Site-Specific Glycation of Human Heat Shock Protein (Hsp27) Enhances Its Chaperone Activity

Abstract

Non-enzymatic posttranslational modifications are believed to affect at least 30% of human proteins, commonly termed glycation. Many of these modifications are implicated in various pathological conditions, e.g., cataract, diabetes, neurodegenerative diseases, and cancer. Chemical protein synthesis enables access to full-length proteins carrying site-specific modifications. One such modification, argpyrimidine (Apy), has been detected in human small heat shock protein Hsp27 and closely related proteins in patient-derived tissues. Thus far, studies have looked into only artificial mixtures of Apy modifications, and only one has analyzed Apy188. We were interested in understanding the impact of such individual Apy modifications on five different arginine sites within the crucial N-terminal domain of Hsp27. By combining protein semisynthesis with biochemical assays on semisynthetic Hsp27 analogues with single-point Apy modification at those sites, we have shown how a seemingly minimal modification within this region results in dramatically altered functional attributes.

Figure 1

Salient features of Hsp27. (A) Domain architecture of Hsp27 highlighting the amino acid sequence within the distal NTD. Arginine residues chosen for the single-point Apy modification are colored red. The functionally important IPV motif (red) within the CTD interacts with the β4−β8 groove (deep blue) of the ACD. The illustration was created utilizing the reported crystal structure (Protein Data Bank entry 4mjh). (B) Generation of argpyrimidine (Apy) via the methylglyoxal (MG) modification of arginine.

Figure 2

Semisynthesis of site-specifically Apy-modified full-length Hsp27. (A) Fmoc/t-Bu-based SPPS utilizing building block 1 generated the site-specifically Apy-containing thioesters 3a–f. (B) Recombinant expression of a fusion construct 4 followed by enzymatic removal of the affinity tag provided the unmodified segment 6 required for the subsequent (C) native chemical ligation (6 M Gdn·HCl, 0.2 M NaP_i, 300 mM MESNa, 50 mM TCEP, pH 7.8, and 37 °C) and desulfurization (6 M Gdn·HCl, 0.2 M NaP_i, 250 mM TCEP, 100 mM MESNa, 20 mM VA-044, pH 7.2, and 37 °C) to furnish the final products 8a–f, which were further characterized by ESI-MS and RP-HPLC.

Figure 3

Biophysical characterization of semisynthetic Hsp27 analogues. (A) Overlay of the far-ultraviolet circular dichroism spectra of folded wild type 8a and site-specifically Apy-modified Hsp27 8b–f. For the sake of clarity, the identity of the Hsp27 variants is also included. (B) Size-exclusion chromatography profile of folded Hsp27 samples 8a–f. The elution volume and the molecular weight (in kilodaltons) of the standard proteins used for calibration are shown as triangles.

Figure 4

Biochemical characterization of semisynthetic Hsp27 analogues. (A) Citrate synthase (CS; 2 μM) was incubated at 45 °C in the absence or presence of 0.45 μM Hsp27 analogues 8a–f. (B) Glyceraldehyde 3-phosphate dehydrogenase (GAPDH; 3 μM) was incubated at 45 °C in the absence or presence of 0.6 μM Hsp27 analogues 8a–f. (C) Malate dehydrogenase (MDH; 2 μM) was incubated at 45 °C in the absence or presence of 0.25 μM Hsp27 analogues 8a–f. (D) Chaperone activity expressed as percentage protection with respect to amorphous aggregation in the absence of a chaperone.

Figure 5

Biophysical characterization of semisynthetic Hsp27 analogues 8a and 8g. (A) Overlay of the far-ultraviolet circular dichroism spectra of folded wild type 8a and penta-Apy-modified Hsp27 8g. (B) Size-exclusion chromatography profile of folded Hsp27 samples 8a and 8g. The elution volume and molecular weight (in kilodaltons) of the standard proteins used for calibration are shown as triangles.

Figure 6

Comparison of in vitro chaperone activity with those of the client proteins. (A) Citrate synthase (CS; 2 μM) was incubated at 45 °C in the absence or presence of 0.45 μM Hsp27 analogues 8a, 8d, and 8g. (B) Glyceraldehyde 3-phosphate dehydrogenase (GAPDH; 3 μM) was incubated at 45 °C in the absence or presence of 0.6 μM Hsp27 analogues 8a, 8e, and 8g. (C) Malate dehydrogenase (MDH; 2 μM) was incubated at 45 °C in the absence or presence of 0.25 μM Hsp27 analogues 8a, 8d, and 8g. (D) Chaperone activity expressed as percentage protection with respect to amorphous aggregation in the absence of a chaperone.

Figure 7

In vitro chaperone activity assay of 8a, 8d, and 8g against tau as a model client protein known to undergo amyloidogenic aggregation. The aggregation was monitored via ThT fluorescence (excitation at 444 nm and emission at 490 nm) as well as scanning electron microscopy. (A) Activity comparison between 8a and 8d at two different doses (1:1 and 1:2 tau:8a or tau:8d). (B) Activity of penta-modified version 8g (1:1 tau:8g). (C) Scanning electron microscopy (SEM) image from the assays with tau only (induced). (D) Image from the assays with tau and 8a. (E) Image from the assays with tau and 8d. (F) Image from the assays with tau and 8g. Scale bars of 200 nm. Black arrows mark the tau fibrils.

Figure 8

(A) Impact of Apy modification on the chaperone activity of Hsp27 against an amyloidogenic client protein. Single-site modification (n = 1) causes no apparent change in the activity as the hydrophobic binding patch (colored dark teal) is still available for binding to the client protein (colored teal), while multiple modifications (n = 5) presumably cause certain structural reorganizations of Hsp27 resulting in the abolishment of anti-amyloid chaperone activity. The illustration was created utilizing the reported crystal structures of Hsp27 (Protein Data Bank entry 4mjh) and Aβ(1–42) (Protein Data Bank entry 2mxu). (B) Titration curves for 8a, 8d, and 8g. A chaperone concentration of 4.4 μM was mixed with increasing concentrations of bis-ANS (1–12 μM), and the fluorescence emission intensity was recorded at 495 nm using an excitation wavelength of 385 nm. The data are represented as an average of three measurements, including the standard error of the mean.