Chemoenzymatic and Synthetic Approaches To Investigate Aspartate- and Glutamate-ADP-Ribosylation

Abstract

We report here chemoenzymatic and fully synthetic methodologies to modify aspartate and glutamate side chains with ADP-ribose at specific sites on peptides. Structural analysis of aspartate and glutamate ADP-ribosylated peptides reveals nearquantitative migration of the side chain linkage from the anomeric carbon to the 2″-or 3″-ADP-ribose hydroxyl moieties. We find that this linkage migration pattern is unique to aspartate and glutamate ADP-ribosylation and propose that the observed isomer distribution profile is present in biochemical and cellular environments. After defining distinct stability properties of aspartate and glutamate ADP-ribosylation, we devise methods to install homogenous ADP-ribose chains at specific glutamate sites and assemble glutamate-modified peptides into full-length proteins. By implementing these technologies, we show that histone H2B E2 tri-ADP-ribosylation is able to stimulate the chromatin remodeler ALC1 with similar efficiency to histone serine ADP-ribosylation. Our work reveals fundamental principles of aspartate and glutamate ADP-ribosylation and enables new strategies to interrogate the biochemical consequences of this widespread protein modification.

Preserving ester-linked modifications reveals glutamate and aspartate mono-ADP-ribosylation by PARP1 and its reversal by PARG

“…2d ). This indicates that the lability of Asp/Glu-ADPr is influenced by a combination of factors—specifically, time, pH, and temperature—a conclusion that is corroborated by a recent study 20 .…”

“…2A ). In contrast, in WT cells the mono-ADPr signal was mostly unaffected by boiling, reflecting the stability of Ser-ADPr at high temperatures, as noticed previously 14 , 17 , 20 . Importantly, protein extraction and immunoblotting efficiency, assessed via ponceau, PARP1, GAPDH and H3 staining, was not affected by the omission of boiling.…”

“…These observations indicate that PARG can hydrolyze mono-Asp/Glu-ADPr in cells, a surprising finding in light of several earlier studies which showed that in biochemical reactions PARG cleaves poly-ADPr but cannot remove the last ADP-ribose attached to protein targets 13 – 16 , 18 , 19 . To resolve the discrepancy between our findings in cells and previous results from biochemical assays, we conducted a biochemical characterization of PARG, encouraged by a recent report suggesting that recombinant PARG can remove mono-Asp/Glu-ADPr from in vitro modified peptides 20 and by the complete removal of PARP1 automodification observed in PARG zymograms 21 . First, we used automodified PARP1E988Q as a source of mono-Asp/Glu-ADPr.…”

“…Consistently, in our own studies we were unable to detect mono-Asp/Glu-ADPr by western blotting 14 , 17 . Yet we were intrigued by a small but significant mono-ADPr signal detected by immunofluorescence in DNA damage-treated HPF1 KO cells 17 and by a recent report on the lability of in vitro produced Asp/Glu-ADP-ribosylated peptides 20 . This led us to wonder if the absence of a detectable mono-Asp/Glu-ADPr signal might result not from absence of this modification in cells, but rather from the high chemical lability of its ester linkage compared to the O-glycosidic bond of Ser-ADPr (Fig.…”

“…By contrast, PARG degrades poly-ADPr and, according to the current consensus, is unable to remove the initial protein-ribose bond, leaving a mono-ADPr remnant on its targets 6 , 13 – 19 . However, this view has been recently challenged by evidence that recombinant PARG can cleave mono-ADPr from peptides 20 . This is supported by an earlier finding that PARG can completely degrade poly-ADPr, including the initial ADP-ribose moiety 21 .…”