Marine Biology (2018) 165:88 https://doi.org/10.1007/s00227-018-3348-5
METHOD
Carbonic anhydrase activity in seaweeds: overview and recommendations for measuring activity with an electrometric method, using Macrocystis pyrifera as a model species Pamela A. Fernández1,2 · Michael Y. Roleda3,4 · Ralf Rautenberger3 · Catriona L. Hurd5 Received: 9 February 2018 / Accepted: 5 April 2018 © Springer-Verlag GmbH Germany, part of Springer Nature 2018
Abstract Carbonic anhydrase (CA) plays an important physiological role in all biological systems by accelerating the interconversion of CO2 and H CO3−. In algae, CA is essential for photosynthesis: external CA ( CAext) dehydrates H CO3−, enhancing the supply − of CO2 to the cell surface, and internal CA ( CAint) interconverts H CO3 and C O2 to maintain the inorganic carbon (Ci) pool and supply CO2 to RuBisCO. We first conducted a literature review comparing the conditions in which CA extraction and measurement have been carried out, using the commonly used Wilbur–Anderson method. We found that the assay has been widely modified since its introduction in 1948, mostly without being optimized for the species tested. Based on the review, an optimized protocol for measuring CA in Macrocystis pyrifera was developed, which showed that the assay conditions can strongly affect CA activity. Tris–HCl buffer gave the highest levels of CA activity, but phosphate buffer reduced activity significantly. Buffers containing polyvinylpyrrolidone (PVP) and dithiothreitol (DTT) stabilized CA. Using the optimized assay, CAext and CAint activities were readily measured in Macrocystis with higher precision compared to the non-optimized method. The CAint activity was 2 × higher than CAext, which is attributed to the Ci uptake mechanisms of Macrocystis. This study suggests that the CA assay needs to be optimized for each species prior to experimental work to obtain both accurate and precise results.
Introduction Carbonic anhydrase (CA) is a zinc metalloenzyme that catalyzes the interconversion of C O2 and H CO3− in many liv+ ing organisms: H2 O + CO2 ↔ H + HCO−3 . The presence Responsible Editor: K. Bischof. Reviewed by J. Beardall, A. Flores-Moya, and J. M. Mercado. * Pamela A. Fernández
[email protected] 1
Department of Botany, University of Otago, PO Box 56, Dunedin 9054, New Zealand
2
Centro i ~ mar, Universidad De Los Lagos, Camino Chinquihue Km 6, Puerto Montt, Chile
3
Norwegian Institute of Bioeconomy Research, Kudalsveien 6, 8027 Bodø, Norway
4
The Marine Science Institute, College of Science, University of the Philippines Diliman, Quezon City, Philippines
5
Institute for Marine and Antarctic Studies (IMAS), University of Tasmania, Hobart, TAS 7001, Australia
of CA is ubiquitous among eukaryotes and prokaryotes, having been reported in terrestrial plants, algae, animals, and bacteria (Waygood 1955; Bowes 1969; Weaver and Wetzel 1980; Badger and Price 1994). CA is fundamental in all biological systems, because it accelerates the formation of CO2 with typical reaction rates between 104 and 106 reactions per second, which facilitates the high biological demands of living organisms. The uncatalyzed reaction proceeds too slowly to fulfil such high biological rates (e.g., photosynthetic rates) (Sültemeyer et al. 1990; Badger and Price 1994; Sharma et al. 2009; Sharma and Bhattacharya 2010). In animals, CAs play crucial roles in a wide range of metabolic processes, e.g., catabolic respiration and calcification (Lionetto et al. 2016b). In algae and terrestrial plants, CAs are essential in the photosynthetic process allowing, for example, the rapid supply of CO2 to the enzyme, ribulose1,5-bisphosphate carboxylase/oxygenase (RuBisCO) (Surif and Raven 1989; Raven 1995), and for other metabolic reactions such photorespiration, calcification, and intracellular pH homeostasis (Badger 2003; Hofmann et al. 2013; Bi and Zhou 2016).
13
Vol.:(0123456789)
88
Page 2 of 12
So far, six different and unrelated CA families (i.e., α-, β-, γ-CA, δ-, ζ-, and η-CA) have been identified among the different kingdoms (Lionetto et al. 2016a; DiMario et al. 2018). The most investigated CA is α-CA, which is mostly present in animals, but it is also found in fungi, bacteria, algae, and green plants (Lionetto et al. 2016a, b). Recent genomic analyses have shown that different CA types (i.e., α-, β-, and γ-CAs) might occur in a single species (Deng et al. 2012; Moroney et al. 2011; Heinrich et al. 2012; Ye et al. 2014). In micro- and macroalgae, CAs can be found on the outer surface of the cell wall, i.e., external CA (CAext), and in different intracellular locations, i.e., internal CA (CAint), depending on the species (Bowes 1969; Graham and Smillie 1976; Tsuzuki et al. 1984; Miyachi et al. 1983; Cook et al. 1986, 1988; Surif and Raven 1989; Giordano and Maberly 1989). CAext increases the transport of C O2 through the plasma membrane by passive diffusion after the conversion of H CO3− to C O2, especially at a seawater (SW) pH of ~ 8.07 where most of the Ci exists as H CO3− (Bowes 1969; Miyachi et al. 1983; Tsuzuki 1983). CAint may have distinct functions, depending on which compartment it is located (Sültemeyer 1998). Since C O2 and H CO3− generate different pH values in a solution, changes in their concentrations affect the intracellular pH. Cytosolic CA helps to maintain the equilibrium between Ci concentrations and internal pH, e.g., converting C O2 to H CO3− to maintain the − internal Ci pool as H CO3 and avoid any C O2 leakage outside of the cell when the concentration gradient is higher inside the cell (Fernández et al. 2014), mitochondrial CA potentially limits any C O2 leakage from the cytosol to the medium (Raven 2001), and the chloroplast CA is responsible for the supply of C O2 to the active site of RuBisCO (Badger and Price 1994). In all these cases, CAs are crucial to the proper functioning of carbon dioxide concentrating mechanisms (CCMs) in both micro-and macroalgae, helping them to overcome the constraints of RuBisCO oxygenase, e.g., photorespiration (Badger 2003). The reaction catalyzed by CA has been extensively studied in animals, including humans. Therefore, it is not surprising that the first methods used to measure CA activity in algae were adapted from those developed for animals. In the early studies in the 1940s and 1950s, three methods were described for measuring CA activity in animals: manometric, colorimetric, and electrometric. In the manometric methods, CA activities are determined either by the rate of CO2 produced from the dehydration of HCO3− or by the increased rate of CO2 uptake (Roughton and Booth 1946). In the colorimetric and electrometric methods, CA activity is determined from the hydration of C O2, measured by the rate of pH shift using either indicators including phenol red and bromothymol blue (Waygood 1955; Datta and Shepard 1959) or pH electrodes, respectively (Wilbur and Anderson 1948). In the latter two methods, when C O2—saturated
13
Marine Biology (2018) 165:88
water (0 °C) is mixed with an ice-cold alkaline buffer (e.g., Veronal, pH 7.95–8.15), the pH drops rapidly because of the catalyzed hydration of C O2, resulting in the production of one proton. These two methods were the most commonly used in the early CA studies with both micro- and macroalgal species (Nelson et al. 1969; Bowes 1969; Atkins et al. 1972; Graham and Smillie 1976). More recently, other methods have been developed to estimate CA activity such as mass spectroscopy, isotope disequilibrium, fluorometric technique, modified pH drift method, and the use of CA-specific inhibitors (Mercado et al. 1997a; van Hille 2001; Rost et al. 2007; van Hille et al. 2014, Mustaffa et al. 2017). However, most of these CAs have been indirectly estimated from, e.g., the utilization of HCO3−, fluxes of H CO3− and CO2 across the plasmalemma, or decreases in photosynthesis when specific CA inhibitors are applied. However, the electrometric method is the only assay that provides a direct quantitative estimation of CA activity in algae. Interest in measuring CA activities in macroalgae has substantially increased in the past 10 years because of its central role in the acclimation and adaptation to ocean acidification (OA) (Cornwall et al. 2017; Raven et al. 2017; Griffiths et al. 2017; Fernández et al. 2015; Rautenberger et al. 2015). By 2100, seawater pH will decrease by about 0.4 units with a concomitant increase in C O2 and H CO3− by 200 and 9%, respectively (The Royal Society 2005; IPCC 2013). These significant changes in both Ci availability and the proportion of each Ci form may trigger, enhance, or down-regulate different Ci acquisition mechanisms, which will affect the photosynthetic and growth responses of many marine macroalgae to OA (Hepburn et al. 2011; Mackey et al. 2015). CAext is the main mechanism for Ci acquisition in macroalgae such Saccarina latissima, S. japonica, Sargassum henslowianum, Sargassum fusiforme [= Hizikia fusiforme], and Gracilaria sp. (Andría et al. 1999; Yue et al. 2001; Zou et al. 2003, 2011; Bi and Zhou 2016). With the projected changes in C O2 and H CO3− concentrations for the year 2100, some studies have shown a decrease in the C Aext activity, e.g., by 50% in S. fusiforme (Zou et al. 2003), 17% in Laminaria digitata (Klenell et al. 2004), and about 40–60% in Alaria esculenta and S. latissima (Gordillo et al. 2015). Conversely, CAext may be either up-regulated (Olischläger et al. 2012) or remained unchanged under high CO2 concentrations (Olischläger and Wiencke 2013). These contrasting results suggest that the responses of C Aext to OA are species-specific and may even depend on the experimental conditions under which CA was measured. Most assays determining the activities of CA in macroalgae today are based on the electrometric Wilbur–Anderson method (Wilbur and Anderson 1948). However, modifications such as changes in the chemical composition of the buffer (e.g., type of buffer, components, molarity, and pH) and those in the reaction time have been introduced to the original electrometric
Page 3 of 12
Marine Biology (2018) 165:88
method (Table 1) (Bowes 1969; Graham and Smillie 1976; Miller and Colman 1980; Ramazanov and Semenenko 1988; Giordano and Maberly 1989; Haglund et al. 1992a). Our review of the literature (Table 1) reveals that a large range of assay conditions have been used in different studies, but these have not been optimized for each species. Selecting one assay over another without optimization puts the accuracy of the results into question. The giant kelp Macrocystis pyrifera (hereafter called Macrocystis) is widely distributed along the northern and southern Pacific seaboard, forming extensive underwater forests that provide habitat, food, and shelter to numerous marine organisms, including commercially important fish and invertebrates (Steneck et al. 2002; Graham et al. 2007). Despite its significance in coastal marine ecosystems, surprisingly little was known about its mechanisms of Ci acquisition until two mechanisms were recently identified by Fernández et al. (2014): (1) via the external conversion of H CO3− using C Aext − and (2) direct HCO3 uptake through an anion exchange (AE) port. Although CA activities (only total CA) had been previously measured in Macrocystis (Huovinen et al. 2007; Rothäusler et al. 2011), using a modified Wilbur–Anderson method (Haglund et al. 1992a), it has not been optimized for this species, which makes the comparison between these studies difficult. The understanding of Ci uptake mechanisms in macroalgae under current seawater conditions is critical for predicting the effects of OA on their photosynthesis and growth (e.g., Fernández et al. 2014; Cornwall et al. 2017). Here, we report the steps that we undertook to optimize (i.e., reduce the variation between measurements) the Wilbur–Anderson method for Macrocystis to measure the activities of both C Aext and C Aint. Our comprehensive literature review (Table 1) revealed the most commonly used assay conditions, and we used this as the basis to test the effects of different buffer solutions, their components, and reaction times on the extraction and measurement of CAs in Macrocystis. As a result, an optimized CA assay was developed for Macrocystis sporophytes. The final assay mixture consists of a Tris–HCl buffer (pH 8.5) containing both polyvinylpyrrolidone (PVP) and dithiothreitol (DTT); and the time required for a pH change of 0.4 units (reaction time) in a catalyzed and uncatalyzed reaction was always measured in a pH range from 8.3 to 7.9. This assay can also be adapted to other (brown) macroalgae.
Materials and methods Seaweed collection Adult sporophytes of Macrocystis were collected during low tide from the upper subtidal in Aromoana (45°47′S, 170°43′E), Otago Harbour, in New Zealand between October
88
2012 and March 2013. From 11 individuals (n = 11), the first pneumatocyst-bearing lamina below the apical scimitar were taken. The collected blades were kept moist and dark in an insulated container for transport to the laboratory, 40 min away. In the laboratory, the blades were gently cleaned of any visible epiphytes and rinsed with filtered (0.5 µm pore size) natural seawater (NSW).
CA method optimization For the optimization of the CA assay, 0.8–1.0 g of fresh tissue, which was excised 2 cm above the base of the collected blades (meristematic zone), was used. The tissue was frozen in liquid N2 and ground to a fine powder using chilled mortar and pestle. Sub-samples of 0.06–0.08 g were stored at − 80 °C and utilized within a week for the optimization of the CA assay described below. After protocol optimization, the precision of the method was determined using different individual Macrocystis sporophytes (n = 7). A disc of 0.06–0.08 g was excised from each individual Macrocystis blade, and was frozen in liquid N2 and separately stored at − 80 °C for 3 days until subsequent analyses of total, external, and internal CA activities. Assay optimization The protocol for measur ing total CA activity (CAtotal = CAext + CAint) in Macrocystis was optimized using a modified version of the electrometric method described by Wilbur and Anderson (1948) (Graham and Smillie 1976; Haglund et al. 1992a; Table 1). A step-wise optimization was made as described below. The three buffer solutions commonly used in CA studies are Veronal, Tris, and phosphate (see Table 1). Veronal buffer is difficult to obtain, because it is highly regulated due to its hypnotic properties. Therefore, Tris–HCl and phosphate (Na2HPO4/NaH2PO4) as the most common buffers were used to determine their effect on CA activity. The time required for a pH change of 0.4 units (reaction time) was measured in a pH range from 8.3 to 7.9 after Giordano and Maberly (1989) and Haglund et al. (1992a). PVP (0.3% w/v) and DTT (2 mM) were added to the buffers (Table 2) to avoid interferences by phenolic compounds and oxidations of CA in the extract, respectively. PVP is commonly used to absorb the polyphenolic compounds in brown macroalgae and DTT prevents protein thiols from the irreversible oxidation, resulting in structural changes (Graham and Smillie 1976; Hurd et al. 1995; Toth and Pavia 2001). The molarity of the phosphate buffer was increased from 25 mM [as recommended by Miller and Colman (1980)] to 200 mM, because the addition of PVP, DTT, and EDTA to a 25 mM phosphate buffer led to a substantial drop in pH from 8.3 to 7.4 (data not shown).
13
13
Wilbur and Anderson (1948) Bowes (1969) Atkins et al. (1972)†
Graham and Smillie (1976)
Miller and Colman (1980) Ramazanov and Semenenko (1988) Giordano and Maberly (1989)
Haglund and Pedersén (1992)
Haglund et al. (1992a) Haglund et al. (1992a) modified
Current study
1 2 3
4
5 6 7
8
9 10
11
Veronal K2HPO4 a Veronal b Tris–borate a Veronal b Tris–borate a K2HPO4 a,b Tris a K2HPO4 b Tris–borate a Veronal b Tris a,b Tris a,b Tris a,b Tris a,b Tris a,b Tris a,b
AA ascorbic acid, BSA Bovine serum albumin, TX triton X–100
8.15 7 8.2 8.3 8.2 8.3 8.36 8.5 8.36 8.3 8.2 8.5 8.5 8.5 9.0 9.0 8.5
Buffer used as extraction buffer for total CA extraction and also used as assay buffer
– – – – – 2c – – – 2c – 20c – – – – 0.3
‡‡ Drop of 0.1–0.2 pH units measured within of a pH range between 8.6 and 8.0
‡ Drop of 0.4 pH units measured within the pH range from 8.1 to 7.1 or ††: from 8.1 to 7.4
† Colorimetric assay
Mercaptoethanol
d
Polyclar AT (insoluble PVP)
c
b
BSA (%w/v)
TX (%v/v)
– – – 100d – 25 – 15 – 25 – 25d 25 – – – 2
– – – 1 – 5 – 5 – 5 – 5 5 5 5 5 5 – 25 – – – 25 25 25 25 25 20
–
– – –
– – – – – 0.5 – – – 0.5 – – – – – – –
– – – – – 0.1 – – – 0.1 – – – – – – –
‡ †† 8.5–7.5 8.5–8.1 8.3–7.9
‡‡
8.3–7.8 ‡ 8.2–7.8
8.0–6.3
8.0–6.3 7.0–5.0 8.0–6.3
Page 4 of 12
Buffer used as an assay buffer during CA measurements ( CAext and/or CA total)
a
AA (mM)
30 4 25 100 25 300 25 50 25 300 18 50 50 100 50 50 50
EDTA (mM)
a
DTT (mM)
pH range
PVP (%w/v)
Molarity (mM)
Buffer
pH
Reaction time
Buffer properties
Buffer properties, and pH range at which CA reaction was measured are described
Described by
#
Table 1 Electrometric methods utilized to determine CA in macroalgae
88 Marine Biology (2018) 165:88
Page 5 of 12
Marine Biology (2018) 165:88 Table 2 Extraction buffers tested on the carbonic anhydrase (CA) assay of Macrocystis pyrifera
88
Buffer
Concentration (mM)
pH
PVP (% w/v)
DTT (mM)
Na–EDTA (mM)
Ascorbic acid (mM)
Phosphate Tris–HCl
200 100
8.3 8.5
0.3 0.3
2 2
5 5
– 20
To determine the sensitivity of the buffer and effects on the reaction times, CA activity was first measured using purified CA from bovine erythrocytes (Sigma-Aldrich, St Louis, MO, USA). The activity of purified CA, expressed as Wilbur–Anderson units (WA), defined as (Tb − Ts) × T−1 s , was insensitive to different buffers, exhibiting comparable activity with either Tris–HCl or phosphate buffers, but longer uncatalyzed reaction times were observed using Tris–HCl than phosphate buffer. CA activity was then measured using Macrocystis subsamples. Frozen ground sub-samples (0.06 ± 0.02 g) were individually transferred into a 20-chilled glass vial, containing 10 mL of the extraction buffer (0–2 °C), either 200 mM phosphate (pH 8.3; n = 8) or 100 mM Tris–HCl (pH 8.5; n = 6). The ratio of algal fresh weight (FW, in g) to volume of buffer (in mL) of 0.006:1 was selected to give the highest CA activities according to a preliminary experiment. After the addition of the buffer, each chilled glass vial, containing the algal sub-samples, was vortex mixed for 20 s. To keep the extraction buffer temperature stable at 0–2 °C, the sample was dipped in ice for 5 s, for every 10 s vortex mixing. Thereafter, the temperature of the ice-cold extract was kept constant while sitting on ice and continuously stirred at 900 rpm, throughout the reaction. Temperature and pH were simultaneously measured using an ROSS electrode (Orion 8107BNUMD) coupled to Orion 3-Stars Plus pH Benchtop meter (Orion, Thermo Scientific, San Jose, CA, USA). The pH was measured in NBS scale (pHNBS). When pH stabilized at 8.3, the reaction was started by adding 5 mL of icecold CO2-saturated MilliQ water. The time taken for the pH to drop by 0.4 units, from 8.3 to 7.9, was recorded. The CO2-saturated ultrapure MilliQ water (18.3 MΩ cm) was prepared by bubbling pure C O2 1 h before the assay and kept cold in ice. During the assay, a sustained slow C O2 bubbling through a rubber tubing into the rubber stopper sealed 250 mL glass flask was maintained to keep a C O2-saturated MilliQ water. The 5 mL of C O2-saturated water was collected through a glass syringe and connected to another line of rubber tubing through the rubber stopper. The pH of the solution was monitored throughout the experiment. The relative enzyme activity (REA) was determined using Eq. 1: ( ) Tb − 1, REA = (1) Ts
where Tb and Ts are the times in seconds required to drop by 0.4 pH units in the uncatalyzed reaction (Tb, buffer without algae) and in the enzyme-catalyzed reaction of the sample (Ts), respectively. REA was standardized to the sample’s FW (REA g−1 FW). The use of Tris–HCl buffer resulted in higher total CA activities compared to the phosphate buffer (see “Results” section). Thereafter, the molarity of, and the components in, the Tris–HCl buffer was further optimized for the second and third steps of the optimization process, respectively. Comparison between 50 and 100 mM Tris–HCl buffer (pH 8.5) showed optimal activity at lower concentration (see Results); and it was used for the third step optimization, where different components (e.g., DTT and PVP) were included or excluded into the 50 mM Tris–HCl buffer (see Table 3). Several tests showed that Buffer III (50 mM Tris–HCl containing DDT, PVP, Na-EDTA, and ascorbic acid) resulted in the lowest coefficient of variation (% CV; see result) and this was used for the fourth step of the optimization process, i.e., the pH range at which the reaction time is recorded. For the fourth step of the optimization process, two of the most commonly used pH ranges were selected, i.e., the time taken for a linear pH drop by: (1) 0.4 units from pH 8.3–7.9 (Reiskind et al. 1988; Giordano and Maberly 1989; Haglund et al. 1992a; Mercado et al. 1999, 2001, 2002; Table 1) and (2) 1.0 unit drop from 8.3 to 7.3 (Mercado et al. 1997b, 1998; Gómez et al. 1998; Kevekordes et al. 2006; Table 1). The time taken for a pH change from 8.3 to 7.9 was first recorded. From the same sample, the pH was allowed to drop further to 7.3, and the total time for a pH change from 8.3 to 7.3 was recorded. The pH range 8.3–7.9 was optimal (see “Results” section) and used for the subsequent analyses (the same used in the previous determinations).
Table 3 Components in a 50 mM Tris–HCl (pH 8.5) extraction buffer to evaluate their effects on the carbonic anhydrase (CA) assay of Macrocystis pyrifera Medium
DTT (mM)
PVP (% w/v)
Na–EDTA Ascor(mM) bic acid (mM)
Buffer I Buffer II Buffer III
– – 2
– 0.3 0.3
5 5 5
15 15 15
13
88
Page 6 of 12
External and internal CA activity measurements After the protocol for measuring CAtotal was optimized, CAext and CAint activity were measured using the optimized method described above. From seven blades, each from an individual Macrocystis sporophyte, one disc of 0.074 ± 0.001 g FW was excised; the same disc was used for both CAext and C Aint measurements, and C Atotal activity was calculated from the sum of CAext and CAint. A preliminary experiment showed no statistical differences between CAtotal activity measured from one disc (14.05 ± 3.46 REA g−1 FW, n = 4) and calculated from the sum of both CA activities (5.28 ± 2.04 REA g−1 FW CAext + 10.01 ± 1.34 REA g−1 FW CAint, n = 4) (Student’s t test, t = − 0.902, df = 7, P = 0.397). For CAext measurements, whole frozen discs (average weight = 0.074 ± 0.001 g FW) were individually rinsed with MilliQ water for 10 s and transferred into a 20 mL glass vial, containing 10 mL of extraction buffer. The enzymatic reaction was started by adding the C O2-saturated MilliQ water. The CA activity was measured by recording the reaction time required to lower the pH by 0.4 units as optimized above. After CAext was measured, each disc was ground to fine powder and used for measuring C Aint activity. Individually pre-chilled vials were used for each measurement to avoid interference with the subsequent CA measurements. CA activities measured using the optimized protocol were compared to the first CA activities obtained using a nonoptimized method (buffer without DDT and PVP). The % CV between samples was compared.
Marine Biology (2018) 165:88
REA g−1 FW, and 4.01 ± 1.69 REA g−1 FW, respectively (Student’s t test, t = 4.409, df = 12, P =
