English

• Chemical protein synthesis has emerged as a powerful tool for investigating proteins that are inaccessible through recombinant DNA technology, such as those with post-translational modifications, containing non-natural amino acids, or toxic to protein expression systems [1-7]. This process typically involves merging multiple peptide fragments produced through solid-phase peptide synthesis (SPPS) to achieve the full-length molecule (Scheme 1) [8,9]. Among the various methods employed for fragment merging, the one-pot approach, based on kinetically controlled native chemical ligation (NCL) (Scheme 1A) [10,11] or stepwise hydrazide oxidation [12,13], is particularly advantageous. By bypassing time-consuming and product-loss-prone purification steps, the kinetically controlled NCL significantly enhances the convenience and efficiency of synthesis [14-17].

  Scheme 1

  

  Scheme 1.  One-pot protein synthesis. (A) Conventional approach. (B) Streamlined approach developed in this study. P1, P2 and P3 represent peptides, R1 and R2 denote amino acid side chains, and Ar1 and Ar2 represent aryl groups. * indicates potential additional steps.

  DownLoad: Full-Size Img PowerPoint

  The kinetically controlled NCL has facilitated the synthesis of many proteins [18-20]. However, this approach involves an undesirable step: the thiol-thioester exchange, which activates inactive alkylthioesters with arylthiol additives, such as 4-mercaptophenylacetic acid (MPAA) (Scheme 1A) [21]. To ensure the completion of the thiol-thioester exchange reaction, a large excess of MPAA, typically over 100 equiv., is required [22]. This large amount of MPAA can complicate both reaction monitoring and the subsequent desulfurization step [23], which is necessary when non-Cys-based NCL is used for synthesis. To facilitate efficient desulfurization, an additional step is required to remove MPAA, potentially compromising the overall yield [24,25]. Efforts have been made to address these issues by simplifying the additive removal step or eliminating it through the development of thiol alternatives that do not interfere with desulfurization [26-32]. Nonetheless, the necessity for an additional thiol-thioester exchange step using a large excess of thiol additives still persists. We envisioned that by harnessing the distinct reactivity of arylthioesters, it might be possible to kinetically control the reaction sequence without needing to activate the less-reactive thioester, thus enabling a shorter and more convenient protein synthesis procedure featuring one-pot ligation and desulfurization (Scheme 1B) [33].

  The key to this advancement lies in identifying two arylthioesters that can both directly participate in NCL reactions while exhibiting sufficient reactivity differences to prevent competition in a one-pot synthesis setup (Scheme 1B). Back in 2006, Kent and colleagues evaluated the ligation activity of various arylthioesters [22]. Their efforts led to the discovery of the useful thiol additive MPAA. However, the ligation activity results for the tested arylthioesters were heavily influenced by the thiol-thioester exchange rate required to generate them, not accurately reflecting their intrinsic ligation capabilities. Consequently, these findings cannot inform the development of the envisioned one-pot approach. To realize this vision, a more precise assessment of arylthioester ligation activity is essential.

  We utilized the ligation reaction between 17 representative peptide arylthioesters 1a-q (Fig. 1A and Figs. S1–S17 in Supporting information) and a Cys-containing peptide 2 (Fig. 1A and Fig. S18 in Supporting information) to achieve this goal. The arylthioesters, each with a C-terminal Ala residue, were synthesized using our previously adopted epimerization-free method, employing commercially available arylthiols with pKa values ranging from approximately 3 to 7 (Scheme S1 in Supporting information) [34,35]. Peptide 2 was directly prepared using SPPS. The ligation reaction was initiated by mixing each peptide arylthioester with 1 equiv. of peptide 2 in a buffer containing 6 mol/L Gn·HCl, 100 mmol/L Na₂HPO₄, 50 mmol/L TCEP, at pH 7.5 and 25 ℃ [11]. The final concentration of each peptide was 1.0 mmol/L [36]. Samples were withdrawn at various intervals during the ligation reaction, and the ligation product was quantified based on the UV absorbance obtained in the UPLC-MS analysis. The ligation activity of each arylthioester was determined by measuring the amount of desired product generated one minute after initiating the ligation reaction, corresponding to the end of the initial linear phase of the reaction-time curve (Fig. 1B and Fig. S19 in Supporting information) [22].

  Figure 1

  

  Figure 1.  Assessment of the ligation activity of arylthioesters. (A) Structures of 17 representative arylthioesters, with the pKa values of their corresponding arylthiols shown in blue within parentheses. (B) Illustration of the method for evaluating the ligation activity of arylthioesters.

  DownLoad: Full-Size Img PowerPoint

  In this research design, peptide arylthioesters were no longer generated in situ through the thiol-thioester exchange reaction, thereby eliminating variability in their formation rates from influencing ligation activity. Additionally, the ligation activity was assessed using commonly employed short model peptide fragments [37,38], 1 and 2, facilitating comparison with other related research results. Moreover, these fragments were specifically designed to contain Ala and Cys at the ligation junction. Since these two amino acids have small side chains, this design is anticipated to circumvent factors such as steric effects that may interfere with the ligation activity [34,39].

  Comparison of the ligation activity of all the arylthioesters revealed a surprising, yet somewhat expected trend, almost completely reversing the findings of Kent and colleagues [22]. Our results indicated that arylthioesters derived from arylthiols with smaller pKa values exhibited faster reaction activity (Fig. 2).

  Figure 2

  

  Figure 2.  Comparison of the ligation activity of 17 arylthioesters. Inset: Pearson correlation for the 15 arylthioesters highlighted in blue (arylthiol pKa: 4.0–7.0).

  DownLoad: Full-Size Img PowerPoint

  Specifically, the peptide arylthioester 1b, derived from 4-nitro-thiophenol with a pKa value of approximately 4.68, displayed the highest reaction activity. As the pKa values of the arylthiols increased gradually, the corresponding arylthioesters exhibited a gradual decrease in ligation activity. Impressively, our analysis uncovered a strong correlation between the ligation activity of the middle 15 arylthioesters (1b-p) and the pKa values of their corresponding arylthiols, with a Pearson correlation coefficient of 0.88. This finding suggests that the reactivity of arylthioesters in NCL is primarily determined by the pKa values of their corresponding arylthiols. Although some highly reactive arylthioesters may experience minor hydrolysis during the ligation process, the variation in the rate of this side reaction follows a trend similar to that of the ligation rate. Therefore, the primary role of pKa in determining ligation activity is unlikely to be significantly affected by the presence of hydrolysis. The reversed trend observed in the previous study reflects the activity of the arylthiols in the thiol-thioester exchange step, rather than the NCL step.

  Another notable observation is the unusual change in the ligation activity of 1a and 1q, whose corresponding arylthiols have pKa values at the two extremes of the pKa range (Fig. 2, highlighted in yellow). To uncover the reason behind this variation, we calculated the free energy profiles corresponding to the NCL reactions of arylthioesters 1b, 1p, and 1q using the density functional theory (DFT) method. Since the widely accepted mechanism for NCL involves two steps. A reversible transthioesterification step and a subsequent intramolecular S-N acyl transfer [40], and the second step is consistent for all variants, our investigation focused only on the transthioesterification step. To simplify this investigation [40], the structures of the three arylthioesters, 1b, 1p, and 1q, were reduced to Ac-Ala-S-Ph(4-NO₂) (1b’), Ac-Ala-S-Ph(2,5-diCH₃) (1p’), and Ac-Ala-S-Ph(2,6-diCH₃) (1q’) for the calculations. Similarly, the structure of peptide fragment 2 was reduced to H—Cys-NHMe 2′ (Fig. 3).

  Figure 3

  

  Figure 3.  DFT-calculated free energy profiles for the transthioesteri-fication step explaining the observed unusual ligation activity of arylthioesters.

  DownLoad: Full-Size Img PowerPoint

  Comparison of the initial reaction complexes formed by 1b’ and 1p’ with 2′ revealed that the distance between the attacking S atom of 2′ and the carbonyl C atom in 1p’ is longer than that between the same S atom and the carbonyl C in 1b’ (Fig. 3 upper panel, 3.513 and 3.395 Å, respectively). This suggests that the approach of the S atom to the carbonyl group in 1p’ is more restricted, likely due to steric hindrance conferred by the 2-CH₃ group on the aromatic ring. This, combined with the difference in energy required for the subsequent breaking of the original C-S bond in 1b’ and 1p’, leads to distinct activation free energies of 55.7 kJ/mol and 66.1 kJ/mol, respectively, explaining the observed difference in the ligation activity of 1b and 1p (Fig. 3 lower panel). Comparison of the reactions of 1p’ and 1q’ with 2′ revealed that an additional steric effect caused by the 6-CH₃ group on the aromatic ring further hinders the attack of the S atom of 2′ on the carbonyl C in 1q’, resulting in a much higher activation free energy of 80.3 kJ/mol, consequently significantly slowing the ligation activity. The irregular ligation activity of 1a is likely due to a similar reason.

  With a large set of intrinsic ligation activity data in hand, we could now explore the execution of a one-pot synthesis as outlined in Scheme 1B. Instead of directly pursuing a protein synthesis, we first verified the feasibility of this approach through the synthesis of an artificial peptide 5, which was designed based on the sequences of model peptides 1 and 2. This design allowed us to conveniently utilize the already available synthetic fragments, thereby facilitating a quicker test (Fig. 4). To our happiness, by leveraging the reactivity difference between the arylthioesters containing C-terminal Ala-S-Ph(4-NO₂) and Ala-S-Ph(2,6-diCH₃), we successfully accomplished the synthesis of peptide 5. The initial ligation of 1.3 equiv. of 1b and 1.0 equiv. of 2 proceeded smoothly and efficiently, completing in 1 h at 25 ℃. Subsequently, the second ligation was initiated by adding 2.0 equiv. of a new fragment, labeled as 4, to the reaction mixture, and this step was completed in 12 h at 37 ℃. After all ligation reactions were finished, a desulfurization buffer containing VA-044 and tBuSH was directly added to the reaction solution to produce the desired product 5 at 37 ℃ in 10 h. A satisfactory 56% isolated yield was achieved by directly purifying the reaction mixture using HPLC.

  Figure 4

  

  Figure 4.  Feasibility test for the envisioned one-pot protein synthesis approach by leveraging the differences in arylthioester ligation activity conferred by C-terminal Ala-S-Ph(4-NO₂) and Ala-S-Ph(2,6-diCH₃). The lower portion shows the LC-MS characterization of the artificial peptide 5, which has calculated m/z values of [M + 2H]²⁺, 1014.52 and [M + 3H]³⁺, 676.69.

  DownLoad: Full-Size Img PowerPoint

  Finally, we demonstrated the robustness of this one-pot method through the synthesis of a real-world protein. Consistent with the primary focus of our laboratory, we chose the unglycosylated form of human mucin-like protein 1 9, consisting of 70 amino acids and containing a Cys69 to Ser mutation, as the synthetic target [41-43]. In its natural form, this protein bears various glycans on its Ser and Thr residues. The method for the synthesis of 9 closely mirrored that for 6 (Fig. 5). After the iterative ligation steps involving fragments 6,7, and 8, direct one-pot desulfurization and HPLC purification of the reaction mixture yielded the desired product 9 with a pleasing 32% isolated yield.

  Figure 5

  

  Figure 5.  One-pot chemical synthesis of human mucin-like protein 1 9. Its calculated m/z values are [M + 4H]⁴⁺, 1723.10; [M + 5H]⁵⁺, 1378.68; [M + 6H]⁶⁺, 1149.07 and [M + 7H]⁷⁺, 985.06.

  DownLoad: Full-Size Img PowerPoint

  In summary, this study focuses on improving the highly sought-after multi-fragment, one-pot chemical protein synthesis. By successfully synthesizing previously challenging-to-produce peptide arylthioesters with systematically varied structures, we were able to determine the differences in their ligation activity and reveal an interesting reversal trend compared to that of previous research, i.e., the increase in the ligation activity of arylthioesters closely correlates with the decrease in the pKa values of their corresponding arylthiols. Furthermore, DFT calculation findings suggest that molecular forces such as steric hindrance and C-S bond dissociation energy may together contribute to the observed ligation activity differences of the arylthioesters. Finally, the usefulness of these discoveries is effectively validated through the exploitation of the distinct reactivity of arylthioesters in the one-pot synthesis of an artificial peptide and human mucin-like protein 1. Overall, this research opens a new avenue for one-pot chemical protein synthesis. Future research in this direction is expected to continually ease the process for synthesizing proteins and modified proteins, thus more rapidly advancing their research and application.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Min Fu: Writing – review & editing, Visualization, Investigation, Formal analysis, Data curation. Ruihan Wang: Writing – review & editing, Visualization, Software, Methodology, Investigation, Formal analysis, Data curation. Wenqiang Liu: Investigation, Data curation. Sen Zhou: Investigation, Formal analysis, Data curation. Chunhong Zhong: Investigation, Data curation. Yaohao Li: Investigation, Conceptualization. Pan He: Data curation. Xin Li: Investigation, Formal analysis, Data curation. Shiying Shang: Writing – review & editing. Zhongping Tan: Writing – review & editing, Writing – original draft, Supervision, Resources, Project administration, Funding acquisition, Conceptualization.

  Acknowledgments

  The authors thank the CAMS Innovation Fund for Medical Sciences (CIFMS, No. 2021-I2M-1-026), the National Key R & D Program of China (No. 2018YFE0111400), the NIH Research Project Grant Program (No. R01 EB025892), the National Natural Science Foundation of China (the Training Program of the Major Research Plan, No. 91853120), the National Major Scientific and Technological Special Project of China (Nos. 2018ZX09711001-005 and 2018ZX09711001-013), the State Key Laboratory of Bioactive Substance and Function of Natural Medicines, Institute of Materia Medica, and the Biomedical High Performance Computing Platform, Chinese Academy of Medical Sciences, the Chinese Academy of Medical Sciences and Peking Union Medical College for funding and support.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110542.

  
  
• 1. [1]

     P.E. Dawson, S.B.H. Kent, Annu. Rev. Biochem. 69 (2000) 923–960. doi: 10.1146/annurev.biochem.69.1.923

  2. [2]

     T.W. Muir, Annu. Rev. Biochem. 72 (2003) 249–289. doi: 10.1146/annurev.biochem.72.121801.161900

  3. [3]

     B.L. Nilsson, M.B. Soellner, R.T. Raines, Annu. Rev. Biophys. Biomol. Struct. 34 (2005) 91–118. doi: 10.1146/annurev.biophys.34.040204.144700

  4. [4]

     S.S. Kulkarni, J. Sayers, B. Premdjee, R.J. Payne, Nat. Rev. Chem. 2 (2018) 0122.

  5. [5]

     L. Liu, Isr. J. Chem. 59 (2019) 64–70.

  6. [6]

     Z.Q. Sun, H. Liu, X.C. Li, Chem 10 (2024) 767–799.

  7. [7]

     S. Dong, J.S. Zheng, Y. Li, et al., Sci. China Chem. 67 (2024) 1060–1096. doi: 10.1007/s11426-024-1950-1

  8. [8]

     P.E. Dawson, T.W. Muir, I. Clark-Lewis, S.B.H. Kent, Science 266 (1994) 776–779. doi: 10.1126/science.7973629

  9. [9]

     L. CF, J.P. Tam, J. Am. Chem. Soc. 116 (1994) 4149–4153.

  10. [10]

      D. Bang, B.L. Pentelute, S.B. Kent, Angew. Chem. Int. Ed. 45 (2006) 3985–3988. doi: 10.1002/anie.200600702

  11. [11]

      Z.P. Tan, S.Y. Shang, S.J. Danishefsky, Angew. Chem. Int. Ed. 49 (2010) 9500–9503. doi: 10.1002/anie.201005513

  12. [12]

      G.M. Fang, Y.M. Li, F. Shen, et al., Angew. Chem. Int. Ed. 50 (2011) 7645–7649. doi: 10.1002/anie.201100996

  13. [13]

      G.M. Fang, J.X. Wang, L. Liu, Angew. Chem. Int. Ed. 51 (2012) 10347–10350. doi: 10.1002/anie.201203843

  14. [14]

      J.Y. Lee, D. Bang, Biopolymers 94 (2010) 441–447. doi: 10.1002/bip.21379

  15. [15]

      L.R. Malins, R.J. Payne, Curr. Opin. Chem. Biol. 22 (2014) 70–78.

  16. [16]

      C. Zuo, B. Zhang, B. Yan, J.S. Zheng, Org. Biomol. Chem. 17 (2019) 727–744. doi: 10.1039/c8ob02610f

  17. [17]

      V. Agouridas, O. El Mahdi, V. Diemer, et al., Chem. Rev. 119 (2019) 7328–7443. doi: 10.1021/acs.chemrev.8b00712

  18. [18]

      S. Tang, Y.Y. Si, Z.P. Wang, et al., Angew. Chem. Int. Ed. 54 (2015) 5713–5717. doi: 10.1002/anie.201500051

  19. [19]

      C.L. Lee, H. Liu, C.T. Wong, et al., J. Am. Chem. Soc. 138 (2016) 10477–10484. doi: 10.1021/jacs.6b04238

  20. [20]

      D.L. Huang, Y. Li, J. Liang, et al., J. Am. Chem. Soc. 142 (2020) 8790–8799. doi: 10.1021/jacs.0c01561

  21. [21]

      G. Hayashi, K. Nakatsu, S. Suzuki, Y. Nakajima, Amino Acid. Pept. Protein. 45 (2024) 1–26. doi: 10.1039/bk9781839169328-00001

  22. [22]

      E.C. Johnson, S.B. Kent, J. Am. Chem. Soc. 128 (2006) 6640–6646. doi: 10.1021/ja058344i

  23. [23]

      T. Moyal, H.P. Hemantha, P. Siman, et al., Chem. Sci. 4 (2013) 2496–2501. doi: 10.1039/c3sc50239b

  24. [24]

      P. Siman, O. Blatt, T. Moyal, et al., ChemBioChem 12 (2011) 1097–1104. doi: 10.1002/cbic.201100033

  25. [25]

      L.M. Gatzemeier, F. Meyer, U. Diederichsen, T.F. Outeiro, Chem. Eur. J. 29 (2023) e202300649.

  26. [26]

      R.E. Thompson, X. Liu, N. Alonso-García, et al., J. Am. Chem. Soc. 136 (2014) 8161–8164. doi: 10.1021/ja502806r

  27. [27]

      Y.C. Huang, C.C. Chen, S. Gao, et al., Chem. Eur. J. 22 (2016) 7623–7628. doi: 10.1002/chem.201600101

  28. [28]

      K. Ohkawachi, D. Kobayashi, K. Morimoto, et al., Org. Lett. 22 (2020) 5289–5293. doi: 10.1021/acs.orglett.0c01383

  29. [29]

      S. Suzuki, Y. Nakajima, N. Kamo, et al., Molecules 28 (2023) 3655. doi: 10.3390/molecules28093655

  30. [30]

      K. Sakamoto, S. Tsuda, M. Mochizuki, et al., Chem. Eur. J. 22 (2016) 17940–17944. doi: 10.1002/chem.201604320

  31. [31]

      K. Sakamoto, S. Tsuda, H. Nishio, T. Yoshiya, Chem. Commun. 53 (2017) 12236–12239.

  32. [32]

      J. Zhang, H.Y. Liu, S. Teng, et al., Chem. Commun. 59 (2023) 6513–6516. doi: 10.1039/d3cc00983a

  33. [33]

      K.M. Koeller, C.H. Wong, Chem. Rev. 100 (2000) 4465–4494.

  34. [34]

      M. Fu, P. He, S. Zhou, et al., Chin. Chem. Lett. 35 (2024) 109434.

  35. [35]

      S.Y. Shang, Z.P. Tan, S.J. Danishefsky, Proc. Natl. Acad. Sci. U. S. A. 108 (2011) 5986–5989. doi: 10.1073/pnas.1103118108

  36. [36]

      T. Durek, P.F. Alewood, Angew. Chem. Int. Ed. 50 (2011) 12042–12045. doi: 10.1002/anie.201105512

  37. [37]

      R.J. Payne, S. Ficht, W.A. Greenberg, C.H. Wong, Angew. Chem. Int. Ed. 47 (2008) 4411–4415. doi: 10.1002/anie.200705298

  38. [38]

      G.L. Thomas, Y.S. Hsieh, C.K. Chun, et al., Org. Lett. 13 (2011) 4770–4773. doi: 10.1021/ol2017356

  39. [39]

      T.M. Hackeng, J.H. Griffin, P.E. Dawson, Proc. Natl. Acad. Sci. U. S. A. 96 (1999) 10068–10073.

  40. [40]

      C. Wang, Q.X. Guo, Y. Fu, Chem. Asian J. 6 (2011) 1241–1251. doi: 10.1002/asia.201000760

  41. [41]

      B. Ma, X.Y. Guan, Y.H. Li, et al., Front Chem. 8 (2020) 622.

  42. [42]

      G.J. Strous, Crit. Rev. Biochem. Mol. Biol. 27 (1992) 57–92. doi: 10.3109/10409239209082559

  43. [43]

      S.S. Dhanisha, C. Guruvayoorappan, S. Drishya, P. Abeesh, Crit. Rev. Oncol. Hematol. 122 (2018) 98–122.

• 

  Scheme 1  One-pot protein synthesis. (A) Conventional approach. (B) Streamlined approach developed in this study. P1, P2 and P3 represent peptides, R1 and R2 denote amino acid side chains, and Ar1 and Ar2 represent aryl groups. * indicates potential additional steps.

  下载: 全尺寸图片 幻灯片
  

  Figure 1  Assessment of the ligation activity of arylthioesters. (A) Structures of 17 representative arylthioesters, with the pKa values of their corresponding arylthiols shown in blue within parentheses. (B) Illustration of the method for evaluating the ligation activity of arylthioesters.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Comparison of the ligation activity of 17 arylthioesters. Inset: Pearson correlation for the 15 arylthioesters highlighted in blue (arylthiol pKa: 4.0–7.0).

  下载: 全尺寸图片 幻灯片
  

  Figure 3  DFT-calculated free energy profiles for the transthioesteri-fication step explaining the observed unusual ligation activity of arylthioesters.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Feasibility test for the envisioned one-pot protein synthesis approach by leveraging the differences in arylthioester ligation activity conferred by C-terminal Ala-S-Ph(4-NO₂) and Ala-S-Ph(2,6-diCH₃). The lower portion shows the LC-MS characterization of the artificial peptide 5, which has calculated m/z values of [M + 2H]²⁺, 1014.52 and [M + 3H]³⁺, 676.69.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  One-pot chemical synthesis of human mucin-like protein 1 9. Its calculated m/z values are [M + 4H]⁴⁺, 1723.10; [M + 5H]⁵⁺, 1378.68; [M + 6H]⁶⁺, 1149.07 and [M + 7H]⁷⁺, 985.06.

  下载: 全尺寸图片 幻灯片
•