A shelf stable Fmoc hydrazine resin for the synthesis of peptide hydrazides

Abstract C‐terminal hydrazides are an important class of synthetic peptides with an ever expanding scope of applications, but their widespread application for chemical protein synthesis has been hampered due to the lack of stable resin linkers for synthesis of longer and more challenging peptide hydrazide fragments. We present a practical method for the regeneration, loading, and storage of trityl‐chloride resins for the production of hydrazide containing peptides, leveraging 9‐fluorenylmethyl carbazate. We show that these resins are extremely stable under several common resin storage conditions. The application of these resins to solid phase peptide synthesis (SPPS) is demonstrated through the synthesis of the 40‐mer GLP‐1R agonist peptide “P5”. These studies support the broad utility of Fmoc‐NHNH‐Trt resins for SPPS of C‐terminal hydrazide peptides.


| INTRODUCTION
The hydrazide moiety has been ubiquitous in the field of peptide chemistry for more than 50 years, [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17] primarily due to its utility as a precursor to acyl azides, used for amide couplings [1][2][3][4][5][6][7][8][9][10][11][12][13]18] and thioester formation [14][15][16][17] to enable Native Chemical Ligation. [19] Our group recently published mild conditions for formation of peptide α-thioesters from hydrazides via the Knorr pyrazole synthesis. [17] This method avoids the need for strong oxidizing conditions, enabling broader compatibility with common synthetic peptide functionalities, and has been widely adopted in a broad range of applications. [20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37] The increased use of C-terminal α-hydrazides in peptide chemistry creates a demand for robust synthetic tools for accessing these moieties. A number of methods have been published to allow access to C-terminal hydrazides on both synthetic [15,16,38] and expressed [15,39,40] peptides and proteins. The current synthetic methods suffer from one of two drawbacks. They either require a two part deprotection procedure involving hydrazinolysis followed by global deprotection, [38] or rely on time-of-use formation of hydrazine resins from trityl chloride resins, where the hydrazine form is not shelf stable. [16] The chloride form may also need to be regenerated prior to use, making solid phase peptide synthesis (SPPS) of peptide hydrazides significantly less convenient than standard amides or acids. [41][42][43] Chemical protein synthesis requires the solid phase synthesis of long and hydrophobic peptide fragments, which has been addressed through the use of higher swelling polyethylene glycol (PEG) containing resins. [44][45][46][47][48] When using PEGylated trityl chloride resins (e.g., TentaGel ® and ChemMatrix ® ), the yields of synthetic peptides are often significantly lower than expected and decrease over time after resins are first opened. [49] In addition, while there has been a move toward increased automation and real-time monitoring of peptide synthesis and purification workflows, regenerating and hydrazine loading a trityl chloride resin prior to each synthesis represents a manual process that must be performed in a fume hood, and cannot be monitored other than by initiation of synthesis. We therefore sought to develop a procedure for regenerating and loading trityl chloride resins that, upon receipt from the manufacturer, would allow for immediate loading evaluation and long term stable storage in a state that is competent for automated synthesis of peptide α-hydrazides without any extraneous steps. 9-Fluorenylmethyl carbazate (Fmoc-NHNH 2 ) was identified as an ideal reagent for preparation and long term storage of peptide-hydrazide synthesis resins (Figure 1), and the relative stability and regeneratability of TentaGel ® resins loaded with Fmoc-NHNH 2 as compared to chloride or hydrazine was explored.    ð Þ , where w is the weight of the resin used in milligrams. [50] Reported loadings are the arithmetic mean of the triplicate loading tests ± 1 standard deviation, and error bars in figures represent ± 1 standard deviation.

| Peptide synthesis and cleavage
All peptides were synthesized by conventional Fmoc/tBu chemistry using either [ Cleavage reactions were carried out for 1 h at room temperature and TFA removed by nitrogen flow or rotary evaporation. Crude peptide was recovered by precipitation in diethyl ether. TFA is a corrosive and hazardous acidic reagent that fumes profusely under standard atmospheric conditions. It should only be used in a properly ventilated fume hood alongside proper gloves, lab coats, and eye protection. It represents an environmental hazard and must not be disposed of in drains.
While evaporation under a nitrogen stream has been a common technique, the use of rotary evaporation followed by appropriate hazardous waste disposal reduces the environmental impact of TFA usage.    Table S3).

| Peptide characterization and purification
In order to establish whether this stability enhancement is unique to the Fmoc-NHNH 2 loaded resins as compared to previously reported hydrazide loaded resins [16] sample D from Figure 2  with the losses believed to be caused by incomplete coupling or partial deprotection of the Fmoc-glycine. Meanwhile, the resin stored as the free hydrazine (D2) had a final loading of 0.0224 ± 0.001 mmol/g, just 9.6% of its initial value ( Figure S3, Table S4). This indicates the Fmoc group is required to maintain stable loading of the resin.
Another strategy that could help to protect resins from hydrolysis, and that would allow rapid initiation of syntheses would be to maintain pre-measured and swollen resins. To determine if this is a viable strategy for Fmoc-NHNH-Trityl resins a supply of Tentagel ® XV trityl chloride with a manufacturer reported loading of 0.22 mmol/g was regenerated and substituted with Fmoc-NHNH 2 . After thoroughly washing with DMF and DCM and drying with Et 2 O and a vacuum desiccator the loading was measured to be 0.221 ± 0.007 mmol/g. The dry resin was split into eight fractions, all of which were stored together in a À10 C freezer for 30 days. Prior to measuring loading retention each resin was  Table S5).
We further investigated the stability of the resins at room temperature without vacuum storage. Two separate samples of resin measured at 0.209 ± 0.003 and 0.213 ± 0.002 mmol/g, respectively were stored in a vacuum desiccator for 7 days, then removed and stored, open to air, at room temperature, for a further 23 days. These resins were then thoroughly washed and dried with Et 2 O followed by overnight desiccation, and their loadings were re-tested in triplicate, showing loading retentions of greater than 95% (Figure 4, Tables S6 and S7).
To demonstrate the capacity of Fmoc-NHNH-Trityl resins for the synthesis of large peptides, the 40 amino acid GLP1-R agonist peptide P5 [51] was synthesized at 0.1 mmol scale using HATU/DIEA conditions on a Fmoc-NHNH-Trityl Tentagel ® XV (0.221 ± 0.007 mmol/g). conditions. We hope this will encourage more groups to explore the myriad applications of hydrazide containing peptides, and that resin manufacturers might make available trityl chloride resins that are preloaded with Fmoc-NHNH 2 to make these methods available to the broadest possible scope of researchers.

ACKNOWLEDGMENTS
Financial support for this work was provided by NIH (R21GM132787) and MJB was supported by the Skaggs Fellowship in Molecular Medicine. The authors wish to thank Wolfgang Rapp for his assistance in identifying conditions for the drying of Tentagel ® resins, and for his helpful input on the stability of PEG containing trityl chloride resins.

DATA AVAILABILITY STATEMENT
The data that supports the findings of this study are available in the supplementary material of this article.