Discovering Targets of Non-enzymatic Acylation by Thioester Reactivity Profiling

Abstract

Non-enzymatic protein modification driven by thioester reactivity is thought to play a major role in the establishment of cellular lysine acylation. However, the specific protein targets of this process are largely unknown. Here we report an experimental strategy to investigate non-enzymatic acylation in cells. Specifically, we develop a chemoproteomic method that separates thioester reactivity from enzymatic utilization, allowing selective enrichment of non-enzymatic acylation targets. Applying this method to cancer cell lines identifies numerous candidate targets of non-enzymatic acylation, including several enzymes in lower glycolysis. Functional studies highlight malonyl-CoA as a reactive thioester metabolite that can modify and inhibit glycolytic enzyme activity. Finally, we show that synthetic thioesters can be used as novel reagents to probe non-enzymatic acylation in living cells. Our studies provide new insights into the targets and drivers of non-enzymatic acylation, and demonstrate the utility of reactivity-based methods to experimentally investigate this phenomenon in biology and disease.

Keywords: Warburg effect; acetylation; acylation; epigenetics; glycolysis; malonylation; metabolism; non-enzymatic; reactivity-based protein profiling; thioester.

Figure 1

Design of a reactivity-based approach for profiling non-enzymatic acetylation. (a) Structure of acetyl-CoA bound to the prototypical lysine acetyltransferase GCN5L2 (PDB 1Z4R). Yellow lines indicate hydrogen bonds made between Gcn5 and the acetyl-CoA cofactor. (b) Reactivity-based profiling reagent 1.

Figure 2

Assessing the reactivity of a thioester reporter in complex proteomes. (a) Strategy for ex situ (lysate) and in situ (cellular) labeling using thioester 1. (b) Dose-dependent ex situ labeling of cancer cell lysates by thioester 1 (A549, 15 h, 0, 100, 200, 400, 800 μM). (c) Labeling by 1 is competed by acetyl-CoA (1 h pre-incubation with 0, 100, 200, 400, or 800 μM acetyl-CoA, then 400 μM 1 for 15 h). (d) Dose-dependent in situ labeling of cancer cell by thioester 1 (A549, 15 h, 0, 100, 200, 400, 800 μM).

Figure 3

Assessing the targets of a thioester reporter in complex proteomes. (a) Strategy for ex situ (lysate) and in situ (cellular) labeling of thioester 1 targets. Following click chemistry to biotin-azide, tagged proteins are subjected to on-bead tryptic digest and identified by LC-MS/MS. (b) Validation of targets by affinity capture/immunoblot. Left: In situ enrichment of Raji proteins by thioester 1 (15 h; 0, 100, 400 μM). Right: Ex situ enrichment of Raji proteins by thioester 1 is competed by acetyl-CoA (30 min, 1 mM acetyl-CoA; then 15 h, 100 μM 1). (c) Spectral count data for universally identified thioester-reactive proteins enriched by thioester probe 1 (15 h, 400 μM). Data represents the output of individual LC-MS/MS experiments. Listed proteins were enriched ≥2-fold in each cancer cell line specified, using both in situ and ex situ labeling protocols, respectively. Functionally related proteins are grouped according to color. Complete lists of enriched targets are given in Supplementary Table S1. (d) Terms strongly enriched during gene ontology analysis of universally identified thioester-reactive proteins.

Figure 4

Thioester reactivity-mediated inhibition of glycolytic enzyme activity. (a) Effect of cytosolic acyl-CoA thioesters (200 μM) on recombinant GAPDH activity after 30 min pre-incubation. (b) Inhibition of recombinant GAPDH by malonyl-CoA (200 μM) is time-dependent and correlates with increased lysine malonylation. (c) Gel-based comparison of acetyl-CoA, malonyl-CoA and succinyl-CoA reactivity. Acyl-CoAs (200 μM) were incubated with BSA at 37 °C for 6 h prior to western blotting. Enzyme activity represents the average of ≥3 replicates, with significance analyzed by unpaired Student’s t test (ns = not significant, * = P < 0.05, ** = P < 0.01, and *** = P < 0.001)

Figure 5

Exploring the effects of cytosolic thioester reactivity on glycolytic enzymes. (a) Strategies for direct (left) and indirect (right) manipulation of enzyme malonylation using malonyl-NAC and orlistat, respectively. (b) Effects of malonyl-NAC (1 mM) and orlistat (25 μM) on cellular malonylation in A549 cells. (c) Effects of malonyl-NAC (1 mM) and orlistat (25 μM) on GAPDH activity in A549 cells. (d) Effects of malonyl-NAC (1 mM) on pyruvate kinase activity in A549 cells. Enzyme activity represents the average of ≥3 replicates, with significance analyzed by unpaired Student’s t test (ns = not significant, * = P < 0.05, ** = P < 0.01)

Figure 6

Exploring the effects of cytosolic thioester reactivity on cellular glucose metabolism. (a) Schematic of glycolysis. Bold green and red arrows reflect expected changes in abundance of metabolites that lie upstream and downstream of glycolytic enzymes targeted by non-enzymatic acylation. Gene names in italic refer to high confidence targets of non-enzymatic acylation where malonyl-NAC may intervene. (b) Effects of malonyl-NAC (1 mM, 24 h) on glucose-6-phosphate (G6P) and fructose-1,6-bisphosphate (FBP) levels in A549 cells. (c) Effects of malonyl-NAC (1 mM, 24 h) on lactate levels in A549 cells. (d) Effects of malonyl-NAC (1 mM, 24 h) on glucose-derived acetyl-CoA (M+2) in A549 cells. Metabolomic measurements represent the average of ≥3 replicates, with significance analyzed by unpaired Student’s t test (ns = not significant, * = P < 0.05, ** = P < 0.01)