Site-selective peptide bond hydrolysis and ligation in water by short peptide-based assemblies

Significance

The emergence of self-assembled peptide microphases with modest catalytic capabilities underscores their potential role as intermediates en route to extant enzymes. However, the mechanism enabling these structurally minimal units to break and form the very bonds that constitute them remains puzzling. Our findings suggest that short peptide-based aggregates indeed have capabilities to catalyze the challenging events of peptide bond-breaking and ligation in aqueous milieu; a feature that is exclusively reserved for evolved enzymes with long sequences. The development of such synthetically amenable constructs will offer insights into the intricate networks of reactions that led to the diversity of life.

Abstract

The evolution of complex chemical inventory from Darwin’s nutrient-rich warm pond necessitated rudimentary yet efficient catalytic folds. Short peptides and their self-organized microstructures, ranging from spherical colloids to amyloidogenic aggregates might have played a crucial role in the emergence of contemporary catalytic entities. However, the question of how short peptide fragments had functions akin to contemporary complex enzymes to catalyze cleavage and formation of highly stable peptide bonds that constitute the backbone of all proteins remains an unresolved yet fundamentally important question in terms of the origins of enzymes. We report short-peptide-based spherical assemblies that demonstrated residue-specific cleavage and formation of peptide bonds of diverse peptide-based substrates under aqueous environment. Despite the short sequence length, the assemblies utilized the synergistic collaboration of four residues which included the catalytic triad of extant serine proteases with a nonproteinogenic amino acid (quinone moiety), to facilitate proteolysis, ligation, and a three-step (hydrolysis–ligation–hydrolysis) cascade. Such short-peptide-based catalytic assemblies argue for their candidacy as the earliest protein folds and open up avenues for biotechnological applications.

Considerations of the origins of chemical evolution, beginning with Darwin’s nutrient-rich warm pond hinges on the emergence of a diverse chemical inventory generated from a series of chemical reactions. Conditions conducive to the random condensation of Miller–Urey amino acids into longer oligomers have led to complex arrays of polypeptide-based spherical colloids, suggesting an approach to supramolecular assemblies from a simpler inventory (1, 2). Interestingly, prions, protein-only infectious agents have shown the ability to undergo mutation and selective amplification as a population of assembled polypeptide amyloid phases independent of nucleic acids (3, 4). Thus, simpler chemical processes leading to the formation of short peptides and their physical self-assembly into diverse catalytic polypeptide phases, from particles to amyloid fibers (5, 6), reveal alternative routes to rudimentary catalysts that may have preceded modern-day enzymes (7–18). Directed evolution and protein reprogramming strategies have demonstrated that the incorporation of derivatized or nonproteinogenic amino acids into the active site can enhance binding affinity, broaden substrate scope, and augment enantioselectivity, and such alternative tools might have been explored to confer catalytic abilities in shorter protein sequences during early evolution (19, 20).

However, a formidable challenge for these short-peptide-based catalysts lies in their ability to catalyze the formation and degradation of the extraordinarily stable amide bonds (half-life of approximately ~350 to 600 y) that constitute their very core. The constant turnover by catalytic degradation and formation of peptide bonds would have allowed for a diverse pool of peptide sequences to be tested before the emergence of longer functional proteins. Nonetheless, efforts to confer catalytic competencies and specificities for ligating/cleaving kinetically stable amide linkages by short peptides are scarce (21, 22) due to the challenge of aligning multiple catalytic residues in a productive orientation within the limitation of short sequences.

Herein, we show that the short-peptide constructed from the combination of the proteinogenic and nonproteinogenic amino acids can manifest the characteristics of cross-β aggregates to manipulate highly stable peptide bonds (hydrolysis and ligation) with remarkable specificity in an aqueous medium [90% v/v, water in acetonitrile (ACN)] without the need of a harsh chemical environment. The short peptide sequence featured the nucleating core “¹⁷LVF²⁰F” of Aβ (1 to 42) with histidine as the C-terminal while N-terminal contained aspartic acid and serine condensed with 3,4-dihydroxyphenylacetic acid as derivatized L-3,4-dihydroxyphenylalanine amino acid (diphenol-SDLVFFH, Fig. 1 A and B). While the peptide sequence had the catalytic triad (SDH) of extant serine proteases (Fig. 1A) (23), N-termini quinone could provide the necessary enthalpic gain required for substrate–catalyst binding, a strategy extensively employed by diverse enzymes in the form of quinone–guanidinium complexes for uphill proton transfer reactions (Fig. 1A) (24, 25).

Fig. 1.

The design, characterization, and proteolytic activity of a catalytic short peptide. (A) The catalytic motifs (shown in the box) used to construct the short peptide were derived from the conserved catalytic triad of serine proteases (PDB ID: 1HAZ, 21), consisting of Ser (S195), His (H57), and Asp (D102), and the Quinone–Asp–Arg complex in the cytochrome–bf6 complex with TDS (PDB ID 4H13, 22). (B) Schematics depicting the diphenol functionalized-SDLVFFH peptide that incorporates the nucleating core (¹⁷LVF²⁰F) of Aβ (1 to 42) with the catalytic residues and forms spherical morphologies with a negative surface potential and cleaves amide bonds at the carboxy-terminal of arginyl residues, (C) TEM and (D) Cryo-EM images of the peptide assembled at pH 8.0 (Q-SDH), (E) Confocal laser scanning microscopy image of the assemblies in the presence of Rhodamine-6G dye, (F) Atomic force microscopy image of the Q-SDH assembly, with a height profile shown in the Inset. (G) The structure of the substrate L-BAPA and HPLC traces (extracted at 380 nm) showing the time course generation of 4-nitroaniline (pNA) as a product of hydrolysis. 4-nitrophenol is used as an internal standard (IS). (H) Bar diagram shows comparison in % hydrolytic conversion of L-BAPA in the presence of Q-SDH compared to background (Ctrl) and disassembled Q-SDH by HFIP treatment. Hydrolysis % was calculated from at least three independent measurements.

Results and Discussion

To this end, 3,4-dihydroxyphenylacetic acid functionalized-SDLVFFH peptide was self-assembled at pH 8.0. The characteristic reddish-brown color of the dissolved peptides at pH 8.0 with broad absorbance and disappearance of the absorption peak at 281 nm confirmed the formation of quinone functionalized-SDLVFFH (Q-SDH, Fig. 1B and SI Appendix, Fig. S9). The oxidized form (m/z = 1,010.64, Q-SDH) of the synthesized peptide (m/z = 1,012.45) was confirmed through LC–MS (SI Appendix, Figs. S10 and S11). Q-SDH accessed spherical self-assemblies (~600 nm to 1 µm size) as was observed under transmission electron microscopy (TEM) and further corroborated with dynamic light scattering measurements (Fig. 1 C and D and SI Appendix, Fig. S12). Atomic force microscopy measurements also showed spherical assemblies with consistent Gaussian increase in height profiles (range of ca. 20 to 25 nm, Fig. 1F). Attenuated total reflectance-Fourier transform infrared (ATR-FTIR) spectrum and occurrence of positive and negative band in circular dichroism (CD) spectroscopy at 192 nm and 206 nm, respectively, indicated the formation of an ordered array of twisted β-sheet-like structures (SI Appendix, Figs. S13 and S14) (26). The assembled microphases exhibited a negative surface potential [zeta potential (ζ) = −18.2 mV, SI Appendix, Fig. S15] which imparted colloidal stability to the spherical aggregates. Confocal laser scanning microscopy measurements demonstrated that the negatively charged Q-SDH assemblies showed a preference for binding with positively charged dye (Rhodamine-6G) comparison to negatively charged fluorescein (Fig. 1E and SI Appendix, Fig. S16).

For investigating the proteolytic activity, we used a trypsin-specific chromogenic peptide substrate, Nα-benzoyl-L-arginine-4-nitroanilide hydrochloride (L-BAPA). We hypothesized that the positively charged guanidinium moiety of L-BAPA could exploit electrostatic interactions with the negatively charged catalytic surface of the peptide assemblies. Subsequently, it would enhance the formation of a quinone–guanidinium complex, thereby providing an enthalpic advantage for substrate–catalyst binding (Fig. 1 A and G). Incubating the assemblies with the substrate at 37 °C in an aqueous medium resulted in the generation of 4-nitroaniline (pNA) which was monitored via high-performance liquid chromatography (HPLC) (pH of the reaction medium = 6.5, Fig. 1G). HPLC analyses revealed a hydrolytic activity with the significant conversion of 13.6 ± 6.2% from L-BAPA to pNA after 72 h (Fig. 1H and SI Appendix, Figs. S17–S19, minuscule background hydrolysis of ~0.5% was seen in the absence of Q-SDH, Fig. 1H and SI Appendix, Fig. S20). The catalytic rate constant (k_cat) for Q-SDH with L-BAPA was found to be (5.4 ± 3.3) × 10⁻⁷ s⁻¹ (SI Appendix, Fig. S21). After being subjected to 72 h of incubation at 37 °C and pH 6.5, the assemblies, although smaller in dimension, preserved their spherical morphology, as confirmed by TEM and DLS analyses (SI Appendix, Fig. S22). Additionally, hexafluoro-2-propanol (HFIP) mediated disassembled peptide exhibited a significant reduction in hydrolysis (1.01 ± 0.01% after 72 h), underscoring the importance of cooperative colocalization of the catalytic components on the microphase surface (Fig. 1H and SI Appendix, Fig. S23). Notably, previous reports have either employed harsh chemical conditions or mechanical activation of the amide bonds when working in aqueous environment to cleave peptide bonds (27, 28).

In Nature, different classes of serine proteases, including trypsin, chymotrypsin, and elastases, display proteolytic activity specific to the carboxyl end of cationic (Arg/Lys), hydrophobic (Phe/Tyr), and small neutral (Gly/Ala/Leu) peptide residues, respectively. Intrigued by the significant proteolytic activity displayed by the short peptide-based assemblies, we examined their substrate specificity by exposing them to amides that lacked the positively charged guanidinium group. Instead of arginine residue, a substrate featuring hydrophobic leucine moiety (Nα-benzoyl-L-leucine-4-nitroanilide hydrochloride, L-BLPA) was synthesized (See SI Appendix for methods). Indeed, when L-BLPA was incubated with Q-SDH assemblies under similar conditions, a hydrolytic activity with conversion of only 1.11 ± 0.35% was observed after 72 h (Fig. 2A and SI Appendix, Fig. S24). Furthermore, the affinity of Q-SDH assemblies toward the binding and cleavage of an acidic (anionic) substrate N-succinyl-L-phenylalanine-4-nitroanilide (L-SPPA) was found to be negligible as conversion of 0.21 ± 0.18% was observed after 72 h of reaction (Fig. 2A and SI Appendix, Figs. S25 and S26). This significant decline in the proteolytic activity of Q-SDH toward neutral or acidic amino acid moieties suggested the noteworthy specificity of the binding grooves toward positively charged substrates (Fig. 2B), a strategy efficiently used by evolved enzymes for the regulation of complex biological processes.

Fig. 2.

Residue selective catalytic specificity of Q-SDH assemblies. (A) Bar diagram illustrating the % hydrolytic conversion for three substrates, cationic L-BAPA, neutral L-BLPA, and anionic L-SPPA after 72 h of incubation in presence of Q-SDH assemblies, (B) Schematic depicting the catalytic specificity of the Q-SDH assemblies toward positively charged substrates, % hydrolytic conversions of site-selective amide hydrolysis for synthetic peptides (C) Bz-RAAGAG, (D) Bz-AAGRAG, (E) Bz-RAARAG, and (F) Bz-RAAKAG in the presence of Q-SDH assemblies after 72 h at 37 °C. Sites that were found to be hydrolyzed are marked with broken lines and % occurrence of the corresponding observed fragments in the solution are written below the respective sequences.

Next, we investigated the proteolytic capabilities of Q-SDH toward synthesized inactivated peptides. Arginine or lysine residues were incorporated at specific positions to evaluate the cleavage specificity of Q-SDH. Two heptapeptide variants based on the trypsin substrate L-BAPA were designed, where each variant featured arginine at different positions (Bz-RAAGAG, Bz-AAGRAG, Fig. 2 C and D). When Q-SDH assemblies were incubated with Bz-RAAGAG for 72 h in 4-(2-Hydroxyethyl)piperazine-1-ethane-sulfonic acid (HEPES) buffer (50 mM, pH 8.0 as very low conversions were observed in water medium at pH 6.5), a site-selective cleavage was observed as the carboxyl-terminal peptide bond of the arginyl residue (10.6 ± 4.7%) got exclusively hydrolyzed despite the presence of seven amide bonds in the substrate (Fig. 2C and SI Appendix, Figs. S27 and S28). It is worth noting that in the absence of Q-SDH assemblies, the peptide Bz-RAAGAG did not suffer any detectable hydrolysis under similar reaction conditions, which suggested the high thermodynamic stability of the substrate (SI Appendix, Fig. S28). We also observed similar site-selective cleavage for another variant Bz-AAGRAG which was cleaved exclusively at the carboxy-terminal position (7.9 ± 1.1 %) of the 5th position arginyl residue (almost no background hydrolysis was observed, Fig. 2D and SI Appendix, Figs. S29 and S30).

Further, in order to probe the selectivity toward cationic residues, two dicationic substrates were synthesized (Bz-RAARAG and Bz-RAAKAG, Fig. 2 E and F). Strikingly, we observed almost complete digestion of the parent substrate (Bz-RAARAG) with exclusive selectivity for the 2nd and 5th positions featuring cationic guanidinium moieties. The fragments observed were Bz-R-COOH (18.4 ± 7.5%, cleaved from carboxyl end of the 2nd position), Bz-RAA-COOH (14.7 ± 10.9%, cleavage at the amine end of the 5th position), and Bz-RAAR-COOH (64.3 ± 18.4%, cleavage at the carboxyl end of the 5th residue, Fig. 2E and SI Appendix, Figs. S31 and S32). Interestingly, the catalytic assemblies showed a clear preference toward the C-termini (~83%) versus the N-termini (~15%) as seen in contemporary trypsin. When the 5th position arginyl residue was mutated with lysine (Bz-RAAKAG), we found significant peptide bond hydrolysis (66.2 ± 9.2%) with similar specificities (Fig. 2F and SI Appendix, Figs. S33–S35). Bz-RAAKAG was broken primarily at the carboxyl end of the 5th lysine residue (Bz-RAAK-COOH, 35.9 ± 10.9%), followed by the carboxyl end of the 2nd arginyl residue (Bz-R-COOH, 20.4 ± 12.2%), and moderate hydrolysis at the N-terminal of the 5th lysine residue (Bz-RAA-COOH, 9.9 ± 0.6%). These findings underpinned the high catalytic proficiency with remarkable site selectivity of the short-peptide-based assemblies that resembled the attributes of extant proteases.

To discern the role of various catalytic residues toward the proteolytic activity, we conducted sequential site-directed mutations of the catalytic units while keeping the nucleating core intact (Fig. 3A). These mutations resulted in four distinct control sequences (Fig. 3A). Mutation of the terminal quinone with an acetyl moiety (Ac) led to Ac-SDLVFFH (Ac-SDH, SI Appendix, Fig. S36), which accessed fibrillar morphologies (pH 8.0, Fig. 3B and SI Appendix, Fig. S37, CD and FTIR suggested β-sheet like arrangement). Interestingly, the absence of the quinone moiety resulted in a dramatic loss of catalytic activity toward L-BAPA (0.99 ± 0.39% after 72 h, Fig. 3F and SI Appendix, Figs. S38 and S39). Despite featuring negatively charged surfaces, the significantly reduced activity by Ac-SDH suggested the role of quinone–guanidinium interaction for the catalyst–substrate complexation.

Fig. 3.

Dissection of the catalytic core of Q-SDH. (A) Mutation of the four catalytic centers by acylation or alanine residue and corresponding % hydrolysis toward L-BAPA after 72 h, (B–E) Morphologies of the mutated assemblies of Q-SDH assembled at pH 8 and observed under TEM, (F) Bar diagram for the % hydrolytic conversion of the L-BAPA via all the mutated assemblies every 24 h over the course of 3 d.

Similar point mutations of Ser, Asp, and His were performed by substituting them with alanine (Ala) to ensure minimal steric or electronic perturbations (Fig. 3A). Q-ADLVFFH (Q-ADH) accessed spherical morphologies similar to what was observed for Q-SDH (pH 8.0, Fig. 3C and SI Appendix, Figs. S40–S43). However, despite the morphological resemblance, Q-ADH exhibited a considerably reduced proteolytic activity (2.44 ± 0.48%) compared to Q-SDH after 72 h of reaction (Fig. 3F and SI Appendix, Figs. S44 and S45), underpinning the importance of serine in the catalytic triad. Furthermore, mutants of Asp and His residues (Q-SAH and Q-SDA respectively, Fig. 3A) assembled into fibrous networks when incubated at pH 8.0 (Fig. 3 D and E and SI Appendix, Figs. S40, S46, S47, S50, and S51). Notably, while Q-SAH showed a drastic decline of conversions (1.94 ± 1.14%, Fig. 3F and SI Appendix, Figs. S48 and S49), Q-SDA was completely catalytically inactive (0.87 ± 0.49%, SI Appendix, Figs. S52 and S53). This finding mirrors the observations derived from site-directed mutagenesis studies on natural proteases, where a single His residue mutation alone is shown to render the catalytic triad inactive (29, 30). Further, mutating aspartic acid residue with positively charged amino acids, lysine (diphenol-SKLVFFH), and arginine (diphenol-SRLVFFH) did not show hydrolytic activity toward the anionic substrate L-SPPA (SI Appendix, Figs. S54 and S55). In combination, mutations in the catalytic core established the synergistic role of all four catalytic sites in proteolytic activity.

Next, we sought to integrate a systems chemistry (31–34) approach to probe the catalytic prowess of short peptides toward amide bond formation (35). Based on the van’t Hoff’s principle of reversibility for all biocatalysts, various serine and cysteine proteases have demonstrated notable efficacies in ligation reaction by catalyzing the condensation reactions between acyl bond donors and acceptor pairs (36). The synthesis of a full-length ribonuclease A (124 residues) from six esterified peptides is an exquisite example of the usefulness of protein ligases (37). We expected that the acyl-catalyst intermediate (Fig. 4A, catalyzed by Q-SDH) generated during the reaction could undergo deacylation competitively by either water to yield hydrolysis product (acid) or by another amino-nucleophile (peptide based) to generate ligated product. The ratio between the aminolysis and hydrolysis products [ligated/hydrolysis (L/H)] is decisively determined by the kinetic constraints (35). For this reason, peptide-esters are commonly employed as acyl donors for ligation reactions due to the possibility of higher temporal accumulation of the corresponding acyl intermediate state (36). To this end, we investigated the hydrolytic capability of Q-SDH assemblies toward a trypsin-specific inactivated ester bond of Nα-Benzoyl-L-arginine ethyl ester hydrochloride (L-BAEE, Fig. 4A). Q-SDH demonstrated a promising catalytic conversion of the L-BAEE ester bond, reaching 52.7 ± 7.2% after 72 h (Fig. 4B and SI Appendix, Figs. S56 and S57) underpinning the crucial catalytic role played by the triad within Q-SDH, while previous endeavors in engineering short peptide-based hydrolases have predominantly focused on the hydrolysis of activated carboxylic esters to demonstrate their hydrolytic capabilities. Apart from few examples of X-ray crystal structures that suggest the presence of high-energy tetrahedral intermediates, other pieces of evidence have been primarily indirect via analogies with active site inhibition theory and site-selective mutation-based activity assays (21, 38, 39). In the absence of the crystalline nature of the peptide catalyst, we probed the intermediates via mass spectrometry and found evidence for the proposed acyl-catalyst intermediate by arresting the reaction at liquid N₂ temperature and pH 5, as these conditions are reported to facilitate the stabilization of intermediates (SI Appendix, Fig. S58, see SI Appendix for details) (39). The accumulated tetrahedral intermediate is posed to react with the acyl-acceptors yielding peptide bond through a double displacement ping-pong mechanism (40).

Fig. 4.

Peptide ligation activity of Q-SDH assemblies. (A) Mechanistic pathways illustrating the kinetically controlled peptide coupling mediated by Q-SDH assemblies using a peptide-ester (L-BAEE), (B) % Hydrolytic conversions of peptide-ester L-BAEE in the presence and absence of peptide assemblies at 24 h intervals over 3 d, (C) Structures of the acyl acceptors investigated in this study, along with the corresponding % of hydrolysis and % of ligated products at 37 °C and −15 °C (parenthesis) after 72 h, (D) Cascade events involving peptide-ester hydrolysis, peptide ligation, and subsequent peptide hydrolysis using AARAG as the acyl acceptor with Q-SDH assemblies after 72 h.

Initially, we probed the ligation of leucineanilide (LA, 2 mM) with L-BAEE (300 µM, ligation of artificial amino acid moieties provides opportunities for therapeutics and bioconjugation). LC–MS analysis revealed a significant ~21.2% ligated product (between L-BAEE and LA) along with 52.0% hydrolysis of L-BAEE and a L/H ratio of 0.41 (Fig. 4C and SI Appendix, Fig. S59, background reaction without Q-SDH yielded negligible ligated product formation). Further, drawing from previous studies that employed trypsin and chymotrypsin-catalyzed peptide synthesis in supercooled fluids and frozen aqueous systems for kinetically controlled reactions (41), we sought to investigate the reaction yields under −15 °C conditions (See SI Appendix for details). The enhanced proton mobility and reduced water activity in a frozen aqueous medium was expected to facilitate the suppression of hydrolysis products in favor of ligation. Indeed at −15 °C, the ligated product (24.4%) significantly surpassed hydrolysis (7.7%), with an increase in L/H ratio (3.18) by ~eightfold (Fig. 4C and SI Appendix, Fig. S60).

Further, to explore the ligation potential of Q-SDH assemblies for oligomeric peptides containing natural amino acid residues, we investigated coupling of L-BAEE with synthetic pentapeptides (NH₂-AAKAG, Fig. 4C) featuring positively charged residue due to their already established affinity toward the negatively charged Q-SDH assemblies. However, we observed a modest yield of 1.7% for the ligated product and low hydrolytic conversion of 15.2%, with a L/H ratio of 0.11 at 37 °C over 72 h (Fig. 4C and SI Appendix, Fig. S61, at −15 °C the L/H ratio increased slightly to 0.16, SI Appendix, Fig. S62). This decrease in hydrolysis could be attributed to the competitive binding between the positively charged peptide-ester (L-BAEE) and the synthetic peptide substrate (AAKAG) for the negatively charged Q-SDH assemblies. Next, we used NH₂-AARAG (AARAG) as an acyl acceptor which yielded significant amount of ligated product (17.1%, Bz-RAARAG) with ~33.4% hydrolysis product (L/H ratio of 0.5 was observed, Fig. 4 C and D and SI Appendix, Figs. S63 and S64). It would be important to note here, we earlier showed in the discussion that this ligated product (Bz-RAARAG) had a high propensity to undergo hydrolysis in the presence of Q-SDH assemblies (Fig. 2C). This could be the reason for the significantly higher amount of ligated product seen in case of Bz-RAARAG (with L-BAEE), as the subsequent hydrolytic degradation of Bz-RAARAG to other fragments (Figs. 2C and 4D) was expected to shift the equilibrium toward the right-hand side (at −15 °C, lower L/H ratio of 0.25 was observed possibly due to the impediment of degradation of Bz-RAARAG, SI Appendix, Fig. S65). Indeed, we also detected the formation of the fragment Bz-RAAR-COOH that resulted from the degradation of Bz-RAARAG (Bz-RAAR-COOH formed at a yield of 9.9% with respect to the formed Bz-RAARAG, Fig. 4D and SI Appendix, Figs. S63 and S64). In combination, these results establish the realization of a kinetically challenging three-step cascade events constituting a network of peptide-ester hydrolysis-peptide formation-peptide degradation, catalyzed by remarkably specific and efficient short peptide-based assemblies; a phenomenon until now reserved for highly evolved extant enzymes.

We have combined the insights from contemporary enzymes to probe two challenging events en route to the evolutionary journey of modern biochemistry: catalytic breaking and the formation of the highly stable peptide bond in an aqueous environment (ca. 90% v/v, water in ACN). Despite the limitation of the short length, the spherical assemblies of reported synthetic constructs utilized four of their eight residues for the synergistic collaboration of the catalytic triad of extant serine proteases and quinone moieties to exhibit residue-selective proteolytic and ligation activities. Furthermore, the short peptides could facilitate three-step cascade events of hydrolysis–ligation–degradation that might have been advantageous for early minimal enzymes for complexification of pool of chemicals with high kinetic stability. The need for a high catalyst-to-substrate ratio suggests that the sequence space of these peptides needs further exploration to improve their practical catalytic applicability. Nonetheless, these minimal self-assembled catalysts demonstrate the ability to catalyze the challenging chemical transformations involving amide bonds that constitute their very core, a feat previously considered achievable only by modern enzymes. As we expand the sequence space to harness their full catalytic potential, these ultrashort peptides offer advantages over the fragile natural enzymes that work in specific conditions and feature synthetically flexibility making them adaptable for biotechnological applications.

Materials

Fluorenylmethyloxycarbonyl (Fmoc) and Boc protected amino acids, DIPEA (N,N-diisopropylethylamine), DIC (N,N′-diisopropylcarbodiimide), HBTU, HOBt, Trifluoroacetic acid (TFA), benzoic acid, L-leucine methyl ester, piperidine were purchased from Sigma Aldrich. Oxyma was purchased from Nova Biochem. L-BAPA was purchased from SRL. N-succinyl-L-phenylalanine-4-nitroanilide (L-SPPA) and L-BAEE were purchased from Sigma Aldrich. All solvents and Fmoc-Rink amide MBHA Resin were purchased from Merck. MQ-water was used throughout the experiments.

Experimental Methods.

Peptide assembly.

Peptide assemblies and their HFIP treatment was performed as reported previously (42, 43). Briefly, synthetic peptide powders were dissolved in 0 to 40% acetonitrile–water for various peptides containing 0.1% TFA (final concentration 2.5 mM). Then, pH was adjusted to 8.0 by drop-wise addition of 0.1 N NaOH. The homogeneous solution was kept for 15 d at 4 °C and checked under TEM. The assemblies of diphenol-SKLVFFH and diphenol-SRLVFFH were set up under similar conditions and pH was adjusted to 6.0. For HFIP treatment, 15-d-old assemblies were lyophilized and obtained powder was incubated with HFIP for 2 h. Homogeneous clear solution was sonicated for 30 min and was dried again using N₂ bubbling. The obtained brown powder was used for experiments.

TEM.

Peptide assemblies were diluted to 250 μM using water followed by the adsorption on a TEM grid for 5 min. Excess peptide solution was wicked off with the filter paper. It was followed by the addition of 5 μL of 1 % (w/v) Uranyl acetate and incubated for 3 min. Samples were then kept in a desiccator under vacuum for 25 min. TEM micrographs were recorded with a JEOL JEM 2100 with a Tungsten filament at an accelerating voltage of 200 kV. Cryo-TEM imaging was performed on a JEM-2100Plus microscope, operating at 120 kV. A 6.5 µL droplet was placed on a lacey carbon copper grid (Agar Scientific). The grid was held by tweezers mounted on a Cryoplunge 3 (Gatan). The specimen was blotted and plunged into a liquid ethane reservoir cooled by liquid nitrogen. The vitrified samples were transferred to a Gatan cryoholder through a cryotransfer stage cooled by liquid nitrogen. During observation of the vitrified samples, the cryoholder temperature was maintained below −180 °C. The images were recorded with a CCD camera.

Circular dichroism.

CD spectra were recorded using a JASCO J-810 circular dichroism spectrometer fitted with a Peltier temperature controller to maintain the temperature at 25 °C. Then, 200 μL of the peptide samples (500 µM) were placed into a quartz cuvette with 10 mm path length. Each spectrum was obtained by scanning wavelength from 400 nm to 190 nm at a scanning rate of 100 nm/min. Three successive wavelength scans were taken to average for each sample.

Fourier-transform infrared spectroscopy.

Sample aliquots were dried as a thin film and infrared spectra spectra were acquired using Bruker (model no: Alpha) in ATR mode (Platinum ATR) at room temperature and averaging 256 scans with 4 cm⁻¹ resolution. Background spectra were subtracted from each sample spectrum.

HPLC.

HPLC studies were performed in Waters HPLC system [2535 quaternary pump for analytical LC (equipped with 2489 UV-Vis and QDa detector)]. X-bridge® T3 C18 5 µm, 4.6 × 250 mm analytical column was used for monitoring the hydrolysis of L-BAPA maintaining a flow rate of 1 mL/min with a gradient from 15 to 60% acetonitrile in water (both solvents contain 0.1% TFA) with a total run time of 40 min. The formation of hydrolyzed product pNA was monitored at 380 nm. Increase of peak areas was noted, and from a standard plot of substrates with known concentration, conversions were calculated. To probe the L-BAEE and peptide hydrolysis and ligation activities, Spherisorb C18 S5 ODS2 5 µm, 4.6 × 250 mm analytical column was used with appropriate flow rates, depending on the nature of the substrates. Hydrolysis of L-BAPA, L-SPPA, and L-BLPA was monitored by extracting the chromatograms at 380 nm. Hydrolysis and Ligation of other peptides were monitored by extracting the chromatograms at 245 nm.

Confocal microscopy.

The images were recorded in Olympus Laser Scanning Confocal System Model FV3000 [part of the Atomic Force Microscope with Rheological Measurement and Confocal Imaging Unit Facility, supported by Swarnajayanti (SB/SJF/2020-21/08)]. Negatively charged dye fluorescein (λ_ex = 488 nm laser line, 20 µM) and positively charged dye Rhodamine-6G (λ_ex = 561 nm laser line, 20 µM) were incubated with 1 mM peptide assemblies for 15 min. For differential binding study, both fluorescein and Rhodamine-6G were incubated with peptide assemblies. To show the differential binding study, the peptide catalyst was incubated with fluorescein for 5 min followed by the addition of Rhodamine-6G and system was kept for 15 min. The solutions were then cast on a glass slide and were enclosed with a coverslip before imaging.

Mass spectrometry of reaction intermediate.

The catalyst (600 µM) and 1.5 mM L-BAEE were mixed at pH 5.0 and frozen rapidly in liquid nitrogen (39). The frozen mixture was allowed to melt partially and diluted with cold methanol to obtain the mass spectra.

TFA removal.

Diphenol-SDLVFFH was dissolved in 100 mM HCl solution (1 mg/mL) and sonicated for 10 min and lyophilized (44). The procedure was repeated twice, and the resulting peptide powder was assembled at pH 8.0 following the procedure outlined in the section of peptide assembly.

Catalytic Assays.

Hydrolysis reactions.

Standard conditions for screening catalytic activity were as follows: 300 µM of the substrate (L-BAPA, L-BAEE, L-BLPA, and L-SPPA, 20 mM in dimethylsulfoxide) was incubated with 600 µM of the catalytic peptide [40% ACN: H₂O v/v (containing 0.1% TFA), stock of 2.5 mM at pH 8.0) at 37 °C in MQ-water (final pH 6.5) for 72 h, along with an internal standard (IS) (pNP)]. Then, 150 µL aliquots were taken out every 24 h and diluted to half concentration in a 50% ACN:H₂O solution containing 0.1% TFA. The mixture was then analyzed using LC–MS, and the peak areas of all components were adjusted to fit the concentration of the IS. Control experiments were conducted using 300 µM of the substrate under similar conditions but in the absence of peptides (the pH of the ACN-H₂O containing 0.1% TFA was adjusted to pH 6.5 using 0.1 N NaOH). Synthetic peptide substrates were tested for hydrolysis at pH 8.0 HEPES buffer (50 mM). The conversion percentage of L-BAPA, L-BAEE, L-BLPA, and L-SPPA were calculated based on the calibration curves of their respective reactants and generation of pNA or acid components. For synthetic peptides, conversion % was calculated using the calibration curve of Bz-RAAKAG (for reactant) and appearance of hydrolysis acid products.

Ligation reactions.

Standard conditions for screening catalytic ligation activity were as follows: First, 300 µM of the substrate (L-BAEE) was incubated with 600 µM of the catalytic peptide and 2 mM acyl acceptors at 37 °C in pH 8.0 HEPES buffer (50 mM) for 72 h (total volume of 300 µL). At 72 h, 150 µL aliquots were taken out and diluted to half concentration in a 50% ACN: H₂O solution. The mixture was then analyzed using LC–MS, with peaks extracted at 245 nm for analysis. Control experiments were conducted in the absence of peptides, where 300 µM of the substrate (L-BAEE) was incubated with 2 mM of acyl acceptors at 37 °C in MQ-water for 72 h, using pH 8.0 HEPES buffer (50 mM). The conversion percentage for hydrolysis was calculated based on the peak areas of L-BAEE and the hydrolysis products, including ligated products. In the case of Bz-RAAKAG and Bz-RAARAG, a calibration plot of Bz-RAAKAG was used due to its similar chromophoric nature. For the ligated product of leucine anilide, the disappearance of the acyl ester and the appearance of its hydrolyzed product were utilized to calculate the % formation of the ligated product.

The catalyst, the acyl bond donor, and the acyl acceptor were added to precooled HEPES pH 8.0 buffer to achieve final concentrations of 600 µM, 300 µM, and 2 mM, respectively (total volume of 300 µL). Subsequently, the mixtures were rapidly shaken and transferred into liquid nitrogen for 20 s to achieve rapid freezing. The tubes were then incubated at −15 °C. Finally, the reactions were analyzed by RP-HPLC, with peaks extracted at 245 nm for analysis. For this purpose, at defined time intervals, 150 µL aliquots were withdrawn and diluted to 300 µL with a quenching solution of 50% ACN: H₂O (0.1% TFA). For control reactions for spontaneous hydrolysis and aminolysis of the acyl donor ester, parallel reactions without catalysts were analyzed. The extent of spontaneous hydrolysis of L-BAEE was found to be less than 10%.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

All study data are included in the article and/or SI Appendix.