1. Bioinformatics Analysis
Previous studies in our lab demonstrated that the salt-induced PPIase activity in the mycelia of a halotolerant strain of P. oxalicum was accompanied by enhanced expression of the cyclophilin gene PoxCYP1828. To further understand the role of this cyclophilin, we carried out cloning of its cDNA, followed by a heterologous expression, purification and characterization of the protein. In silico analysis revealed that the open reading frame (ORF) of PoxCYP18 consists of 522 nucleotides encoding a protein of 173 amino acid (AA) residues, with predicted molecular mass and pI values of 18.91 kDa and 8.87, respectively. Comparative in silico analysis of PoxCYP18 cDNA with the genome sequence of P. oxalicum revealed that the full-length gene comprises 1001 bps and contains three introns of 260, 118 and 101 bps, and four different exons of 58, 170, 107 and 187 bps. This genomic organization of the PoxCYP18 gene was also supported by the size of PCR amplicon obtained with gene-specific primers using genomic DNA (1001 bp) and cDNA (522 bp) of P. oxalicum as templates (Supplementary table 1; Supplementary Fig. S1a). Computational analysis further revealed that the CLD in PoxCYP18 ranges from 16 to 172 AAs, consistent with the length in other orthologues 28 PoxCYP18 contains two cysteine residues (Cys-45 and Cys-170), and all nine active site residues required for CsA binding and PPIase activity are conserved (Supplementary Fig. S2). Analysis with the software Peptide Property Calculator indicated that compared to acidic (9.83%) and basic amino acid residues (13.29%), the neutral (39.88%) and hydrophobic residues (36.99%) are present in greater proportion. PoxCYP18 contains 20 positively (Arg + Lys) and 17 negatively (Asp + Glu) charged residues. The whole protein contains about 17.34% alpha helix, which might impart its structure a greater stability (Supplementary Fig. S1d). Phylogenetic analysis of PoxCYP18 sequence with other fungal cyclophilins clustered these proteins in a single clade (Supplementary Fig. S1c), depicting maximum similarity (90.2%) with Aspergillus niger orthologue, CYPA (Supplementary Fig. S1b, Supplementary table 2).
2. Cloning of PoxCYP18 cDNA and biochemical characterization of the protein
The PoxCYP18 gene was PCR amplified from the cDNA of P. oxalicum using gene-specific primers that resulted in an amplicon of 522 bp which was cloned into pET-28a(+) expression vector. The nucleotide sequence of the recombinant plasmid was confirmed by Sanger sequencing (accession no. MZ407579). The recombinant PoxCYP18, containing a six-histidine tag in its N-terminus, was heterologously expressed in Escherichia coli BL21(DE3)pLysS, followed by purification on Ni-NTA-agarose. SDS-PAGE analysis showed the presence of a single band of approximately 22 kDa, implying that the recombinant PoxCYP18 was purified to homogeneity (Fig. 1a). The purified PoxCYP18, after establishing its identity by immunoblotting with anti-His antibody (Fig. 1a), was used for estimation of PPIase activity by studying changes in the kinetics of chymotrypsin-mediated cleavage of the test peptide 29. These studies revealed that compared to the uncatalyzed reaction (0.0140 s− 1), the first-order rate constant was higher (0.077 s− 1) in the presence of PoxCYP18 (Fig. 1b), and it increased with the amount of protein (Supplementary Fig. S3), signifying that this cyclophilin is an active PPIase. Bovine serum albumin (BSA), used as a negative control, had no significant effect on the reaction rate. The activity of FKBPs and cyclophilins is inhibited specifically by FK506 and CsA, respectively, with no cross-inhibition reported 8. The presence of CsA resulted in a dramatic decrease in the rate constant of PoxCYP18 catalyzed reaction (0.0135 s− 1; Fig. 1b; Supplementary Fig. S4a), with an inhibition constant of 5.043 nM (Fig. 1c), indicating abrogation of the PPIase activity. On the contrary, the addition of up to 2 µM FK506 had no significant effect on the enzymatic activity of PoxCYP18 (0.0735 s− 1; Fig. 1b; Supplementary Fig. S4a), further implying that this PPIase is a true cyclophilin.
3. Role of cysteine residues in redox regulation of enzyme activity:
The PPIase activity of cyclophilins can be regulated in a redox-dependent or independent manner. Redox mechanisms involving disulfide bond formation between cysteine residues have been proposed to play an important role in controlling PPIase activity of different divergent and non-divergent cyclophilins 17,30. The existence of a four or more amino acid long additional stretch in CLDs, which forms a protruding loop and corresponds to position 48–54 (XXGKXLH) in the wheat cyclophilin TaCYPA-1 (Supplementary Fig. S2), is a characteristic structural feature of the divergent cyclophilins distinguishing them from the non-divergent cyclophilins 31,32. Divergent loop cyclophilins are generally observed in plants, such as TaCYPA-1 in wheat 33, CsCYP in Citrus sinensis34 and Catr1 in Catharanthus roseus35. Two invariable cysteine residues (Cys-40 and Cys-168), and a conserved Glutamic acid (Glu-83) are also observed in TaCYPA-1, CsCYP and Catr1 that are unique to divergent cyclophilins 36. On the contrary, the non-divergent cyclophilins, such as hCYPA, SmCYPA and Cpr1 from Schistosoma mansoni and Saccharomyces cerevisiae, respectively, lack the additional loop 30. While the PPIase activity of a divergent cyclophilin, CsCYP, containing two cysteine residues at positions − 40 and − 168, is controlled by a loop displacement mechanism mediated through the formation of a disulfide bond between these residues 34, the activity of a non-divergent cyclophilin, SmCYPA in S. mansoni, which contains four cysteine residues at positions − 69, -122, -126 and − 168, (Supplementary Fig. S2), is modulated through disulfide bond formation between Cys-122 and Cys-126 30, with oxidation resulting in loss of activity in both the proteins. The PoxCYP18 is a non-divergent cyclophilin and consists of two highly conserved cysteine residues at Cys-45 and Cys-170 positions (Supplementary Fig. S2). CuSO4 has been employed as an oxidizing agent to study redox regulation in different cyclophilins 30,31,37. In our study, the presence of 300 µM CuSO4 resulted in 95% abrogation of the PPIase activity of PoxCYP18, with an inhibition constant (Ki) of 39.44 µM (Fig. 2a-b). The PPIase activity of PoxCYP18 was not affected by either EDTA, a metal chelating agent, or DTT, that is used to reduce the disulfide bonds (Fig. 2a). To further understand the nature of Cu2+-induced inhibition, the Cu2+-treated PoxCYP18 was incubated with EDTA (1 mM) and DTT (10 mM) before the estimation of PPIase activity. The Cu2+-induced inhibition of PPIase activity of PoxCYP18 was partially reversed by EDTA and DTT since compared to 5% residual activity in the presence of Cu2+ alone, up to 70% and 75% residual activity was observed, respectively, when these two compounds were also added along with Cu2+ (Fig. 2a). Since PoxCYP18 consists of two cysteine residues at positions 45 and 170, we carried out a titration experiment of reducing versus oxidizing agent to investigate whether the loss of activity by Cu2+ was due to alteration in the structure. These studies revealed that migration of PoxCYP18 on 12% SDS-PAGE was redox-dependent (Fig. 3). Two closely moving bands of approx. 19 and 22 kDa were observed on 12% SDS-PAGE when PoxCYP18 was analyzed in the absence of Cu2+ and DTT (Fig. 3). The 19 kDa and the 22 kDa bands are likely the oxidized and reduced forms of PoxCYP18 since the former was observed in the presence of Cu2+ alone (Fig. 3), and the latter at high concentrations of DTT (20 and 50 mM) (Fig. 3). These observations suggest that the recombinant PoxCYP18 produced in E. coli, under the conditions used, is a mixture of reduced and oxidized forms of this protein. SDS-PAGE analysis also revealed the presence of two additional bands of 44 and 46 kDa, respectively, for the native protein. Immunoblotting with anti-His antibodies confirmed the identity of all the bands as either monomers of PoxCYP18 (19 kDa and 22 kDa) or its dimers (44 and 46 kDa) (Supplementary Fig. S5).
4. Generation of PoxCYP18 variants
To investigate the functional significance of cysteine residues in redox regulation of PoxCYP18, we generated site-directed mutants of PoxCYP18 by substituting cysteine residues with serine at positions − 45 (PoxCYP18C45S), -170 (PoxCYP18C170S), and at both − 45 and − 170 (PoxCYP18C40S/C170S). The recombinant mutant proteins were purified and validated by SDS-PAGE analysis (Fig. 1a), followed by western blotting with anti-His antibody (Supplementary Fig. S5). Enzymatic analysis revealed that relative to the uncatalyzed control, the first-order rate constant was significantly higher in the presence of PoxCYP18C45S (0.0657 s− 1), PoxCYP18C170S (0.0741 s− 1) and PoxCYP18C40S/C170S (0.0350 s− 1), respectively (Fig. 1b), while BSA had no significant effect. The rate of reaction increased with an increase in the amount of the respective mutant proteins (Supplementary Fig. S3). These observations indicate that the observed PPIase activity was specifically due to the presence of mutant cyclophilins. The kcat/Km values for PoxCYP18C45S (1.18 × 107 M− 1s− 1) and PoxCYP18C170S (1.40 × 107 M− 1s− 1) were similar to the native PoxCYP18 (1.46 × 107 M− 1s− 1). However, kcat/Km of the double mutant PoxCYP18C40S/C170S was significantly lower (6.4 × 106 M− 1s− 1), implying that together these two cysteine residues play an important role in catalysis. The presence of CsA resulted in a dramatic decrease in the rate constant for all three mutant proteins (Supplementary Fig. S4b-4d) with Ki values of 6.47 nM, 7.85 nM and 6.08 nM for PoxCYP18C45S, PoxCYP18C170S, and PoxCYP18C40S/C170S, respectively (Fig. 1c) being comparable to PoxCYP18 (5.04 nM).
To investigate how the substitution of cysteine residues in PoxCYP18 affects redox regulation, the PPIase activity of these mutant proteins was also studied after treatment with Cu2+. It was observed that the substitution of Cys-45 had no significant effect on sensitivity to Cu2+, since Ki for PoxCYP18C45S (33.4 µM) was similar to the Ki for the native PoxCYP18 (Ki = 39.4 µM) (Fig. 2b). On the contrary, substituting cysteine at position 170 alone or at both 45 and 170 together resulted in a substantial decrease in sensitivity to Cu2 + compared to the native PoxCYP18 (39.4 µM). This was evident since the Ki for Cu2+ increased more than 3-fold for PoxCYP18C170S (123.57 µM) and PoxCYP18C40S/C170S (163.57 µM) (Fig. 2b), suggesting that compared to Cys-45, Cys-170 plays an important role in Cu2+ sensitivity of this cyclophilin. Contrary to the native PoxCYP18, the addition of EDTA and DTT did not revert the Cu2+-induced inhibition of any of the PoxCYP18 mutants (Fig. 2a).
The effect of mutating the cysteine residues on structural changes was studied by electrophoretic analysis. It was observed that compared to native PoxCYP18, the mutant proteins depicted an altered migration pattern on SDS-PAGE (Fig. 3). Only a single band, that corresponded to a reduced band of native PoxCYP18 (22 kDa), was observed for all the mutant PoxCYP18 proteins under both oxidizing and reducing conditions (Fig. 3). The PoxCYP18C45S and PoxCYP18C170S depicted bands of 44 and 46 kDa, respectively, which correspond to the two bands observed for native PoxCYP18 (Fig. 3). Immunoblotting confirmed the high molecular weight bands of 44 and 46 kDa as dimers of these mutants (Supplementary Fig. S5). These bands, however, disappeared under reducing conditions implying the role of cysteine residues in the formation of dimers (Fig. 3). The bands of 44 and 46 kDa appear to be the dimers due to inter-monomer disulfide bond between − 170 cysteine residues of PoxCYP18C45S, and − 45 cysteine residues of PoxCYP18C170S, respectively, since loading of both these mutants together depicted the presence of two bands compared with one each when these proteins were analyzed separately by SDS-PAGE (Fig. 3). The fact that the double mutant PoxCYP18C45S/C170S did not depict a dimer band under any of the conditions (Fig. 3) further supports this inference.
5. Effect of NaCl and temperature on enzyme activity
As the strain of P. oxalicum, the source of PoxCYP18, is a halotolerant fungus and can tolerate up to 3 M NaCl, we therefore also analyzed the stability of this protein under high salt and temperature (45°C, 50°C and 60°C) conditions. In vitro biochemical assays revealed that PPIase activity of PoxCYP18, PoxCYP18C45S and PoxCYP18C170S was stable in up to 1.8 M NaCl, with the significant decrease observed only at 2.4 M (72% residual activity) and 3.0 M (22% residual activity: Fig. 4a). However, the double mutant PoxCYP18C45S/C170S, showed greater sensitivity to salt since the decrease in activity was observed from 1.2 M NaCl onwards (Fig. 4a). Thermostability analysis demonstrated the presence of approx. 70% residual PPIase activity after 2 h at 45°C, compared to 57% and 25% after 30 min at 50°C and 60°C, respectively. Retention of approximately 38%, 22% and 11% residual activity even after 6 h at 45°C, 50°C and 60°C, respectively, suggests that PoxCYP18 is relatively thermostable (Fig. 4b). On the contrary, all the mutants (PoxCYP18C45S, PoxCYP18C170S and PoxCYP18C45S/C170S) of this protein lost their PPIase activity by 30 min at 45°C, indicating the role of cysteine residues in the stability of this cyclophilin.
6. Stress tolerance in E. coli
The cyclophilins have been implicated in abiotic stress tolerance of bacteria, fungi and plants 27,28,38,39 but the role of PoxCYP18 has not been studied in stress adaptation. Therefore, in the present study, we investigated the role of PoxCYP18 in heat and salt stress using E. coli as a model. The protective role of PoxCYP18 in E. coli was explored in response to salt stress (500 mM NaCl) by inducing the protein with IPTG and following the growth of cultures at 37°C by taking absorbance at 600 nm (A600) until stationary phase, and also by spot analysis. It was observed that compared to control (E. coli cells transformed with empty vector), the PoxCYP18 expressing cells showed a higher number of colonies, signifying enhanced tolerance to salt stress. Though the cells expressing mutant proteins PoxCYP18C45S, PoxCYP18C170S and PoxCYP18C45S/C170S exhibited lesser growth compared to PoxCYP18, it was still higher than the control cells containing the non-recombinant vector. These results indicate that the mutants also imparted tolerance under salt stress, albeit to a lesser extent than the native PoxCYP18 (Fig. 5). The protective role of PoxCYP18 in E. coli was also studied against heat stress (47°C). Relative to control, no significant difference was observed at high temperatures in the growth of E. coli cells expressing PoxCYP18 and PoxCYP18C45S. However, the growth was adversely affected at both 37°C and 47°C when the E. coli cells were transformed with PoxCYP18C170S and PoxCYP18C45S/C170S, implying that these mutants have deleterious effects (Fig. 5).
7. 3D structure modeling and molecular dynamics (MD) simulation
The PPIase activity of cyclophilins is governed in a redox-dependent as well as independent manner. The redox-dependent mechanisms are different for the divergent and non-divergent cyclophilins, and involve precise control of PPIase activity through the formation of disulfide bonds between cysteine residues 17,30. The results of the present study revealed that the activity of PoxCYP18 is regulated through redox mechanisms, prompting us to generate 3D models of PoxCYP18 and its mutants (PoxCYP18C45S, PoxCYP18C170S and PoxCYP18C45S/C170S), followed by 50 ns MD simulations for gaining better insight into the possible structural features that may be responsible for this modulation. 3D modeled structure of PoxCYP18 (Fig. 6a) shows the presence of a CLD (16–172 residues) consisting of eight β sheets and two α helixes. All the nine active site residues responsible for CsA binding and PPIase activity are conserved (Fig. 6b), and the two cysteine resides that are well placed (1.89 Å) and oriented towards each other (Fig. 6B), indicate a possibility of forming disulfide bond under oxidizing conditions. Importance of these two cysteine residues in maintaining the overall stability of the PoxCYP18 structure and their effect on the dynamics of the active site can be understood from their presence in two important segments. The first segment (yellow color, Fig. 6b) contains Cys-45 and the active site residues Arg60, phe65, Met66 and Gln68, while the second segment (cyan color; Fig. 4.29b) contains Cys-170 and the active site residues Ala106, Phe118, Trp126, Leu127 and His131. Most of the active site residues (Arg60, Met66 and Gln68) in the first segment are part of the beta-sheet stabilized by a strong network of H-bonds. On the other hand, the majority of the active site residues (Ala106, Trp126, Leu127 and His131) in the second segment are part of the loop region making them more prone to fluctuations, thereby, rendering more dynamics to the active site. The presence of cysteine residues at the ends of these two segments might be significant for maintaining the overall active conformation of the protein and also for controlling the active site dynamics, specifically involving the residues Trp126, Leu127 and His131. His131is essential for PPIase activity and changes in this residue have been reported to completely abrogate the catalytic activity of this enzyme 40.
Although PoxCYP18 is a non-divergent cyclophilin, the arrangement of Cys-45 and Cys-170 residues is similar to that observed for divergent cyclophilin, CsCYP, containing two cysteine residues at positions − 40 and − 168 34 (Supplementary Fig. S2). To study the role of these residues in PoxCYP18, three mutant structures PoxCYP18C45S, PoxCYP18C170S and PoxCYP18C45S/C170S were also modeled and subjected to 50 ns MD simulations (along with the native PoxCYP18) using Gromacs. The average structures of all the PoxCYP18 variants obtained though the MD simulations were superimposed (Fig. 6c), which indicated that most of the secondary structure elements are conserved between the native and the mutant forms. However, some of the loop regions were observed to vary significantly, including the two loop regions belonging to the segments harboring the active site residues (for example Ala106, Trp126, Leu127 and His131, Fig. 6c, highlighted in red circles) and cysteine residues at their ends. Relative to PoxCYP18, fluctuations in these loop regions of all three mutants were further confirmed by the root mean square fluctuation values of the amino acids during the time of simulation. The superimposition of the active site residues (Fig. 6c) suggested that the average orientation of PoxCYP18C45S active site residues, specifically the Trp126 (green color), makes the active site less accessible for substrate binding. The Trp126 residue (blue color) in the double mutant, PoxCYP18C45S/C170S, seems to orient further inwards towards the active site, rendering it even less accessible to the substrate. Such conformations of the active site residues in PoxCYP18C45S and PoxCYP18C45S/C170S might explain the decrease in catalytic efficiencies (1.18 × 107 M-1s-1 and 6.4 × 106 M-1s-1, respectively) relative to the PoxCYP18 (1.46 × 107 M-1s-1). On the contrary, in spite of some of the residues showing greater fluctuations in the loop regions (residue no. 115–135, Fig. 6d), PoxCYP18C170S is able to keep the active site residues (cyan) well-oriented, which may enable the substrate to access the active site (Fig. 6c), thus leading to comparable catalytic efficiency (1.40 × 107 M-1s-1) with the native PoxCYP18 (1.46 × 107 M-1s-1). These analyses imply that the cysteine resides not only makes the PoxCYP18 protein responsive to redox conditions, but also plays an important role in maintaining the overall structure of the active site open for access to the substrate.
The changes in the overall compactness of the 3D structures were also analyzed by measuring the radius of gyration (Supplementary Fig. S6). According to this analysis, the overall structure of the mutants PoxCYP18C45S, PoxCYP18C170S and PoxCYP18C45S/C170S showed some fluctuations in their 3D structures during the time of simulation. As compared to PoxCYP18, which maintains the overall native structure, all three mutants tend to deviate more frequently from their overall starting shape. In the native PoxCYP18, the formation of a disulfide bond between the two cysteine residues under oxidized conditions may make the structure more compact and stable as compared to the reduced state, explaining the altered migration pattern of the mutant proteins on SDS-PAGE. This hypothesis, however, requires validation at the structural level by further biophysical characterization of these proteins.