pH-mediated control of anti-aggregation activities of cyanobacterial and E. coli chaperonin GroELs

Abstract

In contrast to Escherichia coli, cyanobacteria have multiple GroELs, the bacterial homologues of chaperonin/Hsp60. We have shown that cyanobacterial GroELs are mutually distinct and different from E. coli GroEL with which the paradigm for chaperonin structure/function has been established. However, little is known about regulation of cyanobacterial GroELs. This study investigated effect of pH (varied from 7.0 to 8.5) on chaperone activity of GroEL1 and GroEL2 from the cyanobacterium Synechococcus elongatus PCC7942 and E. coli GroEL. GroEL1 and GroEL2 showed pH dependency in suppression of aggregation of heat-denatured malate dehydrogenase, lactate dehydrogenase and citrate synthase. They exhibited higher anti-aggregation activity at more alkaline pHs. Escherichia coli GroEL showed a similar pH-dependence in suppressing aggregation of heat-denatured lactate dehydrogenase. No pH dependence was observed in all the GroELs when urea-denatured lactate dehydrogenase was used for anti-aggregation assay, suggesting that the pH-dependence is related to some denatured structures. There was no significant influence of pH on the chaperone activity of all the GroELs to promote refolding of heat-denatured malate dehydrogenase. It is known that pH in cyanobacterial cytoplasm increases by one pH unit following a shift from darkness to light, suggesting that the pH-change modulates chaperone activity of cyanobacterial GroEL1 and GroEL2.

In general, proteins need to fold into defined 3D structures to be functional. In vitro, some proteins spontaneously fold to their native state (1). But in the cellular environment, due to overcrowding proteins/macromolecules and various protein stresses, this spontaneous folding becomes hampered, and misfolding and/or aggregation of proteins frequently occur. To avoid these dangers, cells have a complex network of molecular chaperones, which prevent aggregation and promote efficient folding of proteins (2, 3).

The molecular chaperone GroEL is a bacterial member of the chaperonin/Hsp60 family. In general, GroEL consists of 14 identical subunits of ∼57 kDa which form 2 heptameric rings stacked back to back (4). GroEL assists folding of a non-native protein in an ATP-dependent way with the aid of the co-chaperonin GroES (5). One unfolded/misfolded protein enters into the cavity formed by the GroEL heptameric ring and the GroES heptamer covers the mouth of the cavity. Inside the closed cavity, the non-native protein folds without adverse interprotein interaction. ATP binding/hydrolysis regulate the chaperone reaction (2, 6, 7).

The majority of bacterial species including Escherichia coli, the most widely studied model organism, have only one groEL gene that forms an operon with the groES gene (8). However, most cyanobacteria including Synechococcus elongatus PCC7942 have two groEL genes (9, 10). One of them, groEL1, forms an operon with groES, whereas the other one (groEL2) does not. Like E. coli GroEL (11), GroEL1 from S. elongatus PCC7942 is essential (12). In contrast, GroEL2s from the thermophilic cyanobacterium Thermosynechococcus elongatus and the mesophilic cyanobacterium S. elongatus PCC7942 are not essential (12, 13). Overexpression of GroEL1 and GroEL2 increases thermotolerance in Synechocystis sp. PCC6803 (14). We also showed that the non-essential GroEL2 from T. elongatus plays a crucial role under high and low temperature stresses (13). GroEL2 does not complement a temperature sensitive mutation in E. coli GroEL whereas GroEL1 does (10, 15, 16). Biochemical properties of GroEL1 and GroEL2 from S. elongatus PCC7942 are mutually distinct and different from E. coli GroEL (17). In addition, regulatory mechanisms for the expression of the groESL1 operon and groEL2 gene are diversified from each other (18). We postulated that the groEL2 gene has acquired a novel, beneficial function especially under stresses and become preserved by natural selection, with the groEL1 gene retaining the original, essential function (18). However, it is not known by what mechanism GroEL2 plays a role under stress at the molecular level.

The cytosolic pH of cyanobacteria has been reported to change during light-to-dark transitions (19). The present study investigated effect of pH changes on the chaperone activity of GroEL1 and GroEL2 from S. elongatus PCC7942 as well as that of E. coli GroEL. Previously, we showed that both GroEL1 and GroEL2 suppress aggregation of heat-denatured malate dehydrogenase (MDH) efficiently like E. coli GroEL (17). The measurement was performed at pH 8.0. Here, the anti-aggregation activity was assayed at pHs 7.0, 7.5, 8.0 and 8.5 with using heat-denatured MDH, heat-denatured lactate dehydrogenase (LDH), heat-denatured citrate synthase (CS) and urea-denatured LDH as substrates. We found that the activity of both cyanobacterial GroELs to suppress aggregation of the heat-denatured dehydrogenases and CS are remarkably pH-dependent. The activity of E. coli GroEL was also affected by pH when heat-denatured LDH was used as a substrate. We will discuss our results with relevance to cellular function of the two GroELs under light and heat stress.

Materials and Methods

Preparation of GroEL1, GroEL2, DnaK2 from S. elongatus PCC7942 and E. coli GroEL

Escherichia coli BL21 (DE3) harbouring pET21a to which either S. elongatus groEL1 or groEL2 was cloned (17) was cultivated to overexpress C-terminally His-tagged GroEL1 or GroEL2. The recombinant proteins extracted from the cells were purified by Ni Sepharose 6 Fast Flow (GE Healthcare, Tokyo, Japan) column chromatography. GroEL1 or GroEL2 eluted from the column with 20 mM sodium phosphate buffer (pH 8.0) containing 500 mM imidazole and 500 mM NaCl was desalted by Sephadex G-25 (GE Healthcare) equilibrated with 50 mM HEPES−KOH (pH 7.5). Purities of GroEL1 and GroEL2 were comparable to that of E. coli GroEL (Supplementary Fig. S1). Construction of a strain that over-expresses C-terminally His-tagged DnaK2 from S. elongatus and their purification method were described previously (20). Escherichia coli GroEL and GroES were purchased from Takara (Kyoto, Japan).

Anti-aggregation assay with heat-denatured substrates

Fifty millimolar of HEPES−KOH (pH 7.0, 7.5, 8.0 or 8.5) with or without a molecular chaperone (GroEL1, GroEL2, E. coli GroEL or DnaK2) was pre-incubated at 45°C for 3 min. Then pig heart mitochondrial MDH purchased from Oriental Yeast (Kyoto, Japan) was added to a final monomer concentration of 0.2 μM. Suppression of aggregation of MDH was assayed at 45°C for 10 min by continuous monitoring of the absorbance changes at 360 nm with a Shimadzu UV-1800 spectrophotometer (Shimadzu, Kyoto, Japan). Anti-aggregation assay with rabbit muscle LDH, purchased from Oriental Yeast (Kyoto, Japan), was performed in the same way as that with MDH except LDH was incubated at 50°C. The assay with porcine heart CS, purchased from Sigma-Aldrich (Tokyo, Japan), was performed in the same way except 0.4 µM CS was incubated at 48°C in the presence of 2.8 µM or 5.6 µM GroELs.

Anti-aggregation assay with urea-denatured LDH

LDH (final concentration of 10 µM) was incubated in the presence of 6 M urea at 25°C for 20 min with shaking at 1,000 rpm. Then 20 μl of the LDH solution was added to 980 μl of 50 mM HEPES−KOH at various pHs with or without a molecular chaperone. Aggregation of LDH was measured at 25°C for 10 min by continuous monitoring of the absorbance changes at 360 nm.

MDH refolding assay

MDH (final concentration of 0.4 µM) was denatured in the presence of 5.6 µM GroEL1, GroEL2, E. coli GroEL or bovine serum albumin (BSA, a control protein) at 45°C for 25 min in 30 µl of a solution containing 15 mM HEPES−KOH (pH 8.0) and 10 mM dithiothreitol (DTT), and then the MDH solution was kept on ice for 1 min. Thirty microlitres of two times-concentrated refolding solution were added to 30 µl of each of the heat-treated MDH solutions at 25°C to start refolding of the MDH. The two times-concentrated refolding solution contained 150 mM HEPES−KOH (pH 7.0, 7.5, 8.0 or 8.5), 10 mM DTT, 300 mM KCl, 40 mM MgCl₂, 4 mM ATP and 2.8 µM E. coli GroES. The final concentration of MDH was 0.2 µM and those of GroEL1, GroEL2, E. coli GroEL and BSA were 2.8 µM. Aliquot of the refolding solution was withdrawn immediately, 120 min or 240 min after the start of the MDH refolding reaction, and the MDH activity was determined by monitoring decrease of the NADH absorption at 340 nm with an Ultrospec 3100 pro spectrophotometer (GE Healthcare) in medium containing 300 mM HEPES−KOH (pH 7.5), 10 mM DTT, 0.5 mM oxaloacetic acid and 0.28 mM NADH.

Results

Effect of pH on the activity of cyanobacterial GroELs, E. coli GroEL and cyanobacterial DnaK2 to suppress aggregation of heat-denatured MDH

We investigated the effect of pH on the activities of S. elongatus GroEL1, GroEL2 and E. coli GroEL to suppress aggregation of denatured proteins. Prevention of protein aggregation is characteristic among evolutionarily conserved molecular chaperones. Non-native/denatured proteins are captured by molecular chaperones, which usually results in suppression of aggregation of the proteins. MDH was used as a thermolabile model substrate for measurement of anti-aggregation activity of GroELs. MDH was incubated at 45°C in order to facilitate its denaturation and aggregation. The aggregation can be measured by increase in apparent absorbance at 360 nm as a result of light scattering in the MDH solution. Considering that GroELs form a heptamer/tetradecamer ring typically, GroELs were added seven or fourteen times more than the MDH monomer. At pH 7.0, 1.4 and 2.8 µM GroEL1 did not suppress the aggregation of 0.2 µM MDH at 45°C (Fig. 1). Instead, it increased turbidity of the MDH solution furthermore. In contrast to GroEL1, 1.4 µM GroEL2 suppressed the aggregation slightly, and 2.8 µM GroEL2 did it significantly. At pH 7.5, 2.8 µM GroEL1 suppressed aggregation of MDH to the same extent as 1.4 µM GroEL2 did; 2.8 µM GroEL2 suppressed most of the MDH aggregation at this pH. At pH 8.0 and pH 8.5, 1.4 µM GroEL1 or GroEL2 suppressed most of the aggregation of MDH (Fig. 1). Increase in the concentration of GroEL1 or GroEL2 from 1.4 to 2.8 µM did not increase the suppression, furthermore, suggesting that the anti-aggregation activity of each GroEL reached its maximum at or <1.4 µM at pHs 8.0 and 8.5. GroEL2 exhibited higher activity per mole than GroEL1 within a pH range of 7.0 − 8.0.

In contrast to the cyanobacterial GroELs, pH had no effect on anti-aggregation activity of E. coli GroEL; 1.4 µM E. coli GroEL exhibited complete suppression of the MDH aggregation at pHs 7.0, 7.5, 8.0 and 8.5 (Fig. 2). These results showed for the first time that the pH response of cyanobacterial GroEL1 and GroEL2 anti-aggregation activity is different from that of E. coli GroEL. Furthermore, their responses are mutually distinct. Both cyanobacterial GroELs showed a pH optimum of 8.0 − 8.5, but when pH decreases from 8.0 to 7.0, GroEL1 lost its anti-aggregation activity completely, whereas GroEL2 retained some activity even at pH 7.0 (Figs 1 and 8).

We wondered whether the pH responsiveness of GroEL1 and GroEL2 is shared among other chaperones in cyanobacteria. We examined the effect of pH on anti-aggregation activity of cyanobacterial DnaK2. DnaK2 is a bacterial member of the Hsp70/DnaK chaperone family. In the presence of 1.4 µM DnaK2, aggregation of 0.2 µM MDH was completely suppressed at pHs ranging from 7.0 to 8.5 (Fig. 3). Even in the presence of DnaK2 at a much lower concentration (0.2 µM), the MDH aggregation was almost completely suppressed regardless of the pH changes (Fig. 3). Our present study revealed a unique chaperone property of cyanobacterial GroEL1 and GroEL2.

Effect of pH on the activity of cyanobacterial GroELs, E. coli GroEL and cyanobacterial DnaK2 to suppress aggregation of heat-denatured LDH

The pH-dependent anti-aggregation activity of cyanobacterial GroELs was also observed when we performed experiments with 0.2 μM LDH heat-denatured at 50°C. GroEL1 at 1.4 and 2.8 μM did not show the anti-aggregation activity at pH 7.0 and 7.5, whereas it suppressed aggregation of LDH at pH 8.0 and 8.5 (Figs 4 and 8); 2.8 μM GroEL2 suppressed the LDH aggregation at all the pHs examined. At 1.4 μM, GroEL2 did not show the anti-aggregation activity at pH 7.0 and 7.5 although it suppressed the LDH aggregation at pH 8.0 and 8.5 (Figs 4 and 8). Escherichia coli GroEL at 1.4 μM also showed a pH dependence of the activity when LDH was used as another substrate instead of MDH (Figs 5 and 8). It lost most of the activity at pH 7.0 whereas the LDH aggregation was completely suppressed by E. coli GroEL at pH 8.0 and 8.5. In contrast to GroELs, almost or complete suppression of the aggregation was observed at pHs ranging from 7.0 to 8.5 in the presence of DnaK2 at a much lower concentration (0.2 μM) than those of GroELs (Fig. 6).

Effect of pH on the activity of cyanobacterial GroELs and E. coli GroEL to suppress aggregation of heat-denatured CS

In order to confirm that GroEL1 and GroEL2 respond to pH regardless of the type of a protein substrate, we performed the anti-aggregation assay with CS as a substrate. In contrast to MDH and LDH, CS is an acidic protein (see Discussion). CS at 0.4 µM was heat-treated at 48°C because it did not aggregate significantly at 45°C. Dependency of anti-aggregation activity of cyanobacterial GroELs on pH was again observed with the heat-denatured CS. 2.8 and 5.6 μM GroEL1 did not show the anti-aggregation activity at pH 7.0 and 7.5, whereas it suppressed aggregation of CS at pH 8.0 and 8.5 (Figs 7 and 8). At pH 7.0, 2.8 and 5.6 μM GroEL2 did not suppress the CS aggregation. However, at other pHs, GroEL2 showed the anti-aggregation activity. pH had no effect on the anti-aggregation activity of E. coli GroEL (Figs 7 and 8). Escherichia coli GroEL at 2.8 µM suppressed the CS aggregation completely at all pHs examined as it did the MDH aggregation regardless of pH changes (Fig. 2).

Effect of pH on the activity of cyanobacterial GroELs, and E. coli GroEL to suppress aggregation of urea-denatured LDH

Depending on denaturing conditions, various denatured structures are produced (21). In order to test whether the pH dependency of the anti-aggregation activity of GroELs is related to some denatured structures, we measured the activity with urea-denatured LDH as a substrate. High concentration of urea generally yields the most complete unfolding. As shown in Fig. 9, LDH aggregated at 25°C when the LDH solution containing 6 M urea was diluted 50-fold. When GroEL1, GroEL2, or E. coli GroEL at 1.4 μM was present during the dilution, the aggregation was inhibited to a similar extent regardless of the type of GroEL. Their anti-aggregation activities were not affected by pH changes from 7.0 to 8.5.

Effect of pH on refolding of MDH heat-denatured in the presence of GroELs

We examined whether pH affects the chaperone activity of cyanobacterial GroELs and E. coli GroEL to assist in the folding of heat-denatured MDH. A 0.4 µM MDH was denatured at 45°C in HEPES buffer (pH 8.0) containing 5.6 µM GroEL1, GroEL2, E. coli GroEL or BSA (a control protein). Then, refolding of MDH was initiated at 25°C by adding the same volume of a refolding mixture containing GroES, ATP and HEPES buffer whose pH was 7.0, 7.5, 8.0 or 8.5. The MDH activity in the refolding mixture was measured after 2 or 4 h. The molar concentration of each GroEL in the refolding mixture was 14-fold higher than that of MDH. MDH was heat-denatured with one of the GroELs at pH 8.0 in order to suppress MDH aggregation (Fig. 1). When BSA was present during the heat-treatment, reactivation of MDH was negligible at a pH range from 7.0 to 8.5 (Fig. 10A and B). The presence of GroELs resulted in enhancement of the MDH refolding in all the pHs. More than 80% of the initial MDH activity was recovered within 2 h in the presence of E. coli GroEL. In contrast, only 20% was recovered within 4 h in the presence of GroEL1 or GroEL2. These results are consistent with the previous studies which have reported that activities of cyanobacterial GroELs to assist refolding of various substrates in vitro are much lower than that of E. coli GroEL (17, 22). The chaperone activities of GroEL, GroEL1 and GroEL2 in a pH range from 7.0 to 8.5 were kept constant (Fig. 10A and B). No significant pH-dependent activity was observed in any of the GroELs.

Discussion

As described in the Introduction section, there are two kinds of GroEL in S. elongatus PCC7942 whereas only one GroEL is present in E. coli. Amino acid sequence identities between these GroELs are 55 − 62%, and they are acidic proteins with similar sizes (Supplementary Fig. S2). Amino acid sequence comparison showed that E. coli GroEL is equally similar to the cyanobacterial GroEL1 and GroEL2. However, biochemical properties of cyanobacterial GroEL1 and GroEL2 are mutually distinct and different from E. coli GroEL (17, 18). Previous studies have indicated that GroEL2 is a non-classical GroEL, whereas GroEL1 appears to be equivalent to E. coli GroEL (10, 12, 15, 16, 18).

In the present study, we found that the anti-aggregation activity of both cyanobacterial GroEL1 and GroEL2 is regulated by pH when heat-denatured MDH, LDH and CS are used as substrates. They suppressed aggregation of heat-denatured MDH and CS more efficiently at pHs 8.0 and 8.5 than pHs 7.0 and 7.5 (Figs 1 and 7) whereas E. coli GroEL did it almost completely at all the pHs examined (Figs 2 and 7). The aggregation in the presence of GroEL1 at pH 7.0 was significantly enhanced as compared with that in the absence of GroEL1. This enhancement is not due to aggregation of GroEL1 at pH 7.0. GroEL1 (and GroEL2) does not aggregate at pHs 7.0, 7.5 (Supplementary Fig. S3) and other pHs (data not shown). We assume that interaction of GroEL1 with the substrates causes an increase in turbidity of the mixture. A clear pH-dependence of the anti-aggregation activities of GroEL1 and GroEL2 was also observed when heat-denatured LDH was used as another substrate (Fig. 4). A pH-dependence of the E. coli anti-aggregation activity was observed when it was evaluated with heat-denatured LDH (Fig. 5). Thus all the GroELs showed higher activity to suppress aggregation of LDH at pHs 8.0 and 8.5 than at pHs 7.0 and 7.5 (Fig. 8), whereas DnaK2 suppressed aggregation of both heat-denatured MDH and LDH (almost) completely at all the pHs examined (Figs 3 and 6).

In the present study, we used three different substrate proteins, i.e. MDH, LDH and CS for the anti-aggregation assay. The isoelectric points (pIs) of pig heart mitochondrial MDH, rabbit muscle LDH, and porcine heart mitochondrial CS are 8.93, 8.17 and 6.68, respectively. These theoretical pIs were computed by the Compute pI/Mw tool with the amino acid sequence of each protein obtained from the UniProt Knowledgebase. We observed the pH responses (i.e. higher activity at pH 8.0 and 8.5 than at pH 7.0 and 7.5) by cyanobacterial GroEL1 and GroEL2 regardless of the type of a protein substrate, suggesting that the electrostatic property of a substrate protein is not involved in the pH response. Besides, the pI values of various E. coli GroEL substrates are reported to reflect the pIs of proteins in the E. coli cytoplasm (23), indicating that pI of a substrate is not a determining factor in interactions with GroEL. Thus it is inferred that the pH response is a noticeable feature of the GroELs.

Interestingly, all the GroELs did not show a pH-dependent activity to suppress aggregation of urea-denatured LDH (Fig. 9). High concentration of urea generally yields the most complete unfolding when compared with other denaturing conditions including high temperature (21). These results suggest that the pH-dependence is related to higher order structures of denatured proteins rather than some amino acid sequences.

pH did not have any significant effects on the ability of GroELs to assist refolding of denatured MDH (Fig. 10A and B). In the experiments, GroEL was heat-treated with MDH to form an MDH−GroEL complex. The MDH was released from the chaperonin during its refolding. Thus, it appears that pH increase from 7.0 or 7.5 to 8.0 or 8.5 accelerates stable association of GroEL1 or GroEL2 with denatured MDH to form a soluble complex (Fig. 10C) whereas it does not affect release of MDH from the complex to refold to its native structure.

Light or the photosynthetic electron transport (PET) induces alkalization in the cyanobacterial cytosol and chloroplast stroma as well as various physiological changes (Fig. 10D). PET is coupled to proton pumping from cytosol or stroma to the thylakoid lumen. The cytosolic pHs of light- and dark-acclimated S. elongatus PCC7942 are estimated to be 8.4 and 7.3, respectively (19). Similarly, pH in the chloroplast stroma increases by almost one pH unit upon illumination (24). The alkalization of chloroplast stroma turns on the operation of photosynthesis in the chloroplast because the pH optimum of photosynthetic CO₂ fixation is ∼8.1, with no activity below pH 7.3. The proton gradient causes photophosphorylation to produce ATP. It is known that proteins including the large subunit of ribulose bisphosphate carboxylase−oxygenase (Rubisco) are synthesized on chloroplast ribosomes using light energy or ATP generated by PET (25). PET stimulates transcription of genes encoding large and small subunits of cyanobacterial Rubisco (26), indicating that the synthesis of cyanobacterial Rubisco is enhanced in the light. It also increases transcription and translation of GroEL1 and GroEL2 in cyanobacteria (27, 28).

It is well established that the Rubisco large subunit is a substrate of chloroplast chaperonin (a homologue of GroEL) (29). Escherichia coli GroEL assists folding of Rubisco large subunit in vitro in an ATP- and GroES-dependent way (5). Cyanobacterial GroEL1 and GroEL2 facilitate folding/assembly of Rubisco holoenzyme in an E. coli cell although GroEL1 does it much more efficiently than GroEL2 (16). The pH-dependence of the anti-aggregation activities of GroEL1 and GroEL2 may have an implication for their effective interaction with non-native Rubisco and other proteins in the light and/or under light and heat stress to suppress off-pathway aggregation of the proteins.

Compared with GroEL1, GroEL2 retained more anti-aggregation activity at pH 7.0 and 7.5 (Fig. 8). This property may favour the function of GroEL2 under stresses (13) as follows. Under severe stress where PET is inhibited (30), the cytoplasmic pH is expected to drop. GroEL2 may maintain its chaperone activity under the conditions to contribute to stress tolerance of a cell.

Supplementary Data

Supplementary Data are available at JB Online.

Acknowledgements

T.A. is a recipient of a Japanese government scholarship for study in Japan. T.A. would like to extend her sincere gratitude to Bangladesh Agricultural University for giving her study leave to pursue her higher studies in Japan.

Funding

This work was supported in part by Grants-in-Aid for Scientific Research (C) (no. 18K05407, to H.N.) from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

Conflict of interest

None declared.

References

Abbreviations:

Abbreviations

Abbreviations
 
• BSA

  bovine serum albumin;

 
• CS

  citrate synthase;

 
• DTT

  dithiothreitol;

 
• LDH

  lactate dehydrogenase;

 
• MDH

  malate dehydrogenase;

 
• PET

  photosynthetic electron transport.