivalents (assuming similar molar absorption coefficients). The second MCYST, giving the characteristic absorbance maximum at 238 nm but an uncharacteristic LC-MS spectrum with high fragmentation (G. Jones, personal communication), is yet to be conclusively identified (Section 2.3.4) and was not included in the measurement of total MCYST. At all specific growth rates this compound constituted less than 15% of total MCYST-LR equivalents. Thus, the values presented for MCYST in this study refer to MCYST-LR.


78

Notably, MCYSTs were not detected in the extracellular medium of any of the cultures, regardless of the specific growth rate. With a detection limit of 1.0 pmol on-column using HPLC, extracellular MCYST could not have exceeded a concentration of 5 nM, and was therefore always less than 1% of total culture MCYST. Cell MCYST quota (QMCYST) ranged from 0.052 – 0.116 fmol cell-1, showing a positive linear correlation with specific growth rate (r = 0.952, Fig. 3.4). Extrapolation of the fitted regression suggests that QMCYST reaches a minimum value at µ = 0 (QMCYSTmin) and a maximum value (QMCYSTmax) at µ max (Fig. 3.4). The linear relationship between QMCYST and µ can be described in terms of the predicted minimum and maximum cell quotas (QMCYSTmin and QMCYSTmax) and the maximum specific growth rate (µmax) thus;  QMCYST max − QMCYST min   + QMCYST min QMCYST = µ ×  µ max  

(Equation 3.6)

Since µ max cannot be achieved in chemostat cultures, this parameter was determined from analysis of separate batch culture data and estimated to be 1.2 d-1. Using this value, QMCYSTmin and QMCYSTmax were subsequently calculated, from linear regression analysis, to be 0.050 ± 0.004 (SE) and 0.129 ± 0.006 (SE) fmol cell-1 respectively. The MCYST production parameters determined under the conditions examined in this study are summarised in Table 3.3. MCYST expressed per unit dry weight (µg MCYST (mg dry weight)-1) increased significantly with increasing specific growth rate (Fig. 3.6A). However, because cell dry weight decreased with increasing specific growth rate (Table 3.1), the increase in MCYST:dry weight ratio was more than five-fold (1.18 to 6.47 µg (mg dry weight)-1) compared with the less than three-fold increase for QMCYST. MCYST expressed per unit Lowry protein (µg MCYST (mg protein)-1) was greater at high µ than low µ, but reached a maximum at intermediate µ (Fig. 3.6B). Expressed as a proportion of total protein amino acid content (µg MCYST (mg total amino acids) -1), MCYST also increased with µ (Fig. 3.6C). MCYST normalised to chlorophyll a was not significantly different over the specific growth rates examined (Fig. 3.6D); the average MCYST:chlorophyll a ratio was 0.59 ± 0.03 (SE) on a mass basis (g:g) or 0.53 ± 0.02 (SE) on a molar basis (mol:mol). Intracellular MCYST concentration (mol MCYST per unit cell volume) ranged from 0.47 ± 0.34 (SE) mM to 5.5 ± 0.77 (SE) mM over the specific growth rates examined (Fig. 3.6E). There was a strong negative correlation between intracellular MCYST concentration and cell volume (Fig. 3.7) with smaller cells containing significantly higher concentrations of MCYST (P < 0.001). On


79

a per culture volume basis, the culture MCYST concentration increased with µ to 0.58 d-1, thereafter remaining relatively constant at approximately 0.4 µM (Fig. 3.6F). This follows a similar pattern to cell concentrations and culture dry weight (see Table 3.1). It must be noted, however, that this measurement is based entirely on internal MCYST, as none was found in the external medium. The net MCYST production rate (RMCYST, fmol MCYST cell-1 d-1) was determined from the product of µ and QMCYST. Minimum net rates of MCYST production were 0.005 ± 0.0005 fmol cell-1 d-1 and 0.11 ± 0.002 (µg (mg dry weight)-1 d-1 at 0.1 d-1, and maximum net rates were 0.13 ± 0.01 fmol cell-1 d-1 and 6.9 ± 0.07 µg (mg dry weight)-1 d-1 at 1.08 d-1. The net rates of MCYST production (per mg dry weight) reported in previous studies for M. viridis TAC44 (0.175 µg (mg dry weight)-1 d-1) and M. aeruginosa M228-12 (1.13 µg (mg dry weight)-1 d-1) (Watanabe et al. 1989) and for M. aeruginosa UTEX 2388 (0.11 to 0.44 µg (mg dry weight)-1 d-1) (Oh et al. 2000) fall within the range reported here. RMCYST shows a positive correlation with specific growth rate (Fig. 3.8) and can also be described in terms of QMCYSTmin, QMCYSTmax, µ and µmax by the equation:  (µ × QMCYST max ) − (µ × QMCYST min ) RMCYST = µ ×   + (µ × QMCYST min ) µ max  

(Equation 3.7)

This relationship predicts that RMCYST = 0 in cells at stationary phase and that RMCYST reaches a maximum (RMCYSTmax) of 0.155 fmol cell-1 d-1 (9.1 mg g-1 dry weight d-1) at µ max (Table 3.3). The hyperbolic shape of the relationship (Fig. 3.8) results from the absolute difference between QMCYSTmax and QMCYSTmin; the higher the QMCYSTmax:QMCYSTmin ratio, the greater the curvature in the RMCYST versus µ plot.

TABLE 3.1 Effect of specific growth rate on various cellular parameters, protein, total amino acids and chlorophyll a in M. aeruginosa MASH-01A19 grown in N-limited chemostats. 1 Cellular paramters Protein content Total amino acid content 2 Chlorophyll a content

Specific growth rate -1

Cells ml-1

Cell dry weight

Cell volume

Quota

(µg (mg dry

Quota

(µg (mg dry

Quota

(µg (mg dry

(µ, d )

(× 10 )

(pg cell )

(µm3)

(pg cell )

weight) )

(fmol cell )

weight) )

(pg cell )

weight)-1)

0.10

1.17 ± 0.10

43.4 ± 3.2

111 ± 49

6.7 ± 0.4

154 ± 3.0

113 ± 4.0

312 ± 39

ND

ND

0.13

1.36 ± 0.10

37.4 ± 3.5

99.8 ± 41

6.4 ± 0.3

173 ± 10

53.4 ± 5.8

163 ± 18

ND

ND

0.20

1.33 ± 0.02

36.0 ± 1.4

110 ± 48

7.6 ± 0.2

212 ± 6.0

48.2 ± 2.4

153 ± 8.0

ND

ND

0.23

1.38 ± 0.11

37.4 ± 1.2

96.4 ± 37

8.3 ± 0.3

221 ± 6.0

55.9 ± 8.7

172 ± 26

ND

ND

0.27

2.04 ± 0.10

33.0 ± 3.3

84.1 ± 32

8.0 ± 0.5

246 ± 8.0

49.2 ± 4.2

168 ± 14

ND

ND

0.45

2.70 ± 0.36

21.4 ± 3.3

61.7 ± 31

8.3 ± 0.6

397 ± 34

39.7 ± 0.9

136 ± 4.0

0.12 ± 0.004

5.64 ± 0.79

0.49

2.84 ± 0.34

19.9 ± 1.8

75.6 ± 29

6.3 ± 0.2

320 ± 23

39.5 ± 0.3

132 ± 10

0.15 ± 0.004

7.46 ± 0.61

0.59

4.64 ± 0.65

21.2 ± 0.5

89.4 ± 26

4.9 ± 0.3

231 ± 20

24.3 ± 3.3

85.3 ± 12

0.14 ± 0.010

6.71 ± 0.44

0.70

4.10 ± 0.69

20.2 ± 1.8

20.0 ± 4.9

5.8 ± 0.6

286 ± 15

35.4 ± 6.6

112 ± 28

0.18 ± 0.018

8.95 ± 0.64

0.96

3.72 ± 0.81

17.7 ± 0.7

19.2 ± 4.9

6.8 ± 0.5

387 ± 46

ND

ND

0.19 ± 0.013

11.0 ± 1.2

1.08

3.45 ± 0.14

18.0 ± 1.6

39.4 ± 14

8.1 ± 0.6

451 ± 30

ND

ND

0.22 ± 0.006

12.1 ± 0.83

6

-1

-1

-1

-1 3

-1 4

-1

1

Data presented are means of triplicate cultures ± SE except for cell volume; cell volumes are means from all cultures at each specific growth rate ± SD. All data show significant variation over the range of specific growth rates examined (P < 0.05). ND, not determined.

2

Total amino acids obtained from intact cells by hydrolysis in HCl/phenol (see Section 3.3.2).

3

Values are the total of all protein amino acids (listed in Table 3.2) per cell determined as the mean of the triplicate chemostats.

4

Total amino acid weights were calculated assuming asparagine and glutamine constituted 50% of Asx and Glx respectively.

80

TABLE 3.2 Amino acid quotas (fmol cell-1) of whole cells of M. aeruginosa MASH-01A19 at steady-state in N-limited chemostats. 1 Specific growth rate (d-1) 0.10

0.13

0.20

0.23

0.27

0.45

0.49

0.59

0.70

Alanine

10.6 ± 2.5

6.6 ± 1.3

6.2 ± 0.6

7.4 ± 1.9

5.7 ± 0.8

4.7 ± 0.3

4.3 ± 0.8

2.4 ± 1.1

3.3 ± 1.4

Arginine

7.4 ± 1.9

2.1 ± 0.4

1.7 ± 0.1

2.5 ± 0.4

2.2 ± 0.5

1.9 ± 0.4

1.8 ± 0.1

1.2 ± 0.2

1.8 ± 0.5

2.7 ± 0.8

4.4 ± 0.9

3.1 ± 0.0

4.0 ± 1.2

2.1 ± 0.8

2.2 ± 1.0

2.3 ± 1.3

1.0 ± 0.7

1.1 ± 0.8

Cysteine

0.6 ± 0.5

0.3 ± 0.1

0.3 ± 0.0

0.2 ± 0.0

0.3 ± 0.1

0.2 ± 0.0

0.1 ± 0.1

0.1 ± 0.1

0.2 ± 0.0

2

5.6 ± 1.5

6.0 ± 1.1

5.2 ± 0.4

5.7 ± 1.8

3.5 ± 1.0

3.4± 0.6

3.4 ± 1.3

1.6 ± 0.9

1.9 ± 1.1

Glycine

9.7 ± 2.0

4.2 ± 0.7

4.5 ± 0.3

5.3 ± 1.3

4.4 ± 0.9

3.9 ± 0.1

3.7 ± 0.4

2.7 ± 0.4

3.8 ± 0.9

Histidine

1.5 ± 0.4

0.8 ± 0.1

0.8 ± 0.1

0.9 ± 0.2

0.7 ± 0.2

0.6 ± 0.1

0.5 ± 0.1

0.4 ± 0.1

0.7 ± 0.1

Isoleucine

7.8 ± 1.7

2.0 ± 0.3

1.8 ± 0.2

3.5 ± 0.5

2.0 ± 0.3

1.6 ± 0.4

1.5 ± 0.0

1.0 ± 0.2

1.4 ± 0.4

Leucine

11.2 ± 2.3

5.3 ± 0.9

4.8 ± 0.5

5.8 ± 1.6

5.2 ± 0.7

4.6 ± 0.6

4.1 ± 0.2

2.7 ± 0.7

3.8 ± 1.1

Lysine

5.9± 1.3

1.9 ± 0.3

1.8 ± 0.1

2.1 ± 0.6

1.9 ± 0.3

1.5 ± 0.2

1.5 ± 0.1

0.9 ± 0.3

1.3 ± 0.4

Methionine

3.1 ± 0.6

0.3 ± 0.1

0.2 ± 0.1

0.6 ± 0.2

0.5 ± 0.1

0.5 ± 0.1

0.6 ± 0.0

0.5 ± 0.0

0.8 ± 0.1

Phenylalanine

4.3 ± 1.0

1.7 ± 0.3

1.6 ± 0.1

1.8 ± 0.5

1.8 ± 0.3

1.7 ± 0.0

1.6 ± 0.1

1.3 ± 0.1

1.8 ± 0.4

Proline

5.8 ± 1.4

1.7 ± 0.5

1.9 ± 0.1

2.4 ± 0.4

2.4 ± 0.5

2.1 ± 0.0

1.9 ± 0.2

1.7 ± 0.2

2.2 ± 0.5

Serine

9.2 ± 1.8

3.4 ± 0.6

3.3 ± 0.6

4.3 ± 0.9

3.6 ± 0.6

2.9 ± 0.5

2.5 ± 0.2

1.7 ± 0.5

2.4 ± 0.8

Threonine

7.0 ± 1.4

1.9 ± 0.8

1.5 ± 0.8

2.0 ± 1.0

2.6 ± 0.9

2.1 ± 0.4

2.1 ± 0.1

1.6 ± 0.5

2.3 ± 0.7

Tryptophan

0.8 ± 0.4

0.0 ± 0.0

0.6 ± 0.5

0.1 ± 0.7

0.0 ± 0.0

1.4 ± 0.4

1.4 ± 0.3

1.3 ± 0.8

1.1 ± 0.7

Tyrosine

4.9 ± 1.1

1.5 ± 0.4

1.4 ± 0.2

1.8 ± 0.4

1.8 ± 0.5

1.5 ± 0.2

1.2 ± 0.0

1.0 ± 0.1

1.4 ± 0.3

Valine

9.4 ± 2.0

4.7 ± 0.8

4.0 ± 0.4

4.7 ± 1.3

4.5 ± 0.7

3.6 ± 0.6

3.0 ± 0.1

2.0 ± 0.5

2.8 ± 0.8

Asx

Glx

2

1

The values are the amount of each amino acid per cell in each of three replicate chemostats at each specific growth rate ± SD.

2

Due to deamidation during hydrolysis, asparagine and glutamine are measured as aspartate (aspartate + asparagine; Asx) and glutamate (glutamate + glutamine; Glx) respectively.

81

82

Fig. 3.2. The effects of heat treatment (used in the separation of M. aeruginosa colonies for cell counting) on cell volume. There is a small, but significant (P < 0.05), decrease in cell volume (approximately 4 µm3) after heat treatment. Error bars indicate ± SD.

Fig. 3.3. Time course of cell concentrations (cells

ml-1)

in

batch

cultures

of

M.

aeruginosa MASH-01A19 grown with either 2.0 mM NO3- (|

QRUPDO 0/$ PHGLXP RU

0.2 mM NO3- (

DV XVHG LQ WKH FKHPRVWDW

cultures).

Fig. 3.4.

Relationship between specific

growth rate (µ, d-1) and intracellular MCYST quota (QMCYST, fmol cell-1) of M. aeruginosa MASH-01A19

grown

in

N-limited

chemostats. Error bars indicate ±SEM (n = 3).

The solid line shows cell quotas

calculated from experimental specific growth rates (r = 0.952) using Equation 3.6, and the dashed line represents the extrapolation of this relationship to µ max (where QMCYST = QMCYSTmax) and µ = 0 (where QMCYST = QMCYSTmin).

83

TABLE 3.3 Derived MCYST production parameters for M. aeruginosa MASH-01A19 grown in N-limited chemostats at 26 °C and constant irradiance (40 µmol photons m-2 s-1). 1 Measurement

1

QMCYSTmin

QMCYSTmax

RMCYSTmax

Cell quota basis

0.050 fmol cell-1

0.129 fmol cell-1

0.155 fmol cell-1 d-1

Dry weight basis

0.71 µg (mg dry weight)-1

7.6 µg (mg dry weight)-1

9.12 µg (mg dry weight)-1 d-1

QMCYSTmax and RMCYSTmax were calculated using µ max = 1.2 d-1, as derived from batch culture experiments.

84

85

Fig. 3.5.

Photomicrographs of N-limited chemostat

grown cells of M. aeruginosa MASH-01A19 grown at µ = 0.27 d-1 (A) and µ = 0.59 d-1 (B) showing the differences in cell volume.

Scale bars are 10 µm in both

photomicrographs.

Cells at both specific growth rates

have been heat treated in order to disrupt colonies (for the purpose of cell counting) prior to the determination of cell volume.

86

87

Fig. 3.6. Relationship between specific growth rate (µ, d-1) and MCYST expressed as a ratio to: (A) dry weight (µg MCYST (mg dry weight)-1); (B) protein (µg MCYST (mg protein)-1); (C) total amino acids (µg MCYST (mg total amino acids)-1); and (D) chlorophyll a (mol MCYST:mol chlorophyll a) in M. aeruginosa MASH01A19 grown in N-limited chemostats. MCYST is also expressed on an intracellular concentration basis (mM) (E) and as a total culture concentration (µM) (F). Error bars are SEM (n = 3).

88

89

Fig. 3.7. Intracellular MCYST concentration (mM) in cells of M. aeruginosa MASH01A19, grown in N-limited chemostats, as a function of cell volume (µm3). As shown in Table 3.1, cell volume decreases with increasing specific growth rate.

Fig. 3.8. The net MCYST production rate (RMCYST, fmol MCYST cell-1 d-1) as a function of specific growth rate (µ, d-1) in M. aeruginosa MASH-01A19 grown in Nlimited chemostats. Error bars are ±SEM (n = 3).

The solid line shows net rates of

MCYST production calculated from the estimated QMCYSTmin and QMCYSTmax values (Table 3.3) and µ max (1.2 d-1) using Equation 3.7.

90


3.5

91

Discussion

This study shows that the intracellular MCYST quota of M. aeruginosa increases with the specific growth rate in a consistent and predictable manner (Fig. 3.4). Various other cellular parameters are also linked to the specific growth rate (Table 3.1). Of interest is the inverse relationship between specific growth rate and cell volume (Table 3.1). Thus, rapidly growing cells are much smaller than their slow-growing counterparts and also contain a higher quota of MCYST. As a result, the intracellular concentration of MCYST in small rapidly growing cells is much greater than in large, slow-growing cells (Figs 3.6E, 3.7). These findings highlight the importance of determining intracellular MCYST quotas in experiments examining MCYST content and production. The cell quota of a number of other cellular components (viz. protein, amino acids and chlorophyll a) also varies as a function of specific growth rate (Table 3.1). However, determining MCYST as a ratio to any of these components provides little understanding of the underlying cellular physiology and, on this basis, the data are potentially misleading. This is evidenced by the observation that with increasing specific growth rate MCYST increased linearly as a ratio to cell dry weight and total amino acid content (Figs 3.6A, 3.6C). Also, MCYST generally increased but reached a maximum at intermediate specific growth rate as a ratio to protein content (Fig. 3.6B). However, MCYST was invariant as a ratio to chlorophyll a (Fig. 3.6D).

The model put forward in Equation 3.6 proposes that QMCYST at any specific growth rate is dependent upon the constants QMCYSTmax, QMCYSTmin and µmax in N-limited cultures (Table 3.3, Fig. 3.4). The parameters QMCYSTmax and QMCYSTmin determine a fixed range of intracellular MCYST quotas. This implies that toxigenic strains of M. aeruginosa will always contain, at least, a minimum QMCYST, and that they will not exceed a maximum QMCYST determined by the nutrient saturated µ max (for the given temperature and light growth conditions). Although a specific growth rate of zero cannot be achieved in a chemostat, the QMCYSTmin determined in this experiment is very similar to that observed for N-limited batch cultures of M. aeruginosa MASH01A19 at stationary phase, where QMCYST remained stable for at least two weeks (Orr & Jones 1998). Orr & Jones (1998) found that QMCYSTmax ranged from 0.13 to 0.16 fmol cell-1, again similar to that predicted from this chemostat study (Table 3.3).

Collectively, the data are

consistent with the qualitative generalisation that MCYST production is constitutive and that toxigenic strains remain so under a variety of growth conditions (Sivonen & Jones 1999) but that quantitatively, the amount of MCYST produced is subject to some form of regulation linked to the specific growth rate. In support of this conclusion, data from other researchers suggest that minimum and maximum cell quotas of MCYSTs exist in other strains, and even in different Microcystis spp. (e.g. Watanabe et al. 1989; Song et al. 1998). In addition, the maximum MCYST:dry weight ratio observed in this study (7.6 µg (mg dry weight)-1) is very similar to that


92

reported in late exponential phase for the original strain MASH 01 (the parent culture of MASH01A19 used in this study) by Bolch et al. (1997) (i.e. 7.24 µg (mg dry weight)-1) indicating a conserved process of toxin production in this strain for several years.

Böttcher et al. (2001) recently found QMCYST remained constant but that µ increased with increasing irradiance in turbidostat experiments. This might appear to contradict the result presented here. However, light-limited turbidostats differ from chemostats in that, regardless of the irradiance, µ is equal to light-limited µ max since turbidostats are always nutrient saturated. Therefore, in turbidostats, QMCYST always equals nutrient saturated QMCYSTmax. Thus, on the basis of the study reported in this chapter, the findings of Böttcher et al. (2001) suggest nutrient saturated µ max increases with increasing irradiance while QMCYSTmax remains constant. In contrast, batch culture studies under nutrient replete conditions over a range of temperatures reveal that QMCYSTmin decreases in response to increasing temperature (see Section 4.7). Studies on the response of QMCYSTmin and QMCYSTmax to physical conditions limiting growth (temperature, irradiance) are examined in Chapter 4 of this thesis. The constants QMCYSTmin, QMCYSTmax and µmax also determine the net rate of MCYST production (RMCYST, Equation 3.7) in addition to describing the cell quota of MCYST. RMCYST is the product of QMCYST and µ and, as a consequence, Equation 3.7 predicts no net MCYST production at µ = 0 (or stationary phase in batch culture). Also, RMCYST is constrained by the maximum cell division rate, as QMCYST will not exceed QMCYSTmax which is achieved at µ max (i.e. RMCYSTmax = µ max × QMCYSTmax).

This is inconsistent with the regression model advanced by Oh et al. (2000),

however, which predicts net production of MCYST at zero specific growth rate (0.082 µg (mg dry weight)-1 d-1) in P-limited chemostat grown cells of M. aeruginosa UTEX 2388. This implies that at zero specific growth rate, MCYST production continues (uncoupled from growth), resulting in increasingly toxic non-dividing cells. However, no data were presented to support this prediction. A similar proposal was examined experimentally by Pirt & Righelato (1967) with respect to penicillin synthesis at zero specific growth rate in Penicillium chrysogenum (Authority), a heterotrophic organism. They concluded that P. chrysogenum continues to produce penicillin at zero specific growth rate provided it is supplied with a continued 'maintenance ration' of energy and/or carbon input (Pirt & Righelato 1967). Even under these conditions, however, penicillin production eventually ceased (Pirt & Righelato 1967). Thus, the continued production of MCYST proposed by Oh et al. (2000) after growth has ceased would require considerable inputs of energy and nutrients for continued cellular synthesis of MCYST.

Although

photosynthesis could provide the requirements for carbon and energy, provided an adequate amount of light was available, the continued production of MCYST under nutrient limitation is unlikely. Thus, the net rate of MCYST production is eventually likely to fall to zero at zero


93

specific growth rate, as predicted in this study (Fig. 3.8; Equation 3.7), even under conditions of P-limitation described by Oh et al. (2000). Oh et al. (2000) also found that MCYST:dry weight ratio (µg (mg dry weight)-1) correlated negatively with µ in P-limited chemostats. This result conflicts with the findings presented here (Fig. 3.6A). However, cyanobacterial dry weight is affected differentially by N and P limitation (Ahlgren 1985), demonstrating that the physiological regulation of dry weight production is quite different under various nutrient limitations. It is possible that the observed variation between the results presented here and those of Oh et al. (2000) may result from differential dry weight changes under N and P limitation. Hence, simple comparisons of MCYST:dry weight ratio data cannot be made between cultures grown under different nutrient limitations. The near absence of MCYST cell quotas from the existing literature makes it almost impossible to compare the results presented here with those of other studies.

MCYST is commonly expressed as a proportion of dry weight (Table 1.2), although several studies have expressed MCYST as a proportion of culture volume (e.g. Orr & Jones 1998), protein (e.g. Utkilen & Gjølme 1995), chlorophyll a (e.g. Lyck et al. 1996), biomass volume (e.g. Hesse & Kohl 2001) and in units of toxicity (e.g. van der Westhuizen & Eloff 1985). However, the data presented in this chapter suggest that the use of many of these parameters as denominators of MCYST content can provide a misleading indication of the cellular physiology of MCYST production (see Fig. 3.6). The data presented in Table 3.1 indicate that the cellular constituents measured in this study can vary independently of one another as a function of specific growth rate. The data in Table 3.1 also indicate that potential bias may arise if MCYST is expressed as a proportion of total cellular protein. As measured by the Lowry method, there is more protein per cell at most specific growth rates than would be expected from analysis of total cellular amino acids (Table 3.1). Thus, one or both of these methods provides a misleading analysis of cellular protein content. Lowry et al. (1951) showed that a disadvantage of their protein determination method was that different proteins had different absorption coefficients and that different results may be obtained depending on the protein used as a standard. Furthermore, the influence of the amino acid tyrosine in determining protein content can be substantial (Lowry et al. 1951). Thus, the different results obtained for total protein content by the Lowry method and by total protein amino acid content (Table 3.1) are likely to result from deficiencies in the former.

The likely changes in protein complement that may arise under variable growth

conditions (e.g. Grossman et al. 1994), suggest that protein determined by the Lowry method is an unsatisfactory indicator of growth on which to base changes in MCYST content. Thus, the conclusions made by several workers on the effects of variable growth conditions on MCYST expressed as a proportion to protein (e.g. Utkilen & Gjølme 1992, 1995; Lyck et al. 1996; Bickel et al. 2000) must be considered in this light.


94

Analysis of the intracellular amino acid complement indicates that cells contain more than sufficient amino acids to account for the intracellular MCYST quota. Specifically, the total pool (soluble and non soluble) of the amino acids alanine, arginine, Asx, Glx and leucine (the method of analysis used does not differentiate between D and L isomers), which are constituents of MCYST-LR, have pool sizes at least ten-fold greater than the MCYST quota (Table 3.2; Fig. 3.4). In addition, the amino acid serine, which has been shown to be incorporated into the amino acid dehydroalanine (Pearce & Rinehart 1979) and is a suggested precursor of Mdha in MCYSTs (Rinehart et al. 1994) also exceeds the amount of MCYST by more than ten-fold (Table 3.2; Fig. 3.4). The amino acid phenylalanine which is required for the production of the Adda moiety of MCYST (Moore et al. 1991) is also present in relatively large quantities compared with the MCYST quota (Table 3.2; Fig. 3.4). Notably, however, the amino acid methionine is present in amounts as low as three-fold that of the MCYST quota (Table 3.2; Fig. 3.4, µ = 0.2 d-1). The important role of this amino acid as a methyl group donor in MCYST biosynthesis has been identified by Moore et al. (Moore et al. 1991). In addition, the requirement of methionine in the production of S-adenosylmethionine (SAM), proposed to be utilised in Adda biosynthesis and during MCYST construction (Tillett et al. 2000) suggests that there may be a high demand for methionine during MCYST synthesis. The intracellular free amino acid quota was too small to measure. This implies that methionine, which is required for a number of other cellular processes in addition to MCYST synthesis, must be subject to very rapid turnover, perhaps indicating that it could play an important role in controlling the production of MCYSTs. Aspects of this idea are investigated in S-limited batch cultures of M. aeruginosa MASH-01A19 in Section 4.4.

Shi et al. (1995) found that MCYST was associated with the thylakoids of M. aeruginosa PCC 7820, suggesting a close physical association between MCYSTs and the photosynthetic machinery of the cell. The constant ratio of MCYST:chlorophyll a (1:2, mol:mol, Fig. 3.6D) found in this study supports this contention and suggests that MCYST synthesis and/or function could be linked to photosynthetic processes. In addition, Lee et al. (2000) recently found that the concentration of MCYST in cultures of M. aeruginosa UTEX grown at different N:P ratios maintained a constant ratio with the chlorophyll a concentration (approximately 1:2 by weight). The absence of reports of major perturbations in the photosynthetic activity of M. aeruginosa PCC 7806 after knocking out MCYST production (Dittmann et al. 1997), however, suggests that MCYSTs are not essential in photosynthesis. Nevertheless, the mutant of strain PCC 7806 which is deficient in MCYST synthetase has slightly altered thylakoid structure and also exhibits irregularities in the structure of gas vesicles (E. Dittmann and T. Börner, personal communication). The observed relationship between chlorophyll a and MCYST by Lee et al. (2000) led them to conclude that the MCYST content of cultures could be monitored by in vivo chlorophyll a fluorescence. However, the results obtained for other cell constituents in this experiment (Table 3.1; Fig. 3.6) suggest that further analyses are required before any one cellular


95

constituent could be employed as a measure of MCYST content. Any functional relationship between MCYSTs and photosynthesis is speculative at present.

The range of cell volumes between low and high specific growth rates (Table 3.1) is consistent with the range reported for batch cultures of M. aeruginosa grown at different irradiances by Krüger & Eloff (1981).

The same authors suggest that size is a likely indicator of the

physiological state of a cell, with stressed cells larger in size. Perhaps the large cell volumes observed at low specific growth rates in the experiment reported here (Table 3.1) indicate that these cells were physiologically stressed compared with their smaller counterparts found at high specific growth rates. Since QMCYST is highest in cells at high specific growth rates (Fig. 3.4) this supports the generally held view that MCYST production is greatest when conditions are favourable for growth (Sivonen 1996; Sivonen & Jones 1999).

The observation that smaller cells contain a higher MCYST concentration than larger cells (Fig. 3.7) may have implications for toxicity towards grazing zooplankton.

Some daphnids are

sensitive to toxic strains of M. aeruginosa (Rohrlack et al. 1999). Thus, if the intake of M. aeruginosa is determined by the volume of cells daphnids consume, it is conceivable that zooplankton feeding on rapidly growing M. aeruginosa could ingest a greater number of smaller cells, thus receiving a considerably larger dose of toxin. As not all daphnids are sensitive to toxic M. aeruginosa (Matveev et al. 1994), changes in intracellular concentration of MCYST could potentially result in a qualitative change in daphnid populations. This is speculation, however, and further work is required to determine the importance of Microcystis spp. cell size and toxin ingestion rates in daphnids.

3.6

Implications of results from chemostats with respect to batch cultures

Given the observed relationship between µ and QMCYST, Equation 3.6 predicts that QMCYST can be determined for any value of µ if µ max, QMCYSTmax and QMCYSTmin are known. However, obtaining these constants in a chemostat study is time-consuming. In theory, it should be possible to determine these parameters in batch culture, which would provide a more practical approach with less complicated apparatus. Despite the shortcomings of batch cultures mentioned previously (Section 3.1), simple analysis of batch cultures throughout the entire batch culture growth cycle from exponential phase to stationary phase should enable the determination of the constants QMCYSTmax and QMCYSTmin for specified growth conditions. As QMCYSTmin represents QMCYST at nutrient-limited stationary phase in a batch culture (i.e. µ = 0), and QMCYSTmax represents QMCYST during nutrient-saturated exponential growth (i.e. µ = µ max), these parameters can be determined from a single batch culture experiment (see Fig. 3.9).


96

Fig. 3.9. Theory for the determination of the constants QMCYSTmax and QMCYSTmin from batch cultures. (A) An idealised time course of log cell concentration of M. aeruginosa growing in batch culture showing six separate phases of growth: I, lag; II, accelerating growth; III exponential growth; IV, decelerating growth; V stationary; VI, decline. (B) An idealised time course of relative growth rates (specific growth rate, µ, as a fraction of the maximum specific growth rate, µ max) of M. aeruginosa grown in batch culture. It shows that the relative growth rate changes with time depending on the phase of growth in the batch culture. The circles in panel A mark the proposed occurrence of the MCYST production parameters QMCYSTmax and QMCYSTmin during a batch culture growth cycle. QMCYSTmax occurs shortly after inoculation when cells are in exponential growth phase (II), when µ = µ max (i.e. relative growth rate = 1). QMCYSTmin occurs in batch cultures at stationary phase (V), when the relative growth rate = 0. Data from Orr & Jones (1998) also indicate that QMCYSTmin is maintained during the decline phase (VI). In the decelerating growth phase of batch culture (IV), when the relative growth rate declines from 1 to 0, values of QMCYST will vary between QMCYSTmax and QMCYSTmin. The determination of intracellular quotas of MCYST at these times should permit the relationship between µ and QMCYST to be calculated according to Equation 3.6 (see Section 3.4.2). Panel A adapted from Pirt (1975).

The suggested approach to determining the parameters QMCYSTmax and QMCYSTmin in batch culture described in Fig. 3.9 differs from that used in previously published batch culture studies investigating MCYST production. Firstly, it emphasises the requirement to determine MCYST at an intracellular level (i.e. QMCYST). Secondly, it recognises that QMCYST may vary significantly depending upon the phase of growth. The data presented in this chapter indicate that the greatest variation in QMCYST will occur during the decelerating growth phase (IV in Fig. 3.9) as QMCYST decreases from QMCYSTmax to QMCYSTmin. Several previous batch culture studies have used a more simplified approach in which MCYST content has been determined at a single time point during the culture cycle (e.g.

:DWDQDEH 2LVKL /XNDþ $HJHUWHU .

Watanabe & Oishi


97

(1985), for example, chose to determine the toxicity of M. aeruginosa batch cultures (grown at different irradiance, temperature, N and P) at an unspecified time after exponential phase had finished. In theory, this could represent any one of phases IV, V and VI as shown in Fig. 3.9. Although their data showed that particular factors did affect toxicity, this would not be unexpected given the predicted variation in MCYST content during these phases of growth. Furthermore, if growth conditions are such that QMCYSTmax and QMCYSTmin differ significantly, individual values of QMCYST obtained during decelerating growth phase could also differ. Therefore, the determination of the relationship between QMCYST and µ requires that frequent sampling of batch cultures is carried out in order to both determine accurate values for µ and to observe any changes in QMCYST. A further question arising from Fig. 3.9 is the fate of MCYST in cultures during decline phase (phase VI in Fig. 3.9). Previous studies have suggested that MCYSTs are released into the external medium upon cell lysis as cultures decline (see the review by Sivonen & Jones 1999). Interestingly, no MCYST was detected in the external medium of the chemostat cultures during the experiment described in this chapter, perhaps indicating a low death rate during continuous culture. However, it might be expected that the external MCYST concentration would rise in batch culture during decline phase. This possibility is examined in the following chapter of this thesis by analysing both the intracellular and extracellular MCYST in batch cultures.

The findings presented in this chapter quantitatively demonstrate that under N-limited growth, QMCYST in M. aeruginosa is a function of µ. As µ is controlled by the intracellular quota of the limiting nutrient (QN) under N-limited growth (see Equation 3.5 (Section 3.2) and Droop 1973) the results are consistent with the regulation of QMCYST by QN. Whether or not this is the case, application of the proposed relationship between QMCYST and µ to a variety of growth conditions in batch cultures should allow quantitative analysis of the effects of environmental factors on the intracellular MCYST quota of M. aeruginosa. The evaluation of this concept is the subject of the ensuing chapter of this thesis.

98

CHAPTER 4 The influence of growth conditions on the intracellular MCYST quota of M. aeruginosa MASH-01A19 in batch culture

Chapter 4 MCYST production in batch cultures

4.1

99

Introduction

The previous chapter presented the theory that the minimum and maximum intracellular MCYST quotas (QMCYSTmax and QMCYSTmin) of M. aeruginosa can be determined from simple analysis of batch cultures, avoiding the use of complicated chemostat apparatus and time consuming data collection. Excluding the work presented in Chapter 3 and a small number of studies in the literature (Utkilen & Gjølme 1992, 1995; Lyck et al. 1996; Oh et al. 2000; Böttcher et al. 2001), most previous studies of MCYST production in M. aeruginosa have been conducted almost exclusively in batch culture (Table 1.2). A considerable number of batch culture studies have been reported, describing the content of MCYSTs in cyanobacteria grown under a variety of environmental conditions (see Table 1.2 and Sivonen & Jones 1999 for a review). However, few of these studies relates the internal cell quota of MCYST (QMCYST) in M. aeruginosa to the conditions of growth. Orr & Jones (1998) showed that the specific rate of MCYST production (µ MCYST) in M. aeruginosa MASH-01A19 correlated positively with the specific cell division rate (µ c).

This was established by determining QMCYST in cells supplied with a range of NO3-

concentrations in batch culture. The information gathered by Orr & Jones (1998) is unique, being the first report to describe the effect of an environmental factor on the MCYST content of individual M. aeruginosa cells. Despite the seminal importance of this work, few studies since (e.g. Böttcher et al. 2001) have examined changes in intracellular MCYST content of M. aeruginosa in relation to varied environmental conditions, and none has examined such changes in batch culture.

There is some consensus in the literature that conditions favouring growth also favour MCYST production. However, the roles played by specific environmental factors in controlling MCYST production have not been clearly defined (Sivonen & Jones 1999). For example, the role of iron in MCYST production remains unclear despite numerous investigations /XNDþ $egerter 1993; Utkilen & Gjølme 1995; Lyck et al. 1996; Bickel et al. 2000). As pointed out in Section 1.6.6 and the previous chapter, researchers have presented MCYST production/content data in a variety of forms, creating difficulty in unifying the existing literature. Thus, the true effects of growth conditions on MCYST production cannot be clearly determined. For example, Utkilen & Gjølme (1995) showed that interpretations vary if MCYST is expressed as a ratio to dry weight or protein content, since the effects of environmental factors on dry weight composition (i.e. protein and carbohydrate content of biomass) can differ from their effects on MCYST production (Bickel et al. 2000). This is also highlighted by the data presented in the previous chapter. Given the pool of data suggesting that MCYST production is affected by environmental conditions (Sivonen & Jones 1999), it can be hypothesised that changes in growth conditions will affect the proposed


100

relationship between growth rate and QMCYST (Equation 3.6, Chapter 3). More specifically, variation of the constants QMCYSTmax and QMCYSTmin would be indicative of any response in MCYST production in M. aeruginosa cells to an alteration in growth conditions. In this chapter the intracellular MCYST quota of M. aeruginosa MASH-01A19 is examined in batch cultures under a variety of growth conditions to determine their effect on MCYST production parameters.

MCYST concentrations within cell material can vary significantly over the growth cycle in batch culture (e.g. Watanabe et al. 1989; Orr & Jones 1998; Song et al. 1998). However, many workers have reported variations in MCYST content between treatments, based upon single sampling points over an entire batch culture cycle (e.g. Watanabe & Oishi 1985). Like Sivonen and coworkers (Sivonen 1990; Rapala et al. 1997; Rapala & Sivonen 1998) and Orr & Jones (1998), attention is given here to the determination of MCYST throughout the phases of growth in batch cultured cells so that comparisons can be made between treatments at specific stages of growth. Since the determination of accurate specific growth rates relies upon sampling frequency (Section 3.1) cultures are also analysed at regular intervals for changes in cell concentrations and biomass. Thus, in this chapter, reasonable comparison of the data obtained for batch cultures can be made with the findings described for chemostats in Chapter 3. The data presented here reveal that QMCYST varies with growth rate in batch culture, as observed for chemostats, and that specific environmental factors affect QMCYST either indirectly through their effects upon growth rate or more directly through effects upon MCYST synthesis.


4.2

101

Growth and intracellular MCYST quota of M. aeruginosa MASH-01A19 grown under standard conditions in batch culture

4.2.1

Introduction

In order to determine the effects various factors have upon MCYST production in M. aeruginosa cells, it is imperative that information first be collected regarding the production of MCYST under defined standard conditions. This also allows for the determination of general growth characteristics of cells in batch culture and provides information for the design and sampling regimes of subsequent experiments.

The data presented here describe the growth of M.

aeruginosa MASH-01A19 under a specific set of conditions (referred to as standard growth conditions), and the changes in MCYST content that occur as cultures age. In particular, the MCYST production parameters presented in Chapter 3 (viz. QMCYSTmax, QMCYSTmin and RMCYST) and maximum MCYST yield (total intracellular + extracellular MCYST at stationary phase) are estimated as determinants of MCYST production capabilities of this strain in standard batch cultures.

The study presented in this section specifically addresses the hypothesis that the

relationship between growth rate and QMCYST, observed in chemostat culture (Chapter 3), also applies to batch cultured cells of M. aeruginosa as they progress from rapid growth under nutrient replete conditions to zero growth in stationary phase. Despite the differences between batch and chemostat culture systems, the data reveal that the theoretical MCYST production parameters (QMCYSTmin and QMCYSTmax) can be determined from specific cell division rates in batch culture.

4.2.2

Standard growth conditions

The axenic Microcystis aeruginosa strain, MASH-01A19 (Bolch & Blackburn 1996; Bolch et al. 1996) was provided by the CSIRO Marine Laboratories culture collection (CSIRO Marine Laboratories, Hobart, Tasmania, Australia). The MCYST production capabilities of this strain are described in Chapters 2 and 3. For its standard growth and maintenance cultivation, modified MLA medium (Table 2.1) was used. Standard cultures (250 ml) were maintained at 26 ± 1 °C and constant irradiance (40 ± 5 µmol photons m-2 s-1, PAR - supplied by a bank of overhead cool white fluorescent tubes) in 500 ml conical culture flasks stoppered with cotton wool. Cultures were aerated by occasional swirling. For the determination of MCYST production under standard conditions, triplicate culture flasks were inoculated to an initial cell concentration of approximately 0.1 × 106 cells ml-1.

The total MCYST content (of both cells and external

medium), cell concentrations (cells ml-1) and biomass concentration (dry weight ml-1) were determined (see Section 4.2.3) in these cultures at regular intervals until stationary phase was


102

reached (approximately 25 to 30 days). Culture flasks were rotated through their positions under the fluorescent lights after sampling to minimise effects caused by any differences in irradiance.

4.2.3

Sampling and analysis

Cell counting and dry weight analysis Immediately after inoculation and thereafter at approximately three-day intervals, cultures were sampled under sterile conditions for cell concentration (cells ml-1) and culture dry weight (mg ml-1). For the determination of cell concentration, approximately 1.0 ml of culture was collected under sterile conditions from each culture flask and treated as described in Section 3.3.2, before counting the cells using a haemocytometer. At the same sampling time, culture dry weight (mg ml-1) was determined by filtering samples from between 10 and 100 ml of culture (depending on visible culture density) through preweighed glass fibre filters (47 mm diameter Whatman GF/C) to collect cell material. The volume of culture collected was accurately recorded so that the number of cells collected for MCYST extraction could be calculated. Filtrates were collected for subsequent analysis of MCYST in the external growth medium. Filters were then dried overnight (approximately 16 h) in a vacuum desiccator at room temperature, and weighed again to determine dry weight. Dry filters were kept in the dark under vacuum at 4 °C until ready for MCYST extraction (Section 3.3.2). Using culture dry weight data (mg ml-1) and culture cell concentrations (cells ml-1), dry weight per cell (pg cell-1) values were calculated at each sampling time.

Determination of specific growth rate Specific growth rate (as specific cell division rate, µ c) was determined from cell concentrations at successive sampling times according to Equation 4.1.  (ln N 2 / ln N 1) µc =    (t 2 − t1) 

(Equation 4.1)

N1 and N2 are cell concentrations (cells ml-1) at sampling times, t1 and t2 (in days) and the term µ c therefore has units of reciprocal time (d-1). For the determination of µ max (maximum specific cell division rate), additional triplicate cultures were incubated under standard conditions and sampled frequently (approximately 6 to 12 h intervals) after inoculation to ascertain the specific cell division rate during exponential growth. The term µ c differs from the term µ (specific growth


103

rate) used in chemostats (Chapter 3) in that it is determined from cell concentrations in batch culture. MCYST extraction and analysis MCYSTs were extracted from the culture material collected onto preweighed glass fibre filters with 80% (v/v) methanol, according to the method described in Section 3.3.2.

These are

subsequently referred to as 80% (v/v) methanol extracts. MCYST concentrations in 80% (v/v) methanol extracts and in cell free culture filtrates were determined by HPLC as described previously (Section 3.3.2). As reported in Chapter 2, M. aeruginosa MASH-01A19 contains both MCYST-LR and an unknown compound with MCYST characteristics (see HPLC chromatogram, Section 2.3.4, Fig. 2.6) which has not been completely characterised. Due to the unknown identity of this compound it has not been included in the analysis of total MCYST in this strain for the experiments described in this chapter. However, amounts of the unknown compound in extracts or culture filtrates never exceeded 25% of total MCYST-LR equivalents at any time during a batch culture growth cycle. As for the previous chapter, the values presented for MCYST in this and subsequent batch culture studies refer specifically to MCYST-LR.

Intracellular MCYST quota (QMCYST) was calculated at each sampling time from the culture cell concentrations, the MCYST content of 80% (v/v) methanol extracts, and the volume of culture analysed. MCYST per unit dry weight was calculated based upon dry weight ml-1 of culture and the amount of MCYST in 80% (v/v) methanol extracts. Linear regression analysis of positive growth rates and QMCYST data from individual cultures was carried out to determine the QMCYST value when µ = 0, referred to as QMCYSTmin throughout this chapter.

The same regression

procedure was used to determine the QMCYST value at the highest experimentally observed µ c, hereafter referred to as QMCYSTmax. Since there is evidence that QMCYST remains constant when batch cultures are in decline (Fig. 4.9; also Orr & Jones 1998), negative µ c values (see Fig. 3.9, Section 3.6) were not included in linear regression analysis for the determination of QMCYSTmax and QMCYSTmin as the slope of the proposed relationship falls to zero at µ c < 0. The net rate of MCYST production (RMCYST, Chapter 3) was determined from the product of specific cell division rate (µ c) and QMCYST. Net rates of MCYST production calculated using this method were not significantly different from those calculated for exponentially growing batch cultures according to the method of Anderson et al. (1990, Equation 2) (P = 0.9).

The specific MCYST production rate (µ MCYST, Orr & Jones 1998) was also determined from the culture MCYST concentration (total intracellular + extracellular MCYST) using Equation 4.1. In this instance, N1 and N2 represent culture MCYST concentrations at sampling times t1 and t2 (in


104

days). Accordingly, µ MCYST, like µ c, has units of d-1 and should not be confused with the net MCYST production rate, RMCYST, which has units of fmol MCYST cell-1 d-1 (refer to page xii for definitions of these terms). Other analyses The concentrations of dissolved PO43-, NO2- and NO3- in the culture medium were determined by anion exchange HPLC. Filtrates of cultures collected at the time of sampling were analysed for dissolved anions directly by separation on a PRP-X100 anion exchange column (150 x 4.1 mm, I.D., Hamilton). Anions were eluted isocractically using 4.0 mM p-hydroxybenzoic acid (pH 8.9) : methanol (97.5 : 2.5) at a flow rate of 2.0 ml min-1 and detected by monitoring the eluant with an Alltech 350 Conductivity Detector.

Anions were quantified by comparison with standard

solutions of each anion. A typical anion analysis chromatogram can be found in Appendix 3. Using this method, the ions BO33- and SO42- coeluted and could not be quantified separately. The pH of cultures was measured directly in an aliquot from each culture flask, removed at the time of sampling, using a calibrated pH meter.

4.2.4

Results

Growth Growth of M. aeruginosa MASH-01A19 under standard growth conditions (26 °C, 40 µmol photons m-2 s-1, in modified MLA medium - Table 2.1) is characterised by a triphasic growth curve with typical exponential, decelerating and stationary phases of growth (Fig. 4.1). Generally, standard cultures inoculated at 0.05 - 0.2 × 106 cells ml-1 from either exponential phase or stationary phase stock cultures took approximately 25 days to reach stationary phase and showed little or no lag phase. Under these conditions maximum cell yields reached up to 4 × 107 cells ml-1, while maximum biomass reached up to 0.5 mg ml-1. The maximum observed specific cell division rate (µ max) for cultures grown under standard conditions was 1.21 d-1 (1.7 doublings d-1 or a generation time of approximately 14 h) although exponential growth was only observed for the first three days in standard cultures (approximately 5 doublings). Thereafter, for the next 20 days or so, batch cultures were in the decelerating growth phase of the batch culture growth cycle (Fig. 4.1; see also Fig. 3.9, Section 3.6) where growth rates range from µ max to µ = 0. Cultures displayed a significant decrease in dry weight per cell (pg cell-1) shortly after inoculation (Fig. 4.2). In general, dry weight per cell was highest in freshly inoculated cultures, reaching a


105

maximum of 30.5 pg cell-1 under standard growth conditions. As cultures approached stationary phase, dry weight per cell decreased to a minimum of 10.6 pg cell-1.

Culture pH increased significantly (P < 0.05, Fig. 4.3) during the exponential growth phase in batch cultures, despite the presence of 3.0 mM CHES/NaOH (pH 8.0), an organic buffer, in the growth medium. As cultures approached stationary phase the pH decreased, approaching values similar to freshly inoculated medium. In the absence of CHES, the pH increased dramatically after inoculation, resulting in culture death (data not shown).

Nutritional requirements Anion analysis of culture filtrates (Fig. 4.4) showed that the cells removed >95 % of the soluble NO3- from the external medium by approximately 12 days after inoculation. During the same period, the external PO43- concentration declined to below detectable limits, but rose again to approximately 10% of the initial concentration by the time cultures reached stationary phase. Low concentrations of NO2- ( 0.05). In batch cultures, QMCYSTmin is the theoretical minimum MCYST quota of nutrient deplete cells in stationary phase (µ c = 0, Fig. 3.9). Similarly, QMCYSTmin in nutrient limited chemostats is the value for QMCYST extrapolated to µ = 0 for nutrient limited cells (Fig. 3.4). Thus, QMCYSTmin for both culture systems is predicted for nutrient limited cells. For a specific growth-limiting condition, similar QMCYSTmin values may therefore be expected in batch and chemostat cultures. However, as there is little evidence to suggest that any particular nutrient limits growth in the standard batch cultures at stationary phase, it is not possible to conclude which factor determines QMCYSTmin under the standard conditions used in this study. QMCYSTmax, calculated from standard batch cultures (Table 4.1), is significantly greater (P < 0.05) than the value predicted from N-limited chemostats (Table 3.3). The QMCYSTmax in batch cultures under standard conditions is determined for nutrient replete cells growing at µ max (i.e. cells have access to luxury supplies of nutrients), whereas QMCYSTmax in chemostats (Chapter 3) is predicted from the data obtained with N-limited cells. It follows that, if MCYST production is reliant upon the concentration of nutrients supplied to cells, exponentially growing batch cultured cells would be expected to produce greater intracellular amounts of MCYST than nutrient limited chemostat cells growing at high growth rates. This conclusion is supported by the general view that high levels of some nutrients (e.g. N and P) support greater MCYST production (Sivonen & Jones 1999). Moreover, Orr and Jones (1998) found that QMCYST determined at early stages in batch cultures (when growth rates were high and therefore QMCYST approximated QMCYSTmax) correlated positively with the concentration of N in the culture medium. Therefore, N supply may have a significant affect on the maximum MCYST content of cells. However, as noted previously, the factor limiting growth under standard batch culture conditions, as cells enter stationary phase, is unknown.

Importantly, as cultures progress from stationary phase into death phase, QMCYST remains constant in batch cultures (Figs 4.7, 4.9), confirming that QMCYST reaches a minimum at µ c = 0 (Chapter 3; Orr & Jones 1998) and that it remains at this value even when µ c becomes negative as cultures decline. This can only be observed in batch cultures, as µ cannot be set to zero in continuous cultures (Chapter 3). A constant QMCYSTmin in stationary phase cells of M. aeruginosa MASH01A19 was also described by Orr & Jones (1998) supporting the idea that toxic strains of cyanobacteria remain toxic even when cell division stops (Sivonen & Jones 1999). Thus, the


115

slope of the proposed relationship between specific growth rate and QMCYST falls to zero when cultures are in decline phase. Presumably MCYST synthesis ceases in non-dividing cells under standard conditions, while MCYST is lost from dying cells through lysis at the same rate as the death rate. In this case extracellular MCYST could serve as a measure of the death rate in MCYST-producing cyanobacterial populations.

This is supported by the evidence that

extracellular MCYSTs occur predominantly at the end of a bloom in freshwater lakes (Park et al. 1998). Assuming that extracellular MCYST is derived solely from lysed cells it can be inferred that the biological role of MCYSTs in living M. aeruginosa cells, if any, is exclusively intracellular (e.g. in photosynthesis; see Chapter 3). In batch cultures, MCYST expressed on a dry weight basis (µg MCYST (mg dry weight)-1) does not display the same relationship with specific cell division rate as seen in chemostats under N limitation (Chapter 3).

Instead, MCYST:dry weight increased with increasing specific cell

division rate up to 0.2 - 0.3 d-1, thereafter remaining constant (Fig. 4.10). This observation is possibly due to differences in the relative rates of biomass accumulation (e.g. carbon assimilation) and MCYST production under the conditions prevailing in batch cultures compared with chemostats. In batch cultures, high cellular rates of biomass accumulation may occur shortly after inoculation, when specific cell division rates are also high. This is supported by the finding that individual cell weight is highest in batch cultures when the cells are growing rapidly (Fig. 4.2), yet the converse is true for chemostats (Table 3.1). Thus, the observed differences in MCYST:dry weight between batch cultures and chemostats could be explained by differences in dry weight accumulation and not necessarily changes in MCYST production. An analysis of the rates of photosynthesis under the two different culture conditions would be required to support this proposal.

In the previous chapter, N-limited chemostat cells exhibited a decrease in both cell dry weight (pg cell-1) and cell volume (µm3) as the growth rate increased. Orr & Jones (1998), in their study on N-limited batch cultures of the same strain, also reported that the cell dry weight increased as the specific cell division rate decreased. In contrast, cell dry weight correlated positively with µ c (i.e. decreased with time, Fig. 4.2) in the standard batch culture conditions examined in this study. Presumably then, the observed relationship between growth rate and MCYST:dry weight in chemostats (Fig. 3.6A) is due to the effects of N supply on dry weight accumulation in cells. Consequently, the finding that the intracellular concentration of MCYSTs is highest in fastgrowing cells of M. aeruginosa MASH-01A19 (Chapter 3) may be specific to N-limited cultures. Indeed, the converse may be true for standard batch cultured cells. This, however, requires further investigation, as cell volume data were not gathered during the experiments reported here.


116

In any event, if MCYST is expressed on a dry weight basis, these results highlight the importance of determining whether biomass accumulation in cultures is directed into cell division or cell mass, especially if comparisons are to be made between cultures grown under different conditions. The net MCYST production rate (RMCYST, fmol MCYST cell-1 d-1), as a function of specific cell division rate, follows the same pattern in both batch and chemostat cultures (Fig. 4.11 and Fig. 3.8). Although this is, to some extent, due to autocorrelation (RMCYST being a product of QMCYST and µ c), the results from both chemostat and standard batch cultures add support to the hypothesis that the net rate of MCYST production is linked to the rate of cell division. As with the chemostat results (Chapter 3) the curvature of the line relating RMCYST and µ c (Fig. 4.11) results from the absolute difference between QMCYSTmax and QMCYSTmin (as described by Equation 3.7 in Chapter 3). The net rates obtained from the product of µ c and QMCYST were not significantly different to those calculated using the method of Anderson et al. (1990) which makes allowances for exponentially growing cells, thereby lending support to the theoretical model expressed in Equation 3.7. Additional information about MCYST production can be gleaned from plots of RMCYST and µ c. It needs to be remembered that µ c is a measure of the cellular rate of cell production according to Equation 4.1, while RMCYST is the cellular rate of MCYST production. The curvature of the RMCYST versus µ relationship is not only indicative of the variation between QMCYSTmax and QMCYSTmin (as noted in Chapter 3), but the curvature towards the µ c axis suggests that cell division is the more predominant process (see Fig. 4.11). As a result, MCYST production under standard batch and N-limited chemostat conditions is, for the most part, occurring more slowly than cell division. The curved relationship also suggests the rate of MCYST production does not increase at the same rate as cell division, contrary to the suggestion by Orr & Jones (1998). Thus, QMCYST results from the relative rates of cell division and MCYST production.

The specific rate of MCYST production, (µ MCYST, Orr & Jones 1998) displays a linear relationship with µ c, with a slope not significantly different from 1, in batch cultured cells grown under standard conditions (Fig. 4.12). Orr & Jones (1998) also found that the specific rates of cell division and MCYST production are equal and that MCYST production ceases when µ c = 0 in Nlimited batch cultures of M. aeruginosa MASH-01A19.

As pointed out in Section 3.1, a

consequence of equal rates of cell division and MCYST production is that the intracellular content of MCYST must remain constant over the range of specific cell division rates from µ c = 0 to µ c = µ max. However, the data for QMCYST under standard batch culture conditions (Fig. 4.9) indicate that this is not the case. Orr & Jones (1998) also reported that QMCYST varied between exponential and stationary phase cells and noted that this variation should produce a curvilinear relationship


117

between µ c and µ MCYST, rather like that observed for RMCYST (Fig. 4.11). The finding here, that there is a strong linear correlation between µ c and µ MCYST (Fig. 4.12) despite significant variation in QMCYST (Fig. 4.9; Table 4.1), indicates that comparison of specific rates of cell division and MCYST production provides little information about intracellular MCYST content. This is probably because µ MCYST is determined from total MCYST concentration in cultures (total intracellular + total extracellular MCYST). Since it is likely that MCYSTs are released into the external medium upon cell death or lysis, rather than by excretion/secretion from living cells (Sivonen & Jones 1999), there is a disproportionate accumulation of MCYST in the medium compared with cells. This is evidenced by the fact that total MCYST concentration within a culture remains constant when decline phase has been reached (Fig. 4.5). Thus, µ MCYST does not share a strict relationship with intracellular MCYST production, especially as cultures approach stationary phase (where death rates are likely to increase). Nonetheless, by comparing specific rate data, Anderson et al. (1990) determined that the specific saxitoxin production rate in Alexandrium spp. occurred between 0.5 and 2 times that of the specific cell division rate. Similarly, the findings presented here (Fig. 4.12), and those of Orr & Jones (1998), indicate that the specific rate of MCYST production also occurs within a range of 0.5 to 2-fold of the specific cell division rate.

The data presented here show that the intracellular MCYST quota has specific QMCYSTmin and QMCYSTmax values. Under standard growth conditions, these values exhibit little variation between cultures (see Section 4.3) but QMCYSTmax and QMCYSTmin are significantly different from one another (Table 4.1). The data show that QMCYST varies between QMCYSTmax and QMCYSTmin depending upon the phase of growth. This finding supports the hypothesis that the effects of environmental factors on QMCYST can be studied by sampling batch cultures at regular intervals, eliminating the need for complex culture systems involving chemostats and extended time periods for data collection. In addition, the study reported in this section provides a basis for comparing the effects of changes in the environment on MCYST production at a cellular level in M. aeruginosa.


4.3

118

The effect of Pi supply on intracellular MCYST quota of M. aeruginosa MASH-01A19 in batch culture

4.3.1

Introduction

The effect of P supply on MCYST production by cyanobacteria has been studied by a number of authors (Table 1.2).

Generally, these studies have concluded that MCYST content of

cyanobacterial biomass increases with increasing P supply (Sivonen & Jones 1999), although there have been reports to the contrary (Codd & Poon 1988). There is no information, however, regarding the intracellular MCYST content of cyanobacteria supplied with different levels of P. Given the findings presented in the previous chapter, and the general consensus that MCYST content of cyanobacterial cells increases with increasing P, a study was undertaken to test the hypothesis that the intracellular MCYST quota (QMCYST) of M. aeruginosa MASH-01A19 increases as the P supply increases. This section examines the relationship between specific cell division rate and QMCYST in batch cultured cells supplied with different concentrations of Pi in the growth medium so that the effect of P on QMCYST, QMCYSTmin and QMCYSTmax could be determined. The findings presented here suggest that while sufficient Pi supply results in high yields of MCYST in batch cultures, P limitation results in an increased intracellular MCYST quota in cells at stationary phase.

4.3.2

Growth conditions, sampling and analysis

Cultures were grown under standard conditions (Section 4.2) in modified MLA medium (Table 2.1) except that Pi (as K2HPO4) was supplied at initial concentrations of 0, 0.02, 0.2 and 2 mM. The initial atomic N:P ratios were therefore 100:1, 10:1 and 1:1 in those cultures which were supplied with Pi. The concentration of 0.02 mM Pi is the same as that used in modified MLA medium (Table 2.1) and reported as standard culture conditions in Section 4.2. Cultures (250ml) were grown in 500 ml conical flasks, stoppered with cotton wool plugs, and were inoculated with exponential phase cells of M. aeruginosa MASH-01A19 to an initial cell concentration of 0.2 × 106 cells ml-1. Cultures were aerated by occasional swirling. Each treatment was conducted in triplicate. Cultures were sampled at approximately three-day intervals for the determination of culture cell concentrations (cells ml-1) and culture dry weight concentrations (mg ml-1) and both intracellular and extracellular MCYST concentrations as described in Section 4.2.

After

sampling, cultures were rotated through their positions under the fluorescent lights to minimise any effects caused by differences in irradiance. MCYST in cultures was analysed until 24 days post-inoculation.

Specific cell division rate was determined as described in Section 4.2.3


119

(Equation. 4.1) and MCYST content is expressed both as a proportion of dry weight and per cell (QMCYST). QMCYSTmin and QMCYSTmax values for all four Pi treatments were calculated from regression analysis of QMCYST and specific cell division rate data obtained from individual cultures as described in the previous section. From these values, mean QMCYSTmin and QMCYSTmax values were calculated for each Pi treatment.

Using these values for QMCYSTmax and QMCYSTmin and the experimentally

observed value for µ max in each treatment, net rate of MCYST production values (RMCYST, fmol MCYST cell-1 d-1) were calculated according to Equation 3.7 (Chapter 3). In addition, the ratio of QMCYSTmax:QMCYSTmin was determined as a measure of the variation in MCYST content of cells for each treatment. Maximum culture MCYST concentration (intracellular + extracellular MCYST in stationary phase cultures, measured at 24 days) was also determined as a measure of gross MCYST production capabilities of cultures supplied with various concentrations of Pi. Data were statistically analysed for variation within and between treatments using ANOVA and linear regression analysis (Excel 97, Microsoft) and treatment effects were analysed using general linear model univariate analysis (SPSS v 10.0.5, SPSS Inc.).

4.3.3

Results

Growth Growth of M. aeruginosa MASH-01A19 was very sensitive to the concentration of Pi in the growth medium (Figs 4.13, 4.14). In the absence of added Pi, very little growth was observed with cell concentrations increasing to only approximately 6.0 × 106 cells ml-1 (approximately 4.4 doublings) and the dry weight biomass only reaching 0.14 mg ml-1 after 24 days (Figs 4.13, 4.14). Based upon cell concentrations, best growth was observed in cultures grown at 0.02 mM Pi (1/10 normal MLA concentration; Bolch & Blackburn 1996) (Fig. 4.13) although greater dry mass was achieved at 0.2 mM Pi (Fig. 4.14). Cultures grown at a concentration of 2 mM Pi showed good growth up to nine days after inoculation, but after a short period of decline both the cell concentration and culture dry weight increased only marginally thereafter (Figs 4.13, 4.14). Cultures grown at all concentrations of Pi exhibited maximum observed specific cell division rates between 0.53 and 0.61 d-1 (Table 4.2; cf. µ max = 1.21 d-1, Section 4.2). As described for standard cultures (Section 4.2.4), cell dry weight (pg cell-1) decreased in standard cultures (0.02 mM Pi) from 28 pg cell-1 shortly after inoculation to 15 pg cell-1 at stationary phase (Fig. 4.15). Cultures grown in the absence of added Pi showed no significant variation (P > 0.05, linear regression analysis) in cell dry weight throughout the culture period, averaging 24 pg cell-1


120

(Fig. 4.15). The dry weight of cells in cultures supplied with 2 mM Pi decreased from 34 to 21 pg cell-1 until nine days post-inoculation, but thereafter increased to 66 pg cell-1 as the cultures began to decline (Figs 4.13, 4.15). As cells grown at 2 mM Pi recovered from this decline (Fig. 4.13) cell dry weight decreased again to reach 37 pg cell-1 at stationary phase (Fig. 4.15). In all treatments with added Pi, dry weight cell-1 varied significantly throughout the culture period (P < 0.05, linear regression analysis).

It was noted that, two days after inoculation, cells in cultures supplied with 2 mM Pi had adopted a clumping colonial habit and that these colonies were resting on the bottom of the culture flask. Cultures supplied with less Pi exhibited the normal homogenous appearance of this strain grown under standard conditions. Cultures with no added Pi began to yellow after two days.

MCYST content and production In all cultures, the culture MCYST concentration (total intracellular + extracellular MCYST) increased as biomass increased over time (Fig. 4.16) with the maximum culture MCYST concentration for each treatment coinciding with the maximum cell concentration. The culture MCYST concentrations were highest in cultures supplied with 0.02 mM Pi throughout the experiment (Fig. 4.16) resulting in a maximum MCYST concentration of 2.5 µM in these cultures after 24 days (Table 4.2). In contrast, cultures not supplied with Pi produced little MCYST and reached a maximum MCYST concentration of just 0.5 µM after 24 days (Fig. 4.16; Table 4.2). The maximum MCYST concentration in cultures supplied with 0.2 mM Pi was also significantly lower than that obtained for 0.02 mM Pi cultures (P < 0.05, Fig. 4.16). MCYST in the external medium increased with time in all treatments (Fig. 4.21). The lowest extracellular concentration of MCYST was found in cultures not supplied with Pi (reaching a maximum of 23% of total MCYST). Cultures grown in the presence of 2 mM Pi showed the greatest extracellular MCYST concentration with up to 95% of all MCYST occurring in the external medium at 18 days after inoculation (Fig. 4.21). A rapid increase in the concentration of extracellular MCYST coincided with a decrease in the cell concentrations in this treatment (Figs 4.13 and 4.21). Cultures supplied with 0.02 and 0.2 mM Pi showed a constant increase in the concentration of extracellular MCYST (Fig. 4.21). For cells grown at 0.2 mM Pi up to 40% of total MCYST was found in the external medium compared with 30% for those supplied with 0.02 mM (Fig. 4.21).


121

Normalised to culture dry weight (i.e. µg MCYST (mg dry weight)-1), MCYST content decreased after inoculation in all treatments (Fig. 4.17).

Cultures grown at 0.02 mM Pi maintained

MCYST:dry weight ratios higher than those in other treatments (Fig. 4.17), and also reached higher MCYST:dry weight maxima (Figs 4.17, 4.18). Plotted as a function of specific cell division rate, MCYST:dry weight displayed a positive linear relationship with µ c, although this was less clear in the culture supplied with 2 mM Pi (Fig. 4.18). The intracellular MCYST quota (QMCYST) in all cultures generally decreased with time (Fig. 4.19). Plotted as a function of specific cell division rate, QMCYST increased significantly with increasing cell division rate in all cultures (P < 0.05, Fig. 4.20). There was a significant effect of Pi supply on QMCYST (P < 0.05, linear model univariate analysis) with QMCYSTmin values significantly higher in cultures not supplied with Pi compared with other treatments (P < 0.05, Table 4.2; Fig. 4.20). QMCYSTmax values were significantly lower in the 2 mM Pi treatment compared with all other treatments (P < 0.05, Table 4.2). As a result, ratios of QMCYSTmax:QMCYSTmin were significantly higher in the 0.02 and 0.2 mM Pi treatments compared with cultures grown with 2 mM Pi or without Pi (P < 0.05, Table 4.2). The conditions that supported the high QMCYSTmax:QMCYSTmin ratios (0.02 and 0.2 mM Pi) also supported best growth and lower extracellular MCYST concentrations (Figs 4.13, 4.14, 4.21).

As reported for standard batch cultures (Section 4.2), the net rate of MCYST synthesis (RMCYST) correlated positively with specific cell division rate for all Pi treatments and passed through the origin (Fig. 4.22). Maximum net rates of MCYST production ranged from 0.070 ± 0.008 (SE) to 0.088 ±0.001 (SE) fmol cell-1 d-1 but were not significantly different between the four Pi treatments (P > 0.05; Table 4.2).

122

Fig. 4.13. Time course of the effect of Pi


concentration on cell concentrations (cells

concentration on culture dry weight (mg ml-1)

ml-1) in batch cultures of M. aeruginosa

in batch cultures of M. aeruginosa MASH-

MASH-01A19.

01A19.

Data presented are for

Data presented are for cultures

cultures without added Pi (

P0 3i (¹), 0.2 mM Pi (| DQG P0 3i ().

without added Pi (



concentration on cell dry weight (pg cell-1) in

concentration

batch cultures of M. aeruginosa MASH-

concentration (intracellular + extracellular

01A19.

MCYST, µM).


without added Pi (

mM Pi (|

P0 3i (¹), 0.2

DQG mM Pi ().

Error bars

indicate ±SEM (n = 3).

Fig. 4.17. Time course of the effect of Pi concentration on MCYST content expressed as a proportion of culture dry weight (µg mg-1) in batch cultures of M. aeruginosa MASH-01A19.



P0 3i (¹), 0.2 mM Pi (| DQG P0 3i (). Error bars indicate ±SEM (n = 3).

P0 3i (¹), 0.2 mM Pi (| DQG P0 3i ().

on

culture

MCYST



P0 3i (¹), 0.2 mM Pi (| DQG P0 3i ().

123

124

Fig. 4.18. MCYST:dry weight (µg MCYST

Fig. 4.20.

(mg dry weight)-1) as a function of specific

(QMCYST, fmol cell-1) as a function of specific

cell division rate (µ c) in cultures supplied

cell division rate (µ c, d-1) in batch cultures of

with different concentrations of Pi.

MASH-01A19

Data


supplied

with

different

shown are single values determined from

concentrations of Pi.

individual batch cultures.

represents QMCYST and µ c from individual cultures.

Each data point

The line is drawn between

QMCYSTmin (µ c = 0) and QMCYSTmax (µ c = µ max) using the values in Table 4.2 for each treatment. The coefficients of determination (r2) for each treatment are shown.



concentration on the intracellular MCYST

concentration on the extracellular MCYST

-1

quota (QMCYST, fmol cell ). Data presented are for cultures without added Pi (

mM Pi (¹), 0.2 mM Pi (| DQG 2.0 mM Pi (). Error bars indicate ±SEM (n = 3).

concentration (µM) in the growth medium. Data presented are mean values from triplicate cultures without added Pi ( 0.02 mM Pi (¹), 0.2 mM Pi (| DQG P0 3i ().

125

126

Fig. 4.22

Net MCYST production rates

(RMCYST, fmol cell-1 d-1) in cultures supplied with different concentrations of Pi. The solid lines indicate RMCYST values determined using Equation 3.7 (Chapter 3) from the QMCYSTmax,

QMCYSTmin

and

µ max

values

determined for each treatment (Table 4.2).

127

TABLE 4.2 MCYST production parameters for M. aeruginosa MASH-01A19 grown with different concentrations of Pi in batch culture. Concentration

Max. culture MCYST

µ max

QMCYSTmin1

QMCYSTmax1,2

RMCYSTmax1,2

(d-1)

(fmol cell-1)

(fmol cell-1)

(fmol cell-1 d-1)

0

0.53

0.092 ± 0.009

0.145 ± 0.008

0.077 ± 0.004

1.6

0.54 ± 0.03

0.02

0.53

0.035 ± 0.008

0.166 ± 0.002

0.088 ± 0.001

4.7

2.46 ± 0.07

0.2

0.61

0.033 ± 0.004

0.142 ± 0.018

0.087 ± 0.011

4.3

1.70 ± 0.11

2.0

0.61

0.048 ± 0.004

0.114 ± 0.013

0.070 ± 0.008

2.4

2.05 ± 0.21

of Pi supplied (mM)

QMCYSTmax/QMCYSTmin

concn (µM)

1

Values given are mean regression estimates (QMCYST v µ c) from triplicate cultures ± SE.

2

QMCYSTmax and RMCYSTmax are determined at the observed µ max.

3

Maximum MCYST concentration is determined from the sum of intracellular and extracellular concentrations of MCYST in cultures at stationary phase (24 days). Data shown are the means of triplicate cultures ± SE.

128


4.3.4

129

Discussion

Cultures grown with 0.02 mM Pi, the concentration used in standard cultures, displayed similar growth and MCYST production to those described in the study reported in Section 4.2, suggesting that the MCYST production parameters are consistent and reproducible indices. These data confirm that QMCYSTmin and QMCYSTmax values are constant for given conditions of growth. This is despite differences in inoculum size between experiments (0.2 × 106 cells ml-1 in this study, compared with 0.1 × 106 in section 4.2) and differences in observed maximum specific cell division rates.

Poor growth of M. aeruginosa in cultures without exogenous Pi is indicative of P limitation. However, M. aeruginosa has a great capacity to store P as polyphosphate and can utilise this under P stress to continue cell division for four to five cell doublings (Kromkamp et al. 1989). In this study, cultures without added Pi attained four doublings in cell concentration before stationary phase was reached. In the absence of added Pi, polyphosphate is broken down so that cell division can continue despite the absence of extracellular P (Grossman et al. 1994). Cells grown under these conditions are likely to display signs of P stress, which include the production of a suite of extracellular phosphatases (Bhaya 1996), and decreases in the levels of chlorophyll a, nucleic acids and in the rate of photosynthesis (Healey 1982). Therefore, P limitation is likely to induce cellular physiological changes including changes in protein complement and soluble metabolites. Although very little growth occurred in cultures grown without added Pi, presumably the presence of polyphosphate in these cells provided for a range of measurable growth rates before cultures reached stationary phase. Nonetheless, QMCYST correlated positively with µ c in cultures not supplied with exogenous Pi. Thus, MCYST content of cells is dependent upon growth rate in both P-limited (Fig. 4.20) and N-limited cells (Chapter 3; Orr & Jones 1998). However, QMCYSTmin was greater in P-limited cells compared with other Pi treatments (Table 4.2). Sivonen (1999) reported that a reduction in cell number and an increase in the concentration of MCYST in the external medium, as occurred midway through the incubation period in the cultures treated with 2 mM Pi (Fig. 4.21), are indicative of cell lysis and the release of MCYST. At 2 mM Pi, up to 95% of the MCYST occurred in the extracellular medium as the cultures underwent a period of decline in cell numbers and dry biomass concentrations (Figs 4.13, 4.14, 4.21). Since the Pi was supplied as K2HPO4 and K+ can inhibit growth of M. aeruginosa (Zehnder & Gorham 1960; McLachlan & Gorham 1961; Parker et al. 1997), this suggests that the deleterious effects of 2 mM K2HPO4 (cell lysis and release of MCYST into the medium) could be attributable to high concentrations of K+ (4.0 mM). Thus, the proposed use of K+ in controlling


130

the growth of Microcystis spp. (Parker et al. 1997), although likely to minimise cell growth, may also promote the loss of MCYST into the surrounding medium. Thus, application of K+ to water supplies for the control of nuisance Microcystis spp. blooms must therefore be carried out when cell concentrations are low, in order to avoid the release of high concentrations of MCYSTs into the water.

Low concentrations of Pi (0.02 mM) supported the maximum culture MCYST concentration (Table 4.2). The final MCYST concentration in cultures supplied with 0.02 mM Pi (standard cultures, Section 4.2) was approximately five-fold greater than cells grown without added Pi (Table 4.2). This finding is consistent with the general view that MCYST production increases as the P supply increases (Sivonen & Jones 1999) and is also consistent with the suggestion that MCYST production is affected indirectly by environmental factors through their effects on cell division (Orr & Jones 1998). That is, conditions providing good growth also support high yields of MCYST. The lower maximum culture MCYST concentrations in cultures supplied with 0.2 and 2 mM Pi cultures were associated with decreased growth, perhaps due to growth inhibition by K +.

It is significant that QMCYSTmin is higher in cultures grown without added Pi. In the previous chapter it was suggested that changes in environmental growth conditions could affect the MCYST production parameters (i.e. QMCYSTmax and QMCYSTmin). The data on the effect of Pi nutrition on QMCYSTmin (Table 4.2) provide evidence that this parameter can change in response to altered environmental conditions, specifically in response to P stress. QMCYSTmin in cultures not supplied with Pi is significantly higher than QMCYSTmin in standard cultures, indicating that Plimited stationary phase cells of M. aeruginosa are potentially very toxic. This finding contradicts the generally held view that P-deficient cells are less toxic than those supplied with P (Sivonen & Jones 1999), but is consistent with the opposing finding by Codd & Poon (1988), that low P supply results in slightly higher toxicity. On a dry weight basis, however, MCYST content is greater in cultures supplied with 0.02 mM Pi compared with cultures without added Pi at high growth rates (Figs 4.16, 4.18). This finding is consistent with the general consensus within the literature, but highlights the potential problems associated with determining MCYST content on a dry weight basis at a fixed point in the growth cycle. Thus, these data provide a possible explanation for the higher cellular toxicity at low Pi presented by Codd & Poon (1988). However, studies by most workers on the effects of P supply on MCYST production (e.g. Codd & Poon 1988; Sivonen 1990; Rapala et al. 1997; Rapala & Sivonen 1998; Bickel et al. 2000; Oh et al. 2000) do not provide data on cell number and cell dry weight (Table 1.2), so that an evaluation of their data in relation to the findings reported in Table 4.2 is not possible. Nonetheless, the data in


131

Table 4.2 show, for the first time, a clear effect of P supply on the MCYST content of M. aeruginosa cells in batch cultures. As a proportion of dry weight, MCYST showed a positive correlation with specific cell division rate in all cultures (Fig. 4.18). This is inconsistent with the findings of Oh et al. (2000) who reported that the MCYST:dry weight content of P-limited chemostat grown cells of M. aeruginosa UTEX 2388 correlated negatively with growth rate. Oh et al. (2000) also showed, however, that the carbon fixation rate (measured as both as g C g-1 chlorophyll a h-1 and mg C g-1 h-1) and carbon content (as a proportion of dry weight) was highest at high growth rates compared with low growth rates.

This suggests that in their study cell dry weight may have been

considerably greater at high growth rates, resulting in low MCYST:dry weight ratio at high growth rates. However, cell dry weight varied little in P-limited cultures in this study (Fig. 4.15). Differences in growth media composition and irradiance between the P-limited study here (no added Pi) and that of Oh et al. (who used 6 µM Pi and 160 µmol photons m-2 s-1) may contribute to different cell biomass (specifically carbon) accumulation between the two investigations and so provide misleading information about MCYST:dry weight content.

However, here again,

comparison of the study by Oh et al. (2000) with the data presented for intracellular MCYST quota here (Table 4.2) is not possible since these authors provide MCYST data as a proportion of dry biomass only. The high value for QMCYSTmin observed in cultures without added Pi provides important information about the control of MCYST production at a cellular level. Consideration must be given here to the concept of QMCYSTmin as the intracellular MCYST content in stationary phase, or non-dividing, cells (Chapter 3). The data in Table 4.2 indicate that P-limited cells contain more MCYST than cells limited by some other factor (e.g. N-limited cells, Chapter 3) when they reach stationary phase. This implies that cell division is more sensitive to P-stress than is MCYST production. In this event, MCYST production could continue for a short time after cell division stops with the result that P-limited stationary phase cells will have a larger cell quota of MCYST than stationary phase cells limited by some other nutrient. Despite the higher values for QMCYSTmin in P-limited cultures compared with P-sufficient cultures, maximum culture MCYST concentration was lowest in cultures grown without added Pi. This is mainly due to P-limited biomass production. Thus, whilst P-limited blooms of M. aeruginosa are not likely to produce a large yield of MCYST, the concentrated scum material of a P-limited M. aeruginosa bloom will be potentially very toxic (on an intracellular basis) once growth has ceased.

Since P limitation is common in many freshwater systems (e.g. Lange 1971) this

information could prove useful in predicting the potential hepatotoxicity of some cyanobacterial blooms.


4.4

132

The effect of SO42- supply on intracellular MCYST quota of M. aeruginosa MASH-01A19 in batch culture

4.4.1

Introduction

In cyanobacteria, sulphur (S) is required for a variety of cellular processes, including the synthesis of many important metabolites (e.g. the amino acids cysteine and methionine, glutathione, coenzyme A, thioredoxin). S deficiency in cyanobacteria results in pigment breakdown (Collier et al. 1999), polyphosphate formation (Lawry & Jensen 1979) and altered thylakoid composition (Jensen & Rachlin 1984; Arino et al. 1995) and can result in qualitative and quantitative changes to organic S compounds of cells (Funteu et al. 1997). The importance of S in cyanobacterial physiology has been well researched and a number of molecular responses of cyanobacteria to S deficiency have been elucidated (Bhaya 1996). For example, cyanobacterial cells are capable of increasing the number of SO42- transport proteins during periods of S starvation, increasing the maximum SO42- uptake rate by twenty-fold (Bhaya 1996). S-deficient cyanobacteria also develop mechanisms for scavenging S from unusual sources including sulphate esters in the external growth medium (through the induction of arylsulphatases) and organic thiols (Schmidt 1998).

The effect of S-limitation on MCYST production has not been examined despite evidence that S is required for several aspects of MCYST biosynthesis (Arment & Carmichael 1996; Nishizawa et al. 1999). Although the S needed for MCYST synthesis is not incorporated into MCYSTs, theory predicts that the amino acid activation (Rinehart et al. 1994; Arment & Carmichael 1996) and methylation processes (Moore et al. 1991) during MCYST-LR production would alone involve approximately 12 mol of S per mol of MCYST-LR. The importance of S is further emphasised by the small pool sizes of the S-containing amino acids in N-limited chemostat grown cells (Table 3.2). Thus, in theory, conditions that support MCYST production would enhance the demand for S metabolites within the cell. In this event S-limitation in M. aeruginosa will lead to a decrease in QMCYST. However, given the findings on QMCYST in N-limited cells in chemostats (Chapter 3) and for cells grown under varying Pi supply in batch culture (Section 4.3), QMCYST in S-limited cells would be expected to depend on the growth rate. These possibilities were examined by growing M. aeruginosa MASH-01A19 in S sufficient (200 µM) or S deficient (2.12 µM) MLA media in batch culture and monitoring intracellular MCYST content (QMCYST) over the entire batch growth cycle to determine the effects of S limitation on MCYST production. The results show that S limitation leads to a low intracellular MCYST quota and suggest that S plays a significant role in MCYST production.


4.4.2

133


Cultures flasks (500 ml) containing 250 ml of either adequate SO42- (200 µM MgSO4 - the concentration used in modified MLA medium for growing cells under standard conditions - see Table 2.1) or limited SO42- (2.12 µM) were inoculated with exponential phase cells of M. aeruginosa MASH-01A19. Thus, initial N:S ratios were 10:1 and 1000:1 respectively. Initial cell concentrations were approximately 0.1 × 106 cells ml-1. The Mg2+ concentration in the SO42-limited cultures was maintained by supplying 198 µM MgCl2 and 2.0 µM MgSO4. The additional 0.12 µM SO42- was supplied as trace elements (see Table 2.1).

Each SO42- treatment was

conducted in triplicate. Cultures were grown under standard conditions (Section 4.2) at an irradiance of 40 ± 5 µmol photons m-2 s-1 and flasks were moved sequentially through their positions under the lights after each sampling, to minimise variation in irradiance experienced by each culture.

The cultures were sampled at approximately three-day intervals for cell

concentration, dry weight and intracellular and extracellular MCYST according to the methods in Section 4.2. Cultures grown under standard conditions (200 µM SO42-) are referred to as control cultures in this section.

Positive specific cell division rates (µ c) were used to calculate the parameters QMCYSTmin and QMCYSTmax from individual cultures according to the method described previously (Section 4.2). These parameters are presented as the means of triplicate cultures for each treatment. Net MCYST production rates (RMCYST) were also calculated from µ c and QMCYST as described previously (Section 4.2).

As for previous experiments, data were statistically analysed for

variation within and between treatments using ANOVA and linear regression analysis (Excel 97, Microsoft), and treatment effects were analysed using general linear model univariate analysis (SPSS v 10.0.5, SPSS Inc.).

4.4.3

Results

Growth Cultures supplied with low S (2.12 µM) exhibited poor cell growth compared with the control culture containing adequate S (Fig. 4.23). At low S, the maximum cell concentration after ten days was only 6.4 × 106 cells ml-1 whereas the corresponding value for control cultures was 2.8 × 107 cells ml-1 (Fig. 4.23). Maximum dry biomass was also much greater (2.8-fold) in control cultures compared to the low S treatment (Fig. 4.24). Poor growth in the low S treatment was exemplified by yellowing of cultures, coinciding with the attainment of maximum cell concentrations. Cell dry weight in the low S treatment declined slightly from 31 pg cell -1 in the first few days after inoculation but thereafter remained constant, averaging 21 pg cell-1 (Fig. 4.25).


134

Cell dry weight in control cultures, however, decreased to approximately 11 pg cell-1 during the course of the experiment (Fig. 4.25). Maximum observed specific cell division rates were not significantly different (P > 0.05) between treatments.

MCYST content and production The culture MCYST concentration (total intracellular + extracellular MCYST) increased in both control (adequate S) and S-limited cultures over time (Fig. 4.26). When the experiment was terminated after 28 days, the maximum MCYST concentration in control cultures was almost seven times greater than the maximum MCYST concentration attained in S-limited cultures (Fig. 4.26; Table 4.3). However, maximum MCYST concentration in S-limited cultures was achieved after only ten days in culture (Fig. 4.26). The concentration of extracellular MCYST in the growth medium increased steadily with time in both control and S-limited cultures (Fig. 4.27). In both instances, the extracellular MCYST concentration continued to increase after maximum total MCYST concentration (Fig. 4.26) and maximum biomass (Figs 4.23, 4.24) were attained. However, despite differences in growth and MCYST production in control and S-limited cultures, approximately 35% of total MCYST was found to be in the external medium after 28 days in both treatments (Figs 4.26, 4.27). These data indicate that cell lysis was contributing to extracellular MCYST after growth had ceased. As a proportion of dry weight, MCYST decreased as a function of time in both S-sufficient and Slimited cultures (Fig. 4.28). However, by six days after inoculation, MCYST:dry weight in Slimited cultures was significantly lower than that in control cultures, and remained so for the duration of the experiment (Fig. 4.28). Plotted as a function of specific cell division rate, MCYST:dry weight was also significantly lower in S-limited cultures at low growth rates (approximately two-fold) compared with the control cultures, although at high growth rates the values between treatments were more similar (Fig. 4.29). The level of S supply had a significant effect on QMCYST (P < 0.05, general linear model univariate analysis). Nonetheless, QMCYST displayed a strong linear correlation with µ c in both S-sufficient and S-limited cultures (Fig. 4.31). Both the QMCYSTmin and QMCYSTmax values (determined from linear regression analysis of the data - using positive values of µ c) were significantly lower in the low S treatment compared with the S-sufficient control (P < 0.05, ANOVA, Table 4.3; Figs 4.30, 4.31). Despite this, there was no significant difference between QMCYSTmax:QMCYSTmin ratios (P > 0.05) from each treatment (Fig. 4.31; Table 4.3).

Maximum net MCYST production rates

(RMCYSTmax; calculated from QMCYSTmax and µ max) were significantly higher in the control cultures (0.126 ± 0.003 (SE) fmol cell-1 d-1) compared with the low S treatment (0.083 ± 0.010 (SE) fmol cell-1 d-1) (P < 0.05; Table 4.3), indicating a decreased net rate of MCYST production under S limitation.

135

Fig. 4.23. Time course of cell concentrations (cells

ml-1)

in

batch

cultures

of

M.

aeruginosa MASH-01A19 grown with either 200 µM SO42- (| RU µM SO42- (

Fig. 4.24.

Time course of dry weight

accumulation (mg ml-1) in batch cultures of M. aeruginosa MASH-01A19 grown with either 200 µM SO42- (|

RU

µM SO42-

(

Fig. 4.25. Time course of dry weight per cell (pg cell-1) in batch cultures of M. aeruginosa MASH-01A19 grown with either 200 µM SO42- (|

RU

µM SO42- (

'DWD

presented are mean cell dry weights from triplicate cultures ± SE (n = 3).

136

137

Fig. 4.26.

Time course of total culture

Fig. 4.27.

Time course of extracellular

MCYST concentration (intracellular and

MCYST

extracellular MCYST, µM) in batch cultures

cultures of M. aeruginosa MASH-01A19

of M. aeruginosa MASH-01A19 grown with

grown with either 200 µM SO42- (|

either 200 µM SO42- (|

µM SO42- (

(

RU

µM SO42-

'DWD SUHVHQWHG DUH WKH PHDQ FXOWXUH

MCYST

concentrations

from

triplicate

culture

concentration

(µM)

in

batch

RU

'DWD SUHVHQWHG DUH WKH PHDQ

MCYST

concentrations

from

triplicate cultures ± SE (n = 3).

cultures ± SE (n = 3).

Fig. 4.28. Time course of MCYST expressed

Fig. 4.29. MCYST expressed as a proportion

as a proportion of culture dry weight (µg

of culture dry weight (µg MCYST (mg dry

-1

MCYST (mg dry weight) ) in batch cultures

weight)-1) and plotted as a function of

of M. aeruginosa MASH-01A19 grown with

specific cell division rate (µ c, d-1)in batch

either 200 µM SO42- (|


(

'DWD SUHVHQWHG

MCYST

RU

µM SO42-

are the mean culture

concentrations

from

triplicate

grown with either 200 µM SO42- (| µM

2-

SO4

(

'DWD

RU

SUHVHQWHG

DUH

cultures ± SE (n = 3).

measurements from individual batch cultures.

Fig. 4.30.

Fig. 4.31.

Time course of intracellular


MCYST quota (QMCYST, fmol cell-1) in batch

(QMCYST, fmol cell-1) plotted as a function of


specific cell division rate (µ c, d-1) in batch

grown with either 200 µM SO42- (|


µM SO42- ( culture

RU

2.12

'DWD SUHVHQWHG DUH WKH PHDQ

MCYST

concentrations

triplicate cultures ± SE (n = 3).

from

grown with either 200 µM SO42- (| µM

SO42-

(

'DWD

RU

SUHVHQWHG

DUH

measurements from individual batch cultures. The lines are drawn between QMCYSTmax and QMCYSTmin values listed in Table 4.3.

The

data show strong correlations with these fitted lines (| r2 = 0.87; r2 = 0.74).

138

TABLE 4.3 MCYST production parameters for M. aeruginosa MASH-01A19 supplied with different concentrations of SO42- in batch cultures. Concentration of

Max. culture MCYST

µ max

QMCYSTmin1

QMCYSTmax1,2

RMCYSTmax1,2

(d-1)

(fmol cell-1)

(fmol cell-1)

(fmol cell-1 d-1)

2.12

0.77

0.031 ± 0.002

0.107 ± 0.013

0.083 ± 0.010

3.5

0.32 ± 0.03

200

0.85

0.050 ± 0.002

0.148 ± 0.004

0.126 ± 0.003

3.0

2.17 ± 0.06

SO42- supplied (µM)

QMCYSTmax/QMCYSTmin

concn (µM)

1


2


3

Maximum MCYST concentration is determined from the sum of intracellular and extracellular concentrations of MCYST in cultures at stationary phase (28 days). The value shown is the mean ± SE of triplicate cultures.

139


4.4.4

140

Discussion

The concentration of SO42- used in the S deficient cultures (2.12 µM) is similar to that which causes S limitation in other cyanobacteria (approximately 1.0 µM, Schmidt 1998). As a result, symptoms of S deprivation in cells supplied with only 2.12 µM SO42-, such as the degradation of C-phycocyanin (and possibly phycoerythrin, Schmidt 1998) and an overproduction of SO42transport proteins (Bhaya 1996), can reasonably be expected. Since S deficiency is typically associated with decreased pools of organic S metabolites within cells (Giovanelli et al. 1980) then decreased rates of cellular processes dependent on these pools, such as protein synthesis, can also be expected. It is evident from the data presented here that limited S supply leads to a low biomass yield (Figs 4.23, 4.24) and a low final yield of MCYST (Fig. 4.26; Table 4.3) in M. aeruginosa MASH-01A19. The data show that the intracellular MCYST quota and MCYST:dry weight ratio are also lower in S-depleted cells compared with S-replete cells (Figs 4.28 to 4.31).

Arment & Carmichael (1996) provided evidence that MCYST production occurs via a thiotemplate mechanism, requiring at least one mol of S per amino acid incorporated into each mol of MCYST (i.e. seven mol S per mol MCYST). In addition, Moore et al. (1991) showed that five mol of methionine were required for the methylation processes involved in the synthesis of MCYST-LR. Under conditions of N limitation in chemostats, cellular pool sizes of methionine were as low as three times that of the intracellular MCYST quota (Table 3.2) even though S supply was adequate. This, along with the likelihood that a multi-enzyme complex is required for MCYST biosynthesis (Nishizawa et al. 1999; Moffitt & Neilan 2000), suggests that M. aeruginosa has a high demand for cellular S for the processes leading to MCYST production. The lower QMCYSTmin value in S-limited cultures compared with cultures supplied with adequate S (Fig. 4.31; Table 4.3) supports the idea that S availability is important in the production of MCYSTs and suggests that MCYST production is compromised under S limitation.

The data presented here support the finding that QMCYST is dependent upon the growth rate (Chapter 3) as there is a strong linear correlation between QMCYST and µ c (Fig. 4.31), despite the evidence of decreased cellular MCYST production under S stress. Thus, the MCYST content of M. aeruginosa can be predicted from the specific cell division rate in S-limited cultures as well as those limited by N (Chapter 3) and P (Section 4.3). In addition, the results presented here show that S-limited cells contain a low intracellular MCYST quota (and low MCYST:dry weight ratio), particularly at low growth rates (Figs 4.29, 4.31; Table 4.3). This supports the hypothesis presented in Section 4.4.1 that S limitation leads to lower QMCYST in this strain, consistent with the proposal that MCYST production slows relative to cell division (or cell division rate increases


141

relative to MCYST production) when S becomes limiting. As argued above, this most likely stems from a decrease in the levels of S metabolites within the cells that are required for MCYST synthesis (e.g. methionine for methylation). As a result, the rate of MCYST synthesis is slowed under conditions of S-limitation, yet cell division is able to continue (although the number of cell divisions is limited by S availability). These events contrast with those that lead to a higher QMCYSTmin in P-limited cultures (Section 4.3). MCYST is low under S-limited conditions on an intracellular quota basis (Figs 4.30, 4.31). The effect of S-deprivation upon MCYST content is even more apparent when MCYST is expressed as a proportion of dry weight (Figs 4.28, 4.29). Here, consideration must be given to the stable cell dry weight in S-limited cultures (approximately 21 pg cell -1; Fig. 4.25). This is similar to the static cell dry weight observed for P-limited cultures in Section 4.3 (Fig. 4.15) and raises the question whether nutrient limitation limits dry weight accumulation (e.g. C-assimilation) by M. aeruginosa. However, even on a dry weight basis, MCYST content is lower under S limitation (Fig. 4.29) than P limitation (Fig. 4.18) in stationary phase cultures. These data show the importance of determining the intracellular MCYST quota as opposed to measuring MCYST on a dry weight basis only. Although a similar qualitative pattern has emerged from this study regarding QMCYST and MCYST:dry weight, the relative quantitative differences in outcomes of MCYST content expressed in both forms is indicative of the potential misinterpretation of MCYST:dry weight data. The MCYST:dry weight ratio grossly overestimates the effect of S limitation on intracellular MCYST content in this case.

The lower QMCYST in S-deprived cells implies that S plays an important role in MCYST synthesis. Requirement of S (as methionine) to support N-methylation and thiolation and to provide Scontaining co-factors (e.g. 4'-phosphopantetheine) for MCYST synthesis (Moffitt & Neilan 2000), together with the demand for S for amino acids for the production of MCYST synthetase itself (Nishizawa et al. 1999), are all potential sources which could be diminished by S limitation. However, the low cellular pool size of methionine found under N limitation in chemostats (Chapter 3) could suggest that this amino acid is specifically limiting MCYST production. With respect to occurrences of toxic blooms, the implications of this study are not clear, however, as S limitation is not reported often in freshwater systems (typical freshwater concentrations of S range between 100 to 200 µM, Schmidt 1998).

Nonetheless, this information provides further

understanding of the physiology of MCYST production and the role S may play in its synthesis. Repetition of this work with analysis of cellular S metabolites (especially methionine) is required to provide further evidence to support the claims made here.


4.5

142

The effect of Fe3+ supply on intracellular MCYST content of M. aeruginosa MASH-01A19 in batch culture

4.5.1

Introduction

The effect of Fe supply on MCYST production in M. aeruginosa has been investigated by a number of authors (Table 1.2). Some of the results are conflicting while others do not allow comparison becaXVH RI WKH ZD\ WKH GDWD ZHUH H[SUHVVHG 3+

showed that low concentrations of Fe

/XNDþ $HJHUWHU (1993), for example,

resulted in higher relative toxin yields (MCYST:dry

weight) in exponentially growing M. aeruginosa PCC 7806, whereas Utkilen & Gjølme (1995) found that low Fe3+ supply decreased the MCYST:dry weight ratio in M. aeruginosa CYA 228/1. Lyck et al. (1996) and Bickel et al. (2000) studied the effect of Fe3+ supply on MCYST content, expressed as a ratio to a number of cellular components (e.g. protein and dry weight) finding that Fe3+ starvation increased MCYST expressed as a proportion to these biomass indicators. Thus, there is no consensus on whether intracellular MCYST production and content of M. aeruginosa are influenced by Fe3+. Moreover, the effect of Fe supply on intracellular MCYST quotas has not been addressed. Since Fe3+ is sometimes limiting in natural freshwater ecosystems (Lange 1971), it has a potential role in influencing the growth and toxicity of blooms of MCYST-producing species of cyanobacteria. Unlike previous studies, the experiment presented here examines the effect of Fe3+ supply on intracellular MCYST content and production in batch cultured M. aeruginosa MASH-01A19. As for the studies presented previously in this chapter, the changes in QMCYST were examined in relation to specific cell division rate over a range of growth phases. Total MCYST concentration in cultures (intracellular + extracellular) was also measured as an indicator of gross MCYST production under Fe3+-limited conditions. The results show that there is a positive correlation between the concentration of Fe3+ supplied to cells and the intracellular MCYST quota of stationary phase cells. Also, final culture MCYST concentration was dependent upon the initial Fe3+ concentration. In addition, MCYST production appears to be retarded during a physiological lag period when cells from stationary phase cultures are transferred into fresh growth medium, though this was not accompanied by a lag in cell division nor an increase in culture dry weight.

4.5.2


Cultures (250 ml) of M. aeruginosa MASH-01A19, containing Fe3+ at concentrations of 0.03, 0.1, 0.3 and 1 µM, were established in modified MLA medium (which normally contains 6 µM Fe3+ Table 2.1). Thus, the atomic ratio of N to Fe3+ in the different growth media varied from 66,667:1 (0.03 µM Fe3+) to 2,000:1 (1 µM Fe3+). Triplicate culture flasks (500 ml) were inoculated to an


143

initial cell concentration of 0.2 × 106 cells ml-1 from a single stock culture that had reached stationary phase.

All Fe3+ treatments were grown under standard conditions of light and

temperature (Section 4.2). The cultures were sampled at approximately three-day intervals for culture dry weight, cell concentration and both intracellular and extracellular MCYST, until 27 days after inoculation, as described previously (Section 4.2). After each sampling, the flasks were rotated through their positions under the light source to minimise any effects from variation in irradiance.

The approach described in Chapter 3 could not be used to determine QMCYSTmin and QMCYSTmax values in the experiments described in this section because QMCYST values at specific cell division rates (µ c) greater than 0.3 to 0.4 d-1 did not increase as µ c increased, as found when N, P or S was the limiting nutrient. Accordingly, values of QMCYST when µ c > 0.3 - 0.4 d-1 were treated as outliers (Cook's D statistic > 1.0). Thus, only QMCYST values at µ c < 0.3 - 0.4 d-1 were used to calculate QMCYSTmin. The cited QMCYSTmin values are the mean of individual regression analyses from three replicate cultures at each Fe3+ treatment. As described in the ensuing sections (4.5.3 and 4.5.4), high cell division rates in Fe-limited cultures were only observed during the first three to six days after establishing cultures. The cells used to inoculate these cultures were stationary phase cells so that immediately after inoculation, when µ c was greatest, the cells were recovering from a previous nutrient limitation. This raises the question whether the cells had the potential to attain much greater QMCYST values than those seen by experiment. Accordingly, theoretical QMCYSTmax values were calculated by extrapolating the regression line established, using µ c values < 0.3 - 0.4 d-1, to the highest experimentally recorded values for µ c. Net MCYST production rates (RMCYST) were calculated from µ c and QMCYST as described previously (Section 4.2). Data were statistically analysed for variation within and between treatments using ANOVA and linear regression analysis (Excel 97, Microsoft) and treatment effects were analysed using general linear model univariate analysis (SPSS v 10.0.5, SPSS Inc.).

4.5.3

Results

Growth The growth of cultures was not strongly influenced by the concentration of Fe3+ (0.03 to 1 µM) with most cultures reaching maximum cell concentrations of 2.5 - 3.0 × 107 cells ml-1 and maximum dry weight concentrations of 0.33 - 0.36 mg ml-1 (Figs 4.32, 4.33). In addition, maximum observed specific cell division rates were between 0.70 and 0.74 d-1 in all cultures (Table 4.4). However, biomass yield in all Fe3+ treatments was slightly lower than that achieved for standard cultures (6 µM Fe3+; Section 4.2). Despite inoculation from a stationary phase stock


144

culture, there was no apparent lag in growth as judged by cell concentrations and culture dry weight (Figs 4.32, 4.33). The dry weight of cells at inoculation (approximately 15 pg cell-1, Fig. 4.34) was similar to that obtained for stationary phase cells in previous studies (Sections 4.2 to 4.4). Shortly after inoculation, cell dry weight increased to approximately 20 pg cell-1 in all treatments, but thereafter decreased to approximately 13 to 15 pg cell-1 in the 0.03, 0.1 and 1 µM Fe3+ treatments by the time the cultures reached stationary phase (Fig. 4.34). The cell dry weight of cultures supplied with 0.3 µM Fe3+ increased after nine days, eventually reaching approximately 20 pg cell-1 after 27 days in culture (Fig. 4.34). This increase in cell dry weight coincided with a brief period of cell death in the 0.3 µM Fe3+ cultures (Figs 4.32, 4.33).

MCYST content and production Fe3+ concentration had a significant effect on the final culture MCYST concentration (total intracellular + extracellular MCYST as measured at 27 days after inoculation) with greater MCYST concentrations in cultures supplied with 0.3 and 1 µM Fe3+ compared with those supplied with 0.03 and 0.1 µM Fe3+ (P < 0.05; Table 4.4; Fig. 4.35). The concentration of MCYST in the external growth medium increased as a function of time in all cultures, reaching approximately 22% of total MCYST (i.e. intracellular + extracellular) in all treatments except for the 0.3 µM Fe3+ treatment (Fig. 4.36). In cultures supplied with 0.3 µM Fe3+, extracellular MCYST increased considerably after 12 days, reaching a maximum of approximately 62% of total MCYST (Fig. 4.36). This increase in extracellular MCYST coincided with a brief period of cell death in these cultures (Figs 4.32, 4.33) and a gradual increase in cell dry weight (Fig. 4.34).

Expressed as a proportion of dry weight, MCYST increased shortly after inoculation in all treatments, reaching maximum values at approximately six days (Fig. 4.37). The initial low MCYST:dry weight values at inoculation differ from those observed in other batch cultures which were inoculated from exponential phase stock cultures but is similar to that observed in stationary phase cultures (see Sections 4.2 to 4.4). After six days, MCYST:dry weight decreased in all treatments, reaching similar minimum values at the conclusion of the experiment, when cells were at stationary phase, as observed in freshly inoculated cultures (Fig. 4.37). MCYST:dry weight generally increased with specific cell division rate (µ c) in all treatments up to about 0.4 d-1 (Fig. 4.38). However, at higher specific cell division rates MCYST:dry weight decreased (Fig. 4.38). Minimum MCYST:dry weight values (observed in stationary phase cultures) were similar in cultures supplied with 0.03 to 0.3 µM Fe3+ (3.2 – 3.8 µg MCYST (mg dry weight)-1) but significantly higher in cultures supplied with Fe3+ at 1 µM (4.8 µg MCYST (mg dry weight)-1; Figs 4.37, 4.38; P < 0.05).


145

The intracellular MCYST quota (QMCYST) was initially low at the time of inoculation (Fig. 4.39), similar to that observed in stationary phase cells from previous studies (Sections 4.2 to 4.4). However, QMCYST increased in all treatments shortly after inoculation, reaching a maximum value at approximately six days (Fig. 4.39). QMCYST then decreased in all cultures to minimum values at stationary phase at the conclusion of the experiment, similar to the QMCYST values of freshly inoculated cultures (Fig. 4.39). Unlike previous experiments, QMCYST increased as a function of specific cell division rate up to approximately 0.3 d-1, thereafter remaining constant or decreasing slightly (Fig. 4.40). Thus, for the purpose of linear regression analysis (in the determination of QMCYSTmax and QMCYSTmin) the QMCYST values determined at maximum observed specific cell division rates were treated as outliers (Cook's D statistic > 1.0). The level of Fe3+ nutrition had a significant effect on QMCYSTmin (P < 0.05) with this parameter correlating positively with the Fe3+ concentration supplied in the growth medium (r = 0.99; Table 4.4). In all cultures the maximum observed intracellular MCYST quotas were obtained at intermediate specific cell division rates (0.3 - 0.4 d-1; Fig. 4.40). As a result, at high cell division rates the QMCYSTmax values calculated from linear regression analysis, by the procedure adopted in earlier studies (Chapter 3; Sections 4.2 to 4.4) are significantly greater than the observed values (Table 4.4; P < 0.05). Accordingly, the maximum observed net MCYST production rates (RMCYST) at high division rates were also relatively low (Fig. 4.41). However, for lower specific cell division rates, RMCYST correlated in a curvilinear fashion with µ c (Fig. 4.41), as described for previous experiments (Sections 4.2 to 4.4).

146

Fig. 4.32. Time course of the effect of Fe3+


concentration on cell concentrations (cells

concentration on culture dry weight (mg

-1

ml ).


supplied with 0.03 µM Fe ( µM Fe 3+

3+

ml-1).


supplied with 0.03 µM Fe3+ ( µM Fe3+

(¹), 0.3 µM Fe3+ (| DQG µM Fe3+ ().

(¹), 0.3 µM Fe3+ (| DQG 1.0 µM Fe3+ ().



concentration on cell dry weight (pg cell-1).

concentration

Data presented are for cultures supplied with


0.03 µM Fe Fe3+ (|

3+

(

DQG

µM Fe

3+

(¹), 0.3 µM

µM Fe3+ (). Error bars

indicate ±SEM (n = 3).

MCYST, µM).

on

cultures supplied with 0.03 µM Fe3+ ( µM Fe3+ (¹), 0.3 µM Fe3+ (| Fe ().

Fig. 4.36. Time course of the effect of Fe3+ on

the

extracellular

concentration of MCYST (µM) in the growth medium.

MCYST


3+

concentration

culture


supplied with 0.03 µM Fe3+ ( µM Fe3+

(¹), 0.3 µM Fe3+ (| DQG µM Fe3+ ().

DQG µM

147

148



concentration on MCYST content expressed

concentration on the intracellular MCYST

-1

as a proportion of dry weight (µg mg ) in

quota (QMCYST, fmol cell-1) in batch cultures.

batch cultures.

Data presented are for cultures supplied with


cultures supplied with 0.03 µM ( Fe3+ (¹), 0.3 µM Fe3+ (|

µM

DQG µM Fe3+

0.03 µM (

( |

µM Fe3+ (¹), 0.3 µM Fe3+

DQG µM Fe3+ ().

Error bars indicate

(). Error bars indicate ±SEM (n = 3).

±SEM (n = 3).

Fig. 4.38

MCYST (µg MCYST (mg dry

Fig. 4.40. QMCYST (fmol MCYST cell-1) as a

weight)-1) as a function of specific cell

function of specific cell division rate (µ c) in

division rate (µ c) in cultures supplied with

cultures

3+

supplied

with 3+

different

different concentrations of Fe . Data shown

concentrations of Fe .

are single values determined for individual

single values determined for individual

replicates of batch cultures.

replicates of batch cultures. The lines are

Data shown are

drawn from QMCYSTmin to QMCYSTmax according to the values for these parameters listed in Table 4.4. Coefficients of determination (r2) for each treatment are shown.

149

150

Fig. 4.41


(RMCYST, fmol cell-1 d-1) as a function of specific cell division rate in batch cultures of M. aeruginosa MASH-01A19 supplied with different concentrations of Fe3+. Data shown are values for individual replicate cultures. The solid lines represent RMCYST values calculated using Equation 3.7 (Chapter 3) from the QMCYSTmax and QMCYSTmin values determined from linear regression analysis (Table 4.4).

151

TABLE 4.4 MCYST production parameters for M. aeruginosa MASH-01A19 supplied with different concentrations of Fe3+ in batch cultures. Concentration

Max. culture MCYST

µ max

QMCYSTmin1

QMCYSTmax1,2

RMCYSTmax2

(d-1)

(fmol cell-1)

(fmol cell-1)

(fmol cell-1 d-1)

0.03

0.72

0.041 ± 0.009

0.172 ± 0.006

0.124 ± 0.028

4.2

1.64 ± 0.25

0.1

0.72

0.044 ± 0.003

0.208 ± 0.001

0.150 ± 0.000

4.7

1.83 ± 0.21

0.3

0.74

0.057 ± 0.007

0.234 ± 0.024

0.173 ± 0.017

4.1

3.01 ± 0.19

1.0

0.70

0.078 ± 0.009

0.176 ± 0.014

0.123 ± 0.010

2.3

2.91 ± 0.11

of Fe

3+

supplied (µM)

concn3

QMCYSTmax/QMCYSTmin

(µM)

1

Values given are mean regression estimates (QMCYST v µ c) from triplicate cultures ± SE. See Section 4.5.2 for the method of determining QMCYSTmax.

2


3


152


4.5.4

153

Discussion

The intracellular MCYST quota of cultures supplied with varied concentrations of exogenous Fe3+ increased as a function of µ c (up to 0.3 - 0.4 d-1), similar to that described for other treatments (Chapter 3; Sections 4.2 to 4.4). Notably, however, both intracellular MCYST quotas and MCYST:dry weight ratios, at highest observed specific cell division rates, were considerably lower than expected given the relationship between QMCYST and µ described in Chapter 3. This indicates that the relationship between QMCYST and µ c does not hold at 0.03 - 1 µM Fe3+ when µ c > 0.3 - 0.4 d-1. However, since low maximum observed QMCYST values occurred shortly after inoculation in all treatments (Fig. 4.39) it is likely that this phenomenon was not associated with Fe3+ limitation due to the low cell concentrations at this time. More likely, the low QMCYSTmax values at high growth rates can be explained by the fact that, in this experiment, cells were transferred to fresh media from a stationary phase stock culture and are likely to have been in a different physiological state compared with exponential phase cells used in previous batch culture experiments (Sections 4.2 to 4.4). This is supported by the low cell dry weights observed at inoculation in all treatments (Fig. 4.34).

Despite no evidence of a lag period in the growth of all cultures (Figs 4.32, 4.33) the data indicate that the physiological condition of cells shortly after inoculation limited MCYST production. Perhaps the initial nutritional status of cells influenced the nutrient uptake rates in a way that inhibits MCYST synthesis. Under conditions of nutrient limitation, for example (i.e. when cells have reached stationary phase), there can be a large increase in specific nutrient transporter proteins within cells, greatly increasing the potential rate of nutrient uptake when cells are transferred to fresh growth medium (Morel 1987). Under the conditions described here, it is possible that nutrient limited cells in stationary phase (used as an inoculum for this experiment) initially redirected their resources into other cellular processes (e.g. the production of transport proteins for the uptake of limited nutrients) at the expense of MCYST synthesis, thus limiting MCYST production. Once nutrient limitation had been overcome, more resources would become available for MCYST synthesis. This would explain the time lag in achieving QMCYSTmax (Fig. 4.39). The apparent absence of a lag in biomass accumulation and cell division, however, indicates that MCYST production is not dependent upon cell division under these conditions. Interestingly, Watanabe et al. (1989) observed a lag in the increase of MCYST:dry weight in batch cultured cells of Microcystis spp., indicating that perhaps their cultures were also inoculated with nutrient-limited cells. However, this is speculation and it is therefore important that the condition of cells prior to inoculation is stated in batch culture experiments.


154

Although QMCYST does not increase with µ c above 0.3 - 0.4 d-1 in Fe-limited cultures, there is a strong linear correlation between QMCYST and µ c in all treatments at specific cell division rates below 0.3 d-1 in all Fe3+ treatments (Fig. 4.40). Moreover, the results show that QMCYSTmin (and for the most part QMCYSTmax) increases with increasing supply of Fe3+ to cultures. This is indicative of a requirement of Fe3+ in the production of MCYSTs by M. aeruginosa MASH-01A19 and is similar to the findings described for S-limited cultures (Section 4.4), but contrary to the results under P limitation (Section 4.3). The findings reported here are consistent with the observations of Utkilen & Gjølme (1995) who discovered that increasing the Fe3+ supply to cells of M. aeruginosa CYA228/1 grown in continuous culture resulted in an increase in MCYST-RR expressed as a ratio of both dry weight and protein. In contraVW

/XNDþ $HJHUWHU (1993) and

Lyck et al. (1996) found that low Fe3+ supply led to an increase in MCYST content. It is very difficult to reconcile such differences with the findings reported here, although Lyck et al. (1996) found that MCYST content increased in response to Fe3+ starvation when expressed as a ratio to either chlorophyll a or protein, but decreased as a ratio to dry weight. Consistent with this, Bickel et al. (2000) found that the protein and carbohydrate proportion of dry weight changed dramatically in response to Fe3+ supply. This suggests that protein and chlorophyll a are not good indicators of changes in MCYST content. So, when MCYST is expressed in terms of dry weight, the findings of Lyck et al. (1996) are quite similar to those observed here. Since previous studies on the effect of Fe on MCYST have not expressed MCYST content as an intracellular quota, and as protein and chlorophyll a content were not determined in the current study, suitable comparisons cannot be made. The apparent cell death that occurred in cultures supplied with 0.3 µM Fe3+, as instanced by the sharp rise in extracellular MCYST (Fig. 4.36), is difficult to explain since this is an intermediate concentration of Fe3+ and cell death was not observed at either lower or higher concentrations of Fe3+. Perhaps a concentration of 0.3 µM Fe3+ results in an intracellular and/or extracellular antagonism with one or more other micro-nutrient elements to create conditions toxic to growth. The rise in extracellular MCYST was accompanied by an apparent rise in cell dry weight in cultures supplied with 0.3 µM Fe3+ (Fig. 4.34).

The observed increase in cell dry weight

presumably results from the collection of large amounts of insoluble cell debris during the determination of culture dry weight, also indicative of cell lysis in these cultures. A rise in the apparent cell dry weight coinciding with a rise in extracellular MCYST appears to be a consistent phenomenon associated with cell death in batch cultures (see Sections 4.3 and 4.6). This supports the suggestion that MCYSTs are released from cells only upon cell death (Sivonen & Jones 1999) and that extracellular MCYST could be a measure of death rates in M. aeruginosa populations, as mentioned previously (Section 4.2.5).

Nonetheless, the surviving cells in the 0.3 µM Fe3+


155

treatment had QMCYST values that follow the proposed relationship with specific cell division rate at specific cell division rates below 0.3 d-1 (Fig. 4.40) and the final culture MCYST concentration appears to be unaffected by the drop in biomass or cell concentrations (Fig. 4.35). The minimum intracellular MCYST quota (QMCYSTmin) increases with increasing Fe3+ concentration in the growth medium, similar to that described for SO42- (Section 4.4). In addition, there is a significant effect of Fe3+ concentration on culture MCYST concentration in stationary phase cultures (Fig. 4.35; Table 4.4), despite similar biomass yields in the different Fe3+ treatments (Figs 4.32, 4.33). These data suggest that Fe3+ could play an important role in the synthesis of MCYST in M. aeruginosa MASH-01A19. Bickel et al. (2000) recently showed that the MCYST:protein ratio of M. aeruginosa NIVA-CYA 228/1 correlated strongly with the adenylate energy charge of cultures under conditions of varied Fe3+ and Pi supply. This suggests that Fe3+ concentration might affect MCYST content of cells indirectly through ATP availability. This is consistent with the known roles of Fe in a number of major physiological processes in cyanobacteria including nitrogen assimilation, photosynthesis, chlorophyll a production and respiration (Boyer et al. 1987), in the photosynthetic electron transport chain (Straus 1994) and in phycobilisome and membrane assembly (Sherman & Sherman 1983).

Shi et al. (1995) found that MCYSTs in M. aeruginosa PCC 7820 were closely associated with thylakoids perhaps indicating that MCYSTs, like Fe, play a role in photosynthesis. The metal ion binding capabilities of MCYSTs have been reported (Humble et al. 1997) raising the possibility that MCYSTs are involved in Fe2+/Fe3+ binding on or near the thylakoids. Indeed, Utkilen & Gjølme (1995) proposed a model in which MCYSTs bind high intracellular concentrations of Fe2+ which result from light-induced Fe3+ reduction.

This model also proposes that MCYST

production is enhanced in the presence of high concentrations of Fe2+ due to the presence of a Fe2+ regulated MCYST synthetase (Utkilen & Gjølme 1995). This is analogous to the Fur repressor system of E. coli in which the control of Fe transport is regulated by Fe-sensitive repressor proteins (Straus 1994). However, due to the apparently constitutive production of MCYSTs in MCYST-producing cells, an enhancer system may be more likely than a repressor system. There is evidence that light quality and irradiance affect the production of MCYST synthetase transcripts (Kaebernick et al. 2000) and it is known that high irradiance increases Fe3+ uptake in toxic strains of M. aeruginosa (Utkilen & Gjølme 1995). Thus, Fe3+ could be a major contributor to the control of MCYST production in M. aeruginosa. However, further evidence is required to confirm the role of Fe3+ in this process.


156

The intracellular quota of M. aeruginosa MASH-01A19 increases with increasing Fe3+ supply. This finding has important implications with respect to M. aeruginosa bloom formation, as high Fe3+ availability could potentially lead to highly toxic surface scums. Fe3+ is often limiting in aquatic environments due to the formation of insoluble Fe-hydroxides at alkaline pH (Straus 1994). However, the extent to which Fe3+ can influence M. aeruginosa bloom toxicity is not clear. This issue is also complicated by the fact that Fe can influence the availability of P through the formation of insoluble Fe-phosphates. The results suggest that MCYST production could be limited in Fe-limited stationary phase cells and that the attainment of maximum toxicity occurs after a time lag when nutrient-limited cells are transferred to a nutrient replete environment. This suggests that the physiological condition of cells at the time of inoculation into nutrient replete growth medium can influence the MCYST content of cells as QMCYST may be initially uncoupled from growth rate.

Care must therefore be taken in extrapolating the results of laboratory

experiments on the effects of environmental variables on MCYST production into field situations, since the outcome of laboratory experiments may be confounded by the physiological condition of the starting material. Although the results here conflict with several previous laboratory studies

YL] /XNDþ $HJHUWHU /\FN et al. 1996; Bickel et al. 2000), this is the first report of the effect of Fe3+ concentration on the intracellular MCYST quota of M. aeruginosa and the effect of Fe3+ on MCYST production and cell growth.


4.6

157

The effect of irradiance on intracellular MCYST quota of M. aeruginosa MASH-01A19 in batch culture

4.6.1

Introduction

There is mounting evidence that light plays a significant role in the production of MCYSTs in M. aeruginosa. Recent genetic analysis has revealed that both irradiance (Nishizawa et al. 1999) and light quality (Kaebernick et al. 2000) influence the expression of MCYST synthetase genes in M. aeruginosa. In general, MCYST production is greatest in cultures grown at irradiances that support best growth (Rapala & Sivonen 1998) and the MCYST content of cultured material (expressed both on a dry weight basis and per unit protein) is highest when exposed to red light (Utkilen & Gjølme 1992). Despite a large pool of literature concerning the effects of irradiance on MCYST production and content of M. aeruginosa (Table 1.2), only Böttcher et al. (2001) provide information on the effect of irradiance on the intracellular MCYST quota. Their data suggest that QMCYSTmax remains relatively constant as irradiance increases (Böttcher et al. 2001) but provides no information on the effect of irradiance on QMCYSTmin (Section 3.5). Thus, a role for irradiance in determining QMCYST has not been clearly established.

For the most part,

researchers have determined the effect of irradiance on MCYST changes as a proportion to dry weight or other biomass indicators (e.g. protein, Utkilen & Gjølme 1992) although both irradiance and light quality can affect the protein complement and photosynthetic activity of cyanobacteria (Wyman & Fay 1987; Grossman et al. 1994). Thus, the existing literature on the effect of irradiance presents MCYST data as a proportion to biomass parameters that are also subject to change under varied light conditions. Here, the influence of different irradiances on the MCYST production parameters (e.g. QMCYSTmax, QMCYSTmin and total MCYST concentration) are examined in batch cultured cells in order to provide a clearer understanding of the role of light in MCYST production by M. aeruginosa MASH-01A19 cells. The results indicate that at intermediate and high irradiance (40 and 65 µmol photons m-2 s-1) QMCYST increases with µ c but at low irradiance (8 µmol photons m-2 s-1) the net rate of MCYST production is suppressed at high specific cell division rates. However, cells grown at low irradiance contain large amounts of MCYST once they reach stationary growth phase. In contrast, cells grown at relatively high irradiance have a significantly lower intracellular MCYST content at both exponential and stationary growth phases compared with cells grown with less light. Low irradiance also supported high production of MCYST on a culture volume basis.


4.6.2

158


Flasks (500 ml) containing 250 ml of modified MLA medium (Table 2.1) were inoculated to an initial concentration of 0.2 × 106 cells ml-1 from exponential phase stock cultures of M. aeruginosa MASH-01A19 grown under standard conditions (40 µmol photons m-2 s-1; Section 4.2). Cultures were incubated at a constant temperature (26 ± 1 °C) and irradiance was varied by positioning culture flasks at various distances from the light source, employing a continuous light cycle with cool white fluorescent lights. Cultures were thus exposed to irradiances of 8 ± 3, 40 ± 5 (standard growth conditions) and 65 ± 5 µmol photons m-2 s-1, as measured at the external surface of the flasks using a Li-Cor quantum light PAR sensor. Each irradiance treatment was conducted in triplicate. Cultures were subsequently sampled at approximately three-day intervals for dry weight, cell concentration and both intracellular and extracellular MCYST as described previously (Section 4.2). After each sampling, the culture flasks were rotated through their positions under the light source to minimise variations in irradiance within each treatment.

For accurate determination of µ max at each irradiance, separate triplicate cultures were set up under the same conditions of irradiance and temperature in modified MLA medium. These cultures were sampled at regular intervals (4 to 16 h) for six days and cell concentrations determined using a haemocytometer. The maximum observed specific rate of cell division was determined for each level of irradiance according to Equation 4.1 (see Section 4.2).

The intracellular MCYST quota parameters (QMCYSTmax and QMCYSTmin) were determined by regression analysis of positive µ c values and QMCYST values obtained from each replicate culture. Due to variation from the proposed relationship at high specific cell division rates in the low irradiance treatment, QMCYSTmax and QMCYSTmin values were calculated as described for low Fe3+ cultures (Section 4.5.2) in this instance. Net MCYST production rates (RMCYST) were calculated from µ c and QMCYST data and compared with RMCYST values determined using the calculated QMCYSTmax, QMCYSTmin values and µ max, as proposed in Equation 3.7. As a measure of the variation in MCYST content of cells in each treatment, the ratio of QMCYSTmax:QMCYSTmin was determined from the calculated parameters. The maximum culture MCYST concentration (intracellular + extracellular MCYST) in stationary phase cultures was also determined as a measure of the gross MCYST production capabilities of cultures exposed to each irradiance.

Data were statistically analysed for variation within and between treatments using ANOVA and linear regression analysis (Excel 97, Microsoft) and treatment effects were analysed using general linear model univariate analysis (SPSS v 10.0.5, SPSS Inc.).


4.6.3

159

Results

Growth Maximum specific cell division rate (µ max) increased with increasing irradiance (Table 4.5). Despite this, a decrease in biomass was observed in cultures grown at 65 µmol photons m-2 s-1 after approximately 12 days (Figs 4.42, 4.43), possibly due to photobleaching and photooxidative death. These cultures recovered from this bleaching period several days later, eventually reaching maximum biomass concentrations of approximately 1.06 × 107 cells ml-1 and 0.19 mg dry weight ml-1 (Figs 4.42, 4.43) when the experiment was terminated at 28 days. Despite this recovery, these cultures were yellow in colour throughout the experiment.

Cultures grown at lower

irradiance were visibly darker green and showed no signs of bleaching throughout the experiment. Biomass concentrations were greatest in cells grown at 40 µmol photons m-2 s-1 (3.34 × 107 cells ml-1; 0.36 mg dry weight ml-1), and slightly lower in cultures exposed to 8 µmol photons m-2 s-1 (2.34 × 107 cells ml-1; 0.33 mg dry weight ml-1) (Figs 4.42, 4.43). Thus, of the irradiances examined, 40 µmol photons m-2 s-1 provided best growth. As the cultures aged, cell dry weight decreased up to a three-fold (30 to 10 pg cell-1) in all treatments although cell dry weight began to increase in cultures exposed to 65 µmol photons m-2 s-1 when cell death occurred (Fig. 4.44).

MCYST content and production Irradiance had a significant effect on the final MCYST concentration (total intracellular + extracellular MCYST determined at 28 days) (P < 0.05) with greatest MCYST concentrations occurring in cultures exposed to just 8 µmol photons m-2 s-1 and lowest MCYST concentrations in cultures exposed to 65 µmol photons m-2 s-1 (Table 4.5; Fig. 4.45). The concentration of MCYST in the extracellular medium increased as a function of time in all treatments (Fig. 4.46). In cultures exposed to 8 and 40 µmol photons m-2 s-1 the maximum extracellular MCYST concentration (at 28 days after inoculation) was approximately 35% of total culture MCYST concentration.

Cultures exposed to 65 µmol photons m-2 s-1 displayed a rapid increase in

extracellular MCYST at day 14 (Fig. 4.46), coinciding with the period of cell death (Figs 4.42, 4.43) and cell dry weight increase (Fig. 4.44). Although maximum extracellular concentrations of MCYST in the 65 µmol photons m-2 s-1 treatment were attained toward the end of the experiment (24 to 28 days, Fig. 4.46), approximately 87% of total MCYST was found in the extracellular medium of these cultures at 14 days.

Expressed on a dry weight basis, MCYST increased shortly after inoculation in cultures exposed to 8 µmol photons m-2 s-1, reaching a maximum of 9.5 ± 0.3 (SE) µg MCYST (mg dry weight)-1 at


160

approximately six days (Fig. 4.47). After this time, however, MCYST:dry weight decreased in these cultures, reaching a minimum of 4.7 ± 0.1 (SE) µg MCYST (mg dry weight)-1 when the experiment was terminated at 28 days (Fig. 4.47). In the 40 µmol photons m-2 s-1 treatment, MCYST:dry weight decreased from the time of inoculation until the experiment was terminated with a final MCYST:dry weight of 3.9 ± 0.1 µg MCYST (mg dry weight )-1 (Fig. 4.47). MCYST:dry weight in cultures exposed to the highest irradiance (65 µmol photons m-2 s-1) decreased over the first 14-17 days, reaching a minimum of just 1.5 ± 0.2 µg MCYST (mg dry weight)-1 after 17 days and thereafter remained constant (Fig. 4.47). MCYST:dry weight values were lowest at low specific cell division rates (as cells approached stationary phase) in all treatments and generally reached highest values at specific cell division rates of about 0.4 to 0.6 d-1 (Fig. 4.48).

The intracellular MCYST quota (QMCYST) decreased with time in all treatments after inoculation (Fig. 4.49). This decrease in QMCYST was most rapid and marked in cultures exposed to high irradiance (65 µmol photons m-2 s-1) and occurred more gradually the lower the irradiance (Fig. 4.49). At the end of the experiment, QMCYST was lowest in high irradiance cultures (0.038 ± 0.007 (SE) fmol MCYST cell-1) compared with 0.054 ± 0.002 (SE) fmol MCYST cell-1 at 40 µmol photons m-2 s-1 and 0.068 ± 0.001 (SE) fmol MCYST cell-1 at 8 µmol photons m-2 s-1 (Fig. 4.49). QMCYST showed strong positive linear correlations with specific cell division rate in cultures exposed to 40 and 65 µmol photons m-2 s-1 (Fig. 4.50). However, at low irradiance (8 µmol photons m-2 s-1) QMCYST remained constant at specific cell division rates greater than 0.4 d-1 (Fig. 4.50) in a manner similar to that observed at low concentrations of Fe3+ (Section 4.4). Accordingly, for the purposes of linear regression analysis, the data obtained at highest specific cell division rates (µ c = 0.67) in cells grown at 8 µmol photons m-2 s-1 were treated as outliers (Cook's D statistic > 1.0). As a result, the QMCYSTmax value calculated by linear regression analysis was significantly higher than the highest observed QMCYST in this treatment (P < 0.05; Table 4.5). Therefore, highest QMCYSTmax values were actually determined in the intermediate irradiance treatment (40 µmol photons m-2 s-1). QMCYSTmin values, calculated from linear regression analysis, were inversely proportional to the irradiance with highest QMCYSTmin values calculated for the low irradiance treatment (Table 4.5). In addition, calculated QMCYSTmin values were significantly different between all treatments (P < 0.05, ANOVA). QMCYSTmax values calculated from linear regression analysis were similar between the 8 and 40 µmol photons m-2 s-1 treatments, but significantly lower in cultures exposed to 65 µmol photons m-2 s-1 (P < 0.05; Table 4.5).


161

Due to the observed variation from the proposed relationship between QMCYST and µ c (Chapter 3) in the 8 µmol photons m-2 s-1 treatment, an additional equation was required to describe the observed relationship between QMCYST and µ c in this case. The data for this treatment were referred to Dr Rouben Rostamian (Department of Mathematics and Statistics, University of Maryland Baltimore County, USA) for analysis and, empirically, the data for the low irradiance treatment can be described by Equation 4.2:  QMCYST max − QMCYST min    QMCYST = QMCYST max +   / 2 × (µ c − µ crit ) − µ crit    

(µ c − µ crit )2 + d 2  

(Equation 4.2)

This equation describes a hyperbolic relationship between QMCYST and µ c in which QMCYST reaches a maximum (QMCYSTmax) at an intermediate growth rate (µ crit). Like µ c, the term µ crit has units of d-1. The term d is the rate of change in QMCYST at

µcrit

and also has units of d-1. Low values of d

result in a sharp transition at µ c = µ crit whereas high values of d result in a gradual change. For simplicity, since there is no data regarding the rate of change in the proposed relationship between µ c and QMCYST at µ crit in cultures exposed to 8 µmol photons m-2 s-1, the term d can be removed and the equation simplified to:  QMCYST max − QMCYST min    / 2 × (µ c − µ crit ) − µ c − µ crit QMCYST = QMCYST max+  µ crit   

[

]

(Equation 4.3)

This formulation of the equation results in a sharp transition at µ crit (see Fig. 4.50). A value for µ crit (0.48 d-1) was estimated for the low irradiance treatment using the linear regression equation and determining the specific cell division rate at which the observed maximum QMCYST (0.137 fmol cell-1, Table 4.5) was reached. Using the observed maximum QMCYST as QMCYSTmax in the above relationship and the QMCYSTmin calculated from linear regression analysis (0.085 fmol cell-1; Table 4.5), values of QMCYST were calculated from Equation 4.3 between µ c = 0 and 0.95 d-1 to give the dotted line in Fig. 4.50. Constrained non-linear regression analysis (SPSS v 10.0.5, SPSS Inc.) using the above values for the constants QMCYSTmin, QMCYSTmax and µ crit resulted in a coefficient of determination (r2) of 0.56 compared with the data obtained experimentally (Fig. 4.50).

162

Fig. 4.42. Time course of cell concentrations

Fig. 4.43. Time course of culture dry weight

(cell ml-1) in batch cultures of M. aeruginosa

concentrations (mg ml-1) in batch cultures of

MASH-01A19 exposed to either 8 (

M. aeruginosa MASH-01A19 exposed to

either 8 (

¹), or 65 µmol photons m-2

Fig. 4.44. Time course of cell dry weight (pg

Fig. 4.45.

Time course of total MCYST

cell-1) in batch cultures of M. aeruginosa


MASH-01A19 exposed to either 8 (

MCYST, µM) in batch cultures of M.

(¹), or 65 µmol photons m s

aeruginosa MASH-01A19 exposed to either

(¹), or 65 µmol photons m s (| -2

-1

-2

-1

s-1 (|

(| (UURU

bars indicate SE (n = 3).

Fig. 4.46. MCYST

8 ( ¹), or 65 µmol photons m-2 s-1 (|

Time course of extracellular concentration

(µM)

in

batch

cultures of M. aeruginosa MASH-01A19 exposed to either 8 ( photons m-2 s-1 (|

¹), or 65 µmol

163

164


Fig. 4.49.

as a proportion to culture dry weight (µg


MCYST (mg dry weight)-1) in batch cultures

cultured cells of M. aeruginosa MASH-

of M. aeruginosa MASH-01A19 exposed to

01A19 exposed to either 8 (

either 8 (

¹), or 65 µmol photons m-2 s-1 (| (UURU EDUV LQGLFDWH 6( n = 3).


¹), or 65 µmol photons m-2 s-1 (| (UURU EDUV LQGLFDWH SE (n = 3).

Fig. 4.48 MCYST expressed as a proportion

Fig. 4.50.

of dry weight (µg MCYST (mg dry

(QMCYST, fmol cell-1), plotted as a function of

weight)-1), plotted as a function of specific

specific cell division rate (µ c) in batch

-1


cell division rate (µ c, d ) in batch cultures of

cultures of M. aeruginosa MASH-01A19.

M. aeruginosa MASH-01A19.

The solid lines are drawn from the QMCYSTmin and QMCYSTmax values determined from linear regression analysis (Table 4.5).

The

2

coefficients of determination (r ) provide a measure of the agreement between the data points and the linear relationship for each treatment.

In the 8 µmol photons m-2 s-1

treatment the dotted line represents QMCYST data calculated using Equation 4.3 (see text) employing the parameters QMCYSTmin = 0.085 fmol cell-1, QMCYSTmax = 0.137 fmol cell-1 and µ crit = 0.48 d-1 (r2 = 0.56).

The arrow

indicates the point of transition between a positive slope and a slope of zero (µ crit).

165

TABLE 4.5 MCYST production parameters for M. aeruginosa MASH-01A19 exposed to different irradiances in batch cultures. Irradiance

Max. culture MCYST

µ max

QMCYSTmin1

QMCYSTmax1,2

RMCYSTmax2

(d-1)

(fmol cell-1)

(fmol cell-1)

(fmol cell-1 d-1)

8

0.95

0.085 ± 0.005

40

1.21

0.050 ± 0.002

0.189 ± 0.006

0.229 ± 0.008

3.8

2.17 ± 0.06

65

1.31

0.031 ± 0.003

0.107 ± 0.001

0.140 ± 0.003

3.5

1.35 ± 0.10

(µmol photons -2 -1

m s )

0.188 ± 0.009 (0.137 ± 0.004)4

0.179 ± 0.009

QMCYSTmax/QMCYSTmin

concn (µM)3

2.2 (1.6)4

2.43 ± 0.08

1


2


3


4

Values in brackets are the observed mean maximum QMCYST ± SE (n = 3) and QMCYSTmax/QMCYSTmin ratio achieved in low irradiance cultures (see Fig. 4.50)

166


4.6.4

167

Discussion

The data presented in this section show that the intracellular MCYST quota of M. aeruginosa MASH-01A19 varies both as a function of specific cell division rate and irradiance in batch cultured cells (Fig. 4.50; Table 4.5).

QMCYSTmin, the minimum intracellular MCYST quota

achieved at stationary phase, is greatest in cultures grown at low irradiance and the value of this parameter decreases with increasing irradiance (Table 4.5).

In contrast, the maximum

intracellular MCYST quota (QMCYSTmax) is greatest in cultures grown at intermediate irradiance (40 µmol photons m-2 s-1) (Fig. 4.50; Table 4.5). Although high irradiance leads to a higher rate of cell division relative to the rate MCYST production (hence low QMCYST), the converse appears to be true for low irradiance, which results in a slowing in the rate of cell division relative to the rate MCYST production (therefore higher QMCYST). Thus, the cellular MCYST content of M. aeruginosa varies in a complex manner with respect to irradiance.

Utkilen & Gjølme (1992) provide an account of the production of MCYSTs by M. aeruginosa CYA 228/1 under varied irradiance in continuous cultures (grown at 0.31 d-1). These authors showed that, as a ratio to dry weight, MCYST reached a maximum in steady-state cells at 40 µmol photons m-2 s-1 and decreased with increasing irradiance. As a ratio to protein, however, MCYST increased with increasing irradiance up to 40 µmol photons m-2 s-1 and remained constant at higher irradiances (Utkilen & Gjølme 1992). Although the findings of Utkilen & Gjølme (1992) highlight the inadequacies of determining MCYST as a ratio to cellular components they support the findings here, that the MCYST content of M. aeruginosa varies as a function of irradiance. More specifically, the two studies agree that MCYST content could be greatest at an irradiance of 40 µmol photons m-2 s-1 (Fig. 4.50; Table 4.5). However, the data presented in this section suggest that this agreement could be fortuitous because if Utkilen & Gjølme (1992) had conducted their study at lower growth rates (i.e. closer to µ = 0) the outcome may have been significantly different with the MCYST content of cultured material greatest at lower irradiances. Thus, the problems associated with determining MCYST content at single growth rates (or single time points in batch cultures) become evident. The data provided in this study show that while QMCYSTmax is greatest in cultures exposed to 40 µmol photons m-2 s-1 (Fig. 4.50; Table 4.5), QMCYSTmin is inversely proportional to irradiance and suggest that stationary phase cells exposed to low irradiance would be very toxic. Although there appears to be a general trend suggesting that greatest MCYST production coincides with irradiances which support best growth, there are some conflicting data regarding the effects of light on MCYST content of cyanobacteria (see discussion by Sivonen 1990). The measurement of MCYST on a cellular level over all phases of growth in


168

batch culture, as described here, suggests that some of this confusion may be due to significant variation in MCYST content within a single culture over time.

That there is an optimum irradiance for QMCYSTmax conflicts with the findings of Böttcher et al. (2001) who found that QMCYSTmax (QMCYST at µ max in turbidostats) remained constant as irradiance increased. They also found relatively lower QMCYSTmax values in cells exposed to low irradiance, although this was not significant. Perhaps this can be explained by the physiological condition of cells in the experiment described here which suggest that cultures exposed to highest irradiance (65 µmol photons m-2 s-1) were stressed (Figs 4.42, 4.43 and 4.46). The cells exposed to high irradiance in the experiment conducted by Böttcher et al. (2001) were possibly better acclimatised to their growth conditions.

Thus, the relatively low QMCYSTmax observed in the experiment

described here could result from physiological stress experienced by the cells. However, further work is required to confirm this.

Expressed as a proportion of dry weight, MCYST responded to each irradiance differently over time (Fig. 4.47). Most notable was an increase in MCYST:dry weight in the low irradiance treatment to a maximum of 9.5 µg MCYST (mg dry weight)-1 at approximately six days after inoculation. This is a high value for MCYST content of dry material, higher than the reported maximum found in cyanobacterial bloom samples (7.3 µg (mg dry weight)-1, Sivonen & Jones 1999). However, at low irradiance, both the intracellular MCYST quota (Fig. 4.49) and the cell dry weight (Fig. 4.44) decrease immediately after inoculation. Thus, the high MCYST:dry weight achieved at low irradiance is likely to be due to a slow rate of dry weight accumulation compared with MCYST production. This is to be expected from cultures exposed to low irradiance, as the rate of CO2 assimilation would be slow. The converse is likely to be the case for cultures exposed to higher irradiances where MCYST:dry weight decreases over time (Fig. 4.47) as the rate of dry matter accumulation is proportionally greater than the rate of MCYST production.

Thus

MCYST:dry weight data in this experiment is more indicative of the relative rates of dry weight accumulation and MCYST production under varied irradiance than it is of the cellular physiological response with respect to MCYST production. This provides further evidence that reference to the culture dry weight is an unsatisfactory basis for understanding how MCYST production by cells responds to irradiance.

The maximum intracellular MCYST quota for cultures grown at low irradiance (8 µmol photons m-2 s-1) is achieved at intermediate growth rates (approximately 0.4 d-1) and does not increase at higher growth rates (Fig. 4.50). This is similar to the result obtained for cells grown at low Fe3+ (Section 4.5). However, in that case cultures were established from a stationary phase stock


169

culture, whereas cultures in this experiment were established from a stock culture in exponential growth phase. Also, the cultures exposed to higher irradiances in this study display the expected linear relationship between QMCYST and µ c (see Chapter 3). Thus, the observed QMCYSTmax at intermediate specific cell division rates in the cells grown at 8 µmol photons m-2 s-1 is due to the low supply of light and not the nutritional condition of cells before inoculation. Equation 4.3 is proposed to describe the relationship between QMCYST and µ c under these conditions. Like Equation 3.6 in Chapter 3, which describes a linear relationship between QMCYST and µ, Equation 4.3 relies upon the constants QMCYSTmax and QMCYSTmin, suggesting that determinable maximum and minimum intracellular MCYST quotas exist within cells grown under these conditions. However, since QMCYSTmax is obtained at growth rates below µ max then the term µ max is not required in this equation. Instead, the relationship suggests a critical specific cell division rate (µ crit) at which the rates of cell division and MCYST production change significantly relative to one another. As a result, in cells grown at 8 µmol photons m-2 s-1 QMCYST increases in a linear fashion as µ c increases from 0 to µ crit, but remains constant when µ c > µ crit. Under the low irradiance conditions examined in this experiment, QMCYST remained constant at high specific cell division rates (µ c > µ crit). This follows the proposed relationship made by Orr & Jones (1998) which suggests that QMCYST should remain constant if the specific cell division rate (µ c, d-1) and the specific MCYST production rate (µ MCYST, d-1) are equal. However, once specific cell division rates decreased below the proposed µ crit, MCYST production decreased relative to specific cell division rate according to the relationship proposed in Chapter 3. The data suggest that at high specific cell division rates at low irradiance, resources are diverted more toward other cellular processes than to MCYST production. A possible reason for this is that cells grown at low irradiance would need to adapt (stock cultures were grown at 40 µmol photons m-2 s-1) by producing photosynthesis-related compounds (e.g. phycobilisomes, Grossman et al. 1994) at the expense of MCYST production. For cells grown at 8 µmol photons m-2 s-1, the high growth rates, > µ crit, occur in the period immediately after inoculation. Thus, perhaps acclimation occurs when µ c declines with time until µ c = µ crit. Thereafter, as µ c continues to decline, resources could be redirected toward MCYST production in the manner proposed by the relationship in Chapter 3. If acclimation to the low irradiance is the reason for the lower than expected QMCYSTmax, this raises the question whether cells from cultures already adapted to low irradiance display a QMCYSTmax similar to that calculated from linear regression analysis of the data (Table 4.5). This question could be addressed by inoculating fresh culture medium with exponentially growing cells preconditioned to low irradiance (8 µmol photons m-2 s-1) and growing them at that irradiance.


170

QMCYST is low in cells exposed to high irradiance (65 µmol photons m-2 s-1; Fig. 4.50; Table 4.5) even though µ max is high under these conditions (Table 4.5). This suggests that cellular MCYST production is suppressed at high irradiance. This is consistent with the findings of several authors (Sivonen 1990; Song et al. 1998) although previous studies on the effect of irradiance have not determined intracellular MCYST quotas. The low QMCYST values obtained at high irradiance are similar to the results obtained in S-limited cultures (Section 4.4). Under the high irradiance conditions described here the cells showed signs of stress (e.g. photobleaching, a decline in biomass (Figs 4.42, 4.43), a release of MCYST to the external medium (Fig. 4.46) and an increase in apparent cell dry weight (Fig. 4.44)). These features are consistent with photooxidative death (Abeliovich & Shilo 1972). Under such conditions cells have an increased demand for organic reducing compounds, such as glutathione.

Perhaps the demand for compounds such as

glutathione could occur at the expense of MCYST, in which case this could explain the similarities in low QMCYST values between this study and that for S-limited cultures (Section 4.4). However, Kaebernick et al. (2000) showed that oxidative stress in M. aeruginosa PCC7806, induced by methylviologen, had a negative effect on the transcription of the mcyB peptide synthetase gene. By analogy, perhaps photooxidative stress, brought on by high irradiance, could also decrease the transcription of genes required for MCYST synthesis, resulting in low QMCYST under these conditions. Although Kaebernick et al. (2000) argue that this is not the case under the high irradiance conditions employed in their study (68 µmol photons m-2 s-1) it is difficult to make comparisons of the physiological condition of the cells used in their study with those described here. Nonetheless, the data in Figs 4.49 and 4.50 indicate that it is likely that MCYST production has been decreased in cells exposed to high irradiance (65 µmol photons m-2 s-1).

In their study on the transcriptional regulation of MCYST synthesis genes, Kaebernick et al. (2000) showed that transcripts of the mcyB MCYST synthetase gene were most abundant at high irradiance (68 µmol photons m-2 s-1) in the early and mid growth phases.

This appears

inconsistent with the findings here. Nonetheless, Kaebernick et al. (2000) did not establish that the MCYST quotas of cells grown under different irradiances correlated with the amount of this transcript.

Thus, there may be post-transcriptional or post-translational control of MCYST

synthesis resulting in the relationship between irradiance and MCYST quotas seen in this study. However, it is apparent from this and other studies (Nishizawa et al. 1999; Kaebernick et al. 2000) that light influences the MCYST content of M. aeruginosa and further study is warranted.

Interestingly, the total MCYST content of cultures is greatest in cultures exposed to lowest irradiance (Fig. 4.45). This complements the finding that the QMCYST of stationary phase cells is inversely correlated with irradiance (Fig. 4.50; Table 4.5).

These observations provoke


171

speculation regarding the role played by MCYSTs in M. aeruginosa. As mentioned in Chapter 3, MCYST is positively correlated with the chlorophyll a content of cells, under N-limited conditions in chemostats. Although chlorophyll a was not measured in the experiment described here (Section 4.6), Walsh et al. (1997) and Woitke et al. (1997) reported that the chlorophyll a content of M. aeruginosa increases as irradiance decreases. Thus, cultures grown at 8 µmol photons m-2 s-1 would contain higher cellular quotas of chlorophyll a than those exposed to high irradiance, which were yellow in colour. Since the cells grown at low irradiance have the highest cell quotas of MCYST, this supports the possibility that the production of MCYST is related to the amount of chlorophyll a, suggesting that MCYST plays some role in photosynthesis. This is consistent with the suggested association of MCYSTs with the thylakoids of M. aeruginosa (Shi et al. 1995). However, further work is required in order to determine the physiological role, if any, played by MCYSTs in the cyanobacteria which produce them.

Caution needs to be exercised in applying the findings reported here to natural cyanobacterial populations. Although the results imply that bloom material at stationary phase in freshwater systems exposed to low irradiance is potentially more toxic than material exposed to higher irradiances, the data reported in this chapter have been gathered from cultures grown under continuous light instead of a light/dark cycle as prevails in the natural environment. Utkilen & Gjølme (1992) found that MCYST (as a ratio to dry weight and to protein) of M. aeruginosa cells decreased with depth (and irradiance) in the water column. However, as cyanobacterial cells control their buoyancy and therefore their vertical position in the water column in response to light (Reynolds & Walsby 1975; Wallace & Hamilton 1999) and since irradiance decreases with depth, the effect of light on MCYST production may be more complex in the natural environment.


4.7

172

The effect of temperature on intracellular MCYST quota of M. aeruginosa MASH-01A19 in batch culture

4.7.1

Introduction

The optimum temperature for the production of MCYSTs (18 - 25 °C, Sivonen & Jones 1999) does not necessarily coincide with temperatures which support maximum growth (Rapala 1998). In addition, several studies have shown that the MCYST complement of some strains, expressed on a dry weight basis, changes with incubation temperature (Table 1.2). Thus, whilst incubation temperature can affect the MCYST content of cyanobacterial cells, little information exists concerning the effect of temperature on the cellular physiology of MCYST production. Since variations in temperature affect growth rate (Goldman & Carpenter 1974; Krüger & Eloff 1981) it is possible that the MCYST production parameters, QMCYSTmax and QMCYSTmin, also vary with temperature. This section describes changes in the intracellular MCYST quota (QMCYST) of M. aeruginosa MASH-01A19 grown over a range of temperatures in batch cultures. Following the approach used in previous sections of this chapter, QMCYST is examined over the range of growth rates experienced throughout the batch culture growth cycle. Thus, the relationship between QMCYST and specific cell division rate (µ c) at each temperature can be compared using the mathematical model presented in Chapter 3 (Equation 3.6). The results suggest that at low temperature (15 °C) QMCYST varies little over a complete batch culture cycle. Conversely, high temperatures (30 °C) lead to a large variation (> five-fold) in Q MCYST over the complete batch cycle with stationary phase cells containing relatively small amounts of MCYST. However, the QMCYST of stationary phase cells was high in cells grown at low temperature compared with those grown at high temperature.

4.7.2


Flasks (500 ml) containing 250 ml of modified MLA medium (Table 2.1) were inoculated to an initial concentration of 0.2 × 106 cells ml-1 from exponential phase stock cultures of M. aeruginosa MASH-01A19 grown under standard conditions (26 °C; Section 4.2). Cultures were incubated at 15 ± 1, 26 ± 1 and 30 ± 1 °C in three constant temperature rooms at a constant irradiance (40 ± 5 µmol photons m-2 s-1) supplied by identical cool white fluorescent lights (NEC) in each room. Each temperature treatment was conducted in triplicate. Cultures grown at 26 and 30 °C were sampled at approximately three-day intervals until approximately 27 days after inoculation. However, due to relatively slower specific cell division rates at 15 °C, these cultures were sampled at approximately six-day intervals for 48 days. At each sampling, dry weight, cell concentration and both intracellular and extracellular MCYST were measured as described


173

previously (Section 4.2). After sampling, the culture flasks were rotated through their positions under the respective light source to minimise variations in irradiance within each treatment.

Maximum specific cell division rates (µ max) at each temperature were determined using separate triplicate cultures grown under the same conditions of temperature and irradiance in modified MLA medium as the cultures described above. These cultures were sampled at regular intervals (4 to 16 h) for six days and cell concentrations determined using a haemocytometer. The maximum specific rate of cell division at each temperature was calculated using Equation 4.1 (see Section 4.2.3).

QMCYSTmax and QMCYSTmin were calculated by linear regression analysis of positive values of µ c and QMCYST data obtained from each replicate culture as described in previous sections of this chapter. Due to variation from the proposed relationship at high specific cell division rates in the low temperature treatment (15 °C), QMCYSTmax and QMCYSTmin values were calculated as described for low Fe3+ cultures (Section 4.5.2) in this instance. Net MCYST production rates (RMCYST) were calculated from µ c and QMCYST data, and compared with RMCYST values determined using the calculated QMCYSTmax, QMCYSTmin values and µ max, as proposed in Equation 3.7. The ratio of QMCYSTmax:QMCYSTmin was also calculated as a measure of the range of MCYST content in cells at each temperature. The maximum culture MCYST concentration in stationary phase cultures was employed as a measure of the gross MCYST production capabilities of cultures at each temperature.

Data were statistically analysed for variation within and between treatments using ANOVA and linear regression analysis (Excel 97, Microsoft) and treatment effects were analysed using general linear model univariate analysis (SPSS v 10.0.5, SPSS Inc.).

4.7.3

Results

Growth Maximum specific cell division rates (µ max) varied significantly with temperature (P < 0.05) with greatest values in cultures grown at 30 °C (1.49 d-1; Table 4.6). Cultures grown at 15 °C, however, sustained a specific cell division rate (µ c) approaching µ max (0.25 d-1; Table 4.6) for approximately 18 days (Fig. 4.51). The highest cell and dry biomass yields were obtained at 26 °C (Figs 4.51, 4.52), but the highest µ max occurred at 30 °C, albeit of short duration. Cultures grown at 30 and 15 °C achieved only moderate cell yields (1.4 × 107 cells ml-1 and 1.1 × 107 cells ml-1 respectively) compared to those grown at 26 °C (3.3 × 107 cells ml-1) when the experiments


174

were terminated (Figs 4.51). Similar results were obtained when growth was expressed as dry weight accumulation with best growth apparent at 26 °C (Fig. 4.52). The culture and cell dry weight data (Figs 4.52, 4.53) and cell concentration data (Fig. 4.51) provide evidence of cell death after approximately 12 days in the cultures grown at 30 °C. At 30 °C, cell dry weight decreased to a minimum of 10.1 ± 0.7 (SE) pg cell-1 shortly prior to the onset of cell death at approximately 12 days, but thereafter the cell weight increased reaching a final cell weight of 19.3 ± 1.0 (SE) pg cell-1 when the treatment was terminated (Fig. 4.53). Cells grown at 26 °C decreased gradually with time, reaching a minimum (10.6 ± 0.4 (SE) pg cell-1) at 24 days after inoculation (Fig. 4.53). In contrast to the higher temperature treatments, cell dry weight in cultures at 15 °C decreased only slightly throughout the entire experiment, reaching a minimum of 27.3 ± 1.4 (SE) pg cell-1 when the treatment was terminated at 48 days (Fig. 4.53).

MCYST content and production The culture MCYST concentration (intracellular + extracellular MCYST) increased over time in all temperature treatments (Fig. 4.54). When the cultures reached stationary phase, greatest culture MCYST concentrations were found in those cultures grown at 26 °C, with lower MCYST concentrations in the 30 °C and 15 °C treatments (Fig. 4.54; Table 4.6). The concentration of MCYST in the extracellular medium also increased as a function of time in all treatments (Fig. 4.55). Cultures grown at 15 °C had a maximum of 30% (0.01 µM) of total MCYST in the external medium after 21 days in culture, with a similar proportion occurring in cultures at 26 °C (35%) after 28 days though the absolute concentration (0.76 µM) was much greater (Fig. 4.55). In contrast, extracellular MCYST in cultures grown at 30 °C had reached approximately 50% of total after just seven days and rose to a maximum of 82% at 21 days after inoculation. This sharp rise in external MCYST coincided with the period of cell death (Figs 4.51, 4.52) and cell dry weight increase (Fig. 4.53) in this treatment, indicating loss of MCYST to the external medium through cell lysis.

As a proportion of culture dry weight, initially MCYST decreased in all temperature treatments after inoculation. This was most evident in cultures grown at 15 and 30 °C (Fig. 4.56). Despite a rapid decrease in MCYST:dry weight in the cultures grown at 15 °C, this parameter increased again after seven days, thereafter remaining constant at approximately 3.5 µg MCYST (mg dry weight)-1 for the remainder of the experiment (Fig. 4.56).

Cultures at 30 °C displayed a

continuous decline in MCYST:dry weight after inoculation, reaching a minimum of 1.2 µg MCYST (mg dry weight)-1 at the conclusion of the experiment (Fig. 4.56). At 26 °C, the decrease


175

in MCYST:dry weight was less dramatic and a final value of approximately 4.0 µg MCYST (mg dry weight)-1 was observed when the treatment was terminated at 28 days. Thus, MCYST:dry weight was significantly affected by temperature (P < 0.05) with the highest MCYST:dry weight ratio at stationary phase occurring in cultures grown at 26 °C (Fig. 4.56). Due to the variation in µ max with temperature (Table 4.6), MCYST:dry weight was plotted as a function of relative growth rate (i.e. µ c/µ max, Fig. 4.57) to facilitate comparison between treatments. MCYST:dry weight varied little as a function of relative growth rate at 15 °C although the lowest proportions were observed at high growth rates (Fig. 4.57). At higher temperatures, MCYST:dry weight showed a positive correlation with relative growth rate (Fig. 4.57).

Expressed on an intracellular basis (QMCYST), MCYST decreased immediately after inoculation in all temperature treatments (Fig. 4.58). As observed when MCYST was expressed as a proportion of dry weight (Fig. 4.56), QMCYST in the 15 °C treatment increased after seven days, thereafter decreasing only slightly to a final value of 0.089 ± 0.005 (SE) fmol MCYST cell-1 when the experiment was terminated (Fig. 4.58). In the higher temperature treatments QMCYST continued to decrease throughout the experimental period with final QMCYST values lower in the 30 °C cultures (0.023 ± 0.001 (SE) fmol cell-1) compared with cultures grown at 26 °C (0.054 ± 0.002 (SE) fmol cell-1) (Fig. 4.58). Again, since µ max varies with temperature (Table 4.6), the QMCYST data for each temperature treatment are plotted as a function of relative growth rate (i.e. µ c/µ max) in Fig. 4.59 to facilitate comparison between treatments. In the two highest temperature treatments (26 and 30 °C) QMCYST increased in proportion to increasing relative growth rate. In the 15 °C treatment, however, QMCYST increased slightly with relative growth rate until µ c/µ max reached approximately 0.7 but at higher relative growth rates QMCYST declined considerably. Accordingly QMCYST values at relative growth rates > 0.7 for 15 °C cultures were excluded from linear regression analysis (Cook's D statistic > 1.0). Thus, QMCYSTmax calculated for cultures grown at 15 °C is significantly higher than the values observed for QMCYST at maximum growth rates (P < 0.05) (Table 4.6; Fig. 4.59).

QMCYSTmin correlated negatively with incubation temperature, with a significantly higher value for this parameter in cultures grown at 15 °C compared with other treatments (Table 4.6; Fig. 4.59). Thus, cells at stationary phase in batch culture contained significantly more MCYST at low temperature compared with high temperature. However, QMCYSTmax was greatest in cultures grown at 26 °C and lower at 15 and 30 °C (Table 4.6). Therefore, the highest QMCYSTmax value occurred at intermediate incubation temperature.

176

Fig. 4.51. Time course of cell concentrations

Fig. 4.52. Time course of culture dry weight

(cell ml-1) in batch cultures of M. aeruginosa

concentrations (mg ml-1) in batch cultures of

MASH-01A19 grown at either 15 ( ¹),

M. aeruginosa MASH-01A19 grown at either

or 30 °C (|

15 ( ¹), or 30 °C (|

Fig. 4.53. Time course of cell dry weight (pg

Fig. 4.54.

cell-1) in batch cultures of M. aeruginosa


MASH-01A19 grown at either 15 ( ¹),

MCYST, µM) in batch cultures of M.

or 30 °C (|

aeruginosa MASH-01A19 grown at either 15

(UURU EDUV LQGLFDWH 6( n

= 3).

Time course of total MCYST

( ¹), or 30 °C (|

Fig. 4.55. MCYST

Time course of extracellular concentration

(µM)

in

batch

cultures of M. aeruginosa MASH-01A19 grown at either 15 ( ¹), or 30 °C (|

177

178


Fig. 4.58.

as a proportion to culture dry weight (µg


MCYST (mg dry weight)-1) in batch cultures

cultured cells of M. aeruginosa MASH-

of M. aeruginosa MASH-01A19 grown at

01A19 grown at either 15 (

either 15 ( ¹), or 30 °C (|

°C (|

(UURU EDUV


(UURU EDUV LQGLFDWH 6( n

¹), or 30

= 3).

indicate SE (n = 3).

Fig. 4.57 MCYST expressed as a proportion

Fig. 4.59.

of dry weight (µg MCYST (mg dry

(QMCYST, fmol cell-1), plotted as a function of

weight)-1), plotted as a function of relative

relative growth rate (µ c/µ max) in batch

growth rate (µ c/µ max) in batch cultures of M.


aeruginosa MASH-01A19 grown at different

grown at different temperatures. The solid

temperatures.

Absolute maximum growth

lines are drawn from the QMCYSTmin and

-1

QMCYSTmax values determined from linear

rates (µ max) for each temperature are 0.25 d -1

-1

(15 °C), 1.21 d (26 °C) and 1.49 d (30 °C).


regression analysis (Table 4.6). coefficients

of

determination

(r2)

The are

determined from these lines using positive µ c data only. Absolute maximum growth rates (µ max) for each temperature are 0.25 d-1 (15 °C), 1.21 d-1 (26 °C) and 1.49 d-1 (30 °C).

179

180

Fig. 4.60


(RMCYST, fmol cell-1 d-1) as a function of relative cell division rate (µ c/µ max) in batch cultures of M. aeruginosa MASH-01A19 grown at different temperatures. Data shown are empirical values from individual replicate cultures. Curves were plotted using Equation 3.7 (Chapter 3) and the values for µ max, QMCYSTmax and QMCYSTmin in Table 4.6. Absolute maximum growth rates (µ max) for each temperature are 0.25 d-1 (15 °C), 1.21 d-1 (26 °C) and 1.49 d-1 (30 °C).

181

TABLE 4.6 MCYST production parameters for M. aeruginosa MASH-01A19 grown at different temperatures in batch cultures. Temperature

µ max

QMCYSTmin1

QMCYSTmax1,2

RMCYSTmax2

(°C)

(d-1)

(fmol cell-1)

(fmol cell-1)

(fmol cell-1 d-1)

15

0.25

0.095 ± 0.005

26

1.21

0.050 ± 0.002

30

1.49

0.024 ± 0.001

0.111 ± 0.003

Max. culture MCYST QMCYSTmax/QMCYSTmin

concn (µM)3

0.028 ± 0.001

1.2

1.24 ± 0.02

0.189 ± 0.006

0.229 ± 0.008

3.8

2.17 ± 0.06

0.130 ± 0.003

0.194 ± 0.004

5.4

1.63 ± 0.03

(0.071 ± 0.003)4

1


2

QMCYSTmax and RMCYSTmax are determined at µ max.

3

Maximum MCYST concentration is determined from the sum of intracellular and extracellular concentrations of MCYST in cultures at stationary phase. The value shown is the mean ± SE of triplicate cultures.

4

The value in brackets is the observed mean QMCYST ± SE (n = 3) at highest µ c in the 15 °C treatment (see Fig. 4.59).

182


4.7.4

183

Discussion

The intracellular MCYST quota of M. aeruginosa MASH-01A19 varies as a function of both specific cell division rate and temperature in batch cultured cells (Fig. 4.59; Table 4.6). The maximum intracellular MCYST quota (QMCYSTmax) was highest at an intermediate incubation temperature (26 °C) whereas the minimum intracellular MCYST quota (QMCYSTmin) increased with decreasing temperature (Table 4.6). Despite this, very low QMCYST values were observed at high specific cell division rates in cultures grown at 15 °C (Fig. 4.59). Nonetheless, for the most part, QMCYST increases proportionally with µ c at all temperatures examined (Fig. 4.59) according to the relationship described for N-limited chemostats (Chapter 3, Equation 3.6).

The growth of M. aeruginosa MASH-01A19 varies considerably with incubation temperature, with µ max increasing as temperature increases (Table 4.6). This concurs with the results of an extensive study by Krüger & Eloff (1981), which revealed that the thermal optimum for growth rate in disparate strains of M. aeruginosa fell within the range of 29 - 30.5 °C (at an irradiance of 20 µmol photons m-2 s-1). A similar increase in growth rates was also observed with increasing temperature in a study by Coles & Jones (2000). Despite this, optimum biomass yields are apparent in cultures grown under standard growth conditions (26 °C) (Figs 4.51, 4.52). This suggests that biomass production is not necessarily related to the specific cell division rate and is limited at the highest and lowest temperatures examined (15 and 30 °C) compared with the standard culture conditions (26 °C). Perhaps the individual processes contributing to biomass production are limited by different cellular reactions at different temperatures (Pirt 1975). Indeed, Rhee (1982) found that the rate of cell division is not always coupled to the rate of photosynthesis (i.e. dry weight accumulation). This is due to the differing activation energies of particular cellular processes, as described by the Arrhenius equation (Goldman & Carpenter 1974), and suggests that at particular temperatures biomass is directed either into cell division or cell size. In such cases major cellular processes, or master reactions (Sorokin 1960), govern which process is dominant.

At 15 °C, for example, cell dry weight remains almost constant throughout the

experiment whereas it decreases in cultures at higher temperatures (Fig. 4.53). Perhaps this indicates that the process of dry weight accumulation (e.g. photosynthetic C assimilation) is more dominant than cell division at 15 °C while the converse is true at higher temperatures. However, since dry weight biomass is indicative of the relative rates of photosynthesis and respiration, this may also suggest that rates of respiration are considerably higher in cultures grown at 26 and 30 °C.

Highest QMCYSTmax values and the greatest maximum culture concentration of MCYST were obtained at the intermediate growth temperature, 26 °C (Table 4.6; Fig. 4.54). This is consistent with other reports that MCYST content of cyanobacteria is maximal at temperatures between 18


184

and 25 °C (Sivonen & Jones 1999). The results here suggest that greatest MCYST production coincides with greatest biomass yield but not with the highest specific cell division rate (Figs 4.51, 4.52; Table 4.6). Despite the suggestion that QMCYST is dependent upon specific cell division rate (Chapter 3), conditions that support high specific cell division rates (e.g. 30 °C, Table 4.6) will not necessarily support high final concentrations of MCYST. This is also apparent from the previous study on the effect of irradiance on MCYST production (Section 4.6). The results of the temperature study show that although QMCYST is proportional to specific cell division rate, QMCYST also depends on the effect of temperature on QMCYSTmax and QMCYSTmin. In other words, QMCYST results from the relative rates of cell division and MCYST production, which can be independently affected by temperature. These ideas are exemplified by a relatively low QMCYSTmax despite a high µ max for the 30 °C treatment compared with a high QMCYSTmax and lower µ max at 26 °C (Table 4.6).

MCYST production is low at the highest observed specific cell division rates in the low temperature (15 °C) treatment (Fig. 4.59). Since the 15 °C cultures were inoculated from an exponential phase stock culture, and the highest µ c values at 15 °C occur shortly after inoculation, the low QMCYST at high µ c (Fig. 4.59) must result from low temperature. This implies that the production of MCYST is temperature dependent. The synthesis of MCYSTs involves a large number of enzymic processes (Rinehart et al. 1994), any of which could be affected by low temperature. Any single enzyme reaction in this process that is slowed relative to specific cell division rate at 15 °C could limit the production of MCYST within the cell and therefore decrease QMCYST. This would become particularly evident at high specific cell division rates if the rate of cell division is less affected by low temperature than the rate of MCYST production. Since changes in QMCYST result from the relative changes in the rates of MCYST synthesis and cell division (see Section 4.2.5), the data suggest that at highest specific cell division rates in the 15 °C treatment MCYST production must be occurring at a slow rate relative to cell division. This results in an intracellular MCYST quota significantly lower than that expected from linear regression analysis (Fig. 4.59). This means that more resources are diverted into cell division at high specific cell division rates and low temperature than into MCYST production, perhaps indicating that MCYST content of cells is of little importance under these conditions.

The QMCYST of cells grown at 15 °C exhibited much less variation with relative growth rate than cells grown at high temperature (Fig. 4.59). Indeed, QMCYST varied only 1.2-fold between the calculated QMCYSTmax and QMCYSTmin values (Table 4.6). Since the absolute difference between QMCYSTmax and QMCYSTmin determines the degree of curvature of a plot between the net MCYST production rate (RMCYST) and µ c (Chapter 3, Equation 3.7; Section 4.2.5), RMCYST is therefore almost constantly proportional to µ c at 15 °C (Fig. 4.60). The only exceptions are the data obtained at highest µ c values (see above). Thus, MCYST is produced at a rate very similar to the


185

rate of cell division at most specific cell division rates at 15 °C. This is the outcome predicted by Orr & Jones (1998) if the specific rates of MCYST production and cell division are the same. Furthermore, since the calculated QMCYSTmin value at 15 °C (0.095 fmol MCYST cell-1) is unusually high, it appears that cells under these conditions are diverting more resources into MCYST production compared with cells grown at higher temperatures. Perhaps this suggests that there is a greater demand for MCYST production in cells as they begin to reach stationary phase at low temperature and is consistent with the idea that MCYST production may be enhanced under particular stress conditions /XNDþ $HJHUWHU . The findings of van der Westhuizen & Eloff (1985) also suggest that high toxicity is related to slow growth rates under some conditions. This may result by chance, in that poor growth conditions result in a relatively slow cell division rate but MCYST production is not significantly affected (thus a high QMCYSTmin). Alternatively, resources could be specifically diverted into MCYST production under poor growth conditions, perhaps conferring the cells with a competitive advantage. If the latter is true, this suggests there is a specific role for MCYSTs in the cells that produce them. High temperature (30 °C) results in a very low Q MCYSTmin (0.024 fmol MCYST cell-1; Table 4.6) implying that cultures at stationary phase also contain relatively low total MCYST concentrations (Table 4.6; Fig.4.54). Similar results have been obtained for Anabaena sp. strains 90 and 202A1 grown at 30 °C (Rapala et al. 1997). This information suggests that MCYST production is slow in cells grown at high temperature, especially as cell division rates approach zero. Thus, there appears to be a diversion of cellular resources away from MCYST production under these conditions.

Relatively high extracellular concentrations of MCYST were found in cultures grown at 30 °C (Fig. 4.55), suggesting that considerable cell lysis occurs under these conditions.

This is

supported by a decrease in the cell concentration and culture dry weight (Figs 4.51, 4.52) and an increase in the apparent cell dry weight (Fig. 4.53) in these cultures after 17 days. Therefore, the high specific cell division rates at 30 °C appear to be offset by high death rates, resulting in low biomass yields at 30 °C (Figs 4.51, 4.52). The cessation of growth, which would have otherwise contributed to the total MCYST production, is therefore a probable contributor to the low total culture MCYST concentration in this treatment (Fig. 4.54; Table 4.6).

It is interesting to note that, on a dry weight basis, MCYST displays a different pattern compared with MCYST expressed on a cellular basis (QMCYST): the MCYST:dry weight value for 26 °C cultures at stationary phase exceeds that of cells grown at 15 °C (Fig. 4.56) whereas the opposite is the case for QMCYSTmin (Fig. 4.58). These results highlight the different interpretations possible if MCYST is expressed as a proportion of different biomass indicators. More specifically, the data indicate that the rate of MCYST production at 15 °C is slow relative to the rate of dry weight


186

accumulation compared with 26 °C. Conversely, the rate of MCYST production is fast relative to the rate of cell division at 15 °C compared with 26 °C. Previous studies on the effect of temperature on MCYST production have predominantly expressed MCYST content on a dry weight basis (Table 1.2) and therefore provide little information on the cellular response of MCYST production to temperature.

The data presented here clearly show that significant

variation in QMCYST occurs within a single culture over the batch culture growth cycle at high temperature, while at low temperature this parameter varies only slightly (Table 4.6). Furthermore, since the growth rate varies during the course of a batch culture, results obtained from once-sampled cultures (e.g. Watanabe & Oishi 1985) do not take into account the physiological changes that occur as cells progress from exponential to stationary phases.

The data presented here suggest that temperatures supporting maximum biomass production are likely to support greatest MCYST concentrations within a bloom of M. aeruginosa.

The

temperatures attained in summer in the epilimnion of temperate lakes (approximately 20-25 °C) (Welch 1952) are likely to support high biomass and high MCYST production. Thus, greatest MCYST production is likely to occur in temperate regions where cyanobacterial blooms are common (Carmichael et al. 1985). Higher temperatures (25-30 °C) which are typically reported for tropical freshwaters (Ruttner 1963) are likely to support high growth rates of M. aeruginosa but the data in Figs 4.52 and 4.54 predict that both biomass yields and overall MCYST concentrations under these conditions in the field may be relatively low. Consistent with this, Mahakhant et al. (1998) found that the MCYST:dry weight ratio of M. aeruginosa blooms in tropical Thailand are low compared with blooms of the same species in temperate Australian waters. Low temperature also leads to a low biomass yield and low total MCYST concentration (Table 4.6).

In the field, temperature, like irradiance (see Section 4.6 and Utkilen & Gjølme 1992), decreases with depth in stratified temperate waters in summer (Welch 1952). The importance of these effects on growth, biomass and MCYST production within the water column of freshwater bodies remains to be determined.

*

*

*

*

*


187

Analysis of intracellular MCYST quotas in batch cultures of M. aeruginosa confirms the findings presented in Chapter 3 and supports the theory that MCYST production parameters (viz. QMCYSTmax, QMCYSTmin and RMCYSTmax) can be determined from relatively simple measurements of batch cultures. In addition, these parameters are affected by the conditions of growth, implying that environmental conditions have some control over MCYST production. QMCYSTmax, QMCYSTmin and RMCYSTmax affect QMCYST either indirectly, though their effects upon growth rate, or directly through their influence on MCYST biosynthesis. Moreover, the studies reported in this chapter reveal that QMCYST can vary as much as almost five-fold within a single batch culture over time (at 30 °C; Section 4.7). This finding is particularly important in that previous studies have reported variations of a similar magnitude between treatments although changes in MCYST content have not been followed throughout the culture period (e.g. Codd & Poon 1988). The influence of specific growth conditions on QMCYST is summarised in the ensuing chapter.

188

CHAPTER 5 Concluding Remarks

Chapter 5 Concluding remarks

189

The data reported in this thesis provide a unifying view of the disparate outcomes of previous studies on MCYST production by cyanobacteria (Table 1.2).

The experiments reported in

Chapters 3 and 4 show that the intracellular MCYST quota (QMCYST) of M. aeruginosa MASH01A19 varies in a predictable manner in response to growth rate under most of the growth conditions examined. More specifically the data show that QMCYST increases as the growth rate increases, such that fast-growing cells contain more MCYST. Importantly, the data show that MCYST production (expressed on a cellular level) is constitutive and suggests that hepatotoxic strains of M. aeruginosa will remain so regardless of the environmental conditions. However, cellular MCYST production is subject to several levels of control as QMCYST results from both the relative rates of MCYST production and cell division. Thus, factors affecting growth rate also affect QMCYST. The evidence presented in Chapters 3 and 4 suggests that the response of QMCYST to the factors light, temperature N and Pi is mediated through their effect on growth rate (Table 5.1). However, when Fe or S are limiting, both MCYST production and biomass production are affected, with the result that QMCYST is low (Chapter 4), suggesting that both S and Fe may have roles in MCYST biosynthesis as well as growth.

Cell volume is inversely proportional to growth rate in N-limited chemostats (Chapter 3) and cell dry weight decreases with time in batch cultures (Chapter 4). Thus, cell size and weight also change in response to growth rate. It is therefore apparent that the convention of expressing MCYST as a proportion of dry weight or biovolume takes no account of these responses. In addition, the relationships between growth rate and the biomass indicators, dry weight, protein and chlorophyll a, suggest that these indicators do not necessarily provide an accurate representation of cell numbers (Chapter 3). Thus, MCYST expressed as a proportion of any one of these biomass indicators provides potentially misleading information about the cellular production of MCYSTs. The data reported in Chapters 3 and 4, therefore differ from most previous studies of the effects of environmental factors on MCYST production (Table 1.2), as they show that changes in MCYST occur on a cellular level rather than a biomass level. The results presented in this thesis, which show for the first time the influence of a variety of growth conditions on the intracellular content and production of MCYSTs, provide not only a novel approach to determining MCYST content and production but indicate how environmental conditions affect the cellular physiology of MCYST production (Table 5.1).

Chapter 4 confirms that, for most growth conditions, QMCYST is dependent upon growth rate and can be calculated from the maximum growth rate (µ max) and the minimum and maximum intracellular MCYST quotas (QMCYSTmin and QMCYSTmax respectively). This is not only indicative of the constitutive nature of MCYST production but provides an index for comparison of MCYST

Chapter 5 Concluding remarks

190

production under varied growth conditions (Table 5.1). Equation 3.6 shows how the intracellular MCYST content can be determined from the values given in Tables 3.3, and 4.1 to 4.7, for various growth conditions.

The effects on the MCYST production parameters by the

environmental factors studied in Chapters 3 and 4 are summarised in Table 5.1. This thesis therefore provides a novel approach to determining the influence of particular environmental factors on MCYST production. Application of this approach to other growth conditions is likely to provide further insight into cellular MCYST physiology. For example, Bickel et al. (2000) have suggested that MCYST production is linked to energy charge in hepatotoxic cyanobacteria; this hypothesis could be tested by determining both MCYST and energy charge at an intracellular level. Further information on the control of MCYST production could be gained by determining QMCYST under conditions which may affect the level of MCYST synthetase transcripts (e.g. Kaebernick et al. 2000).

The results presented in this thesis raise the question whether the relationship between µ and QMCYST obtained for M. aeruginosa MASH-01A19 (Equation 3.6) is likely to apply, in a modified form, to other strains of M. aeruginosa and other cyanobacteria which produce MCYSTs. Due to the apparent link between growth rate and QMCYST, the data indicate that the control of MCYST synthesis in M. aeruginosa MASH-01A19 is closely related to cell division.

Thus, the

mechanism(s) that control(s) MCYST production is/are also possibly controlled by cell division. Since cell division is an intrinsic characteristic of cyanobacterial survival, it would be reasonable to hypothesise that constitutive and regulatory control systems for MCYST are also well conserved amongst hepatotoxic cyanobacteria. Presumably, the greater production of MCYSTs in some strains than in others, and the absence of MCYSTs from some strains, reflects degree of expression of MCYST synthetases and/or their activity (Nishizawa et al. 1999). It would be less likely that different control systems exist in different strains.

The production of different

MCYSTs within the one strain under varied growth conditions is possibly due to some lack of control in the MCYST biosynthetic system (e.g. variable substrate specificity), thus leading to the large number of MCYSTs so far described. However, the overall control of total MCYST production is likely to remain constant. Further research is required to determine if this is the case.

The unknown MCYST from MASH-01A19 comprises a relatively small proportion of total MCYST (approximately 10 % of MCYST-LR) in M. aeruginosa MASH-01A19, yet it shows

JUHDWHU LQKLELWLRQ RI 33 DQG ELQGLQJ LQ DQ DQWLERG\ DVVD\ GHVLJQHG WR GHWHFW 0&