A Versatile "Synthesis Tag" (SynTag) for the Chemical Synthesis of Aggregating Peptides and Proteins

Abstract

Solid-phase peptide synthesis (SPPS) and native chemical ligation (NCL) are powerful methods for obtaining peptides and proteins that are otherwise inaccessible. Nonetheless, numerous sequences are difficult to prepare via SPPS, and cleaved peptides often have low aqueous solubility. To address these challenges, we developed a "Synthesis Tag" consisting of six arginines connected to the target sequence via a cleavable MeDbz linker. "SynTag" effectively improves batch- and flow-SPPS of "difficult sequences", enhances the solubility of the cleaved peptides, and provides direct access to native sequences by hydrolysis, or peptide thioesters for NCL. We demonstrate its utility in the first chemical synthesis of the MYC transactivation domain with a single NCL. We envisage SynTag to become a broadly applicable tool that enables the synthesis and study of previously unattainable peptides and proteins.

Figure 1

Flow-based solid-phase peptide synthesis enabled the development of a broadly applicable synthesis tag (SynTag), that simultaneously improves peptide crude purities and solubilities. (A) Chemical protein synthesis often consists of two steps: stepwise SPPS followed by native chemical ligation. Major challenges are (I) sequence dependence of SPPS, often framed as “difficult peptides” (aggregation), associated with low yield and purity, and (II) low solubility of final peptide fragments for HPLC and NCL. (B) Impact of aggregation and solubility on crude peptide analysis (UHPLC) displayed for four cases: NBDY, Barstar[75–90], Aβ [27–42] and MYC[85–143]. While aggregation leads to increased side products, low solubility leads to characteristic UHPLC peak broadening. Note that only small amounts of Aβ [27–42] could be loaded onto the UHPLC column to minimize further peak broadening, resulting in comparatively low absorbance. (C) Flow-based peptide synthesis allows for in-line UV–vis monitoring and the detection of aggregation through shape change of Fmoc-deprotection peaks. Upon aggregation, the resin starts to agglomerate, reducing coupling efficiencies and impacting crude purities. (D) This work: a versatile SynTag that reduces aggregation during SPPS, improves crude peptide solubility, and delivers either native peptides or substrates for NCL.

Figure 2

A C-terminal sequence of six Arg(Pbf) decreases aggregation in flow- and batch-SPPS. (A) Aggregation as a function of Fmoc deprotection peak broadening by in-line UV–vis (310 nm) in flow-SPPS for various amino acid tags. The first amino acids coupled alter the severity and onset of aggregation for Barstar[75–90]. All peptides were synthesized by AFPS at 90 °C. (B) Aggregation detected by in-line UV–vis directly translates to crude purity of cleaved and deprotected Barstar[75–90] and hGH[176–191]F176Y (abbreviated as hGH[176–191]). Crude purity was determined via UHPLC (214 nm). (C) Adding C-terminal, but not N-terminal [Arg(Pbf)]₆ improves crude peptide purity of Barstar[75–90] synthesized by batch-SPPS at room temperature. (D) C-terminal [Arg(Pbf)]₆ reduces aggregation in various aggregating test peptides as determined by UHPLC (214 nm). These peptides were synthesized by AFPS at 90 °C. (E) C-terminal [Arg(Pbf)]₆ not only decreases aggregation and improves peptide crude purity, but also affords the positively charged “free” Arg₆ after cleavage from the resin and global deprotection, which provides several advantageous features. (F) The C-terminal [Arg(Pbf)]₆-tag improves crude purity of MYC[123–143] 3-fold, potentially enabling the synthesis of longer MYC fragments. *Two peaks with the same mass were observed for hGH[176–191]F176Y -Pro₆.

Figure 3

IR and SSNMR experiments indicate that [Arg(Pbf)]₆ changes the structure of difficult peptides on resin from rigid β-sheets to flexible α-helices. (A) IR and SSNMR analysis indicates a correlation between secondary peptide structure and aggregation during SPPS. (B) Resin-bound and fully protected Barstar[75–90] (no tag) shows only β-sheet SSNMR signals. (C) Upon solvation with DMF, the β-sheet signals remain unchanged. (D) SSNMR spectrum of resin-bound and fully protected Barstar[75–90]-[Arg(Pbf)]₆ shows both α-helix and β-sheet signals for all ¹³C-labeled residues. (E) Upon solvation with DMF, the α-helix signals disappear, yet the β-sheet signals remain.

Figure 4

Common cleavable linkers impact the severity of aggregation, but not its onset. (A) Chemical structures of cleavable linkers used in this study. The gray circle indicates the SPPS resin. (B) Aggregation as a function of Fmoc-deprotection peak broadening by in-line UV–vis (310 nm) in AFPS for various cleavable linkers. No linker and double linker (Rink/HMPB) were used as negative and positive control experiments, respectively. (C) The impact of the linker on aggregation (maximum values for relative peak broadening throughout the synthesis) translates to the purity of the cleaved Barstar[75–90] in most cases. *Cyclization and immediate cleavage from MeDbz were accompanied by side-products originating from the cleavage protocol, not synthesis itself.

Figure 5

SynTag improves the chemical synthesis and purification of difficult peptides, including Aβ42[27–42]. (A) SynTag successfully mitigated the onset and severity of aggregation for the synthesis of PKI-α. (B) Adding SynTag to the C-terminus also improved the crude purity of PKI-α. (C) After MeDbz cyclization, cleavage, and global deprotection, C-terminal SynTag improves solubility and can be hydrolyzed to give the native peptide with a C-terminal carboxylic acid. (D) Aggregating and insoluble Aβ42[27–42] serves as test case to demonstrate that SynTag simultaneously improves synthesis and solubility. (E) Aβ42[27–42] synthesis without tag, with Arg₆ and with SynTag; synthesis was followed by in-line UV–vis analysis. Both Arg₆ and SynTag have the same effect on deprotection peak broadening (aggregation) during synthesis. (F) SynTag markedly improves crude purity of Aβ42[27–42]. Peaks are less broad with SynTag, indicating increased solubility. (G) After HPLC purification of crude Aβ42[27–42] with SynTag, hydrolysis in water at 23 °C for 30 h afforded native Aβ42[27–42]. (H) Analysis of the start and end points of hydrolysis by UHPLC (214 nm) shows that purified and hydrolyzed Aβ42[27–42] elutes later than Aβ42[27–42]-MeNbz-Arg₆ and results in a broadening of the peak, as expected.

Figure 6

SynTag enables the chemical synthesis of MYC transactivation domain (TAD) using a single NCL. (A) After cyclization, cleavage, and deprotection, C-terminal SynTag improves solubility and can be transformed into a thioester for NCL. (B) MYC[1–143] TAD was chosen as a target protein to demonstrate the utility of SynTag in chemical protein synthesis. Solubility issues are expected due to low pI of both envisaged fragments. Furthermore, the C-terminal fragment severely aggregates during synthesis. (C) Synthesis of untagged MYC[86–143] is unsuccessful, and the desired mass was not detected by LC–MS. (D) Successful synthesis of MYC[86–143] by using an Arg₆ tag, demonstrating suppression of aggregation during long peptide synthesis. (E) Arg₆ tag effectively suppresses aggregation of MYC[86–143] during synthesis compared to Rink amide linker alone, as illustrated by in-line UV–vis analysis. (F) HPLC purification yields clean MYC[85–143](D85C)-Arg₆. (G) Synthesis of MYC[1–84] with SynTag improves solubility and yields clean peptide after HPLC purification. (H) Combination of purified peptides in NCL buffer results in complete ligation after 120 min. (I) HPLC purification yields clean MYC[1–143](D85C)-Arg₆ in 35% isolated yield. MYC[1–143](D85C)-Arg₆ structure as predicted using AlphaFold, Arg₆ tag is highlighted in yellow.