Evidence for an Inorganic Carbon-Concentrating Mechanism in the Symbiotic Dinoflagellate Symbiodinium sp.

Abstract

The presence of a carbon-concentrating mechanism in the symbiotic dinoflagellate Symbiodinium sp. was investigated. Its existence was postulated to explain how these algae fix inorganic carbon (C_i) efficiently despite the presence of a form II Rubisco. When the dinoflagellates were isolated from their host, the giant clam (Tridacna gigas), CO₂ uptake was found to support the majority of net photosynthesis (45%–80%) at pH 8.0; however, 2 d after isolation this decreased to 5% to 65%, with HCO₃⁻ uptake supporting 35% to 95% of net photosynthesis. Measurements of intracellular C_i concentrations showed that levels inside the cell were between two and seven times what would be expected from passive diffusion of C_i into the cell. Symbiodinium also exhibits a distinct light-activated intracellular carbonic anhydrase activity. This, coupled with elevated intracellular C_i and the ability to utilize both CO₂ and HCO₃⁻ from the medium, suggests that Symbiodinium sp. does possess a carbon-concentrating mechanism. However, intracellular C_i levels are not as large as might be expected of an alga utilizing a form II Rubisco with a poor affinity for CO₂.

Dinoflagellates of the genus Symbiodinium (=zooxanthellae) are known for their role in a number of symbiotic associations with mainly tropical marine invertebrates, including corals, clams, and sea anemones (Trench, 1987). The algae are either intra- or intercellular and generally associated with the digestive system of the host. The host therefore has a major influence on the supply of inorganic carbon (C_i) to the symbiont. Following carbon fixation by the zooxanthellae, much of the photosynthate is exported to the host and can contribute up to 100% of the host's energy requirements (Klumpp et al., 1992). The supply and fixation of carbon therefore has a major influence on the symbiosis.

In the giant clam (Tridacna gigas) zooxanthellae are found in tubules emanating from the stomach (Norton et al., 1992). These are in close proximity to the hemal sinuses, which contain hemolymph, the clam's blood supply. The hemolymph is the immediate source of nutrients for the zooxanthellae and its composition is affected by the photosynthetic rate of the dinoflagellates resulting in a diurnal variation in a number of parameters (Fitt et al., 1995, D. Yellowlees, personal communication). Thus, during photosynthesis the hemolymph [C_i] can drop from 1.8 to 0.8 mm, with a concomitant increase in pH from 7.3 in the dark to 8.2 at high light levels. During high rates of photosynthesis the hemolymph is supersaturated with O₂, as bubbles are present in hemolymph samples removed from the sinuses. The fluctuations in C_i and pH are probably greater in the tubules themselves, but no measurements have been reported to date.

Recently, we reported that Symbiodinium sp. possesses a form II Rubisco (Whitney et al., 1995), which had previously been reported only in prokaryotic anaerobic, non-sulfur purple bacteria. Like other form II enzymes, dinoflagellate Rubisco has a relatively low discrimination ratio (S_rel) between CO₂ and O₂ (Jordon and Ogren, 1981). Apart from dinoflagellates, all form II enzymes are found in anaerobic bacteria, in which a low S_rel value has no physiological significance. Whitney and Andrews (1998) reported an S_rel of approximately 35 for the form II Rubisco from Amphidinium carterae, a related free-living dinoflagellate. While this is the highest reported S_rel for a form II Rubisco, it is still 40% lower than any form I enzyme. Whitney and Andrews (1998) concluded that this S_rel value would allow dinoflagellates to maintain a positive photosynthetic carbon balance; however, the ratio of oxygenation to carboxylation would utilize light energy very inefficiently. This raises the question of how do dinoflagellates, with a form II Rubisco, survive in an aerobic environment and, in the case of Symbiodinium sp., export significant amounts of photosynthate? One possible mechanism for overcoming the limitations of a form II Rubisco in a potentially unfavorable CO₂/O₂ ratio environment would be the utilization of a carbon-concentrating mechanism (CCM). This would increase internal CO₂ concentrations and minimize the effect of the oxygenation reaction of Rubisco.

A significant number of algae and cyanobacteria have been shown to actively accumulate C_i internally by utilizing a CCM (for review, see Badger et al., 1998). These elevated internal C_i levels allow algae to grow in C_i-limiting environments, produce higher carbon fixation rates, and also reduce the energetically wasteful oxygenation reaction of Rubisco.

There is circumstantial evidence that Symbiodinium sp. does possess a CCM; it has both internal and external carbonic anhydrase (CA) (Yellowlees et al., 1993) and appears capable of both HCO₃⁻ and CO₂ utilization. Zooxanthellae isolated from corals utilize predominantly HCO₃⁻ (Goiran et al., 1996), while those from giant clams appear to utilize CO₂ (Yellowlees et al., 1993). Whether this is due to the different environment within the host or different zooxanthellae strains is not known. Zooxanthellae also exhibit changes in photosynthetic characteristics after isolation from a host, with a decrease in both P_max and K_0.5 photosynthesis, suggesting that there might be changes in the C_i supply to Rubisco (W. Leggat, personal observation).

In addition, the fresh water dinoflagellate Peridinium gatunense has been found to acquire a CCM under C_i-limiting conditions and can maintain internal C_i concentrations between 7- and 80-fold above external levels (Berman-Frank and Erez, 1996; Berman-Frank et al., 1998).

This study was designed to determine if the symbiotic dinoflagellate possesses a CCM and whether the characteristics of the CCM change with time after isolation from the giant clam. We report here the results of our study into the uptake and accumulation of C_i by Symbiodinium sp. The results indicate that zooxanthellae do possess a CCM; however, intracellular C_i concentrations are not as high as might be expected of an alga utilizing a form II Rubisco for photosynthetic carbon fixation.

MATERIALS AND METHODS

Isolation of Zooxanthellae and Culturing of Algae

Giant clams (Tridacna gigas) were obtained from the Australian Centre for International Agricultural Research Giant Clam Project (James Cook University Orpheus Island Research Station, Queensland, Australia) and transported to the open-air aquarium at James Cook University (Townsville). Clams were acclimatized there for at least 3 weeks before experiments were commenced. Clams were then either killed at Townsville or flown to the Australian National University (Canberra, ACT, Australia) before being killed. Zooxanthellae were isolated by blending the mantle of a freshly killed clam in 0.45 μm of filtered seawater. The homogenate was then strained through two layers of cheesecloth and zooxanthellae pelleted by centrifugation at 600g for 2.5 min at 25°C. The zooxanthellae were then washed four more times in filtered seawater before they were suspended in filtered seawater at a cell density of approximately 2 × 10⁶ cells mL⁻¹ and cultured with a photon flux density of 100 μE m⁻² s⁻¹.

A culture of Amphidinium carterae (CS-21) was obtained from the Commonwealth Scientific and Industrial Research Organization Culture Collection of Microalgae (Hobart, Australia) and grown in G media (Loeblich, 1975) at 25°C at a photon flux density of 100 μE m⁻² s⁻¹. The algae were harvested during log-phase growth.

MS Measurements

Steady-State Photosynthesis and C_i Fluxes

All experiments were conducted at 28°C in CO₂-free artificial seawater medium containing 428 mm NaCl and 25 mm 1,3-bis(Tris[hydroxymethyl]methylamino) propane (BTP) (pH 7.0 or 8.0). Experiments were conducted as previously described in Badger et al. (1994). An O₂ electrode chamber was connected to a mass spectrometer via a gas-permeable membrane. The mass spectrometer was sequentially focused on masses 44 (CO₂) and 32 (O₂), and the changes in the concentrations recorded. Estimations were made of the HCO₃⁻ concentration by calibrating at acidic, buffered, and basic pH. Estimates of the flux of CO₂, O₂, and HCO₃⁻ into the cell were made using the equations of Badger et al. (1994).

Zooxanthellae were pelleted at 600g for 2.5 min, and resuspended in CO₂-free artificial seawater medium, so that the Chl a concentration was between 4 and 10 μg Chl a mL⁻¹, as determined by the method of Jeffrey and Humphrey (1975). Artificial seawater medium (5 mL, pH 7.0 or 8.0) was placed in the electrode chamber with 5 μL of acetazolamide (final concentration in cuvette of 50 μm), and 200 μL of zooxanthellae suspension was added. The cuvette was then illuminated (approximately 500 μE m⁻² s⁻¹) for 2 min to acclimatize the algae before the light was shut off and the cuvette was purged with N₂ until the [O₂] was approximately 100 μm. NaHCO₃ was then added and dark measurements taken until equilibrium was achieved. The cells were then illuminated at a saturating light level (500 μE m⁻² s⁻¹) (Chang et al., 1983; Iglesias-Prieto and Trench, 1994) and readings taken until steady state was reached. Dark-light cycles were repeated with increasing HCO₃⁻ concentrations.

H¹³C¹⁸O₃⁻ Exchange

These experiments were also conducted in the mass spectrometer using methods similar to that of Palmqvist et al. (1995). This method gives qualitative information about the presence of a CCM. The loss of ¹⁸O from CO₂ was measured by monitoring the CO₂ masses 49 (¹³C¹⁸O₂), 47 (¹³C¹⁸O¹⁶O), 45 (¹³C¹⁶O₂), and 44 (CO₂). The log enrichment of the ¹⁸O fraction in ¹³C¹⁸O₂ was calculated using the equations of Palmqvist et al. (1995):

Experiments were conducted in 4 mL of artificial seawater medium (pH 8.0) to which was added H¹³C¹⁸O₃⁻ (final concentration 1 mm), and the uncatalyzed exchange was allowed to equilibrate (2 min). Zooxanthellae (100 μL) was then added (approximately 5 μg Chl a mL⁻¹) and left in the dark for 3 min until equilibrium was established. They were then exposed to light for 3 min (500 μE m⁻² s⁻¹), followed by 5 min of darkness. In acetazolamide (AZA) and ethoxyzolamide (EZA) treatments, the inhibitor was added before the labeled bicarbonate to a final concentration of 50 and 500 μm respectively.

Photosynthetic O₂ Exchange

All experiments were conducted at 28°C in CO₂-free artificial seawater medium containing 428 mm NaCl and 25 mm BTP (pH 8.0). This media was maintained CO₂ free by purging with CO₂-free air, and was degassed under vacuum prior to use in the O₂-exchange assays. Cells were maintained in a concentrated suspension at room temperature prior to use. O₂-exchange assays were conducted in a 4-mL cuvette attached to a mass spectrometer via a Teflon semipermeable membrane. A similar method and calculations have been previously described (Canvin et al., 1980; Furbank et al., 1982). The assay and measurements involve the introduction of ¹⁸O₂ into reaction medium that has been depleted of ¹⁶O₂. This was done after the introduction of cells, using a small bubble of ¹⁸O₂ above the reaction medium. During both the dark and light periods, changes in mass 32 (¹⁶O₂) and mass 36 (¹⁸O₂) were continuously monitored, and the rate of change in the concentration of these species was used to calculate gross O₂ evolution, gross O₂ uptake, and net O₂ evolution. Assays were conducted with a cell density of 2 to 4 μg Chl a mL⁻¹, and an O₂ concentration of 250 to 350 μm (higher at the end of the experiment due to net O₂ evolution). Light was provided at the top of the cuvette at 500 μE m⁻² s⁻¹ through a fiber-optic light source. Experiments were performed by adding cells to the cuvette and waiting for a steady-state dark value. Light was then switched on and C_i was added sequentially, allowing 4 to 7 min for a steady-state rate to be achieved at each C_i concentration.

Silicone Oil Centrifugation

This experiment used a method adapted from that of Badger et al. (1980). Freshly isolated or cultured zooxanthellae or A. carterae were pelleted (200g) and resuspended at a density of approximately 15 μg Chl a mL⁻¹ in a solution of 25 mm BTP and 428 mm NaCl (pH 7.0 or 8.0) that had been bubbled with CO₂-free air for 2 d. The zooxanthellae suspension was then placed in an O₂ electrode chamber (Hansatech Instruments, King's Lynn, UK) at light levels of 500 μE m⁻² s⁻¹ until O₂ production had ceased and the cells had reached their CO₂ compensation point. Killing solution (20 μL of 2 n KOH and 10% MeOH) was added to 400-μL plastic microfuge tubes (Eppendorf Scientific, Westbury, NY), and 50 μL of silicone oil (approximately 2:1, 200/20, Wacker Chemie, Munich) was overlaid. The zooxanthellae suspension (200 μL) was then added.

The tubes were illuminated at a photon flux density of approximately 500 μE m⁻² s⁻¹. H¹⁴CO₃⁻ of known specific activity was added so that the final concentration was between 20 and 2,000 μm. Cells were incubated at each C_i concentration for 15 to 20 s, after which the tubes were centrifuged at 15,000g for 15 s to pellet the zooxanthellae. The tubes were snap-frozen and the bottom layer containing the zooxanthellae and the killing solution cut from the tube and resuspended in 500 μL of 0.1 n NaOH. The resuspended solution (200 μL) was added to 200 μL of 0.1 n NaOH or 200 μL of 0.2 n HCl. The acidic sample was placed in a fume hood and heated to 60°C for 1 h so that any unfixed ¹⁴C was evolved. BSC scintillation fluid (4 mL, Amersham-Pharmacia Biotech, Uppsala) was added to both the acid and basic samples and the ¹⁴C counted in a liquid scintillation counter (Wallac 1410, EG&G Wallac, Turku, Finland). The basic samples were assumed to contain both the fixed and unfixed ¹⁴C in the cell, while the acidic sample contained only the fixed C_i. Corrections were made for any extracellular ¹⁴C.

Calculation of Intracellular Volume and pH

Total extracellular and intracellular volumes were calculated using ³H₂O and either [¹⁴C]mannitol or [¹⁴C]dextran (ICN, Costa Mesa, CA). Intracellular pH was measured using 5,5-dimethyl-[2-¹⁴C]oxazolidine 2,4-dione (Badger et al., 1980).

RESULTS

Measurement of Steady-State Photosynthesis and C_i Fluxes before and after Isolation

An initial characterization of C_i uptake processes in the zooxanthellae can be approached by deriving estimates for the ability of the cells to utilize both CO₂ and HCO₃⁻ as carbon sources for photosynthesis. Mass spectrometry techniques (Badger et al., 1994) enable measurements of the net fluxes of CO₂, HCO₃⁻, and O₂ into and out of the cell under steady-state photosynthesis conditions. This approach shows that freshly isolated zooxanthellae assayed at pH 8.0 predominantly take up CO₂ from the external media but are also capable of some HCO₃⁻ uptake (Fig. 1). Net CO₂ uptake is able to support between 45% and 80% of net photosynthesis, with the contribution increasing at higher levels of C_i. Conversely, the contribution of net HCO₃⁻ uptake declines from 55% to 20% over this same range.

Figure 1.

Comparison of net photosynthetic rate (●, ○), CO₂ uptake (▾, ▿), and HCO₃⁻ uptake (▪, □) for Symbiodinium sp. freshly isolated from the giant clam T. gigas (black symbols) and a 2-d-old culture (white symbols) with differing C_i concentrations. Assays were conducted in 25 mm BTP (pH 8.0), 428 mm NaCl, and 50 μm AZA at 28^OC with a photon flux density of 500 μE m⁻² s⁻¹. Data were collected using the mass spectrometry disequilibrium technique described in “Materials and Methods.”

After 2 d of isolation, the rates and patterns of C_i uptake changed significantly to increase the capacity for HCO₃⁻ uptake. The contribution of net CO₂ uptake declined to 5% to 65%, while HCO₃⁻ increased to 95% to 35% over the same range of C_i concentrations. This change in HCO₃⁻ uptake indicates the induction of a HCO₃⁻ uptake system that allows utilization of a greater proportion of the available C_i, particularly at limiting [C_i]. Following isolation, the maximum net photosynthetic rate decreased from 300 ± 20 to 260 ± 10 μmol mg⁻¹ Chl a h⁻¹ (Fig. 1), while the K_0.5 (C_i) was found to be 640 ± 100 and 500 ± 50 μm, respectively. A similar, although larger decrease has previously been observed after isolation (W. Leggat, personal observation).

Flux measurements at pH 7.0 of zooxanthellae isolated for 1 d (Fig. 2) show that there is less HCO₃⁻ uptake capacity at this pH, where CO₂ is a more dominant species. Although HCO₃⁻ does support 40% of photosynthesis at the most limiting [C_i], this rapidly declines as C_i increases. At half-saturating [C_i], the net HCO₃⁻ uptake supports only 10% of net photosynthesis. The K_0.5(C_i) at pH 7.0 was 99 ± 9 μm compared with 433 ± 41 μm at pH 8.0. This indicates that K_0.5(CO₂) declines from around 16 μm to 9 μm from pH 7.0 to 8.0, supporting the occurrence of some increased HCO₃⁻ uptake, but clearly indicating that CO₂ uptake plays a significant role in photosynthesis at both pH values. The net photosynthesis (P_max) declined at pH 7.0 from 230 ± 10 (pH 8.0) to 160 ± 5 μmol mg⁻¹ Chl a h⁻¹. The negative HCO₃⁻ uptake values at pH 7.0 in this figure are most likely an artifact due to the errors involved in the calculation of this value. The main significance of this is that HCO₃⁻ uptake is low compared with CO₂ uptake.

Figure 2.

Comparison of net photosynthetic rates (●, ○), CO₂ uptake (▾, ▿), and HCO₃⁻ uptake (▪, □) for Symbiodinium sp. at pH 7.0 (black symbols) and pH 8.0 (white symbols) after being in culture for 1 d. Assays were conducted in 25 mm BTP and 428 mm NaCl with 50 μm AZA at 28°C with a photon flux density of 500 μE m⁻² s⁻¹. Data was collected using the mass spectrometry disequilibrium technique described in “Materials and Methods.”

Light-Stimulated CA Activity

In searching for evidence for the operation of a CCM, one of the key processes common to both algae and cyanobacteria is active C_i transport, taking both CO₂ and HCO₃⁻ from outside the cell and placing it in contact with localized regions of CA inside the cell. This enables the generation of CO₂ from accumulated HCO₃⁻ so that [CO₂] can be elevated around Rubisco (Badger and Price, 1992). A robust way that this can be measured is through monitoring light-stimulated inorganic exchange processes using ¹⁸O-enriched C_i species (Palmqvist et al., 1994). The active uptake of C_i species promotes the loss of ¹⁸O to unlabeled water due to internal CA activity, and this can be measured by examining the changes in isotopic enrichment of the CO₂ species.

Figure 3 shows the results of such experiments. Control experiments and those with 50 μm AZA showed similar patterns (Fig. 3). The initial rapid decrease in enrichment with the addition of zooxanthellae is due to access of external CO₂ to internal CA. This rate declines as steady-state equilibrium is reached. Upon illumination, there was a distinct light-stimulated decline in enrichment due to light-activated CA activity. AZA (50 μm), which inhibits external CA (Miyachi et al., 1983; Moroney et al., 1985), did not have any effect on the enrichment pattern in the dark or the light, suggesting that there may be little external CA activity associated with these cells, and that it does not inhibit active uptake of C_i.

Figure 3.

Effect of dark/light periods and CA inhibitors on the log enrichment (percentage ¹³C¹⁸O₂) of C_i in the medium during assays of Symbiodinium sp. Assays were conducted in a water jacketed cuvette (28°C) connected to a mass spectrometer. Freshly isolated cells were added to the cuvette (final concentration of 0.125 μg Chl a mL⁻¹) containing 4 mL of CO₂-free media (25 mm BTP, pH 8.0, and 428 mm NaCl) and H¹³C¹⁸O₃⁻ (1 mm) that had achieved chemical equilibrium. After 3 min of darkness, the cuvette was illuminated at a photon flux density of 500 μE m⁻² s⁻¹ for 3 min, followed by another 5 min of darkness, as indicated by the bar at the top of the graph. Where appropriate, the CA inhibitors AZA (▿; final concentration 50 μm) or EZA (+; final concentration 500 μm) were added prior to the addition of the cells. Data are presented as log enrichment as per the formula detailed in “Materials and Methods.”

However, 500 μm EZA, which inhibits both internal and external CA, eliminated light-stimulated CA activity and active C_i uptake (Fig. 3). Patterns similar to those found with AZA and EZA have been observed in Chlamydomonas reinhardtii and Scenedesmus obliquus (Palmqvist et al., 1995). The effects of EZA are most readily interpreted as being due to an inhibition of both C_i transport processes and internal CA. This leads to a situation in which external C_i species are in passive equilibrium with the internal compartment. When light activates photosynthesis under these conditions, there is a competition between Rubisco and C_i hydration processes for CO₂, which actually reduces the exchange of label from CO₂, causing the enrichment to rise rather than fall. This pattern is also typical of algal species that appear to lack a CCM (Palmqvist et al., 1995). The rapid decline in the dark in the presence of EZA is somewhat anomalous and not readily explained. However, this feature was present in all experiments that were conducted using EZA as an inhibitor.

Similar experiments were also conducted on zooxanthellae that had been cultured for 2 d, and the results obtained were similar to those of freshly isolated zooxanthellae (data not shown).

Silicone Oil Centrifugation

Although mass spectrometry allows estimates to be made of C_i uptake, it does not provide information about the intracellular C_i concentration (C_int). However, silicone oil centrifugation provides estimates of both the intracellular pool size and the pH, which influences the equilibrium concentration of C_i within the cell. When measurements were made of the C_int and fixed carbon at both pH 7.0 and 8.0, the P_max at pH 7.0 and 8.0 was 144 ± 7 and 142 ± 9 μmol C fixed mg⁻¹ Chl a h⁻¹, respectively, while the K_0.5 (C_i) was 90 ± 10 and 600 ± 100 μm (data not shown). The internal pH of the zooxanthellae was estimated as 7.62 ± 0.09, while the total intracellular volume of the cells in the assay was 0.12 ± 0.05 μL. At pH 7.0, freshly isolated zooxanthellae had an internal C_i/external C_i (C_int/C_ext) between 1.5 and 3.2 times what would be expected if only passive diffusion of C_i into the cells had occurred (Fig. 4; Table I). At pH 8.0 the results were similar, with the C_int between 2.2 and 4.6 times what would be expected after passive diffusion (Fig. 4, Table I). C_int/C_ext was measured for 5 d after isolation, over this period C_int/C_ext increased to a maximum of 24.2 (7 times passive diffusion) after 1 d and decreased to a maximum of 14.4 (4.2 times passive diffusion) 5 d after isolation (Table I).

Figure 4.

The effect of external C_i on C_int/C_ext at pH 7.0 (A) and 8.0 (B) for Symbiodinium sp. freshly isolated from T. gigas. Passive diffusion indicates the C_int/C_ext expected if only passive diffusion of CO₂ into the cell was occurring, assuming an intracellular pH of 7.62. Assays were conducted using silicone oil centrifugation in 25 mm BTP and 428 mm NaCl at a photon flux density of 500 μE m⁻² s⁻¹ using H¹⁴CO₃⁻ as a substrate as in “Materials and Methods.” Error bars represent ses (n = 3).

Table I.

Comparison of intracellular and extracellular inorganic carbon concentration of Symbiodinium sp. and A. carterae calculated using silicone oil centrifugation at differing pH values

Species	Culture Conditions	External pH	Ci External	Cint/Cext	Estimated Cint/Cext for Passive Diffusiona	Cint/Cext
						Estimated Cint/Cext
			μm
Symbiodinium sp.	Freshly isolated	8.0	200	0.9 ± 0.3	0.41	2.2 ± 0.7
			400	1.9 ± 0.2	0.41	4.6 ± 0.5
			1,000	1.7 ± 0.2	0.41	4.1 ± 0.5
			2,000	1.3 ± 0.1	0.41	3.2 ± 0.2
	Freshly isolated	7.0	20	11.0 ± 1.0	3.47	3.2 ± 0.3
			40	7.1 ± 1.1	3.47	2.0 ± 0.3
			100	5.1 ± 0.6	3.47	1.5 ± 0.2
			200	8.4 ± 0.3	3.47	2.4 ± 0.1
	1-d-old Cultureb	7.0	20	24.2 ± 5.1	3.47	7.0 ± 1.5
			40	23.0 ± 4.5	3.47	6.6 ± 1.3
			100	16.5 ± 1.4	3.47	4.8 ± 0.4
			200	12.9 ± 1.2	3.47	3.7 ± 0.3
	5-d-old Cultureb	7.0	20	14.4 ± 2.4	3.47	4.1 ± 0.1
			40	11.8 ± 1.7	3.47	3.4 ± 0.5
			100	9.0 ± 0.9	3.47	2.6 ± 0.3
			200	6.8 ± 0.7	3.47	2.0 ± 0.2
A. carterae	Culturedc	7.0	20	7.3 ± 2.6	ND	ND
			40	26.4 ± 7.7	ND	ND
			100	17.4 ± 2.2	ND	ND
			200	20.5 ± 5.8	ND	ND

Assays were conducted in 25 mm BTP and 428 mm NaCl at a photon flux density of 500 μE m⁻²s⁻¹.

^aEstimated C_int/C_ext assuming only passive diffusion of CO₂ into the cell and an intracellular pH of 7.62. 

^bSymbiodinium sp. was isolated from T. gigas and maintained in filtered seawater (0.45 μm). 

^cA. carterae was grown in G media (Loeblich, 1975) and harvested in log phase growth. 

The internal C_i was also determined for the free-living dinoflagellate A. carterae. The C_int/C_ext ratio was similar to that found for Symbiodinium, between 7.3 and 26.4 (Table I).

Photosynthetic O₂ Uptake

One of the key roles of the operation of a CCM in photosynthetic organisms is to suppress the oxygenase activity of Rubisco and thus reduce the deleterious effects of O₂ on photosynthesis. This may be particularly so for zooxanthellae possessing a form II Rubisco with low S_rel and a potentially high oxygenase activity (Whitney and Andrews, 1998). Therefore, it was of interest to examine the photosynthetic O₂ uptake associated with photosynthesis and the effects of C_i limitation on potential oxygenase activity.

Figure 5 shows the response of both gross and net O₂ fluxes to varying inorganic carbon at pH 8.0. In control cells, gross O₂ uptake near the C_i compensation point represented about 30% of the maximum O₂ evolution rate at saturating C_i. O₂ uptake was stimulated by the light, increasing from around 50 in the dark to 130 μmol mg⁻¹ Chl a h⁻¹ at the lowest C_i concentration. Increasing C_i actually stimulated O₂ uptake reactions up to around 0.5 mm C_i, increasing O₂ uptake to 45% maximum O₂ evolution. A stimulation of O₂ uptake in the light by C_i has been seen in higher plants (Canvin et al., 1980) and has been interpreted as being due to an activation of Rubisco by CO₂, leading to increased oxygenase activity. Cells were treated with EZA to inhibit CCM activity and induce increased C_i limitation. EZA-treated cells showed a decreased affinity for C_i, and the P_max photosynthesis may have also been reduced. Despite an increased C_i limitation in EZA-treated cells, O₂ uptake was actually decreased at limiting C_i compared with control cells, and again there was evidence for stimulation by increasing C_i. In a number of experiments with control cells, stimulation of O₂ uptake by increasing C_i was always observed, and the maximum O₂ uptake capacity varied between 35% to 45% of maximum O₂ evolution at saturating C_i.

Figure 5.

Photosynthetic O₂ exchange of freshly isolated Symbiodinium sp. in response to external C_i. The experiments were conducted as described in “Materials and Methods” at a cell density of 2.2 μg Chl a mL⁻¹ and a photon flux density of 500 μE m⁻² s⁻¹. Shown are values for gross O₂ evolution (Evol), gross O₂ uptake, and net O₂ evolution. C_i responses are shown for both control (black symbols) and plus 500 μm EZA (white symbols) added just before cells for each treatment, together with a value for O₂ uptake measured in the dark.

It is possible that the zooxanthellal Rubisco may have a low affinity for O₂, as is the case for cyanobacterial and non-green algal form I Rubiscos (Jordon and Ogren, 1981). Thus, the response of O₂ uptake to varying O₂ was examined and is shown in Figure 6. Both close to the C_i compensation point and at near-saturating [C_i], there was no stimulation of O₂ uptake by increasing O₂ from 0.1 to 0.5 mm.

Figure 6.

Photosynthetic O₂ exchange in freshly isolated Symbiodinium sp. in response to external O₂. The experiments were conducted as described in Figure 6 at a cell density of 4.7 μg Chl a mL⁻¹. Values for gross O₂ evolution (Evol), gross O₂ uptake, and net O₂ evolution are shown for each treatment, together with a value for O₂ uptake measured in the dark. O₂ responses are shown both in the presence of 1 mm added C_i (black symbols) and in the absence of added C_i (white symbols). Experiments were conducted from low to high O₂, and the O₂ concentration was increased between points by the introduction of a small bubble of ¹⁸O₂.

DISCUSSION

It has been hypothesized that as Symbiodinium sp. possess a form II Rubisco they must have a CCM that increases CO₂ at the site of carbon fixation. This study was designed to examine the C_i uptake and utilization of Symbiodinium sp. after isolation to determine if this hypothesis is true. We also examined the effect of culturing freshly isolated zooxanthellae in filtered seawater. This was prompted by two previous observations: that the mantle tissue of giant clams contain high levels of CA that may assist in supplying C_i to zooxanthellae in symbiosis (Yellowlees et al., 1993; Baillie and Yellowlees, 1998), and that zooxanthellae P_max and K_0.5 both decrease by approximately one-half after isolation (W. Leggat, personal observation).

Freshly isolated zooxanthellae exhibit a number of characteristics of algae that possess a CCM: they are able to utilize HCO₃⁻ for photosynthesis (Fig. 1), they show light-stimulated CA exchange (Fig. 3), EZA decreases the affinity of photosynthesis for C_i (Figs. 3 and 5), they are able to accumulate a modest amount of internal C_i in excess of a passive accumulation (Table I), and their photosynthetic O₂ uptake shows few characteristics to suggest that there is substantial Rubisco oxygenase activity at limiting C_i (Figs. 5 and 6). These features will be discussed in further detail below.

CO₂ and HCO₃⁻ Uptake

Presently there is some conjecture about what form of C_i is utilized by zooxanthellae. Studies on zooxanthellae isolated from corals found that HCO₃⁻ was the C_i species taken up mostly by zooxanthellae (Goiran et al., 1996); however, studies on Symbiodinium sp. isolated from giant clams suggest that CO₂ is preferentially utilized (Yellowlees et al., 1993). Different clades of Symbiodinium sp. are known to populate different hosts (Rowan and Powers, 1991) and different environments within the one host (Rowan and Knowlton, 1995). This may account for observed differences in the photosynthetic characteristics of the zooxanthellae isolated from different hosts. We found that zooxanthellae from T. gigas do utilize CO₂ predominantly, with net CO₂ uptake supporting 45% to 80% of net photosynthesis at pH 8.0 when first isolated (Fig. 1), and the contribution increasing at higher C_i levels. The contribution of CO₂ is even more pronounced at pH 7.0 (Fig. 2). However, 2 d after isolation, net CO₂ uptake decreased to 5% to 65% over the same C_i range, while net HCO₃⁻ uptake increased (Fig. 1). With isolation there was also a concomitant decrease in P_max by 13% and K_0.5 by 23% (Fig. 1). These changes in C_i uptake and utilization would appear to be the result of the changes associated with moving from a symbiotic to a free-living lifestyle.

Light-Stimulated CA Activity

When given labeled H¹³C¹⁸O₃⁻ as a C_i source, freshly isolated, aged, and cultured zooxanthellae displayed a light-stimulated CA activity that was inhibited by the EZA (Fig. 3). These data clearly show that light is able to stimulate the access of external C_i to internal CA activity, and is consistent with light-stimulated C_i uptake activities. Similar activities have been observed in green and non-green algae (Palmqvist et al., 1994, 1995; Badger et al., 1998) and cyanobacteria (Badger and Price, 1989) that possess CCMs.

Intracellular C_i Accumulation

The initial discovery of CCM activity in both cyanobacteria and green algae was associated with the demonstration that these cells could actively accumulate C_i in the light, and that this accumulated C_i was used to elevate internal [CO₂] around Rubisco. Thus, many attempts to demonstrate a CCM have relied on the ability to measure such accumulation in the light. The experiments conducted in this study with freshly isolated and cultured zooxanthellae (Fig. 4; Table 1) found that the internal C_i was between 1.5 and 4.6 times what would be expected if only passive diffusion of C_i into the cell was occurring. This was true at both pH 7.0 and pH 8.0. Internal C_i concentration measurements were also made for A. carterae, a non-symbiotic marine dinoflagellate, as a comparison. A. carterae was found to concentrate internal C_i approximately 25-fold more than the external C_i (Table I), which is slightly more than Symbiodinium sp.. These values are similar to those found by Burns and Beardall (1987) for other marine algae, in which C_int/C_ext values were between 5.5 and 8.3. Berman-Frank and associates have made the only other report of an intracellular pool for a dinoflagellate, finding that the P. gatunense have internal C_i levels between 7 and 80 times the external medium (Berman-Frank and Erez, 1996; Berman-Frank et al., 1998). However, this algae lives in freshwater and there were no estimates made of the intracellular pH, therefore, it is not possible to determine how much greater the C_int is compared with that which would be facilitated by passive diffusion.

It is surprising that marine dinoflagellates do not show larger accumulation ratios considering they possess a form II Rubisco that may have poor affinity for CO₂. The levels obtained for C_int were comparable to those values previously found for other marine algae that utilize a form I Rubisco (Burns and Beardall, 1987). It may be expected that dinoflagellates would have to concentrate C_i to a greater extent than other algae to overcome the limitations of a relatively inefficient Rubisco. There may be a number of explanations for this. Recently, a thylakoid CA has been found to be central to the operation of the CCM in C. reinhardtii (Karlsson et al., 1995, 1998; Funke et al., 1997), and this CA has been hypothesized to use thylakoid protons to convert HCO₃⁻ to CO₂ (Raven, 1997). If this is the case, then models of such a CCM indicate that proton-facilitated conversion of HCO₃⁻ to CO₂ may actually lead to a depletion of the internal C_i pool relative to passive accumulation, although CO₂ is still elevated within a localized region (Badger et al., 1998). Another possibility is that the measured accumulation of C_i may be in a specific region such as the stroma or even a subregion of the chloroplast. Thus, the actual level of C_i may be much higher in this subregion compared to the values estimated here on a whole-cell basis.

Due to the unstable nature of the dinoflagellate enzyme (Whitney and Yellowlees, 1995) the kinetic parameters of Symbiodinium sp. Rubisco are not known. It is therefore difficult to accurately model the CO₂ concentrations required within the cell to achieve the photosynthetic responses observed in this report. However, some estimates can be made by taking the kinetic properties of the form II Rubisco from Rhodospirillum rubrum and adjusting the K_0.5(CO₂) to achieve an S_rel of 37, which was the value recently obtained for A. carterae (Whitney and Andrews, 1998). This adjustment can be made by simply reducing the K_0.5(CO₂) to 50 to 60 μm in the presence of 21% O₂. A K_0.5(CO₂) of 16 μm at pH 7.0 (Fig. 2) could therefore theoretically be produced by a 3- to 4-fold concentration of external CO₂ within the cell, while at pH 8.0 a 6- to 7-fold concentration could produce a K_0.5(CO₂) of 9 μm (Fig. 2). More careful calculations will have to await a full kinetic characterization of dinoflagellate Rubisco.

Photosynthetic O₂ Uptake

The photosynthetic O₂ uptake displays characteristics that are not consistent with the presence of a large Rubisco oxygenase activity in these cells. Although there was considerable O₂ uptake capacity, representing some 35% to 45% of maximum O₂ evolution, it was not stimulated by increasing O₂ and was inhibited by limiting C_i concentrations (Figs. 5 and 6). The stimulation of O₂ uptake by increasing C_i may be interpreted as being due to activation of Rubisco by increasing internal CO₂, as seen with higher plants (Canvin et al., 1980), but the insensitivity to O₂ is not readily explained. Based on other form II enzymes and cyanobacterial form I Rubiscos with low S_rel values (Badger et al., 1998), it may be expected that the oxygenase activity from zooxanthellae may display a low affinity for O₂, thus exhibiting low oxygenase activity at ambient levels of O₂. The saturation of O₂ uptake by 0.1 mm O₂ at both low and high C_i concentrations is inconsistent with this and is more similar to an O₂ uptake reaction coupled to photosynthetic electron transport, such as the Mehler reaction (Badger, 1985). The photosynthetic O₂ uptake process in zooxanthellae requires further study before an adequate explanation is forthcoming.

A Zooxanthella CCM

Currently, the function of the pyrenoid in eukaryotic algae is not understood, but it has been hypothesized that it may play a role in CO₂ elevation, similar to carboxysomes in cyanobacteria. This hypothesis is based on changes in pyrenoid morphology when cells are transferred from high to low C_i conditions (Ramazanov et al., 1994), localization of Rubisco to the pyrenoid (Osafune et al., 1990), and a general correlation between the presence of pyrenoids and CCMs (Badger et al., 1998). All of these data suggest that the pyrenoid plays some role in C_i accumulation. Recently, electron microscopic examination has found that Symbiodinium sp. also has Rubisco localized within the pyrenoid (D. Yellowlees, personal communication); therefore, it is reasonable to hypothesis that zooxanthellal pyrenoids may also play a role in C_i accumulation.

Recent evidence and speculation suggest that there may be considerable diversity in the mechanistic operation of CCMs in algae. This diversity may include the extent to which a thylakoid-generated proton supply is coupled to the dehydration of HCO₃⁻, as well as the need for a pyrenoid and pyrenoid-located CA to be present (see Raven, 1997; Badger et al., 1998). The evidence presented for a zooxanthella CCM in the present study indicates the presence of active C_i transport, a dependence on internal CA for efficient photosynthesis, and a suppression of photorespiratory O₂ uptake, as well as a well-developed pyrenoid. However, we found no substantial C_i accumulation. Such considerations would favor the operation of a CCM in which C_i is elevated only in a very localized region and/or protons are used to aid conversion of HCO₃⁻ to CO₂, leading to a depressed internal HCO₃⁻ concentration. Obviously, further research must be conducted to determine what type of CCM Symbiodinium sp. possesses. This includes localization of intracellular CA and further study on the characteristics of dinoflagellate Rubisco. This is currently hindered by the extremely unstable nature of the Rubisco after isolation (Whitney and Yellowlees, 1995). Investigations into zooxanthellae isolated from different hosts may also provide further information about zooxanthellae C_i accumulation.