Coupling effects of thiol and urea-type groups for promotion of oxidative protein folding

Shunsuke Okada a, Motonori Matsusaki b, Kenta Arai c, Yuji Hidaka d, Kenji Inaba e, Masaki Okumura *be and Takahiro Muraoka *af
aDepartment of Applied Chemistry, Graduate School of Engineering, Tokyo University of Agriculture and Technology, 2-24-16 Naka-cho, Koganei, Tokyo 184-8588, Japan. E-mail:
bFrontier Research Institute for Interdisciplinary Sciences, Tohoku University, 6-3 Aramaki-Aza-Aoba, Aoba-ku, Sendai 980-8578, Japan. E-mail:
cDepartment of Chemistry, School of Science, Tokai University, Kitakaname, Hiratsuka-shi, Kanagawa 259-1292, Japan
dGraduate School of Science and Engineering, Kinki University, 3-4-1 Kowakae, Higashi-Osaka, Osaka 577-8502, Japan
eInstitute of Multidisciplinary Research for Advanced Materials, Tohoku University, 2-1-1 Katahira, Aoba-ku, Sendai 980-8577, Japan
fInstitute of Global Innovation Research, Tokyo University of Agriculture and Technology, Japan

Received 30th October 2018 , Accepted 26th November 2018

First published on 27th November 2018

Coupling of thiol and urea-type –NHC([double bond, length as m-dash]X)NH2 (X = O or NH) groups is effective in promoting oxidative protein folding. In particular, a thiol compound coupled with a guanidyl (X = NH) group significantly accelerates the rates of folding processes and enhances the yields of native proteins.

Folding of a protein to form its native conformation is essential to acquire its biological functions. The folding process is generally regulated by thermodynamics.1,2 In the case of proteins containing multiple disulfide bonds between cysteine residues, the folding pathways are often complicated due to the formation of intermediates with misbridged or transient disulfide bonds.3–6 Subsequent isomerization reactions that eventually afford native disulfide bonds often limit the rates and yields of oxidative protein folding. These isomerization reactions proceed through reversible exchanges between thiol groups and disulfide bonds. Thus, it is believed that redox-active molecules can assist and accelerate the isomerization and, consequently, the overall folding process.7,8

In the endoplasmic reticulum, glutathione plays a major role in thiol–disulfide exchange reactions, where the reduced form (GSH) cleaves disulfide bonds and the oxidized form (GSSG) allows for the formation of disulfide bonds.9–11 These redox reactions prompt the isomerization of the disulfide bonds to promote the formation of the native conformations of the proteins in vivo. The acceleration of oxidative folding by GSH and other thiol compounds such as cysteine and β-mercaptoethanol (βME) is also demonstrated by in vitro studies.7 It is reported that dithiol,12,13 aromatic thiol,14–16 and selenol17–19 compounds are also effective for the promotion of oxidative protein folding. Here, we report thiol compounds coupled with urea-type groups consisting of –NHC([double bond, length as m-dash]X)NH2 structures and their effects on the promotion of oxidative protein folding.

Compounds GdnSH (X = NH) and UreaSH (X = O) were developed for this study (Scheme 1). These compounds were designed by covalent conjugation of βME with urea or iminourea (i.e., guanidine), to combine redox active and hydrogen-bond forming agents.

image file: c8cc08657e-s1.tif
Scheme 1 Synthetic schemes of GdnSH and UreaSH.

Redox potential E0′ and pKa values of the thiol groups of GdnSH and UreaSH were determined by absorption spectroscopic measurements in buffers with changing pH as well as HPLC analyses monitoring the redox reactions with DTT, as summarized in Table 1. Interestingly, GdnSH and UreaSH showed opposing properties in their E0′ and pKa values compared to GSH. Namely, the pKa of GdnSH is lower than that of GSH20 likely due to the positively charged guanidyl group,8 while that of UreaSH is higher than GSH. Moreover, GdnSH and UreaSH have higher and lower E0′ values than GSH, respectively. In spite of these opposing properties, both GdnSH and UreaSH promote oxidative protein folding, as described below in detail.

Table 1 pKa and redox potential E0′ values of thiol groups of GSH, GdnSH and UreaSH
Compounds pKa E 0′ (mV)
a Reported value in ref. 19.
GSH 9.15 ± 0.04 –256a
GdnSH 8.86 ± 0.02 –237 ± 4
UreaSH 9.20 ± 0.04 –276 ± 7

To validate the capability to introduce disulfide bonds into a reduced and denatured protein, the disulfide-bond formation of fully reduced and denatured ribonuclease (RNase) A was monitored in the presence of disulfide compounds, GSSG, GdnSS and UreaSS ([RNase A] = 8.0 μM; [disulfides] = 0.20 mM). Native RNase A has four pairs of cysteine residues forming disulfide bonds, i.e., C26–C84, C40–C95, C58–110 and C65–C72, and is a widely used model substrate for oxidative protein folding studies.4 At selected time points during incubation for folding, the reactions were quenched by 4-acetamido-4′-maleimidylstilbene-2,2′-disulfonic acid (AMS) that irreversibly couples with thiol groups. The mass increase due to the coupled AMS decreases the electrophoretic mobility of RNase A in 14% SDS-PAGE, which allows for separation of the fully reduced form (R), folding intermediates, and fully oxidized form (namely, a mixture of the native form and a non-native form with four disulfide bonds, represented as N/4S) of RNase A to monitor the progress of the disulfide bond introduction. In the presence of GSSG, the band corresponding to R faded upon extending the incubation time, and almost disappeared after 30 min incubation (Fig. 1A). In turn, the band corresponding to N/4S appeared after 30 min incubation, and intensified by further prolonged incubation. The oxidation process of RNase A in the presence of UreaSS was similar to that of GSSG (Fig. 1B). Interestingly, GdnSS significantly accelerated the oxidation reaction. In the presence of GdnSS, R nearly disappeared after only 5 min incubation. The band corresponding to N/4S appeared after 10 min incubation through a rapid band-shift from initial intermediates to further oxidized species, indicating accelerated disulfide bond introduction (Fig. 1C). The time-course changes of N/4S formation of RNase A were analyzed quantitatively on the basis of the electrophoretic assays (Fig. 1D). It is clear that GdnSS promoted the oxidation of RNase A at the fastest rate, reaching a plateau within the initial 30 min incubation, while the oxidation rate of UreaSS was slower than that of GSSG.

image file: c8cc08657e-f1.tif
Fig. 1 Time course analyses of RNase A oxidation by SDS-PAGE. SDS-PAGE gel images monitoring the oxidation of RNase A (8.0 μM) in the presence of (A) GSSG, (B) UreaSS, (C) GdnSS, (E) GSH and GSSG, (F) UreaSH and GSSG, and (G) GdnSH and GSSG (thiol compounds: 1.0 mM; disulfide compounds: 0.20 mM) in a buffer (50 mM Tris–HCl, 300 mM NaCl, pH 7.5). The oxidation reactions were quenched with AMS after 1, 5, 10, 30, 60 and 90 min incubation. The leftmost and rightmost lanes show the bands corresponding to the markers (M) and native RNase A (N), respectively. R and N/4S represent fully reduced and fully oxidized (a mixture of the native form and a folding intermediate with four disulfide bonds) RNase A, respectively. Time-course changes of the percentages of fully oxidized RNase A (8.0 μM) in the presence of (D) disulfide compounds (0.20 mM; black diamonds: GSSG; red squares: GdnSS; blue circles: UreaSS) or (H) mixtures of thiol (1.0 mM) and disulfide compounds (0.20 mM; black filled diamonds: GSH/GSSG; red filled squares: GdnSH/GSSG; blue filled circles: UreaSH/GSSG) quantified by SDS-PAGE analyses (original data: A–C and E–G). Error bars indicate the means ± SEM of three independent experiments.

It is known that addition of a reductant to an oxidative-folding condition can enhance the efficiency of protein folding to the native structure by facilitating the shuffling of disulfide bonds.7,8 On the basis of the SDS-PAGE assays using AMS (Fig. 1E–G), the oxidation of RNase A in the presence of GSSG and a reductant, GSH, UreaSH or GdnSH, was plotted as a function of the incubation time (Fig. 1H).21

Compared to the GSH/GSSG system, the UreaSH/GSSG system promoted slightly faster oxidation, while the GdnSH/GSSG system exhibited a significantly faster oxidation process. To assess the capabilities of these systems to promote the oxidative folding of RNase A to its native form, recovery of the enzymatic activity was monitored in the presence of GSSG and the thiol compounds (Fig. 2). In the absence of a thiol compound, the recovery of the native form was limited to 15% after 180 min incubation. Addition of GSH accelerated the recovery rate and the activity increased to 34% after 180 min. The UreaSH/GSSG system showed further increased recovery of the native form up to 44%. Of note, the GdnSH/GSSG system allowed for significantly faster and more efficient recovery than the others to afford 63% activity after 180 min incubation. Thus, it is demonstrated that UreaSH and GdnSH promote the oxidative folding of RNase A more efficiently than GSH.

image file: c8cc08657e-f2.tif
Fig. 2 Recovered enzymatic activity of RNase A (8.0 μM) during incubation in the GdnSH/GSSG, UreaSH/GSSG, GSH/GSSG and GSSG only systems (thiol compounds: 1.0 mM; disulfide compounds: 0.20 mM) in a buffer (50 mM Tris–HCl, 300 mM NaCl, pH 7.5). The activity was evaluated by spectroscopic monitoring of the hydrolysis of cCMP to 3′-CMP at 30 °C. Error bars indicate the means ± SEM of three independent experiments.

The folding promoting effects of UreaSH and GdnSH were further investigated using bovine pancreatic trypsin inhibitor (BPTI) containing three disulfide bonds, i.e., C5–C55, C14–C38 and C30–C51. On the basis of the entire folding pathway of BPTI elucidated by Weissman and Kim,22 major accumulated quasi-native intermediates such as N′ and N* fold into native-like structures with two disulfide bonds and lead to the formation of the native structure (N). Oxidative folding of fully reduced and denatured BPTI (30 μM) in the presence of GSSG (0.20 mM) as an oxidant was monitored by reverse-phase HPLC (Fig. 3), which displays the disulfide-linked conformational transitions of BPTI from its reduced (R) form to N via N′, N* and other intermediate species. In the absence of a reductant, BPTI showed spontaneous folding driven by oxidation with GSSG (Fig. S1 in ESI). During the incubation, the fraction of R decreased and the fractions corresponding to N′ and N* intermediates emerged. The fraction of N appeared after 10 min incubation, and the yield of N was 21% after 60 min (Table 2). Addition of GSH (1.0 mM) to the above medium enhanced the folding efficiency of BPTI to afford a yield of N of 24% after 60 min incubation (Fig. 3A and Table 2). Addition of UreaSH allowed for further enhancement, albeit slight, of the yield of N up to 26% (Fig. 3B and Table 2). Remarkably, addition of GdnSH allowed for significant acceleration of the folding and enhancement of the yield (Fig. 3C and Table 2). In the presence of GdnSH and GSSG, the fraction of R decreased rapidly between 1 and 5 min incubation and almost disappeared within 10 min, indicating a fast disulfide-bond introducing reaction. The formation of N was observed even after 5 min incubation, and the yield of N reached up to 51% after 60 min incubation.

image file: c8cc08657e-f3.tif
Fig. 3 Time-course reverse-phase HPLC analyses of oxidative folding of BPTI (30 μM) in the presence of GSSG (0.20 mM) and (A) GSH, (B) UreaSH and (C) GdnSH (1.0 mM). N and R depict native and reduced forms of BPTI, respectively. Eluent buffers: water (containing 0.05% TFA) and CH3CN (containing 0.05% TFA) with a linear gradient; flow rate: 1.0 mL min−1; detection wavelength: 229 nm; temperature: 25 °C.
Table 2 Yields of the native form of BPTI after 60 min incubation
Thiol and disulfide additives Yields of native BPTIa (%)
a Yields were evaluated by RP-HPLC analyses. Errors indicate the means ± SD of three independent experiments.
GSSG 21 ± 0.8
GSH/GSSG 24 ± 0.2
UreaSH/GSSG 26 ± 0.3
GdnSH/GSSG 51 ± 0.2

As described above, GdnSH and UreaSH have opposing properties in the E0′ and pKa values compared to GSH. Nevertheless, both GdnSH and UreaSH allow for promotion of oxidative protein folding, as demonstrated using RNase A and BPTI. On the basis of these results, we considered that the urea-type –NHC([double bond, length as m-dash]X)NH2 group, known to form hydrogen-bonds (X = O, NH) and cation–π interactions (X = NH) with proteins, is likely effective in assisting the folding process. It is known that the capability of guanidine and urea to interact with proteins by such non-covalent interactions allows for increasing solubility of proteins to suppress aggregation.23 Indeed, addition of 5.0 M guanidine–HCl suppressed thermal aggregation of lysozyme in a buffer, while lysozyme formed aggregates in a decreased concentration of guanidine–HCl to 100 mM (20 μM lysozyme in 50 mM Tris–HCl, 300 mM NaCl, pH 7.5, heating to 95 °C, Fig. S2 in ESI). Importantly, addition of GdnSH also suppressed the thermal aggregation of lysozyme, where even 1.0 mM was sufficient. Such a high aggregation suppression efficiency of GdnSH suggests that GdnSH can effectively access and be localized close to the protein molecule to increase its solubility by covalent disulfide-bonding with a cysteine SH-group of a protein in addition to the non-covalent interactions, which is likely an important factor for promotion of oxidative protein folding.

To study further the molecular mechanism of the oxidative protein folding, it is also important to investigate the molecular species that mainly assists the oxidation of a protein. MALDI-TOF MS measurement of a mixture of GdnSH and GSSG in Tris buffer, after 10 min incubation sufficient for equilibration, showed signals corresponding to GdnSS and GdnS-SG, indicating the formation of GdnSS and GdnS-SG in the GdnSH/GSSG system (Fig. S3 in ESI). Based on this result, we monitored the oxidation reaction of RNase A in GdnSH/GdnSS system, where the oxidant is limited only to GdnSS in this condition. Interestingly, GdnSH/GdnSS system showed a very similar oxidation rate of RNase A to GdnSH/GSSG system (Fig. S4 in ESI, blue and red lines), suggesting that GdnSS is involved in the oxidation of RNase A in the GdnSH/GSSG system. This is also consistent with the redox properties that GdnSH has higher E0′ than GSH, where a redox-active compound having higher E0′ is known to work as an oxidizing agent. It is also expected that GdnS-SG has higher E0′ than GSSG because of the GdnS-component included. Hence, it is likely that the oxidation of RNase A is proceeded mainly by GdnSS and, presumably, GdnS-SG as well. Herein, a thiol group with higher nucleophilicity (i.e. lower pKa) performs nucleophilic attack to a disulfide bond in a protein more efficiently, which allows for an effective conversion of the disulfide bond from a non-native form to the native form. Thus, the dual property of GdnSH having strong oxidizability and high nucleophilicity, indicated by its higher E0′ and lower pKa values than GSH and UreaSH, likely allows for the high promotion effect of the oxidative protein folding.

Recent studies reported a link between the redox chemistry and the pathogenesis of misfolding diseases. The protein disulfide isomerase (PDI) family24 is regarded as a novel target for treatment of several neurodegenerative disorders, including Alzheimer's disease and amyotrophic lateral sclerosis. Neurodegenerative disease-related PDI family mutants25 or S-nitrosylation26 cause loss of enzymatic function under the glutathione-controlled redox condition, resulted in accumulation of misfolded proteins. Our findings provide a proof of concept for an oxidative protein folding-accelerating redox molecule, which exerts a notable effect of the guanidyl group on the folding assistance. Thus, a highly reactive redox agent might permit the development of new therapeutic approaches for PDI family-related neurodegenerative diseases and disorders associated with the accumulation of aberrant proteins.

In summary, on the basis of the concept of covalent-coupling between thiol and urea-type –NHC([double bond, length as m-dash]X)NH2 groups, UreaSH and GdnSH were developed. By quantitative analyses of folding using model proteins, RNase A and BPTI, it was demonstrated that GdnSH (X = NH) and UreaSH (X = O) are effective in promoting oxidative protein folding compared to GSH. It was also demonstrated that GdnSH shows a significant promotion and acceleration effect on oxidative protein folding. We believe our concept to couple a thiol group with functional units including hydrogen-bond forming units will expand the designs of redox-active compounds as efficient synthetic promoters of oxidative protein folding.

This work was supported by JSPS and MEXT Grants-in-Aid for Scientific Research Numbers 17H04885 (TM), 17H05147 (TM), 17K15098 (MO), 17H05868 (MO), 17K18123 (KA), 18H03978 (KI), Urakami Foundation for Food and Food Culture Promotion (TM), MEXT LEADER (TM), the Takeda Science Foundation (KI and MO) and the Naito Foundation (MO). This work was performed under the Cooperative Research Program of “NJRC Mater. & Dev.”.

Conflicts of interest

The authors declare no competing interests.

Notes and references

  1. C. B. Anfinsen, Science, 1973, 181, 223 CrossRef CAS.
  2. C. M. Dobson, Nature, 2003, 426, 884 CrossRef CAS.
  3. T. E. Creighton, Biol. Chem., 1997, 378, 731 CAS.
  4. M. Narayan, E. Welker, W. J. Wedemeyer and H. A. Scheraga, Acc. Chem. Res., 2000, 33, 805 CrossRef CAS.
  5. F. Baneyx and M. Mujacic, Nat. Biotechnol., 2004, 22, 1399 CrossRef CAS.
  6. J. L. Arolas, F. X. Aviles, J.-Y. Chang and S. Ventura, Trends Biochem. Sci., 2006, 31, 292 CrossRef CAS.
  7. W. J. Lees, Curr. Opin. Chem. Biol., 2008, 12, 740 CrossRef CAS.
  8. M. Okumura, S. Shimamoto and Y. Hidaka, FEBS J., 2012, 279, 2283 CrossRef CAS.
  9. C. Hwang, A. J. Sinskey and H. F. Lodish, Science, 1992, 257, 1496 CrossRef CAS.
  10. H. Sies, Free Radical Biol. Med., 1999, 27, 916 CrossRef CAS.
  11. P. I. Merksamer, A. Trusina and F. R. Papa, Cell, 2008, 135, 933 CrossRef CAS PubMed.
  12. E. Welker, M. Narayan, W. J. Wedemeyer and H. A. Scheraga, Proc. Natl. Acad. Sci. U. S. A., 2001, 98, 2312 CrossRef CAS.
  13. J. C. Lukesh, III, M. J. Palte and R. T. Raines, J. Am. Chem. Soc., 2012, 134, 4057 CrossRef.
  14. J. D. Gough, R. H. Williams, A. E. Donofrio and W. J. Lees, J. Am. Chem. Soc., 2002, 124(15), 3885 CrossRef CAS.
  15. A. S. Patel and W. J. Lees, Bioorg. Med. Chem., 2012, 20, 1020 CrossRef CAS.
  16. J. C. Lukesh, III, K. A. Andersen, K. K. Wallin and R. T. Raines, Org. Biomol. Chem., 2014, 12, 8598 RSC.
  17. J. Beld, K. J. Woycechowsky and D. Hilvert, Biochemistry, 2007, 46, 5382 CrossRef CAS.
  18. P. S. Reddy and N. Metanis, Chem. Commun., 2016, 52, 3336 RSC.
  19. K. Arai, H. Ueno, Y. Asano, G. Chakrabarty, S. Shimodaira, G. Mugesh and M. Iwaoka, ChemBioChem, 2018, 19, 207 CrossRef CAS.
  20. Reported pKa value of GSH is 9.169 ± 0.115; S.-S. Tang and G.-G. Chang, J. Org. Chem., 1995, 60, 6183 CrossRef CAS.
  21. M. Okumura, S. Shimamoto, T. Nakanishi, Y. Yoshida, T. Konogami, S. Maeda and Y. Hidaka, FEBS Lett., 2012, 586, 3926 CrossRef CAS.
  22. J. S. Weissman and P. S. Kim, Science, 1991, 253, 1386 CrossRef CAS.
  23. L. Ito, K. Shiraki, T. Matsuura, M. Okumura, K. Hasegawa, S. Baba, H. Yamaguchi and T. Kumasaka, Protein Eng., Des. Sel., 2011, 24, 269 CrossRef CAS.
  24. M. Okumura, H. Kadokura and K. Inaba, Free Radical Biol. Med., 2015, 83, 314 CrossRef CAS.
  25. U. Woehlbier, et al. , EMBO J., 2016, 35, 845 CrossRef CAS.
  26. T. Uehara, T. Nakamura, D. Yao, Z. Q. Shi, Z. Gu, Y. Ma, E. Masliah, Y. Nomura and S. A. Lipton, Nature, 2006, 441, 513 CrossRef CAS.


Electronic supplementary information (ESI) available. See DOI: 10.1039/c8cc08657e
These authors contributed equally to this work.

This journal is © The Royal Society of Chemistry 2019