Reversal of tyrosine-linked ADP-ribosylation by ARH3 and PARG

Abstract

ADP-ribosylation is an ancient posttranslational modification with exceptional versatility in terms of breadth of modification targets including at least seven different amino acid side chains, various moieties on nucleic acids, and a variety of small chemical compounds. The spatiotemporal signaling dynamic of the different modification variations is tightly regulated and depends on the writers, erases, and readers of each type. Among these, tyrosine ADP-ribosylation (Tyr-ADPr) has been consistently detected as a novel modification type, but systematic analysis of its potential physiological role, modification establishment, and reversal are still lacking. Here we present a re-analysis of recent ADP-ribosylome data and show that Tyr-ADPr sites are conserved and enriched among ribosome biogenesis and mRNA processing proteins and that these sites are affected by the status of the (ADP-ribosyl)hydrolase ARH3. To facilitate the study of Tyr-ADPr, we establish methodologies for the synthesis of well-defined Tyr-ADPr peptides and with these could show that Tyr-ADPr is reversed both by ARH3 and PARG enzymes. Together, our work lays the foundation for the future exploration of the Tyr-ADPr.

Keywords: (ADP-ribosyl)hydrolase; DNA damage; HPF1; PARP; mass spectrometry; peptide synthesis.

Figure 1

Identification of Tyr-ADPr in ADP-ribosylome data.A, Euler diagram showing Tyr-ADPr site identification in recent publications (16, 17, 18, 19, 22). B, heatmap plot of average abundance for each Tyr-ADPr peptide identified in Hendriks et al. (17). Shown is the average intensity (n = 4 biological replicates) of each identified peptide normalized across experimental conditions. C, heatmap plot of average abundance for each Tyr-ADPr peptide (17) shown is the average intensity (n = 4 biological replicates) of each identified peptide normalized across all conditions. D, bar plot showing average abundance of RPS3A Y155ADPr identified in WT and ARH3^−/− cells untreated or treated with H₂O₂. Data is expressed as mean ± SEM of n = 4 biological replicates. Note, only two data points for WT untreated above the detection limit.

Figure 2

Chemical synthesis of Tyr-ADPr peptide.A, scheme showing synthesis of Tyr-ribosylated building block 5 ready for SPPS. Reagents and conditions: (i) All-Br, DIPEA, DMF. (ii) 20% TFA in DCM, TIS. (iii) TBSOTf, DCM, −50 °C. (iv) Pd(PPh₃)₄, DMBA, DCM. B, scheme showing synthesis of Tyr-ADPr peptide 11 using acid-sensitive peptide segment protection and base-sensitive adenosine protective groups.

Figure 3

ARH3 and PARG reverse HPF1-dependent Tyr-ADP ribosylation.A, CTCF_9-19 (Table S3) was enzymatically modified by PARP1 or PARP2 in the presence or absence of HPF1. B, hydrolysis of Tyr-ADPr on peptide 11 by ARH3 and human macrodomains and detection by chemiluminescence using the AMP-Glo assay (Promega). Samples were measured in triplicates, background corrected, and represent mean ± SD. C, hydrolysis of peptide 11 by representatives of the ARH3-like family (Parus major [PmaARH3], Xenopus tropicalis [XtrARH3], Latimeria chalumnae [LchARH3], Chlamydomonas reinhardtii [CreARH3-like], Fusarium oxysporum sp.f. cubense (race 1) [Foc1ARH], Flavobacterium johnsoniae [FjoARH]) and detection by AMP-Glo assay (Promega). Samples were measured in triplicates, background corrected, and represent mean ± SD. D, hydrolysis of peptide 11 by representatives of the PARG family (‘classical’ members Homo sapiens [hPARG] and Drosophila melanogaster [dmPARG], as well as microbial-type (mPARGs) Thermomonospora curvata [Tc-mPARG] and Flavobacterium johnsoniae [Fjo-mPARG]) and detection by AMP-Glo assay (Promega). Samples were measured in triplicates, background corrected, and represent mean ± SD. E, CTCF_9-19 peptide was enzymatically modified by PARP1/HPF1. The modified peptide was demodified by either ARH3 or PARG using catalytically inactive mutants as controls. F and G, top: Deconvoluted MS spectrum of CTCF_9-19 (F) and PARP1_629-641 (G) mixtures of unmodified (masses = 2012.1 and 2183.19, respectively) and ADP-ribosylated peptides modified on Y15 (mass = 2553.16) and Y634 (mass = 2183.19), respectively. Middle/Bottom: Deconvoluted MS spectra of ADP-ribosylated peptides treated with 1 μM ARH3 (middle) or PARG (bottom) for 3 h at 37 °C. Mass of ADP-ribosyl moiety = 541.06 Da.