Introduction

There has been increasing interest in the interactions between bacteria, algae and their environment, and how those interactions influence alga morphogenesis, nutrient acquisition and community composition as well as bacterial growth and biogeochemical cycles (for example, Croft et al., 2005; Matsuo et al., 2005; Matz et al., 2008; Amin et al., 2009; Barott et al., 2011; Seyedsayamdost et al., 2011). Understanding how these interactions influence the adaption of bacteria and algae to the abiotic environment is important for understanding the physiological processes and evolutionary outcomes of those species as nearly all organisms co-occur with other species. In the present study, we identified how a beneficial interaction between certain bacterial species and the alga Chlamydomonas reinhardtii influences the ability of C. reinhardtii to survive a change in its abiotic environment that is lethal to C. reinhardtii in isolation.

C. reinhardtii, a unicellular eukaryotic microalga found in soil and aquatic environments, grows with CO2 and light, and is considered mesophilic (20–32 °C optimum). An abrupt transfer to a temperature above the typical growth range (for example, 39 °C) results in heat shock, triggering adaptive responses that may or may not allow resumption of growth or survival at the new temperature (Schroda, 2004). Prolonged exposure to incompatible temperatures results in a cessation of metabolism, chlorosis and ultimately cell death. Acclimation to high temperature relies on basal mechanisms to initiate physiological adjustments that permit expression of adaptive mechanisms required for prolonged survival at those and higher temperatures. Similar to other eukaryotes, heat shock transcription factor and heat shock proteins that facilitate proper protein folding and turnover are highly induced upon heat treatment in C. reinhardtii (Schulz-Raffelt et al., 2007).

Many algal species require exogenous cobalamin (vitamin B12), thiamine (vitamin B1) or biotin (vitamin B7) for growth (Croft et al., 2006). As cobalamin is produced only by prokaryotes (Roth et al., 1996; Warren et al., 2002), cobalamin could be acquired by scavenging it from the environment or through a mutualistic or commensal association with cobalamin-producing bacteria (Croft et al., 2005, 2006; Kazamia et al., 2012). Cobalamin is a Co2+-containing, modified tetrapyrrole that functions as a cofactor for enzymes involved in C1 and other metabolic reactions (Roth et al., 1996). In many eukaryotic algae (Croft et al., 2005; Helliwell et al., 2011), cobalamin requirements arise primarily from its use by methionine synthase, responsible for synthesizing methionine and tetrahydrofolate from homocysteine and 5-methyl-tetrahydrofolate (Roth et al., 1996; Warren et al., 2002). C. reinhardtii has only one known cobalamin-dependent enzyme, methionine synthase (METH), and the accessory proteins methionine synthase reductase and AdoCbl synthase required for S-adenosylhomocysteine reduction and adenosylcobalamin synthesis, respectively (Croft et al., 2005; Helliwell et al., 2011). Yet, C. reinhardtii also has a cobalamin-independent methionine synthase (METE) (Croft et al., 2005; Helliwell et al., 2011) that accomplishes the same reaction as METH albeit likely with lower efficiency (Gonzalez et al., 1992). Aside from avoiding dependency on cobalamin for methionine synthesis and the greater ecological flexibility it provides in environments with scarce cobalamin availability, little is known of the advantages or disadvantages of one isoform of methionine synthase over the other in contributing to the ecological success of C. reinhardtii or other algae in natural habitats.

In the course of another study, we observed that C. reinhardtii thermal tolerance was enhanced when cultures were contaminated with bacteria, which led to a series of experiments to elucidate the mechanism of this enhancement. Our data show that certain cobalamin-producing bacterial species when co-cultivated with C. reinhardtii enhance algal thermal tolerance, and that cobalamin-dependent methionine biosynthesis has a critical role in thermal tolerance enhancement. These findings will have implications for better understanding ecosystem functioning as marine algae contribute substantially to primary productivity (Field et al., 1998; Falkowski et al., 2004), and are therefore important factors in global carbon cycling. Consequently, it might be appropriate to include cobalamin availability and cobalamin-producing bacteria and their production and consumption dynamics in efforts to predict and model marine or terrestrial primary productivity as temperatures rise.

Materials and methods

C. reinhardtii and bacterial strains and growth conditions

Wild-type C. reinhardtii strains 21gr (cc1690) and 137c (cc125) were obtained from the Chlamydomonas Resource Center (http://chlamycollection.org/) and were grown phototrophically under continuous illumination (120 μE m−2 sec−1) in minimal salts medium or on agar (1.5%) solidified plates (Geraghty et al., 1990) in a 5% CO2 atmosphere. Bacterial species were grown on TY (Beringer 1974) or rhizobial minimal medium (Broughton et al., 1986) at 25 or 30 °C. Sinorhizobium meliloti wild-type 1021 and its Tn5 transposon insertion mutants cobD−, bluB− and metH− (Campbell et al., 2006; Taga et al., 2007) were generously provided by Michiko Taga (University of California-Berkeley). The cobalamin-deficient cobD− mutant was grown in TY media supplemented with 2 mg l−1 cyanocobalamin. Algal growth media were supplemented with cyanocobalamin (Cbl), methionine (Met), thiamine, biotin or folate at the indicated concentrations. S. meliloti growth media were supplemented with 25 μg ml−1 kanamycin or 20 μg ml−1 gentamycin. Culture supernatant was obtained by centrifugation of a 3-day-old stationary phase S. meliloti culture grown in rhizobial minimal medium, and decanting the culture supernatant, filter sterilizing it, drying it in a speed-vac and re-suspending it in water. The identity of air-borne bacterial contaminants of Chlamydomonas cultures that exhibited thermal tolerance enhancing properties was based on sequencing of the 16S rRNA gene (see Supplementary Information for details).

Thermal tolerance assay

Our thermal tolerance plate assays consisted of growing Chlamydomonas for 0–3 days at 25 °C in the presence or absence of bacteria, or cobalamin or methionine amendments to the culture media before transferring plates (upshift) to higher temperatures. Enhancement of thermal tolerance is defined as a delay in chlorosis at normally lethal temperatures and was assessed 5 days post temperature upshift. Chlamydomonas liquid cultures were adjusted to 1.5 × 106 cell per ml with fresh medium, based on direct microscopy counts. Sinorhizobium were washed twice and re-suspended from plate-grown cultures in a 0.85% saline solution, and serially diluted in a saline solution before mixing equal volumes of bacteria and C. reinhardtii cell suspensions to generate varying inoculum ratios; cell densities were verified by determining colony forming units. These mixtures were inoculated onto plates as co-cultures comprised of the same number of Chlamydomonas cells (5000, unless stated otherwise), but the abundance of S. meliloti cells varied. To minimize cyanocobalamin contamination from the medium, the S. meliloti cobD− mutant was sub-cultured for 1.5 days in TY broth before preparing cells for co-inoculation as described above. Alternatively, C. reinhardtii was streaked onto minimal salts medium plates before inoculating selected areas with bacteria by streaking. For determining viable cell numbers, co-cultures were removed from the plate and serially diluted before plating aliquots onto minimal salts medium or TY plates for enumeration of Chlamydomonas or Sinorhizobium, respectively. Cobalamin production was determined using a modified version of a previously described bioassay (Raux et al., 1996) and epifluorescence microscopy imaging of co-cultures are outlined in Supplementary Information.

Methionine synthase activity assays

Chlamydomonas cells were lysed by sonication in 0.1 M KPO4 buffer containing 10 mM DTT, pH 7.2 and 0.1 mM phenylmethylsulfonyl fluoride, and the mixture was centrifuged at 14 000 g for 60 min at 4 °C. The supernatant provided a crude enzyme prep for methionine synthase activity assays using the method described previously (Drummond et al., 1995), with modification. Protein concentrations were determined using the Bradford assay (Bio-Rad, Hercules, CA, USA) and 0.1–0.2 mg ml−1 BSA-equivalents of crude enzyme prep was used in each assay. Assay substrates include 0.5 mM L-homocysteine (Fisher Scientific, Pittsburgh, PA, USA) and 0.16 mM 5-methyl-tetrahydrofolate monoglutamate (Schircks Laboratories, Jona, Switzerland) for cobalamin-dependent METH or 0.16 mM 5-methyl-tetrahydrofolate triglutamate (Schircks Laboratories) for cobalamin-independent METE activity assays. Cyanocobalamin was provided at a concentration of 0.05 μM for METH activity assays. Assays were performed at 37 °C for 45 min and quenched by adding a 5-N HCl/60% formic acid solution. As the product tetrahydrofolate is unstable in acidic solutions, the reaction mixture was heated at 80 °C for 10 min to form methenyltetrahydrofolate. After centrifugation at 14 000 g for 5 min, the absorbance (350 nm) of methenyltetrahydrofolate in the reaction mixture was measured. Assay mixtures devoid of crude enzyme prep or substrates were used as blanks.

Gene silencing of METH in C. reinhardtii

Artificial miRNA oligonucleotides were designed using the Web MicroRNA Designer platform (http://wmd3.weigelworld.org/cgi-bin/webapp.cgi). The MetHmi-1F and MetHmi-1R oligonucleotides (Supplementary Table S1) targeting C. reinhardtii METH coding sequence (439–448 bp downstream of the start nucleotide base) were annealed and ligated into SpeI-digested pChlamiRNA3int as described previously (Molnar et al., 2009) to generate the artificial microRNA (amiRNA) vector pChlamiRNA3int-METH for transformation with C. reinhardtii strain CC125. Transformants were selected on TAP (Gorman and Levine, 1965) solid medium containing 15 μg ml−1 paromomycin. The mRNA levels of METH were analyzed using qRT-PCR with the primers listed in Supplementary Table S1. The knockdown lines with reduced METH mRNA levels were then examined for their ability to delay chlorosis at elevated temperatures in the absence or presence of cyanocobalamin, methionine or S. mililoti 1021. Procedures for RNA isolation for RT and qRT-PCR are described in Supplementary Information.

Results

Enhancement of thermal tolerance in C. reinhardtii by bacteria

In the course of another study, we observed that C. reinhardtii exhibited delayed chlorosis at normally lethal temperatures when contaminated with bacteria. This observation lead us to examine what features of this bacterial-Chlamydomonas interaction facilitated increased fitness at elevated temperatures. The contaminating bacterial isolates and others available in our collection were co-cultured with C. reinhardtii to assess whether enhancement of thermal tolerance (a.k.a. delayed chlorosis) was species specific using the assay described in the Materials and methods and the footnote to Table 1. Only bacteria of the α-proteobacterial lineage strongly decreased C. reinhardtii chlorosis at elevated temperatures (Table 1, Supplementary Figure S1).

Table 1 Specificity of bacterial enhancement of Chlamydomonas thermal tolerancea
Full size table

As S. meliloti 1021 (recently renamed Ensifer meliloti) most consistently enhanced Chlamydomonas thermal tolerance and this strain has been shown to benefit from its interaction with algae (Kazamia et al., 2012 and unpublished data), we used this strain to further establish the ability of bacterial interactions to delay algal chlorosis and increase survival at elevated temperatures. In these co-cultures S. meliloti forms aggregates between and in close association with C. reinhardtii cells and is not internalized by C. reinhardtii, based on epifluorescence microscopy (Supplementary Figure S2). Our visual assessment of Chlamydomonas chlorosis and survival at 42 °C was validated by C. reinhardtii 21gr chlorophyll content and viability of the co-cultures which was greatly influenced by the S. meliloti inoculum size (Figures 1a,d and e). Shifts in the relative abundances of C. reinhardtii and S. meliloti occurred quickly (Figure 1b) leading to a 5–9-fold greater abundance of S. meliloti by 8 DAI (Figure 1c) in the absence of heat stress. C. reinhardtii death in the 42 °C treatment does not appear to provide sufficient resources to promote S. meliloti growth or survival beyond the levels observed before the temperature upshift (compare Figure 1b with Figure 1d). In contrast, co-cultures with S. meliloti in the absence of thermal stress had no effect on C. reinhardtii, based on chlorophyll content (Figures 1a and e) and survival (Figure 1c), even with large initial S. meliloti inoculum sizes. Ethanol-killed S. meliloti inoculated in a 100-fold excess of C. reinhardtii did not confer thermal protection although highly concentrated sterilized culture supernatants did (Figure 1f), indicating that S. meliloti secretes a metabolite(s) or the component(s) can be released upon cell death.

Figure 1
figure 1

Enhancement of C. reinhardtii thermal tolerance by S. meliloti co-cultivation. (a) S. meliloti 1021 enhances C. reinhardtii 21gr survival at lethal temperatures in a cell density-dependent manner using a co-culture thermal tolerance assay described in the Materials and methods and the footnote to Table 1. Photographs show the extent of chlorosis 5 days after the temperature upshift. (b–d) Viability of C. reinhardtii 21gr and S. meliloti 1021 in co-cultures incubated at 25 °C 3 (b) and 8 (c) days after inoculation (DAI) and at 42 °C for 5 days (d) after the temperature upshift (8 DAI). (e) Chlorophyll content of C. reinhardtii–S. meliloti co-cultures 8 DAI. (f) Co-inoculation of C. reinhardtii 21gr with dead S. meliloti (Sm) cells (1:100 Cr 21gr:Sm 1021), un-concentrated (1X) and concentrated 12-fold (12X) culture medium (M) or S. meliloti culture supernatant (S). Values are the mean±s.e.m. of three independent experiments each comprised of three replications. A separate one-way ANOVA was performed for C. reinhardtii 21gr and S. meliloti 1021 at each temperature and sampling time. Bars with different letters are significantly different based on a Duncan’s test (P⩽0.05).

Full size image

Enhancement of C. reinhardtii thermal tolerance by vitamin B12 or methionine

Many algal species require cobalamin, thiamine or biotin for growth (Croft et al., 2006), and given that only some bacteria or Archaea species produce cobalamin this resource is possibly acquired through an algal–bacterial symbiosis (Croft et al., 2005; Helliwell et al., 2011). C. reinhardtii does not require cobalamin for growth as it has genes for both cobalamin-dependent (METH) and -independent (METE) methionine synthase (Croft et al., 2005; Helliwell et al., 2011). Yet, it’s conceivable one or more of these vitamins could affect algal growth under high temperature. Of the vitamins or related compounds examined, only cyanocobalamin (Table 2 and Figure 2a) significantly delayed C. reinhardtii chlorosis at elevated temperatures, with a minimum requirement of 100 ng l−1 (74 pMol) (Figure 2a). As cobalamin is the methyl-donating cofactor of methionine synthase METH, which catalyzes the synthesis of methionine and tetrahydrofolate from homocysteine and methyltetrahydrofolate, we included methionine and folate as well. The requirement for near mM methionine concentrations (Table 2 and Figure 2) to effectively enhance C. reinhardtii thermal tolerance likely reflects the fortuitous uptake by transporters not specific to methionine as described earlier (Kirk and Kirk, 1978). Folate had no effect on thermal tolerance. Finally, cobalamin or methionine rescue of C. reinhardtii heat sensitivity is not restricted to specific strains, but was observed in a variety of wild-type and mutant strains (Supplementary Table S2) albeit the extent of tolerance varied among strains.

Table 2 Enhancement of C. reinhardtii 21gr thermal tolerance by exogenous supplementsa
Full size table
Figure 2
figure 2

Enhancement of C. reinhardtii thermal tolerance by cyanocobalamin and methionine. (a) Thermal tolerance enhancing effect of C. reinhardtii 21gr on solid media un-amended (MM) or amended with cyanocobalamin (Cbl), methionine (Met) or both cyanocobalamin (1 mg l−1) and methionine (0.2 mg ml−1), based on the thermal tolerance assay. Photographs were taken 8 DAI. Thermal tolerance enhancing effect (TE effect) is defined in Table 1. (b) Thermal tolerance of C. reinhardtii 21gr grown in the presence of cyanocobalamin (1 mg l−1) or methionine (0.2 mg ml−1) and in the absence of a 3-day acclimation period before the temperature upshift. Solid media was un-amended or amended with cyanocobalamin (1 mg l−1) or methionine (0.2 mg ml−1). Photographs were taken 5 DAI. (c) Influence of acclimation time before the temperature upshift on thermal tolerance in the absence or presence of cyanocobalamin (1 mg l−1) or methionine (0.2 mg ml−1) for 1–3 day before the temperature upshift. Photographs were taken 5 days after the temperature upshift. (d) Survival of C. reinhardtii 21gr in liquid culture without or with cyanocobalamin (1 mg l−1), methionine (0.2 mg ml−1) or both amendments without (top) or with (bottom) 42 °C temperature upshift. Values are the mean±s.e.m. of three independent experiments each comprised of five replications.

Full size image

Chlamydomonas thermal tolerance was influenced by when cells were exposed to cobalamin or methionine (Figure 2b) and by the length of the acclimation period (Figure 2c) before a temperature upshift in a cell density-dependent manner (Figures 2b and c). In the absence of any pre-acclimation at 25 °C before a temperature upshift, planktonic cultures grown in the presence of cobalamin or methionine required higher cell densities to exhibit, albeit weak, enhanced thermal tolerance (Figure 2b). In a cell-dependent manner, the longer the acclimation time at 25 °C the better the thermal tolerance (Figure 2c). In contrast, methionine-mediated enhancement of thermal tolerance occurred only after a prolonged (3-day) acclimation period before the temperature upshift (Figure 2c). Cyanocobalamin and to a lesser extent methionine could alleviate heat sensitivity up to 44 °C, but had no effect at higher temperatures (Supplementary Table S3). Cyanocobalamin rescue of chlorosis and heat sensitivity did not rely on large temperature upshifts (for example, 25−42 °C), as we observed similar thermal tolerance enhancement with temperature upshifts from 38 to 42 °C, although chlorosis occurred more slowly in the absence of cyanocobalamin presumably owing to heat-induced metabolic adjustments or reduced metabolic activity of cells grown at higher temperatures. Collectively, these data suggest that better thermal tolerance at high cell densities may be a consequence of earlier entry into a stationary-phase physiological state.

The ability of cyanocobalamin and, to a lesser extent, methionine to enhance C. reinhardtii heat tolerance did not rely on surface growth as similar effects were attained in liquid culture (Figure 2d, Supplementary Table S4), although viability of planktonic cells more quickly diminishes in the absence of cyanocobalamin or methionine following a temperature upshift (Figure 2d) compared with surface-grown cells (Figures 2a–c). A consequence of reduced viability at 42 °C in the absence of cyanocobalamin or methionine is a dramatic reduction in biomass compared with cyanocobalamin treatments (Supplementary Table S4), indicative of scavenging cellular resources to maintain viability.

S. meliloti produced adenosylcobalamin enhances C. reinhardtii thermal tolerance

The preceding data indicated that cyanocobalamin mimics the effect of bacterial-mediated delay in chlorosis and prolonged survival of C. reinhardtii at lethal temperatures. To determine if bacterial-mediated thermal tolerance was due to cobalamin production, we selected the S. meliloti cobalamin-deficient bluB− and cobD− mutants to test this possibility. BluB synthesizes 5,6-dimethylbenzimidazole, the lower ligand of cobalamin while CobD is required for corrin ring synthesis (Campbell et al., 2006; Taga and Walker 2010). We also included a methionine auxotroph (metH− mutant) (Campbell et al., 2006; Taga and Walker, 2010). As expected, both the bluB− and cobD− mutants were dramatically impaired in delaying chlorosis or promoting C. reinhardtii survival (Figure 3). Differential survival of the bluB− and cobD− mutants is likely due to the presence of trace amounts of 5,6-dimethylbenzimidazole in the agar (Campbell et al., 2006), or C. reinhardtii exudates that facilitate bluB− mutant survival (Figure 3c), but neither strain produces sufficient adenosylcobalamin to promote C. reinhardtii heat tolerance. Despite the poor survival of the metH− mutant in the absence of exogenous methionine (Figure 3c), sufficient amounts of cobalamin was produced to facilitate survival of Chlamydomonas at moderate levels. Collectively, these results suggest that cobalamin has a critical role in bacterial-mediated enhancement of C. reinhardtii thermal tolerance. Compared with growth at 25 °C, survival of cobD− and metH− mutants alone or in co-culture with C. reinhardtii were dramatically reduced at 42 °C, implying that adenosylcobalamin and methionine production is also important for heat tolerance of S. meliloti.

Figure 3
figure 3

Adenosylcobalamin biosynthesis by S. meliloti confers thermal tolerance to C. reinhardtii. (a) Chlorosis of C. reinhardtii 21gr when cultivated with S. meliloti strain 1021 and various mutants deficient in adenosylcobalamin lower ligand (bluB−) or corrin ring (cobD−) synthesis, metH− in the thermal tolerance assay. Viability at 25 °C (b) and 42 °C (c) was determined by plate counts 8 DAI. The Cr21gr:Sm 1021 inoculum ratio was 1:100 in all comparisons. Values represent the mean±s.e.m. of two independent experiments, each comprised of three replications. A separate ANOVA was performed for C. reinhardtii 21gr and S. meliloti 1021 and bars with different letters are significantly different based on a Duncan’s test (P⩽0.05).

Full size image

This observation led us to investigate whether the inability of certain bacteria to enhance C. reinhardtii thermal tolerance was due to their inability to produce cobalamin. Many bacterial strains capable of aerobic cobalamin production were unable to enhance Chlamydomonas thermal tolerance (Table 1). Salmonella has served as a model for cobalamin synthesis (Raux et al., 1996; Roth et al., 1996; Warren et al., 2002), but only produces it under anaerobic conditions and it is unclear whether the other Enterobacteriaceae we tested also synthesize cobalamin anaerobically.

Regulation of C. reinhardtii METE and METH by temperature or bacteria

As exogenous cyanocobalamin and methionine rescues C. reinhardtii thermal sensitivity it suggests that either the expression or activity of the cobalamin-independent methionine synthase (METE) is heat-sensitive, whereas cobalamin-dependent (METH) is not. We used quantitative RT-PCR to determine whether temperature, cyanocobalamin or S. meliloti 1021 co-culture affected METE and METH expression. At a non-lethal 25 °C temperature, C. reinhardtii METE is highly expressed (Figure 4a) but in the presence of cyanocobalamin METE is strongly repressed, whereas METH expression is unaffected, which is consistent with other studies (Croft et al., 2005; Helliwell et al., 2011). Importantly, following a 1-h heat shock at 42 °C, METE expression is reduced to <10% of the control but METH expression decreases to about 30% of the control (Figure 4a). As expected the heat shock treatment dramatically induces expression of the heat shock protein gene HSP90A (Schulz-Raffelt et al., 2007) (Supplementary Figure S3).

Figure 4
figure 4

Quantitative real-time PCR of C. reinhardtii METE and METH mRNA. (a) 21gr cells were exposed to a 1-h 42 °C heat shock or maintained at 25 °C in the absence (MM) or presence of cyanocobalamin (Cbl; 1 mg l−1), methionine (Met; 0.2 mg ml−1) or both amendments (Cbl+Met). (b) C. reinhardtii 21gr grown on solid media without or with S. meliloti wild-type 1021 or bluB− mutant. The 21gr:Sm inoculation ratio was ∼1:100. Co-cultures were exposed to a temperature upshift for 2 days before isolating RNA. The constitutively expressed CBLP gene (Schloss 1990) was used as the reference gene and values are normalized relative to 21gr cells grown at 25 °C in MM medium. Bars are the mean±s.d. of two independent experiments. A separate ANOVA was performed for METE and METH and bars with different letters are statistically different based on Duncan’s test (P⩽0.05).

Full size image

We also assessed C. reinhardtii METE and METH expression during its interaction with S. meliloti wild-type and the bluB− mutant 2 days after a temperature upshift. At 25 °C, no significant differences in C. reinhardtii METE expression could be detected without or with S. meliloti wild-type or bluB− mutant co-cultivation (Figure 4b). After an upshift to 38 °C, a temperature near the maximum C. reinhardtii growth temperature, METE expression was greatly reduced in the presence of S. meliloti wild-type and the bluB− mutant, although the decrease was greater in the presence of the wild-type than the bluB− mutant. There was almost no METE detected following a 42°C upshift (Figure 4b). In contrast, following a 38 or 42 °C temperature upshift, METH expression was not significantly affected by the presence or absence of S. meliloti wild-type and the bluB− mutant (Figure 4b). These findings support conclusions inferred in another study (Kazamia et al., 2012) that C. reinhardtii METE expression can be repressed as a consequence of its interaction with a cobalamin-producing bacterial species. The difference between these two studies is the requirement for a higher, non-lethal temperature (Figure 4b) and the supplementation of a bacterial carbon source in the study by Kazamia et al. (2012).

Our methionine synthase activity assays reflect the pool of enzymes isolated from C. reinhardtii cultures exposed to a temperature upshift for varying lengths of time (Figure 5). In both the METH and METE activity assays, the production of tetrahydrofolate was measured spectrophotometrically following its conversion to the more stable methenyltetrahydrofolate. METE activity was greatly reduced in the presence of a heat stress or cyanocobalamin, whereas during heat stress METH activity in the presence of cyanocobalamin was maintained at higher levels than METE (Figure 5). When not grown in the presence of cyanocobalamin and heat stress, METE activity was high, but heat stress caused a precipitous decline in activity. Following the temperature upshift, the METH activity of cyanocobalamin-treated C. reinhardtii decreased over time although these levels were substantially greater than METE activity (Figure 5); these levels may suffice to maintain adequate methionine pools to prolong survival.

Figure 5
figure 5

Effect of temperature on C. reinhardtii methionine synthase activity. METE and METH activity of C. reinhardtii cells grown in the absence (MM) or presence of cyanocobalamin (Cbl; 1 mg l−1), methionine (Met; 0.2 mg ml−1) or both amendments (Cbl+Met). Cultures were grown for 3 days at 25 °C before a 42°C temperature upshift, except the controls were maintained at 25 °C. Crude cell extracts were used in the methionine synthase activity assays and all assays were conducted at 37 °C for 45 min, as described in the Materials and methods. Both assays measure the rate of methenyltetrahydrofolate (CH+-THF) production. Data are the mean±s.e.m. (n=3).

Full size image

Gene silencing in C. reinhardtii

To further our functional analysis of the role of METH in C. reinhardtii thermal tolerance, we took advantage of an amiRNA system to knock down METH expression. For this purpose a plasmid (pChlamiRNA3int-MetH) carrying the strong constitutive PSAD promoter (Molnar et al., 2009) was used to direct expression of an antisense cDNA amiRNA oligonucleotide targeting the 5′-end of METH mRNA. Attempts to reduce METH expression in strain 21gr transformed with this vector were unsuccessful based on screening 60 clones by qRT-PCR and in the chlorosis assay. Five (Hi-1, -2, -4, -12, -21) of the 65 paromycin-resistant transformants of wild-type CC125 screened for reduction in METH expression by quantitative RT-PCR in 3-day-old cultures cultivated at a non-lethal (25 °C) temperature exhibited 10–50% of the METH expression of the empty vector control (Figure 6a). Furthermore, a reduction in METH expression in these five lines was highly correlated with a more rapid onset of chlorosis after temperature upshift even in the presence of cyanocobalamin, methionine or S. meliloti 1021 (Figure 6b). Low levels of METH expression in the five amiRNA lines had no apparent effect on growth and survival at 25 °C in the absence of cobalamin (data not shown). Collectively these data provide strong evidence for the requirement of METH activity for the ability of cobalamin-producing bacteria to enhance C. reinhardtii thermal tolerance.

Figure 6
figure 6

Phenotypes of C. reinhardtii lines expressing artificial microRNA (amiRNA) targeting METH mRNA. Quantitative real-time PCR (a) analysis of METH mRNA in METH-amiRNA expressing lines (Hi). Values are normalized relative to wild-type CC125 cells containing an empty pChlamiRNA3int vector (cc125-int3) in 3-day-old broth cultures grown at 25 °C. Bars are the mean±s.e.m. (n=3). Bars with different letters are statistically different based on a Duncan’s test (P⩽0.05). Chlorosis phenotypes (b) of METH-amiRNA expressing cell lines following a temperature upshift in the presence of cyanocobalamin (Cbl; 1 mg l−1), methionine (Met; 0.2 mg ml−1) or S. meliloti 1021 (Sm 1021; Chlamydomonas: S. meliloti 1021 inoculum ratio of 1:100). Photographs were taken 8 DAI.

Full size image

Discussion

Our findings indicate that interactions with cobalamin-producing bacteria can have profound effects on C. reinhardtii fitness as it responds to changes in environmental temperature and that methionine biosynthesis has a central role in thermal tolerance. This highlights the importance of the presence of co-occurring species in the acclimation of an organism to changes in its abiotic environment. Although concentrated culture supernatants delaying C. reinhardtii chlorosis at high temperature (Figure 1f) implies cobalamin could be scavenged from dead bacteria (Droop, 2007), several lines of evidence suggest otherwise. First, not all cobalamin-producing bacteria are capable of enhancing C. reinhardtii thermal tolerance (Table 1) and second, cobalamin needs to be present for a sufficient length of time before the temperature upshift to be functional (Figures 2b and c). In addition, cobalamin production is regulated by environmental conditions such as oxygen in a species-dependent manner (Jeter et al., 1984). Why some bacteria capable of aerobic cobalamin production enhance C. reinhardtii thermal tolerance and others do not is unclear, but conceivably it’s a consequence of differences in bacterial thermal tolerance or ability to utilize C. reinhardtii resources that support bacterial growth under elevated temperatures. Bacteria we examined that strongly enhance C. reinhardtii thermal tolerance are members of the α-proteobacterial class, which includes the Rhizobiales order. This is consistent with other reports (Hold et al., 2001; Sapp et al., 2007) of specific bacterial lineages, including α-proteobacteria, being associated with various algal species in natural environments. These associations may reflect a certain degree of specificity, possibly a consequence of the suitability of photosynthate algae provide to heterotrophic bacteria (Kazamia et al., 2012), that is, there may be a degree of specificity in this chemical dependency.

The ecology of bacterial–algal interactions is not well understood, but recent studies have highlighted how each partner can influence the physiology of the other. For example, bacterial promotion of algal assimilation of iron in exchange for photosynthate in marine systems (Amin et al., 2009) and C. reinhardtii secretion of substances mimicking the activity of N-acyl-L-homoserine lactone signal molecules used by S. meliloti and other soil bacteria for quorum-sensing regulation of gene expression (Teplitski et al., 2004; Rajamani et al., 2008). None of the S. meliloti proteins differently regulated by C. reinhardtii mimics were involved in adenosylcobalamin biosynthesis (Teplitski et al., 2004). Repression of C. reinhardtii METE expression at non-lethal elevated temperatures (Figure 4b) by adenosylcobalamin-producing S. meliloti provides direct evidence that algae gene expression can be influenced by cobalamin-producing bacteria. At present, it is unclear whether the greater METE repression at higher temperatures is a consequence of more adenosylcobalamin production by S. meliloti, more adenosylcobalamin release into the environment by S. meliloti cell death or more efficient uptake by C. reinharditii cells. Diatoms increase cobalamin acquisition, reduce cobalamin demand or mitigate cellular damage caused by reduced methionine biosynthesis capabilities under low cobalamin conditions (Bertrand et al., 2012). At present, we do not know whether similar strategies are employed by C. reinhardtii or other chlorophyta to alleviate cobalamin scarcity under elevated temperatures or whether they stimulate closer associations with cobalamin-producing bacteria. The minimum cobalamin concentration (74 pMol) required for enhancing C. reinhardtii thermal tolerance is greater than the concentration reported in the few studies examining natural cobalamin concentrations in aquatic and terrestrial environments (Schneider and Stroinski, 1987; Panzeca et al., 2009; Sañudo-Wilhelmy et al., 2012), and greater than that required to support growth of algal vitamin B12 auxotrophs (Croft et al., 2005, Kazamia et al., 2012). Thus, in environments depleted in cobalamin, such as those along the southern California-Baja California coastal ocean (Sañudo-Wilhelmy et al., 2012), alga cobalamin auxotrophs would be scarce or algae would be more efficient in cobalamin acquisition or rely more on closer associations with cobalamin-producing bacteria.

Methionine is central to C. reinhardtii thermal tolerance, given that as little as a 50% reduction in METH expression incapacitates cobalamin-mediated enhancement of thermal tolerance (Figure 6). Methionine is not only required for protein synthesis but also for methylation reactions and methyl cycling, while methionine deficiencies can lead to accumulation of homocysteine to toxic levels. Heat repression of C. reinhardtii METE gene expression could explain decreases in METE pools during heat stress (Figures 4 and 5). In contrast expression of METH was less affected by heat shock, (Figure 4) which likely contributes to the presence of, albeit small, pool sizes of METH during heat stress (Figure 5) over prolonged heat exposure. The effect of heat on C. reinhardtii METE gene expression and METE pools is consistent with those reported for other organisms, including C. reinhardtii (Muhlhaus et al., 2011); it is conceivable that METH is also heat-labile although to a lesser extent than METE. In E. coli, MetE protein is susceptible to heat-induced aggregation and degradation (Mogk et al., 1999), which stimulates the heat shock responses. Thus, C. reinhardtii methionine pools and biosynthesis proteins may be central to growth rate control at elevated temperatures and in activating the heat shock response (Ron and Davis, 1971; Ron 1975; Gur et al., 2002) as well as for maintaining de novo protein biosynthesis capabilities required for photosynthetic machinery acclimation to higher temperatures (Tanaka et al., 2000).

The fitness advantage to C. reinhardtii of the cobalamin-dependent (METH) over the cobalamin-independent (METE) methionine synthase might help explain the retention of METH gene and the apparent independent loss of METE gene in various algal species. A recent survey of cobalamin-dependent enzymes in 15 sequenced algal genomes representative of the Chlorphyta, Chromalveolata and Rhodophyta highlighted the nearly ubiquitous presence of the METH gene, whereas the METE gene was detected in only 50% of the genomes (Helliwell et al., 2011). Organisms with both METE and METH are clearly more flexible in environments that routinely or periodically lack cobalamin availability, but the greater thermal tolerance properties of METH provides for greater plasticity to address changes in environmental temperature despite the cobalamin chemical dependency.

Our findings also have implications for understanding ecosystem responses to global temperature changes. Following an increase in environmental temperature, we could anticipate changes in microbial community composition, including more cobalamin-producers, more efficient cobalamin-scavengers or variants with physiological adaptations circumventing methionine limitations or with more heat tolerant cobalamin-independent methionine synthases (METE). As a consequence of shifts in microbial community structure or metabolic activities, primary productivity, a key component of the global carbon cycle, could be disrupted in the absence of sufficient abundance of cobalamin or cobalamin-producing bacteria. Thus, it might be appropriate to include cobalamin availability and cobalamin-producing bacteria and their production and consumption dynamics in efforts to predict and model marine or terrestrial primary productivity as temperatures rise. Our studies also suggest that optimal algal productivity in engineered systems could be more consistently achieved by relying on cobalamin-dependent (METH) rather than cobalamin-independent (METE) methionine synthase in systems that experience operating temperatures that cause heat stress to algae, thereby decreasing productivity. Consequently, production systems could be operated at higher temperatures in the presence of cobalamin, thereby reducing cooling costs.