Indirect Formation of Peptide Bonds as a Prelude to Ribosomal Transpeptidation

Abstract

The catalytic competency of the ribosome in extant protein biosynthesis is thought to arise primarily from two sources: an ability to precisely juxtapose the termini of two key substrates─3'-aminoacyl and N-acyl-aminoacyl tRNAs─and an ability to ease direct transpeptidation by their desolvation and encapsulation. In the absence of ribosomal, or enzymatic, protection, however, these activated alkyl esters undergo efficient hydrolysis, while significant entropic barriers serve to hamper their intermolecular cross-aminolysis in bulk water. Given that the spontaneous emergence of a catalyst of comparable size and sophistication to the ribosome in a prebiotic RNA world would appear implausible, it is thus natural to ask how appreciable peptide formation could have occurred with such substrates in bulk water without the aid of advanced ribozymatic catalysis. Using a combination of fluorine-tagged aminoacyl adenylate esters, in situ monitoring by ¹⁹F{¹H} NMR spectroscopy, analytical deconvolution of kinetics, pH-rate profile analysis, and temperature-dependence studies, we here explore the mechanistic landscape of indirect amidation, via transesterification and O-to-N rearrangement, as a highly efficient, alternative manifold for transpeptidation that may have served as a prelude to ribosomal peptide synthesis. Our results suggest a potentially overlooked role for those amino acids implicated by the cyanosulfidic reaction network with hydroxyl side chains (Ser and Thr), and they also help to resolve some outstanding ambiguities in the broader literature regarding studies of similar systems (e.g., aminolyzes with Tris buffer). The evolutionary implications of this mode of peptide synthesis and the involvement of a very specific subset of amino acids are discussed.

Figure 1

Underlying chemistry of ribosomal peptide synthesis.

Figure 2

Chemical scheme, example ¹⁹F{¹H}_IG NMR spectrum (376 MHz), and key temporal evolution profiles from the monitoring of a typical aminolysis of MepA-l-PheF (E_m) with l-serinamide ^LS in aqueous solution (this example: H₂O, 20 °C, pH = 7.3, I = 2.0 M). (A) Temporal evolution of experimental mole fractions of E_m,2′, E_m,3′, P_Am, and P_aa. (B) Temporal evolution of experimentally observed 2′/3′ speciation of E_m, with f_{Em,3′} = [E_m,3′]/[E_m]_T = fractional population of the 3′-isomer of E_m over all charge states. (C) Temporal evolution of the mole fraction of E_m,T with fit to the kinetic model: E_m,T → P_Am (k_Am^ψ), E_m,T → P_aa (k_Hyd^ψ). (D) Temporal evolution of the mole fractions of P_Am and P_aa with fit to the kinetic model above. [E_m]_T = [E_m,2′] + [E_m,3′] = total concentration of E_m in all charge states. k_T^ψ = total pseudo-first-order rate constant for the decay of E_m,T. k_Am^ψ = f_{Am}.k_T^ψ = k_Am′_.[^LS]_T = pseudo-first-order rate constant for the aminolysis of E_m,T. k_Am′ = total second-order rate constant for the aminolysis of E_m,T. k_Hyd^ψ = f_{Hyd}. k_T^ψ = pseudo-first-order rate constant for the hydrolysis of E_m,T. [E_m,j′] = [E_m,j′⁽⁰⁾] + [E_m,j′⁽⁺⁾], where [E_m,j′⁽ⁱ⁾] = concentration of j′-regioisomer with a charge of i on the α-amino group. X_{i} = mole fraction of species i.

Figure 3

Kinetic scheme, temporal evolution profiles, and comparative fits to two distinct kinetic models for key species in a typical aminolysis of MepA-l-PheF (E_m) with l-serinamide under a low pH regime (this example: D₂O, 30 °C, pH*(30 °C) = 5.7, I = 2.0 M). Elementary model = E_m,T → P_Am (k_Am^ψ), E_m,T → P_aa (k_Hyd^ψ). Stepwise model: = E_m,T → I_Es (k_Am^ψ), I_Es → P_Am (k_r^ψ), E_m,T → P_aa (k_Hyd^ψ).

Figure 4

Kinetic analysis of the aminolysis of E_m with l-serinamide ^LS, and concurrent hydrolysis, over the range pH* = 5.7–9.3 (20 °C, D₂O, I = 2.0 M, KCl; [E_m]_T,0 = 4.0 mM). (A) pH*–k′_Am profile for [^LS]_T = 600 mM (blue) and 1200 mM (red), with fit to experimental data. (B) pH*–k_Hyd^ψ profile, with same color coding as (A), and fit to experimental data. (C) Empirical selectivity for aminolysis, k_Am^ψ/k_T^ψ, and total pseudo-first-order rate of decay k_T^ψ as a function of pH*(20 °C), with fits to experimental data, for [^LS]_T = 600 mM. (D) as for (C) but with [^LS]_T = 1200 mM. Fitted parameters (k_Am,+, k_Am,0, k_OH^–,+, k_OH^–,0; pK_a*(E_m⁺)) and their standard errors (Table S1, Supporting Information) were determined by global, unweighted nonlinear fitting of the pH*–k′_Am and pH*–k_Hyd^ψ profiles (OriginPro 2024b). Relative uncertainties in k_Hyd^ψ and k′_Am were conservatively estimated to be ±10% (see Supporting Information). Uncertainties in k_T^ψ and k_Am^ψ/k_T^ψ (not fitted directly) were calculated by propagation. See Supporting Information for further discussion.

Figure 5

Simplified structural depictions of E_m⁺ and E_m⁰ and the N-formylated derivative E_m^f. Note that 2′–3′ isomerization (not shown) appears to be limitingly rapid for both E_m and E_m^f, relative to hydrolysis and aminolysis (pH* > 5.5), and so, all rate constants and isotope effects in Tables 1 and 2 are weighted averages. See Supporting Information for details.

Figure 6

Temperature dependence (10–30 °C) of the hydrolysis and aminolysis of E_m with l-serinamide ^LS, as a function of pH* (D₂O, I = 2.0 M, KCl, [E_m]_T,0 ≈ 4 mM). (A) Temperature dependence of the pH*–k′_Am profile (aminolysis), with (unweighted) fits to experimental data. (B) Temperature dependence of the pH*–k_Hyd^ψ profile (hydrolysis), with (unweighted) fits to experimental data. (C) Temperature and pH* dependence of (i) the empirical selectivity for aminolysis k_Am^ψ/k_T^ψ and (ii) the total pseudo-first-order rate of decay k_T^ψ, with fits to experimental data, for [^LS]_T = 1200 mM. (D) Linearized Eyring plots of the second-order rate constants k_Am,+, k_Am,0, k_OH^–,+, and k_OH^–,0. (E) ¹³C{¹H} NMR titrations (δ_13CO) of l-serinamide (100 mM; D₂O, I = 2.0 M, KCl) at different temperatures ([dpK_aH/dT] = −0.025 °C^–1; cf. glycinamide). pH*(10 °C) = 6.38–9.42 (pK_a*(^LS⁺) = 8.17). pH*(20 °C) = 6.01–9.14 (pK_a*(^LS⁺) = 7.94). pH*(30 °C) = 5.69–8.95 (pK_a*(^LS⁺) = 7.69). [^LS]_T = 600–2000 mM. Fitted parameters (k_Am,+, k_Am,0, k_OH^–,+, k_OH^–,0; pK_a*(E_m⁺)) and their standard errors (Table 1, Supporting Information) were determined by global, unweighted nonlinear fitting of pH*–k′_Am and pH*–k_Hyd^ψ (OriginPro 2024b). See the Figure 4 caption and Supporting Information for details of error analysis.

Figure 7

Example ¹⁹F{¹H}_IG NMR monitoring data and kinetic analysis for the aminolysis, and concurrent hydrolysis, of MepA-(l-PheF)₂ (E_Bis) with l-serinamide in D₂O (20 °C, pH* = 5.6–8.6, I = 2.0 M; [^LS]_T = 600–2000 mM; [E_Bis]_T,0 ≈ 4 mM). (A) Temporal evolution profiles, with kinetic fits shown, for all key species under a low pH* regime (this example: pH* = 5.6, [^LS]_T = 2000 mM). (B) Expanded view of (A), with a focus on the intermediate kinetics of E_m,T and I_Es. (C) pH*–k′_Am,Bis profile with (unweighted) fit to experimental data (k′_Am,Bis = k_Am,Bis^ψ/[^LS]_T). (D) pH*–k_Hyd,Bis^ψ profile with (unweighted) fit to experimental data. (E) Co-plot of the pH*–k′_Am (E_m) and pH*–k′_Am,Bis (E_Bis) profiles; dotted line reflects the pH*–k′_Am,Bis profile without any contribution from the aminolysis of the doubly protonated state, E_Bis⁺⁺ (i.e., k_Am,Bis,₊₊ = 0). (F) Co-plot of the pH*–k_T^ψ (E_m) and pH*–k_T,Bis^ψ (E_Bis) profiles ([^LS]_T = 600 mM). Kinetic model for pH* > 6.5 = E_Bis → E_m,T + P_Am (k_Am,Bis^ψ), E_Bis → E_m,T + P_aa (k_Hyd,Bis^ψ), E_m,T → P_Am (k_Am^ψ), E_m,T → P_aa (k_Hyd^ψ). Kinetic model for pH* < 6.5: = E_Bis → E_m,T + I_Es (k_Am,Bis^ψ), E_Bis → E_m,T + P_aa (k_Hyd,Bis^ψ), E_m,T → I_Es (k_Am^ψ), E_m,T → P_aa (k_Hyd^ψ), I_Es → P_Am (k_r^ψ). All steps are (pseudo; ψ) first-order. Fitted parameters (k_Am,Bis,++, k_Am,Bis,+, k_Am,Bis,0, k_{OD^–,Bis,++}, k_OD^–,Bis,+) and their standard errors in Table 4. pK_a*(E_Bis⁺⁺) and pK_a*(E_Bis⁺) constrained during fitting; values taken from ¹⁹F{¹H} NMR chemical shift analysis. See Supporting Information for details.

Figure 8

Structural depictions of E_Bis (simplified) and I_Es in their various charge states. Note that [E_Bis⁺] = [E_Bis,2′+⁺] + [E_Bis,3′+⁺]. For simplicity, we assume that I_Es⁺ and I′_Es⁺ have identical microscopic acidities.