Abstract
The production of biogas (methane) by anaerobic digestion is an important facet to renewable energy, but is subject to instability due to the sensitivity of strictly anaerobic methanogenic archaea (methanogens) to environmental perturbations, such as oxygen. An understanding of the oxidant-sensing mechanisms used by methanogens may lead to the development of more oxidant tolerant (i.e. stable) methanogen strains. MsvR is a redox-sensitive transcriptional regulator that is found exclusively in methanogens. We show here that oxidation of MsvR from Methanosarcina acetivorans (MaMsvR) with hydrogen peroxide oxidizes cysteine thiols, which inactivates MaMsvR binding to its own promoter (PmsvR). Incubation of oxidized MaMsvR with the M. acetivorans thioredoxin system (NADPH, MaTrxR, and MaTrx7) results in reduction of the cysteines back to thiols and activation of PmsvR binding. These data confirm that cysteines are critical for the thiol-disulfide regulation of PmsvR binding by MaMsvR and support a role for the M. acetivorans thioredoxin system in the in vivo activation of MaMsvR. The results support the feasibility of using MaMsvR and PmsvR, along with the Methanosarcina genetic system, to design methanogen strains with oxidant-regulated gene expression systems, which may aid in stabilizing anaerobic digestion.
Keywords: anaerobe, methane, methanogen, oxidative stress, transcription, disulfide
INTRODUCTION
Methane-producing archaea (methanogens) are strict anaerobes that are the rate-limiting step in biogas production [25]. Members of the genus Methanosarcina are particularly important in biogas production, due to their ability to produce methane with multiple substrates, including CO2, methylated compounds (ex. methanol), and acetate, where all other methanogens are restricted to hydrogenotrophic (H2 + CO2) methanogenesis [4]. Environmental perturbations can significantly impact the ability of methanogens to produce methane. For example, many methanogenesis enzymes are oxygen sensitive resulting in a loss of energy conservation and methane production upon exposure to oxygen [9]. The partial reduction of oxygen by flavoenzymes and metalloenzymes results in the production of reactive oxygen species (ROS), including superoxide (O2−) and hydrogen peroxide (H2O2), which target many macromolecules within cells [13]. For example, O2− and H2O2 oxidize Fe-S cluster containing enzymes and thiols of cysteine residues in proteins, causing Fe-S cluster degradation and formation of deleterious disulfides, which leads to loss of metabolism [14]. For anaerobes, such as methanogens, to overcome oxygen exposure requires mechanisms to decrease the production of ROS and/or actively remove ROS, as well as repair oxidatively-damaged proteins. Indeed, many methanogens can survive oxygen exposure; however, methane production is severely inhibited [10,18,1,2]. In particular, members of the Methanosarcinales appear to be some of the most aerotolerant methanogens [1], which is likely due to a large number of putative antioxidant and repair proteins encoded in the genomes of sequenced Methanosarcinales.[5,11,19,7] However, information on the mechanisms used by methanogens, including the Methanosarcinales, to sense and respond to oxygen (oxidative stress) is limited. An understanding of the oxidant-sensing mechanism used by Methanosarcina sp., along with their robust genetic system [12], may lead to the development of strains with increased oxidant tolerance.
Many cells use the formation of disulfides in proteins to monitor the cellular redox state and the presence of deleterious ROS. For example, OxyR is a transcriptional regulator found in Escherichia coli and other bacteria that is used to monitor the presence of H2O2 [26]. Redox-sensing thiols of cysteine residues in OxyR are oxidized in the presence of H2O2, resulting in OxyR activation which increases the expression of H2O2 scavengers, Fe-S cluster repair enzymes, and thiol redox buffer systems. Once H2O2 levels have been reduced, OxyR is inactivated by the reduction of the disulfides by glutaredoxin 1, with reductant supplied by glutathione/glutathione reductase and NADPH [6]. H2O2-sensing transcriptional regulators have also been identified in eukaryotes [3]. Recent evidence has revealed the presence of a redox-sensing transcriptional regulator (MsvR) in methanogens. In Methanothermobacter thermautotrophicus, MsvR regulates expression of F420H2 oxidase (FpaA) by redox-sensitive binding to the promoter region of fpaA [17]. MsvR also regulates its own expression in M. thermautotrophicus. MsvR functions as a negative regulator to repress expression of fpaA and msvR under reducing conditions. Oxidation of MsvR results in the induction of fpaA and msvR. More recently, MsvR from Methanosarcina acetivorans (MaMsvR), a member of the Methanosarcinales, was shown to bind to its own promoter (PmsvR) only under reducing conditions [15]. Formation of disulfide(s) within the C-terminal V4R effector domain of MaMsvR were shown to abrogate binding of MaMsvR to the PmsvR promoter region. For MaMsvR to function as a redox-sensing transcriptional regulator, PmsvR binding by MaMsvR would need to be restored by disulfide reduction to reset the system once oxygen/ROS is removed. Reduction of MaMsvR disulfides and activation of PmsvR binding can be accomplished in vitro by the addition of the disulfide-reducing agent dithiothreitol (DTT); however, the physiological reducing system is unknown. In the majority of organisms, disulfide reduction is mediated by thioredoxin and/or glutaredoxin systems [22].
Methanogens lack glutathione and therefore likely do not possess functional glutaredoxin systems [8,21]. However, recent evidence indicates the majority of methanogens contain thioredoxins (Trx) [20,24], which are small (~12 kDa) proteins that possess a CXXC active site motif necessary for disulfide oxidoreductase activity [22]. In the canonical system, Trx receives reducing equivalents from thioredoxin reductase (TrxR) with NADPH as the electron donor. Thus, the thioredoxin system could serve as the MsvR disulfide reducing system. M. acetivorans contains seven putative Trx homologs (MaTrx1-7) and a single TrxR homolog (MaTrxR). Recent evidence revealed that M. acetivorans contains at least three functional Trxs (MaTrx2, MaTrx6, and Matrx7) and a complete NADPH-dependent thioredoxin system comprised of MaTrxR and MaTrx7 [20]. We show herein that the M. acetivorans NADPH-dependent thioredoxin system can reduce disulfides in oxidized MaMsvR and restore PmsvR -binding activity, indicating that the thioredoxin system is the physiological MaMsvR disulfide reducing system. Thus, it may be feasible to use PmsvR along with the Methanosarcina genetic system to design Methanosarcina strains with oxidant-responsive genes, which may increase the stability of biomethanation.
MATERIALS AND METHODS
Protein purification and manipulation
Recombinant MaTrxR and MaTrx7 were expressed in E. coli and purified to homogeneity as previously described [20]. Strep-tagged MaMsvR was expressed in E. coli and purified to homogeneity as previously described [15]. H2O2-oxidized MaMsvR (MaMsvRox) was prepared by incubation of MaMsvR with 100-fold molar excess of H2O2 in buffer A (20 mM Tris pH 8, 15 mM MgCl2, 120 mM KCl, 12.5 μg/mL heparin, 10% glycerol) for 30 mins. Residual H2O2 was removed by buffer exchange into buffer A using a NAP5 column (GE Healthcare). DTT-reduced samples of MaMsvR were prepared by incubating 100 μM MaMsvRox in buffer A containing 10 mM DTT for 20 min at room temperature. Residual DTT was removed using a NAP5 column. The ability of the thioredoxin system to reduce MaMsvRox was assayed by incubation of 10 μM MaMsvRox with 1 mM NADPH, 0.5 μM MaTrxR, and 2.5 μM MaTrx7 for 1 hour at 37 °C in buffer A. Protein concentrations were determined by both the Bradford assay and using fluorescence with a Qubit protein assay following the manufacturer’s instructions (Invitrogen).
Electrophoretic mobility shift assay (EMSA)
Complimentary 50-bp oligonucleotides containing PmsvR were synthesized (Integrated DNA technologies) and annealed to generate the PmsvR DNA probe used in all EMSAs [15]. DNA-binding reactions were prepared by incubating 100 nM PmsvR with 8 μM MaMsvR in buffer A for 20 min at 37°C. Binding reactions were loaded onto a pre-run 6% polyacrylamide gel in 0.5X TBE buffer and electrophoresed for 75 min at 75 V at 10 °C. Gels were stained using SYBR gold (Life Technologies) and visualized using a Gel-Doc XR+ system (Bio-Rad Technologies).
Quantitation of thiols in MaMsvR
Aliquots of MaMsvR-containing samples used in EMSAs were analyzed for total thiol content using DTNB [23]. MaMsvR was denatured and thiols quantified by the addition of 10 μL of MaMsvR-containing sample to 90 μL of 6M guanidine-HCl in 100 mM KPO4, pH 7.8 containing 175 μM DTNB. Samples were incubated anaerobically for 15 min at room temperature and the absorbance at 412 nm was recorded. The number of thiols per MaMsvR monomer was calculated based on the concentration of TNB using ε412 =13,700 M−1 cm−1 [23]. All samples were analyzed in triplicate. The background amount of thiols contributed by the denatured thioredoxin system was determined in samples containing NADPH, MaTrxR, and MaTrx7, but without MaMsvR.
RESULTS AND DISCUSSION
Reduction of MaMsvR disulfides and activation of MaMsvR DNA-binding by thioredoxin
MaMsvR contains ten cysteine residues, with two located in the DNA-binding domain, four in the V4R domain, and the remaining four located in the linker domain [15]. The cysteines within the V4R domain (C206, C225, C232, and C240) are postulated to function in redox-sensing, whereby thiol-disulfide exchange causes conformation changes which alter the ability of MaMsvR to bind an inverted repeat sequence motif (TTCGN7-9CGAA) upstream of PmsvR. Three of the residues (C206, C232, and C240) are conserved in all MsvR homologs [17]. Specifically, C206 was shown to be critical for redox-sensitive binding of MaMsvR to PmsvR, because a MaMsvR C206A variant was able to bind to PmsvR under non-reducing conditions, whereas the wild-type MaMsvR is unable [15]. Previous results also revealed C225 was not involved in redox-sensing, while C232 and C240 impact MsvR binding to PmsvR, but the precise role of these cysteines is unclear. Thus, C206 is likely, and C232/C240 are possibly, involved in thiol-disulfide formation which serves to control DNA-binding by MaMsvR.
EMSA and thiol quantitation experiments were used to examine the role of thiol-disulfide exchange in controlling DNA binding by MaMsvR. First, MaMsvR was incubated with 100-fold molar excess of H2O2 to generate H2O2-oxidized MaMsvR (MaMsvRox). Quantitation of the thiol content of MaMsvRox under denaturing conditions revealed that four of the cysteines were not oxidized by H2O2 (Table 1), indicating some cysteines are inaccessible to H2O2, and likely do not participate in redox-sensing. Importantly, MaMsvRox was incapable of binding to the PmsvR region as revealed by the lack of shift when examined by EMSA (Fig. 1, lane 2). This result indicates that oxidation of the thiols of six cysteine residues is sufficient to inactive MaMsvR DNA binding. The subsequent treatment of MaMsvRox with DTT resulted in detection of approximately nine thiols (Table 1), consistent with the total number of cysteines present in MaMsvR. Moreover, incubation of MaMsvRox with DTT restored binding to PmsvR (Fig. 1, lane 3). This result is consistent with H2O2 causing the oxidation of six thiols to disulfides, which causes reversible inactivation MaMsvR binding to PmsvR. The remaining four thiols are likely buried within the folded protein and are inaccessible to H2O2 or DTT, and therefore do not participate in thiol-disulfide exchange.
Table 1.
Quantitation of MaMsvR thiols.
| Samplea | Thiols |
|---|---|
| MaMsvRox | 4.0 ± 0.6 |
| MaMsvRox + DTT | 9.0 ± 1.5 |
| MaMsvRox + NADPH/MaTrxR | 4.8 ± 0.1 |
| MaMsvRox + NADPH/MaTrxR/MaTrx7 | 9.9 ± 1.0 |
samples were processed and thiols quantified using DTNB as described in the Materials and Methods
Fig. 1. Activation of MaMsvR PmsvR binding by the M. acetivorans thioredoxin system.
EMSA performed with PmsvR and the addition of the indicated components as described in materials and methods.
Similar experiments were performed to determine if the M. acetivorans thioredoxin system could also activate DNA-binding of MaMsvRox. Incubation of MaMsvRox with NADPH, MaTrxR, and MaTrx7 (complete thioredoxin system) activated binding of MaMsvRox to PmsvR (Fig. 1, lane 6). The complete thioredoxin system alone did not cause a shift of PmsvR in the EMSA (Fig. 1, lane 4) and NADPH/MaTrxR in the absence of MaTrx7 also failed to activate binding of MaMsvRox to PmsvR (Fig. 1, lane 5). Moreover, incubation of MaMsvRox with the complete thioredoxin system resulted in the detection of ten thiols (Table 1), consistent with all the H2O2-generated disulfides in MaMsvR being surface exposed and accessible to reduction by MaTrx7. Taken together these results demonstrate that the M. acetivorans NADPH-dependent thioredoxin system can activate PmsvR binding in oxidized MaMsvR and that MaTrx7 is required for the reduction of disulfides in oxidized MaMsvR. The reduction of MaMsvR by MaTrx7 is the first evidence of thioredoxin playing a role in the regulation of the activity of a transcription regulator in a methanogen. The activation of MaMsvR DNA binding by MaTrx7 also integrates PmsvR regulation by MsvR into the physiology of M. acetivorans, which supports the future use of PmsvR in engineering oxidant-responsive gene expression strains. For example, we have previously demonstrated that increased expression of catalase protects M. acetivorans from H2O2 [16].
Proposed model of MaMsvR regulation in M. acetivorans
Based on results from previous studies [15,17] and herein, we propose the following model (Fig. 2) for the regulation of the PmsvR binding activity of MaMsvR by thiol-disulfide exchange involving the thioredoxin system. Exposure of M. acetivorans to oxidants (ex. H2O2) results in oxidation of critical cysteines in MaMsvR to disulfides. Based on previous studies, C206 plays a crucial role, likely forming an intermolecular disulfide between MaMsvR monomers [15]. However, under the conditions tested here, at least six cysteines are involved in H2O2-induced disulfide formation, which may generate three intra-molecular, six inter-molecular, or some combination of intra/inter-molecular disulfides. Nonetheless, the formation of disulfides likely causes a conformational change in MaMsvR, such that MaMsvR is no longer able to bind PmsvR, which allows for RNAP to bind and transcription to proceed. Removal of oxidant and/or an influx of electron donor would allow for the reduction of MaMsvR disulfides by MaTrx7, with reducing equivalents supplied by MaTrxR and NADPH. The in vitro results presented here demonstrate that MaTrx7 can specifically reduce disulfides in MaMsvR, but we cannot rule out that the additional MaTrxs or other proteins also participate in the in vivo reduction of disulfides in MaMsvR and may do so under different conditions. However, the target specificity and the redox partner(s) of the other MaTrxs is currently unknown [20]. The data presented here link the regulation of MaMsvR to the redox status of M. acetivorans and the availability of reducing equivalents (e.g. NADPH). The results also reveal that methanogens have oxidant sensing systems which are integrated into metabolism in a manner similar to systems identified in bacteria and eukaryotes. Ultimately, due to the thioredoxin-dependent reversible PmsvR binding by MaMsvR it may be feasible to engineer strains with oxidant-inducible genes (e.g. catalase) using PmsvR in an effort to generate oxidant tolerant strains without an increased energy demand that would come from constitutive gene expression.
Fig. 2. Proposed model of MaMsvR activation by the NADPH-dependent MaTrxR-MaTrx7 thioredoxin system in M. acetivorans.

H2O2 causes the oxidation of thiols (SH) to disulfides which inactivates MaMsvR DNA binding, allowing transcription by RNAP. MaTrx7 receives reducing equivalents from NADPH/MaTrxR to reduce the disulfides to thiols and restore MaMsvR DNA binding.
Acknowledgments
This work was supported in part by grant number P20 GM103640 (EAK) and P30 GM103450 (DJL) from the National Institute of General Medical Sciences of the National Institutes of Health, NSF grant number MCB1121292 (DJL), NASA Exobiology grant number NNX12AR60G (DJL), and the Arkansas Biosciences Institute (DJL), the major research component of the Arkansas Tobacco Settlement Proceeds Act of 2000.
References
- 1.Angel R, Claus P, Conrad R. Methanogenic archaea are globally ubiquitous in aerated soils and become active under wet anoxic conditions. ISME J. 2012;6 (4):847–862. doi: 10.1038/ismej.2011.141. ismej2011141 [pii] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Angel R, Matthies D, Conrad R. Activation of methanogenesis in arid biological soil crusts despite the presence of oxygen. PLoS One. 2011;6(5):e20453. doi: 10.1371/journal.pone.0020453. PONE-D-11-04379 [pii] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Cremers CM, Jakob U. Oxidant sensing by reversible disulfide bond formation. J Biol Chem. 2013;288 (37):26489–26496. doi: 10.1074/jbc.R113.462929. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.De Vrieze J, Hennebel T, Boon N, Verstraete W. Methanosarcina: the rediscovered methanogen for heavy duty biomethanation. Bioresource technology. 2012;112:1–9. doi: 10.1016/j.biortech.2012.02.079. [DOI] [PubMed] [Google Scholar]
- 5.Deppenmeier U, Johann A, Hartsch T, Merkl R, Schmitz RA, Martinez-Arias R, Henne A, Wiezer A, Baumer S, Jacobi C, Bruggemann H, Lienard T, Christmann A, Bomeke M, Steckel S, Bhattacharyya A, Lykidis A, Overbeek R, Klenk HP, Gunsalus RP, Fritz HJ, Gottschalk G. The genome of Methanosarcina mazei: evidence for lateral gene transfer between bacteria and archaea. J Mol Microbiol Biotechnol. 2002;4 (4):453–461. [PubMed] [Google Scholar]
- 6.Dubbs JM, Mongkolsuk S. Peroxide-sensing transcriptional regulators in bacteria. J Bacteriol. 2012;194 (20):5495–5503. doi: 10.1128/JB.00304-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Erkel C, Kube M, Reinhardt R, Liesack W. Genome of Rice Cluster I Archaea--the Key Methane Producers in the Rice Rhizosphere. Science. 2006;313 (5785):370–372. doi: 10.1126/science.1127062. [DOI] [PubMed] [Google Scholar]
- 8.Fahey RC. Novel thiols of prokaryotes. Annual review of microbiology. 2001;55:333–356. doi: 10.1146/annurev.micro.55.1.333. [DOI] [PubMed] [Google Scholar]
- 9.Ferry JG. Acetate-based methane production. In: Wall J, Harwood CS, Demain A, editors. Bioenergy. ASM press; Washington D.C: 2008. pp. 155–170. [Google Scholar]
- 10.Fetzer S, Bak F, Conrad R. Sensitivity of methanogenic bacteria from paddy soil to oxygen and desiccation. FEMS Microbiology Ecology. 1993;12:107–115. [Google Scholar]
- 11.Galagan JE, Nusbaum C, Roy A, Endrizzi MG, Macdonald P, FitzHugh W, Calvo S, Engels R, Smirnov S, Atnoor D, Brown A, Allen N, Naylor J, Stange-Thomann N, DeArellano K, Johnson R, Linton L, McEwan P, McKernan K, Talamas J, Tirrell A, Ye W, Zimmer A, Barber RD, Cann I, Graham DE, Grahame DA, Guss AM, Hedderich R, Ingram-Smith C, Kuettner HC, Krzycki JA, Leigh JA, Li W, Liu J, Mukhopadhyay B, Reeve JN, Smith K, Springer TA, Umayam LA, White O, White RH, Conway de Macario E, Ferry JG, Jarrell KF, Jing H, Macario AJ, Paulsen I, Pritchett M, Sowers KR, Swanson RV, Zinder SH, Lander E, Metcalf WW, Birren B. The genome of Methanosarcina acetivorans reveals extensive metabolic and physiological diversity. Genome Res. 2002;12 (4):532–542. doi: 10.1101/gr.223902. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Guss AM, Rother M, Zhang JK, Kulkarni G, Metcalf WW. New methods for tightly regulated gene expression and highly efficient chromosomal integration of cloned genes for Methanosarcina species. Archaea. 2008;2 (3):193–203. doi: 10.1155/2008/534081. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Imlay JA. How oxygen damages microbes: oxygen tolerance and obligate anaerobiosis. Adv Microb Physiol. 2002;46:111–153. doi: 10.1016/s0065-2911(02)46003-1. [DOI] [PubMed] [Google Scholar]
- 14.Imlay JA. Iron-sulphur clusters and the problem with oxygen. Mol Microbiol. 2006;59 (4):1073–1082. doi: 10.1111/j.1365-2958.2006.05028.x. [DOI] [PubMed] [Google Scholar]
- 15.Isom CE, Turner JL, Lessner DJ, Karr EA. Redox-sensitive DNA binding by homodimeric Methanosarcina acetivorans MsvR is modulated by cysteine residues. BMC Microbiol. 2013;13:163. doi: 10.1186/1471-2180-13-163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Jennings ME, Schaff CW, Horne AJ, Lessner FH, Lessner DJ. Expression of a bacterial catalase in a strictly anaerobic methanogen significantly increases tolerance to hydrogen peroxide but not oxygen. Microbiology. 2014;160 (Pt 2):270–278. doi: 10.1099/mic.0.070763-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Karr EA. The methanogen-specific transcription factor MsvR regulates the fpaA-rlp-rub oxidative stress operon adjacent to msvR in Methanothermobacter thermautotrophicus. J Bacteriol. 2010;192 (22):5914–5922. doi: 10.1128/JB.00816-10. JB.00816-10 [pii] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kiener A, Leisinger T. Oxygen sensitivity of methanogenic bacteria. System Appl Microbiol. 1983;4:305–312. doi: 10.1016/S0723-2020(83)80017-4. [DOI] [PubMed] [Google Scholar]
- 19.Maeder DL, Anderson I, Brettin TS, Bruce DC, Gilna P, Han CS, Lapidus A, Metcalf WW, Saunders E, Tapia R, Sowers KR. The Methanosarcina barkeri genome: comparative analysis with Methanosarcina acetivorans and Methanosarcina mazei reveals extensive rearrangement within methanosarcinal genomes. J Bacteriol. 2006;188 (22):7922–7931. doi: 10.1128/JB.00810-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.McCarver AC, Lessner DJ. Molecular characterization of the thioredoxin system from Methanosarcina acetivorans. FEBS J. 2014;281 (20):4598–4611. doi: 10.1111/febs.12964. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.McFarlan SC, Terrell CA, Hogenkamp HP. The purification, characterization, and primary structure of a small redox protein from Methanobacterium thermoautotrophicum, an archaebacterium. J Biol Chem. 1992;267 (15):10561–10569. [PubMed] [Google Scholar]
- 22.Meyer Y, Buchanan BB, Vignols F, Reichheld JP. Thioredoxins and glutaredoxins: unifying elements in redox biology. Annual review of genetics. 2009;43:335–367. doi: 10.1146/annurev-genet-102108-134201. [DOI] [PubMed] [Google Scholar]
- 23.Riddles PW, Blakeley RL, Zerner B. Reassessment of Ellman’s reagent. Methods Enzymol. 1983;91:49–60. doi: 10.1016/s0076-6879(83)91010-8. [DOI] [PubMed] [Google Scholar]
- 24.Susanti D, Wong JH, Vensel WH, Loganathan U, DeSantis R, Schmitz RA, Balsera M, Buchanan BB, Mukhopadhyay B. Thioredoxin targets fundamental processes in a methane-producing archaeon, Methanocaldococcus jannaschii. Proc Natl Acad Sci U S A. 2014;111 (7):2608–2613. doi: 10.1073/pnas.1324240111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Thauer RK, Kaster AK, Seedorf H, Buckel W, Hedderich R. Methanogenic archaea: ecologically relevant differences in energy conservation. Nat Rev Microbiol. 2008;6 (8):579–591. doi: 10.1038/nrmicro1931. nrmicro1931 [pii] [DOI] [PubMed] [Google Scholar]
- 26.Zheng M, Aslund F, Storz G. Activation of the OxyR transcription factor by reversible disulfide bond formation. Science. 1998;279 (5357):1718–1721. doi: 10.1126/science.279.5357.1718. [DOI] [PubMed] [Google Scholar]

