Traceless cysteine-linchpin enables precision engineering of lysine in native proteins

Abstract

The maintenance of machinery requires its operational understanding and a toolbox for repair. The methods for the precision engineering of native proteins meet a similar requirement in biosystems. Its success hinges on the principles regulating chemical reactions with a protein. Here, we report a technology that delivers high-level control over reactivity, chemoselectivity, site-selectivity, modularity, dual-probe installation, and protein-selectivity. It utilizes cysteine-based chemoselective Linchpin-Directed site-selective Modification of lysine residue in a protein (LDM_C-K). The efficiency of the end-user-friendly protocol is evident in quantitative conversions within an hour. A chemically orthogonal C-S bond-formation and bond-dissociation are essential among multiple regulatory attributes. The method offers protein selectivity by targeting a single lysine residue of a single protein in a complex biomolecular mixture. The protocol renders analytically pure single-site probe-engineered protein bioconjugate. Also, it provides access to homogeneous antibody conjugates (AFC and ADC). The LDM_C-K-ADC exhibits highly selective anti-proliferative activity towards breast cancer cells.

Fig. 1. Chemical technology for precision engineering of native proteins.

a Requirement: low-frequency residue-derived proximity regulator to create a unique targetable landscape and opportunity for simultaneous regulation of reactivity, chemoselectivity, site-selectivity, modularity, and protein-selectivity. b Hypothesis: Cys-linchpin-directed modification of Lys residue (LDM_C-K). Step 1, linchpin formation: rapid, intermolecular, chemoselective reaction of F_C with Cys; step 2, cyclization: intramolecular, irreversible, site-selective reaction of F_K with Lys; step 3, F_C to F_X: sequential C-S and C-C bond dissociation; step 4, late-stage modification.

Fig. 2. Development of Cys-based linchpin, F_C.

a Establishing the reactivity for C-S bond formation (Ar, -Ph-4-O(CH₂)₃CO₂Et). b Validating the chemoselectivity for C-S bond formation with Cys. c Establishing C-S/C-C bond dissociation to generate a functional group, F_X, amenable for late-stage transformations.

Fig. 3. Development of LDM_C-K reagents.

a Initial exploration with BLGA to establish the design of F_K and relative reactivity with F_C. b The design and synthesis of potential LDM_C-K and control reagents (for synthesis, see Supplementary Figs. 1–10).

Fig. 4. Peptide modification.

a Control reagent 1l with F_C and Cys-containing peptide 12a results in efficient C-S bond formation. b LDM_C-K reagent 9d mixed with peptide 12b having Cys and Lys at i and i + 4 positions result in C-S bond followed by C-N bond formation rendering the cyclic peptide 14b within 1 h. c LDM_C-K reagent 9d with peptide 12c having thiol protected Cys and Lys at i and i + 4 positions result in no conversion within 1 h. d LDM_C-K reagent 9d with peptide 12d having Cys at i and two Lys residues at i + 4/i + 8 positions result in C-S bond formation followed by site-selective C-N bond formation at i + 4 position. Control experiments to examine competition with e Tyr, f Ser, and g His. h Establishing the complete LDM_C-K workflow. The selected MS for 14i, 15a, 15b, and 15c are given in the inset. Also see Supplementary Figs. 102–106 for detailed data, full XIC, and MS spectra.

Fig. 5. LDM_C-K technology.

a Workflow for C-S bond formation, cyclization, C-S/C-C bond dissociation, and F_C to F_X transformation. b The chemoselective, site-selective, and modular single-site labeling of a native protein (β-lactoglobulin, BLGA, 8a). c The plot of % conversion versus time highlights the progress of Michael addition and cyclization with BLGA in the presence of LDM_C-K reagent 9d. Data are presented as mean values (±SD), n = 3 independent experiments. d Adaptable rigidity of LDM_C-K reagent: Probability distribution for the reagent (9d) and end-to-end distance between electrophilic C1 and C11 atoms in water with (red) or without (black) BLGA. The most probable conformations corresponding to peaks are shown. e Probability distribution of the distance between electrophilic C11 atom of the LDM_C-K reagent 9d and nearby lysine residues (NZ atom) in BLGA. f Interaction of C11 in 9d with the nearby lysine residues. Inset highlights the interaction of Lys101 with the C11 (9d).

Fig. 6. Late-stage installation of various tags on protein.

a Single-site parallel installation of ¹⁹F NMR probe 21, affinity probe 22, and fluorophore 23. b Precise second probe installation through the chemoselective installation of CPM dye (27) on protein bioconjugates (24–26) to render dual-probe conjugates (28–30).

Fig. 7. Protein-selectivity coupled with chemoselectivity and site-selectivity and the purification workflow.

Single-site, single-protein labeling of human serum albumin in a mixture of proteins (MoP), and b cell lysate. Hydrazone formation captures and enriches the bioconjugate by ordered on-resin immobilization allowing recovery of unlabeled proteins. Next, the transoximization and centrifugal spin concentration render analytically pure single-site tagged protein bioconjugate.

Fig. 8. LDM_C-K for homogeneous antibody-fluorophore and drug conjugate (AFC and ADC).

a LDM_C-K reagent 9d renders homogeneous trastuzumab conjugates with specific modifications of K183 and K341. Subsequently, AFC (38) and ADC (39) were prepared by late-stage installation of hydroxylamine derivative of fluorophore (coumarin, 23) and drug (emtansine, DM1, 35a). b Inhibition of cell proliferation by LDM_C-K-ADC (39) as compared to trastuzumab (34), DM1 (35), and Kadcyla (40) in SKBR-3 (HER-2 positive) cancer cell line. The percentage inhibition was calculated using untreated cells as control. c Inhibition of cell proliferation by DM1 (35), ADC (39), and Kadcyla (40) at 0.5 nM concentration in HER-2 positive SKBR-3 as compared to HER-2 negative MDA-MB-231 cells. Data are presented as mean values (±SD), n = 3 biologically independent experiments.