Tunable, biodegradable grafting-from glycopolypeptide bottlebrush polymers

Abstract

The cellular glycocalyx and extracellular matrix are rich in glycoproteins and proteoglycans that play essential physical and biochemical roles in all life. Synthetic mimics of these natural bottlebrush polymers have wide applications in biomedicine, yet preparation has been challenged by their high grafting and glycosylation densities. Using one-pot dual-catalysis polymerization of glycan-bearing α-amino acid N-carboxyanhydrides, we report grafting-from glycopolypeptide brushes. The materials are chemically and conformationally tunable where backbone and sidechain lengths were precisely altered, grafting density modulated up to 100%, and glycan density and identity tuned by monomer feed ratios. The glycobrushes are composed entirely of sugars and amino acids, are non-toxic to cells, and are degradable by natural proteases. Inspired by native lipid-anchored proteoglycans, cholesterol-modified glycobrushes were displayed on the surface of live human cells. Our materials overcome long-standing challenges in glycobrush polymer synthesis and offer new opportunities to examine glycan presentation and multivalency from chemically defined scaffolds.

Fig. 1. Cartoon representations of native proteoglycans vs. our synthetic glycobrushes, which are prepared via two-step, one-pot NCA polymerization.

a Comparison of a representative native proteoglycan bottlebrush (i.e., aggrecan with polypeptide backbone and chondroitin sulfate, and keratin sulfate polysaccharide chains) with our synthetic glycobrushes. b Synthetic route to chemically tunable glycobrush-based dual-catalysis, one-pot NCA polymerization. Chain length, graft density, glycosylation density, and pattern are tuned via NCA monomer feed ratios and equivalents of transition metal catalysts. R’ = CH₂CH₂SCH₃ for Met-linked AMK or CH(CH₃)CH₂CH₃ for Ile-linked AIK. c Functional Ni⁰ catalysts used in this study to install chemical groups of interest at the chain initiation site.

Fig. 2. Glyco-NCAs are quantitatively converted to glycobrush polypeptides via controlled and living polymerization.

a ATR-FTIR traces indicating polymerization of GlcK NCA from PAMK₆₃ macroinitiator to form PAMK₆₃-g-PGlcK₁₁₇ glycobrushes. b Representative GPC/MALS/RI analysis of glycobrushes with varied graft density and chain lengths. c, d Data from GPC/MALS/RI analysis of sidechains cleaved from 100% graft density glycobrushes, M_n (number average molecular weight, black circles) and Ð (weight average molecular weight × number average molecular weight⁻¹, M_w/M_n, red triangles), c is 50% glycosylated sidechains 1 : 1 GlcK:BnE, and d is 100% glycosylated GlcK sidechains. e Cleavage of glycopolypeptide branches at Met residues enables separate analysis of glycobrushes and branches. f Representative example GPC/MALS/RI analysis of a PAMK₆₃ backbone, 100% grafting density, and 100% glycan density glycobrush, PAMK₆₃-g-PGalNAcS₄₃ glycobrush, and cleaved PGalNAcS₄₃ branches.

Fig. 3. Analyses of glycobrush conformations and morphologies.

a CD spectra of various glycobrushes. b CD spectra of linear glycopolypeptide branches. c Hydrodynamic size distribution determined via DLS. d Glycobrush particle volume analysis from AFM images. ****P-value < 0.0001 from a two-sided Mann–Whitney test. n = 30 for (PAMK_0.5-s-PEG₂K_0.5)₃₀₀-g-PGlcK45 and n = 44 for (PAMK_0.5-s-PEG₂K_0.5)₁₅₀-g-PGlcK₄₅. e AFM image of (PAMK_0.5-s-PEG₂K_0.5)₁₅₀-g-PGlcK₃₅. f AFM image of (PAMK_0.5-s-PEG₂K_0.5)₃₀₀-g-PGlcK₄₅.

Fig. 4. Glycobrushes are non-toxic to live human cells and can be slowly degraded by natural proteases.

a Cytotoxicity in HEK293T cells after 24 h incubation with PAMK₆₃-g-PGlcK₂₃ or PAMK₆₃-g-PGalNAcS₃₃ (CCK-8 assay). Data are presented as mean ± SEM. One-way ANOVA adjusted for multiple comparisons was used to compare each treatment to the PBS control. “ns” indicates not significant with α-level of 0.05 and ****P-value < 0.0001 when compared to the PBS control. P-value = 0.5665 and 0.3909 for PAMK₆₃-g-PGalNAcS₃₃ and PAMK₆₃-g-PGlcK₂₃, respectively. b Polyacrylamide gel of glycobrush PAMK₆₃-g-PGalNAcS₃₃ or glycobrush after treatment with various proteases. Digestions were performed at 37 °C for 48 h. Gels were imaged using a periodate-based stain. Studies were performed in four replicates.

Fig. 5. Robust and prolonged glycocalyx engineering using Chol-terminal GalNAcS- glycobrushes.

a Reaction scheme for conjugation of AF594-NHS to glycobrush (PLG_0.8-s-PAIK_0.2)₅₀-g-PGalNAcS₁₃-Chol amino termini, followed by engineering of the glycocalyx of live epithelial cells. Cell images in b, c are 24 h after incubation b with 15 µM AF594-glycobrush-N₃ or c with 15 µM AF594-glycobrush-Chol. d Flow cytometry data of HEK293T cells untreated (blue) or treated with AF594-glycobush-N₃ (magenta) or AF594-glycobrush-Chol (black). Cell images in e–g are paraformaldehyde-fixed HEK293T cells that were incubated with 15 µM AF594-glycobrush-Chol and 30 μg/mL CF488A-transferrin, where e is imaging for AF594-glycobrush-Chol, f is imaging for CF488A-transferrin, and g is the overlay of the images in e and f. Scale bars are 50 µm. Images in b, c, e–g are representative of four separate experiments.