The contribution of PARP1, PARP2 and poly(ADP-ribosyl)ation to base excision repair in the nucleosomal context

Abstract

The regulation of repair processes including base excision repair (BER) in the presence of DNA damage is implemented by a cellular signal: poly(ADP-ribosyl)ation (PARylation), which is catalysed by PARP1 and PARP2. Despite ample studies, it is far from clear how BER is regulated by PARPs and how the roles are distributed between the PARPs. Here, we investigated the effects of PARP1, PARP2 and PARylation on activities of the main BER enzymes (APE1, DNA polymerase β [Polβ] and DNA ligase IIIα [LigIIIα]) in combination with BER scaffold protein XRCC1 in the nucleosomal context. We constructed nucleosome core particles with midward- or outward-oriented damage. It was concluded that in most cases, the presence of PARP1 leads to the suppression of the activities of APE1, Polβ and to a lesser extent LigIIIα. PARylation by PARP1 attenuated this effect to various degrees depending on the enzyme. PARP2 had an influence predominantly on the last stage of BER: DNA sealing. Nonetheless, PARylation by PARP2 led to Polβ inhibition and to significant stimulation of LigIIIα activities in a NAD⁺-dependent manner. On the basis of the obtained and literature data, we suggest a hypothetical model of the contribution of PARP1 and PARP2 to BER.

Figure 1

(A) Nucleosome structure pictured according to Ref. and was created using RCSB PDB online version specifically https://www.rcsb.org/structure/3LZ0. The red nucleobases depict the positions of midward- or outward-oriented lesions. (B) The scheme of preparation of different types of NCPs. The coloured line corresponds to the DNA strand, whereas the green asterisk to the 5′[³²P] or FAM label. (C) Stages of BER with the participating enzymes and substrates tested in this study.

Figure 2

The activation of PARP1 and PARP2 by DNA or an NCP as measured by the amount of a poly(ADP-ribose) synthesised by PARP1 or PARP2 using [³²P]labelled NAD⁺ as a precursor. The graphs in panels (A–D) deal with PARP1 activation. The graphs in panels (E–H) are about PARP2 activation. In all the graphs, the data obtained with a native substrate are indicated by (circle), with AP-NCP by (square) and with gap-NCP by (triangle). The graphs in panels in the left part represent the accumulation kinetics of PAR at 1000 μM NAD⁺; the histograms in panels in the right part represent the total amount of PAR synthesised by PARP1 for 1 min or by PARP2 for 3 min at NAD⁺ concentrations 1–1000 μM or 1–4000 μM, respectively. In all the graphs, the experimental data on DNA substrates are characterised by black, reddish-violet or violet curves; the NCP substrates correspond to the red or yellowish-green curves. The data are presented as an average of at least three independent experiments and are shown as the mean ± SD. The reactions were carried out as described in “Materials and methods”.

Figure 3

The kinetic assay of APE1 activity towards an outward- or midward-oriented AP site in the NCP context in the presence of PARP1 or PARP2. (A) Kinetics for out-NCP and PARP1. (B) Kinetics for mid-NCP and PARP1. (C) Kinetics for out-NCP and PARP2. (D) Kinetics for mid-NCP and PARP2. (E) The influence of PARylation catalysed by PARP1 or PARP2 on the activity of APE1 towards AP site–containing DNA or NCP-based substrates. The ‘control’ label corresponds to the spontaneous cleavage of the AP site; ‘APE1’ indicates the specific cleavage by APE1 under the reaction conditions; ‘+ 0.1 μM PARP’ denotes the specific cleavage in the presence of the indicated protein; and ‘PARylation’ represents the specific cleavage in the presence of the indicated protein and 400 μM NAD⁺. The data are presented as an average of at least three independent experiments and are shown as the mean ± SD. The reactions were carried out as described in “Materials and methods”. All graphs are normalised to spontaneous cleavage of an AP site under the experimental conditions. (F) Separation of the products (on a 10% denaturing polyacrylamide gel) of the reaction of 0.1 μM 5′-FAM–labelled AP-NCP with APE1 under the indicated experimental conditions. In all cases, the APE concentrations of 0.03 and 1 μM were chosen for the reaction conditions involving substrates out-NCP and mid-NCP, respectively.

Figure 4

The activity of Polβ towards gap-NCP substrates by itself (A,B) and during PARylation catalysed by PARP1 (C,D) or PARP2 (E,F) in the absence (A–C,E) or presence of XRCC1 (D,F). The graphs in panels (C–F) represent the data obtained at 10 μM dTTP and 50 nM PARP1 or PARP2. The data are presented as an average of at least three independent experiments and are shown as the mean ± SD. (B) The reaction products of dTMP incorporation by Polβ (lanes 3–6) in the presence of PARP1 (left part) or PARP2 (right part, lanes 7–10) when outward- (upper panel) or midward-oriented (lower panel) 5′-FAM–labelled gap-NCP was employed. Lane 1: substrate AP-NCP, lane 2: substrate gap-NCP. In all cases, Polβ concentrations of 2.5 and 50 nM were chosen as the reaction conditions for substrates out-NCP and mid-NCP, respectively. The reaction products were separated on a 15% denaturing polyacrylamide gel.

Figure 5

The LigIIIα activity on nicked substrate out-NCP. Product separation after the sealing of 5′[³²P]labelled nicked out-NCP by LigIIIα with PARP1 or PARP2 and PARylation in the presence of XRCC1. Lane 1: native out-NCP; lane 2: out-NCP incubated with UDG; lane 3: out-NCP incubated with UDG and APE1; lane 4: out-NCP incubated with UDG, APE1, Polβ and dTTP, resulting in nicked out-NCP; lanes 5–16: nick sealing in the presence of LigIIIα (lanes 5, 8) and PARP1 (lanes 6, 9) or PARP2 (lanes 12, 14) without or with NAD⁺ (lanes 7, 10, 11 and 13, 15, 16, respectively). Lanes 8–10 and 14–15 correspond to lanes 5–7 and 12–13 with an additional treatment (with proteinase K), respectively. Lanes 11 and 16 correspond to lanes 7 and 13 with an additional treatment (with PARG). The data are presented for experiments involving 20 nM NCP, 500 nM LigIIIα, 500 nM XRCC1, 100 nM PARP1 or PARP2 and 100 μM NAD⁺.

Figure 6

The influence of PARP1 or PARP2 and PARylation on the LigIIIα activity in the absence or presence of XRCC1. The ligation magnitude of the 5′[³²P]labelled nicked NCP substrate under the action of LigIIIα was affected by PARP1 or PARP2 and PARylation in the absence or presence of XRCC1. The ligation magnitude was calculated as a percentage of the unreacted primer subtracted from 100%. The data come from experiments involving 20 nM NCP, 500 nM LigIIIα, 100 nM PARP1 or PARP2 and 100 μM NAD⁺. The data are presented as an average of at least three independent experiments and are shown as the mean values ± SD.