Physiological and Morphological Changes Induced by Nutrient ...

3 downloads 0 Views 2MB Size Report
DNA coagulation during transmission electron microscopy preparation. .... oxygen limitation at the higher dilution rate. .... aeration and oxygen transfer rate.
Vol. 56, No. 3

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Mar. 1990, P. 686-692 0099-2240/90/030686-07$02.00/0 Copyright C) 1990, American Society for Microbiology

Physiological and Morphological Changes Induced by Nutrient Limitation of Pseudomonas fluorescens 378 in Continuous Culture ANDERS PERSSON,' GORAN MOLIN,1* AND CLAES WEIBULL2

Department of Applied Microbiology, Chemical Center, Lund University, S-221 00 Lund,' and Department of Microbiology, Lund University, S-223 62 Lund,2 Sweden Received 11 July 1989/Accepted 8 December 1989

Pseudomonasfluorescens 378 was studied in continuous culture at a dilution rate of 0.05 or 0.15 h-1 and under a limitation of carbon/energy, nitrogen, phosphorus, iron(III), or oxygen. Cultures were examined for nutritional consumption, production of biosurfactant AP-6 and lipase, and electron microscopy morphology. Morphological features were lysis and plasmolysis of the cells, vacuoles in the cells, granules in cell nuclei, and DNA coagulation during transmission electron microscopy preparation. Biosurfactant and lipase production were lost after 8 to 15 retention times, but under iron limitation and at low dilution rate they were maintained for more than 30 retention times. Consumption of nutrients varied between different cultures. Between 2.4 and 6.0 g of succinic acid per g (dry weight) was consumed; the highest value was obtained under phosphorus limitation. The uptake of nitrogen was mostly about 0.16 g/g (dry weight), and that of phosphorus varied between 13 and 58 mg/g (dry weight). Phosphorus-limited cells reduced their phosphorus consumption by at least 50% compared with other limitations. Cell morphology varied among different cultures. Up to 25% cell lysis occurred at the higher dilution rate. The frequencies of plasmolysis varied between 0 and 85%. Granules in nuclei were found in 65 to 100% of the cells. Vacuoles appeared mostly in low numbers, but at the lower dilution rate under phosphorus or iron limitation the frequencies increased to between 25 and 85%. At high dilution rate, the DNA coagulated in 30 to 70% of the cells. Multivariate data analysis demonstrated a general difference between the two tested dilution rates; i.e., both nutritional and morphological features differed more between the two tested dilution rates than between the different limitations. Cultures at the lower dilution rate changed more with time; this was especially pronounced for phosphorus or iron limitation. The data analysis also showed a correlation between plasmolysis or vacuoles in the cells and an increased carbon uptake under phosphorus limitation.

Principles of continuous culture were presented in 1950 (15, 18), and the concept has since been a frequently applied research tool. An advantage is that specific growth rate and biomass concentration can be controlled simultaneously and are nearly independent of each other, and by applying simple mathematics, culture characteristics such as yield, maintenance, and kinetic constants can be calculated (22). A primary condition, however, is that the cells are physiologically static. On the other hand, it has long been known that microorganisms can adapt to the environment (19). Thus, cells in a continuous culture may change physiologically with time, e.g., the uptake potential for the carbon source (9) or other limiting nutrients (8). Furthermore, changes occurring under specific conditions do not seem to be randomized (5). Bacterial growth in natural environments is often nutrient limited; i.e., nutritional limitation is a common environmental condition to which bacteria adapt. Thus, the strategy of cells growing under nutrient limitation must be to adapt in a direction that allows efficient growth at low nutritional concentrations (8). Bearing this in mind, we undertook a study of Pseudomonas fluorescens 378 growing in continuous cultures under different limitations. Nutritional consumption and electron microscopic morphology of the cells were followed at two different growth rates (dilution rates), and nutritional and morphological data were compared by multivariate data analysis. We found that the physiological status of the culture changed with time, and this was *

especially pronounced at a low dilution rate under phosphorus or iron limitation. P. fluorescens is a frequently occurring bacterium in soil and water (25) and is well known for its ability to antagonize plant pathogenic fungi, to degradate xenobiotics, and to produce lipases. Strain 378, which has been isolated and classified as described elsewhere (14), is able to produce siderophores, lipase, and biosurfactant AP-6 (21). MATERIALS AND METHODS Organism and medium. P. fluorescens 378 was classified as P. fluorescens biovariant 1, with close resemblance to the type strain (ATCC 13525T; 14); strain 378 possesses biosurfactant-producing ability (20). Stock cultures were stored freeze-dried in skim milk at - 18°C. The organism was grown in synthetic medium as described earlier (20). In continuous cultures, the growth medium was modified as given in Table 1. The modifications were changed in relation to the type of limitation and dilution rate applied. To minimize the risk of infection and precipitation of the medium ingredients, the medium was adjusted to pH 4.0 before heat sterilization. Inoculation procedure. Freeze-dried cultures were recovered in baffled shaker flasks (synthetic medium) at 25°C for 24 h. A new shaker flask was inoculated with 20% of the culture and then incubated for 7 h (25°C). The fermentor was inoculated to 10% of the medium volume. Fermentor equipment. The fermentor (Chemoferm AB, Hagersten, Sweden) was fitted with a six-blade open turbine. Due to foaming problems, the baffles were removed and the air inlet was in the headspace; oxygen tension was measured

Corresponding author. 686

NUTRIENT LIMITATION OF P. FLUORESCENS

VOL. 56, 1990

TABLE 1. Critical substrate components in media used under different growth limitations Limited cultures

Succinic

acid

NH4Cl (g/liter)

(mg/liter)

Carbon Nitrogen Phosphorus Iron Oxygene

3/7b 10 10 10 10

4 0.7 4 4 4

800 800 90/120C 800 800

KH2PO4

Fe(IIl) citrate (mg/liter)

5 5 5 od

5

a Carbon- and energy-limiting component. b 3 g/liter in cultures at high dilution rate nad 7 g/liter at low dilution rate. Lower concentrations of succinic acid and KH2PO4 were applied to avoid oxygen limitation at the higher dilution rate. c 90 mg/liter in cultures at high dilution rate and 120 mg/liter at low dilution rate. See footnote b. d Fe(III) was present only as a contaminant of other substrate components. e Stirrer speed lowered from 500 to 425 rpm.

with a galvanic oxygen electrode. The gas flow allowed the exchange of 1 volume of air per volume of medium per minute. pH was maintained constant by automatic titration with 1 M NaOH or 1 M HCl. Supply medium at pH 4.0 for the feed was stored in 20-liter glass vessels. Experimental design. Continuous cultures with a working volume of 2 liters were run under different limitations (Table 1) at 25°C, pH 7.0, and a stirring speed of 500 rpm. They were started as batch cultures, but after about 24 h the feed was started at a dilution rate (D) of 0.05 or 0.15 h-1. The cultures were run for 21 to 30 days, and samples were withdrawn for analysis at suitable intervals. One experimental series was run as a continuous succession of batch cultures. Thus, as the culture entered stationary growth phase, the fermentor was emptied and residuals served as inoculum when fresh medium was added. Analyses. (i) Suspended biomass. The biomass concentration needed for calculation of yield coefficients was determined as described earlier (20), i.e., by dry-weight measurements of washed cell pellets. (ii) Attached biomass. The proportion of metabolic active biomass attached to surfaces in the fermentor compared with the suspended biomass was estimated from ATP analyses. Glass plates, 10 mm in diameter and fitted to glass rods, were submerged in the culture from the start and withdrawn at timed intervals. After withdrawal, the plates were immediately put into 10% trichloroacetic acid-4 mM EDTA solution. The glass plate was then removed from the rod and sonicated in trichloroacetic acid on ice for four 5-s periods with 10-s pauses between treatments. The sonicator was an MSE Soniprep 150 ultrasonic disintegrator (MSE, Crawley, England). After more than 50-fold dilution in Tris-EDTA solution (0.1 M; pH 7.75), ATP content was determined in an LKB Wallac 1250 Luminometer (Wailac Oy, Turku, Finland), using ATP monitoring reagents and ATP standards from LKB Wallac. ATP of the suspended biomass was also measured by the same principles. (iii) Nutritional components. Nutritional components were determined after centrifugation and filtration of the supernatant (cell-free broth). Ammonium chloride was determined as ammonium ions (3). Phosphate was determined by Swedish Standard SIS 028126, which is based on the method described by Murphy and Riley (16). Succinic acid concentrations were determined by high-pressure liquid chromatography on an Aminex HPX-87H organic acid analysis column (Bio-Rad Laboratories, Richmond, Calif.), using 5 mM H2SO4 as eluent.

687

(iv) Biosurfactants. Production of the surfactant AP-6 (21) was determined, in cell-free samples, by the methods described earlier (20). (v) Lipases. Hydrolytic activity was determined in cell-free samples. A 5-ml amount of an emulsion consisting of triolein (5 ml), gumarabicum (41 ml; 15%) and ice (15 g) was mixed with sodium taurocholat (12 ml; 1.5%), CaCl2 (1 ml; 0.075 M), NaCl (2 ml; 3 M), and deionized water (15 ml), followed by temperature adjustment to 37°C. A 1-ml portion of cell-free supernatant was added, pH was adjusted to 8.00, and the hydrolytic activity was measured by autotitration at pH 8.00 (0.005 M NaOH) over a 2-min period. (vi) Siderophores. Capacity of siderophore production of the cell population was estimated on agar plates with chrome azurol S (24), with the following mineral salts (in milligrams per liter): KH2PO4, 800; MgSO4, 400; CaCl2, 10; MnSO4. H2O, 1; ZnSO4 7H20, 1; CuSO4 5H20, 0.5; CoCl2- 6H2O, 0.5; Na2MoO4 2H2O, 0.1; H3BO3, 0.1; and NH4Cl, 4 g/liter. Glucose or glucose and sucrose were added to a final concentration of 1% as carbon source (heat sterilized separately). The plates were incubated at 25°C for 2 days. Orange halos around colonies indicated the presence of siderophores. Transmission electron microscopy. Samples of the continuous cultures were withdrawn during early and late growth stages. The samples were centrifuged at 10,000 x g for 10 min, washed in Michaelis buffer (pH 6.1), and fixed overnight at 4°C with osmium tetroxide (final concentration, 1%) dissolved in Michaelis buffer. The fixed cells were enrobed in 1% agar at 45°C and stained for 2 h with 1% uranyl acetate in Michaelis buffer. Dehydration with acetone and embedding in epoxy resin (Agar 100; Agar Scientific Ltd., U.K.) followed standard procedures. Thin sections of the embedded material were studied with the Philips EM 300 electron microscope equipped with 30-p.m objective aperture and working at 60 kV. Twenty cells from each preparation were randomly selected for transmission electron microscopy, the only prerequisite being a longitudinal cross section. The cells were investigated for presence of lysis, plasmolysis, granules in nuclei, vacuoles, and coagulated DNA. Multivariate data analysis. Measured data from the different forms of growth limitation were simultaneously compared by using the computer program SIMCA-3B, version TMP (Sepanova AB, Enskede, Sweden). The analysis consisted of a principal-component analysis (equivalent to eigenvektor plots; 10) on the whole data set or on parts of it, without direct reference to class separation. SIMCA has been described elsewhere (2, 27). RESULTS Extracellular secretion. Biosurfactant and lipase production in continuous cultures was followed under different nutritional limitations. The culture had lost its ability to produce biosurfactant when it had lost most of its lipolytic activity (about 80%) (Table 2); the exception was the carbonand energy-limited culture run at the higher dilution rate, in which the lipolytic activity was reduced by only 50%. In most cultures, biosurfactant production was lost after 8 to 15 retention times (Table 2). However, the biosurfactant production was maintained for more than 30 retention times under iron limitation at the lower dilution rate. In a series of semicontinuous batch cultures under nonlimited conditions (growth at maximum rate), about 60% of the biosurfactant production was lost after five batches and >90% was lost after ten batches.

APPL. ENVIRON. MIICROBIOL.

PERSSON ET AL.

688

TABLE 2. Reduction in biosurfactant and lipase production of P. fluorescens 378 during continuous cultures at various nutritional limitations

Dilution rate (h-')

No. of retention times needed to lose biosurfactantproducing abilitya

Carbonb

0.05 0.15

100% higher than in carbon- and energy-limited culture (at the start of the medium flow). Nutritional requirements. The requirement of carbon and energy (succinic acid), nitrogen, and phosphorus of P. fluorescens 378 in continuous culture at different dilution rates and limitations is shown in Table 3. (i) Carbon and energy limitation. No major differences were to be seen between early and late stages of the culture; this trend was the same for all cultures. However, the requirement of succinic acid varied with dilution rate. The requirement of carbon and energy source decreased with increasing dilution rate, while the reverse was true for phosphorus (Table 3). The requirement for the carbon and

affected not only by the dilution rate, but also by the type of limitation. It varied between 2.4 and 6.0 g/g (dry weight), being highest under phosphorus limitation. The carbon source-limited culture at high dilution rate had the highest demand for phosphorus. (ii) Nitrogen limitation. The nitrogen requirement under nitrogen (ammonium) limitation was about 0.15 g/g (dry weight); only minor variations from this value were recorded for most of the other limitations. Under iron(III) limitation, however, the corresponding value was 0.21 g/g (dry weight) (D = 0.15 h-1 [Table 3]). (iii) Phosphorus limitation. The highest recorded consumption of carbon source was found under phosphorus limitation (Table 3). The requirement of phosphorus (phosphate) varied between different types of cultures in the range of 13 to 58 mg/g (dry weight). The demand for phosphorus in phosphorus-limited cells was reduced by at least 50% in comparison with cells of other limitations (Table 3). (iv) Iron(III) limitation. There is a slight trend in the iron-limited culture towards an increasing consumption of the analyzed nutritional elements at the higher dilution rate. The demand of nitrogen also seems to be somewhat higher under iron limitation (Table 3). (v) Oxygen limitation. In the present case, operating under oxygen limitation means that the growth rate is determined by dilution rate and that biomass concentration is affected by aeration and oxygen transfer rate. On several occasions, continuous cultures intended for oxygen limitation converted during running to being carbon source limited. Since oxygen-limited cells did not seem to increase their demand for carbon source (Table 3), it is implied that the efficiency of the culture to utilize oxygen was increasing with culture time. Substrate utilization under oxygen limitation largely follows the same nutritional pattern as the carbon sourcelimited culture. (vi) Attached growth. Based on measurements of ATP of suspended and attached biomass in the fermentor, it was found that a relatively small fraction of the total active biomass, was attached; i.e., 0.5 to 2% of the total biomass was immobilized to surfaces of the fermentor and fermentor equipment. The proportion was in the upper part of the range in the late stage of cultures run at the high dilution rate (D = energy source was

0.15

h-1).

Morphology. Cell lysis, plasmolysis, granules in nuclei,

TABLE 3. Requirement for carbon and energy source (succinic acid), nitrogen (ammonium ions), and phosphorus (phosphate) of P. fluorescens 378 in continuous culture under different nutritional limitations and dilution rates Early/late stagea Limitation

Carbon consumption (g of succinic acid/g of dry biomass)

Nitrogen consumption

Phosphorus consumption

(g of N/g of dry biomass)

(mg of P/g of dry biomass)

h-'

0.15 h-'

0.05 h-'

Carbon

3.4/3.2

2.5/2.4

0.13/0.16

0.14/0.13

26/33

58/49

Nitrogen

4.7/4.1

NDb

0.16/0.14

ND

33/35

ND

Phosphorus

4.5/6.0

5.2/5.5

0.13/0.16

0.19/0.16

16/13

26/16

Iron

3.5/4.1

3.4/3.7

0.19/0.16

0.21/0.21

26/39

33/45

Oxygen

3.9/3.0

2.7/2.9

0.16/0.14

0.17/0.15

29/26

37/49

0.05

0.15

h-1

0.05

h-1

0.15 h-

a Early stage of culture represents the first part of the culture corresponding to a time period of 75 to 350 h (D = 0.05 h-1) or 40 to 240 h (D = 0.15 h-1). Late stage of culture represents the second part of the culture corresponding to a time period of 350 to 750 h (D = 0.05 h-1) or 240 to 500 h (D = 0.15 h-1). Values represent the average of three to four samples withdrawn after different times within the early or late stage of the culture. b ND, No data.

VOL. 56, 1990

NUTRIENT LIMITATION OF P. FLUORESCENS

689

FIG. 1. P. fluorescens 378 cells showing (a, left) coagulated DNA and (a, right) granules in nuclei; (b) vacuoles; (c) plasmolysis; and (d) lysis. Bars, 0.2 pm.

vacuoles, and coagulated DNA in cells of P. fluorescens 378 grown under different nutritional limitations were studied in the transmission electron microscope. The different phenomena are exemplified in Fig. 1 and 2, and the frequencies of occurrence are shown in Table 4. Lysis was most frequent (0 to 25%) at the higher dilution rate. The frequencies of plasmolysis varied between wide limits (0 to 85%), with carbon source-limited culture at low dilution rate showing the maximal value. Granules in nuclei were found in high frequences (65 to 100%) in all samples. Vacuoles appeared mostly in low numbers, but at low dilution rate under phosphorus or iron limitation the frequencies were high (25 to 85%). A high fraction of coagulated DNA (35 to 70%) was found at high dilution rate under carbon, iron, or oxygen limitation. Multivariate data analysis. The data set of Tables 3 and 4 were evaluated by principal-component analysis. Thus, the variables were condensed to two components, preserving as much data variance as possible. Directions in the score and the loading plots are directly related. Hence, the variable responsible for a specific location of an object (continuous culture) in the score plot is correlated by the loading plot; i.e., the parameter(s) (experimentally recorded data) responsible for the obtained deviations is pointed out. Figure 3a shows the score plot of the whole data set; i.e., the continuous cultures at the two dilution rates were the objects, and the experimentally recorded data were the parameters. With the exception of the phosphorus- and iron (D = 0.05 h-1)-limited cultures, there was a clear separation between cultures of high and low dilution rates. The horizontal axis explains 24% of the data variance and the vertical explains 5%; altogether, 29% of the total variance is ex-

plained by the two axes. The phosphorus-limited culture grown at D = 0.15 h-1 almost clustered with those of D = 0.05 h-', and the one grown at D = 0.05 h-1 was separated from all of the others, as was the iron-limited cultures at D = 0.05 h-1. The corresponding loading plot (Fig. 3b) reveals that most variables are involved in the creating of the score plot. The horizontal spreading of the cultures at high dilution rate and that of phosphorus at low dilution rate were primarily caused by requirement of phosphorus, frequency of coagulated DNA and granules in nuclei (being high in the former), and requirement of carbon and energy source, frequency of vacuoles, and reduction in lipase activity (being high in the latter). The vertical spreading of the cultures at high dilution rate and the iron-limited culture of low dilution rate were correspondingly caused by early lysis, plasmolysis, vacuoles, nitrogen requirements, and production ability of AP-6. When morphological data were excluded from the score plot, the two dilution rates separated with the exception of phosphorus limitation (D = 0.15 h-1) and iron limitation (D = 0.05 h-1; Fig. 4). Some 32% of the variance is explained by the horizontal axis, whereas the vertical explains 17%. Phosphorus-limited cultures of both dilution rates were close due to high requirements of carbon and energy source and low requirements of phosphorus. Iron-limited cultures showed high requirements of nitrogen and to some extent production stability of AP-6. The question of early and late stages of the continuous culture was separately scrutinized by setting the early and late stages as objects instead of parameters in the loading plot. This demonstrated larger variations between the early and late stages of phosphorus (both dilution rates)- and

690

PERSSON ET AL.

APPL. ENVIRON. MICROBIOL.

a

TABLE 4. Occurrence of cell lysis, plasmolysis, granules in nuclei, vacuoles, and coagulated DNA in cells of P. fluorescens 378 grown in continuous culture under different nutritional limitations

Clmtureo

limitation

Dilution rate

(h1)

Early/late stage of culture (%)a Lysis Plasmolysis Granules in nuclei Vacuoles Coagulated DNA

Lsi

EFe 231.2 0

c-

c)

.2

o

Carbonb

Phosphorus Iron

Oxygen

0.05 0.15

0/0 85/15 0/20 30/10

70/75 95/95

5/10 5/10

0/0 45/40

0.05 0.15

0/0 20/15

5/0 15/30

80/15

55/85

90/90

0/0

0/0 0/0

0.05 0.15

0/5 10/10

0/45 5/5

70/80 95/5

85/25 0/0

0/20 60/35

0.05 0.15

25/0 10/5

0/0 5/5

65/100 95/90

5/0 0/0

0/0 70/70

*;

2

c.

FIG. 2. P.

fluorescens

378 cell

showing

the thin, rather

homoge-

membrane of the cell vacuole. Compare with the unit membranes of the cell wall and the cytoplasmic membrane. Bar, 0.1 ,.m. neous

PP

EJp

0

Nl

c)

-1.8 -3.2

~~~C

~

C

-1.2 .8 2.8 First principal component

br .1

DISCUSSION Iron(III) limitation is known to stimulate production of siderophores (4, 12), but also the production of P. aeruginosa toxin A and hemagglutamin from the same organism (1), proteinase and lipase (6, 17), and biosurfactant (7). Regarding the lower dilution rate, highest biosurfactant and lipase production were, as well, found under iron limitation, and these abilities were also best preserved under iron

F

0

a Early stage, Sample withdrawn after 100 h of cultivation; late stage, sample withdrawn at the end of the cultivation. b Carbon and energy limitation (succinic acid).

iron-limited cultures (low dilution rate), and smaller but still significant changes in carbon- and oxygen-limited cultures (low dilution rates). Thus, the culture most frequently changed with time at the low dilution rate.

* Fe

2.2

6"'7

1

6A

A1

5

2A

1a5i¢,

7

14A 12

-.05 5

8A

-.252 A

A

13 9

-.45 L

Z 18

4 .A

-. 1 -.3 1 .3 FIG. 3. First and second principal component scores (a) and corresponding loading plot (b). (a) Plotted after principal component analysis of complete data set. Filled symbols represent continuous cultures at high dilution rate (D = 0.15 h-1), and open symbols low dilution rate (D = 0.05 h-1). Fe, Iron-limited culture; C, carbon limitation; P, phosphorus limitation; N, nitrogen limitation; 0, oxygen limitation. Axes show arbitrary units. (b) Numbers refer to the following variables: (1) early and (11) late consumption of succinic acid; (2) early and (12) late phosphorus consumption; (3) early and (13) late nitrogen consumption; (4) production of AP-6; (5) reduction of lipase activity; frequency of (6) early and (14) late lysis; (7) early and (15) late plasmolysis; (8) early and (16) late granules in nuclei; (9) early and (17) late vacuoles; (10) early and (18) late coagulated DNA. Axes show arbitrary units. .

limitation. However, at the higher rate the abilities were best preserved under oxygen or carbon limitation (Table 2). The irreversible loss of the abilities suggests a plasmid connection, and the stability of this hypothetical plasmid(s) is obviously not solely dependent on the iron(III) concentration. The consumption of carbon, nitrogen, and phosphorus found in the present studies is, in general terms, of the same order of magnitude as those presented for Klebsiella aerogenes (22). In many cases, the consumption is affected by dilution rate (growth rate); e.g., the consumption of phosphorus significantly increased at higher dilution rate under carbon or iron limitation (Table 3). It may be pointed out that succinic acid has been shown to be a carbon source that minimizes the production of extracellular oxidation products of P. fluorescens (11) which otherwise might have been responsible for some of the overconsumption. On the other

NUTRIENT LIMITATION OF P. FLUORESCENS

VOL. 56, 1990 N -

ACKNOWLEDGMENTS

0

CO

1.5

P .

0

E.5E

Co

C.)

0

co

P,

0*

-.5 Fe

LiI

I"

-1.5

-

° -2.5 Fe

FeLI

-3.5 -4.7

691

A

-2.7

-.7 1.3 3.3 First principal component

FIG. 4. Score plot after principal-component analysis of the nutritional data only (Table 3). Filled symbols represent continuous cultures at high dilution rate (D 0.15 h-1), and open symbols 0.05 h-1). Fe, Iron-limited culture; represent low dilution rate (D C, carbon limitation; P, phosphorus limitation; N, nitrogen limitation; 0, oxygen limitation. =

=

hand, it is hardly surprising that the dilution rate per se affects the uptake of nutrients. Many organisms can apply different uptake systems for the same nutritional component, a strategy dependent on the availability of the nutrient (8). Furthermore, increased growth rate means an increasing RNA content that may reflect an increasing requirement of phosphorus (26). The present study showed that differences in nutritional uptake are also mirrored in major morphological discrepancies. Considering the fact that the proportions of lysis, plasmolysis, and vacuoles vary (Fig. 1 and Table 4), it is not surprising that the consumed proportions of nutrients also vary. The two other phenomena observed, granules in the nuclei and coagulated DNA, might be regarded as more obscure. However, X-ray analysis indicated that the granules (Fig. 1) had a high calcium content (data not shown). The coagulation of DNA probably reflects an increased permeability of the cell envelope. Under such conditions, the nuclear area may be depleted of substances such as amino acids and peptides, which prevent coagulation in healthy cells. The addition of tryptone or amino acids to the fixation fluid protects the nucleoplasm from coagulation (23). It could be pointed out that cell lysis and the occurrence of coagulated DNA seem to be highly correlated phenomena (Table 4). Phosphorus-limited cells deviated from the others mainly because of a high succinic acid consumption and high frequencies of plasmolysis and vacuoles (Fig. 3). The question arises whether these phenomena are connected. Furthermore, these cells showed a remarkable reduction in their

consumption of phosphate. It is known that phosphorus limitation can decrease the content of phospholipids in the cell membrane. Thus, it has been shown (13) that a marine strain of P. fluorescens was able to exchange the phospholipids for ornithine-containing lipids and an acidic glycolipid. When the chemostat is being used as a tool for biochemical and physiological studies, it is mostly an advantage when the culture is changing as little as possible with time.

The present study with P. fluorescens 378 indicated that such a chemostat should not be allowed to run at low dilution rate, especially when phosphorus or iron limitation is applied.

We are grateful to Roman Wroblewski, Karolinska Hospital, Stockholm, Sweden, for X-ray analysis; Ewa Lie for lipase measurements; and Birgitta Sorenby and Ulla Wulf for skillful technical assistance. Financial support from Berol Kemi AB and the Swedish National Board for Technical Development is acknowledged. LITERATURE CITED 1. Bjorn, M. J., P. A. Sokol, and B. H. Iglewski. 1979. Influence of iron on yields of extracellular products in Pseudomonas aeruginosa cultures. J. Bacteriol. 138:193-200. 2. Blomquist, G., E. Johansson, B. Soderstrom, and S. Wold. 1979. Classification of fungi by means of pyrolysis-gas chromatography-pattern recognition. J. Chromatogr. 173:19-32. 3. Chaney, A. L., and E. P. Marbach. 1962. Modified reagents for the determination of urea and ammonia. Clin. Chem. 8:130-132. 4. de Weger, L. A., R. van Boxtel, B. van der Burg, R. A. Gruters, F. P. Geels, B. Schippers, and B. Lugtenberg. 1986. Siderophores and outer membrane proteins of antagonistic, plantgrowth-stimulating, root-colonizing Pseudomonas spp. J. Bacteriol. 165:585-594. 5. Dykhuizen, D. H., and D. L. Harti. 1983. Selection in chemostats. Microbiol. Rev. 47:150-168. 6. Fernandez, L., C. San Jose, H. Cholette, and R. C. McKellar. 1988. Characterisation of a pyoverdine-deficient mutant of Pseudomonasfluorescens impaired in the secretion of extracellular lipase. Arch. Microbiol. 150:523-528. 7. Guerra-Santos, L., 0. Kappeli, and A. Fiechter. 1984. Pseudomonas aeruginosa biosurfactant production in continuous culture with glucose as carbon source. Appl. Environ. Microbiol. 48:301-305. 8. Harder, W., and L. Dijkhuizen. 1983. Physiological responses to nutrient limitation. Annu. Rev. Microbiol. 37:1-23. 9. Hofle, M. 1983. Long-term changes in chemostat cultures of Cytophagajohnsonae. Appl. Environ. Microbiol. 46:1045-1053. 10. Kowalski, B. R., and C. F. Bender. 1973. Pattern recognition. II. Linear and nonlinear methods for displaying chemical data. J. Am. Soc. 95:686-693. 11. Lee, W. S., J. K. Cooper, and W. H. Lynch. 1984. Membrane enzymes associated with the dissimilation of some citric acid cycle substrates and production of extracellular oxidation products in chemostat cultures of Pseudomonasfluorescens. Can. J. Microbiol. 30:396-405. 12. Meyer, J. M., and M. A. AbdalHah. 1978. The fluorescent pigment of Pseudomonasfluorescens: biosynthesis, purification and physicochemical properties. J. Gen. Microbiol. 107:319328. 13. Minnikin, D. E., and H. Abdolrahimzadeh. 1974. The replacement of phosphatidylethanolamine and acidic phospholipids by an ornithin-amide lipid and a minor phosphorous-free lipid in Pseudomonasfluorescens NCMB 129. FEBS Lett. 43:257-260. 14. Molin, G., and A. Ternstrom. 1986. Phenotypically based taxonomy of psychrotrophic Pseudomonas isolated from spoiled meat, water, and soil. Int. J. Syst. Bacteriol. 36:257-274. 15. Monod, J. 1950. La technique de culture continue. Thdorie et applications. Ann. Inst. Pasteur (Paris) 79:390-410. 16. Murphy, J., and J. P. Riley. 1962. A modified single solution method for the determination of phosphate in natural waters. Anal. Chim. Acta 27:31-36. 17. Nishio, T., T. Chikano, and M. Kamimura. 1987. Purification and some properties of lipase produced by Pseudomonas fragi 22.39B. Agric. Biol. Chem. 51:181-186. 18. Novick, A., and L. Szilard. 1950. Description of the chemostat. Science 112:715-716. 19. Novick, A., and L. Szilard. 1950. Experiments with the chemostat on spontaneous mutations of bacteria. Proc. Natl. Acad. Sci. USA 36:708-719. 20. Persson, A., and G. Molin. 1987. Capacity for biosurfactant production of environmental Pseudomonas and Vibrionaceae growing on carbohydrates. Appl. Microbiol. Biotechnol. 26: 439-442. 21. Persson, A., E. Osterberg, and M. Dostalek. 1988. Biosurfactant

692

PERSSON ET AL.

production by Pseudomonasfluorescens 378: growth and product characteristics. AppI. Microbiol. Biotechnol. 29:1-4. 22. Pirt, S. J. 1975. Principles of microbe and cell cultivation. Blackwell Scientific Publications, London. 23. Ryter, A., and E. Kellenberger. 1958. Etude au microscope electronique de plasmas contenant de l'acide desoxyribonucleique 1. Les nucleoides des bact6ries en croissance active. Z. Naturforsch. Teil B 13:597-605. 24. Schwyn, B., and J. B. Neilands. 1987. Universal chemical assay

APPL. ENVIRON. MICROBIOL.

for the detection and determination of siderophores. Anal. Biochem. 160:47-56. 25. Stanier, R. Y., N. J. Palleroni, and M. Doudoroff. 1966. The aerobic pseudomonads: a taxonomic study. J. Gen. Microbiol. 43:159-271. 26. Tempest, D. W. 1969. Microbial growth. Symp. Soc. Gen. Microbiol. 19:87. 27. Wold, S. 1976. Pattern recognition by means of disjoint principal components. Pattern Recognition 8:127-139.