Structural studies of the PARP-1 BRCT domain

Abstract

Background: Poly(ADP-ribose) polymerase-1 (PARP-1) is one of the first proteins localized to foci of DNA damage. Upon activation by encountering nicked DNA, the PARP-1 mediated trans-poly(ADP-ribosyl)ation of DNA binding proteins occurs, facilitating access and accumulation of DNA repair factors. PARP-1 also auto-(ADP-ribosyl)ates its central BRCT-containing domain forming part of an interaction site for the DNA repair scaffolding protein X-ray cross complementing group 1 protein (XRCC1). The co-localization of XRCC1, as well as bound DNA repair factors, to sites of DNA damage is important for cell survival and genomic integrity.

Results: Here we present the solution structure and biophysical characterization of the BRCT domain of rat PARP-1. The PARP-1 BRCT domain has the globular α/β fold characteristic of BRCT domains and has a thermal melting transition of 43.0°C. In contrast to a previous characterization of this domain, we demonstrate that it is monomeric in solution using both gel-filtration chromatography and small-angle X-ray scattering. Additionally, we report that the first BRCT domain of XRCC1 does not interact significantly with the PARP-1 BRCT domain in the absence of ADP-ribosylation. Moreover, none of the interactions with other longer PARP-1 constructs which previously had been demonstrated in a pull-down assay of mammalian cell extracts were detected.

Conclusions: The PARP-1 BRCT domain has the conserved BRCT fold that is known to be an important protein:protein interaction module in DNA repair and cell signalling pathways. Data indicating no significant protein:protein interactions between PARP-1 and XRCC1 likely results from the absence of poly(ADP-ribose) in one or both binding partners, and further implicates a poly(ADP-ribose)-dependent mechanism for localization of XRCC1 to sites of DNA damage.

Figure 1

Characterization of PARP-1 Constructs. (a) Schematic of PARP-1 indicating the domain structure and constructs used in these experiments. Domain D corresponds to the PARP1 BRCT domain. Functional domains are indicated: ZF1, first zinc finger domain; ZF2, second zinc finger domain; ZF3, third zinc finger domain; WGR, WGR domain. Domains previously demonstrated to interact with XRCC1 are identified with a red asterisk. Molecular masses of constructs, including purification tags are indicated. (b) Schematic of XRCC1 domain structure and constructs used in these experiments. Domain previously identified to interact with PARP-1 is marked with a red asterisk. Molecular masses of constructs, including purification tags are indicated. (c) Gel-filtration chromatogram of domain constructs of XRCC1 and PARP-1. Plots are colored as in (a) (green, blue, red, or light purple) and (b) (black or gray). (d) Gel-filtration chromatogram of mixtures of X1BRCTa and the various PARP-1 domain constructs from panel (a). Chromatograms of the potential complexes are colored based on the scheme of the PARP-1 constructs identified in panel (a). Note that each of the curves in this panel exhibits two maxima that approximately correspond to the maximum positions in panel c and the maximum of the X1BRCTa in panel c (black curve). Molecular mass standards used are represented as colored circles (conalbumin (76.6 kDa), pink; ovalbumin (44.3 kDa), brown; carbonic anhydrase (29.0 kDa), orange; cytochrome C (14.3 kDa), gold) in panels (c) and (d).

Figure 2

Thermal melting of the PARP-1 BRCT domain. Solid line is a fit of the circular dichroism signal to a two-state unfolding model.

Figure 3

Small-angle X-ray scattering of the PARP-1 BRCT domain. (a) SAXS intensity data inset with the probability distribution function. (b) Ab initio molecular envelope (light blue surface representation) of the PARP-1 BRCT domain, generated from SAXS intensity data, superimposed with the NMR solution structure (ribbon representation).

Figure 4

NMR solution structure of the PARP-1 BRCT domain. (a) ¹H-¹⁵N HSQC spectra of the PARP-1 BRCT domain (assigned resonances indicated). Resonances corresponding to side-chain Asn and Gln residues are not labelled. (b) Ribbon diagram of the ten lowest energy structures in the NMR-generated ensemble of the PARP-1 BRCT domain. (c) Ribbon representation of the PARP-1 BRCT domain with labelled secondary structure elements. (d) Electrostatic surface representation of the PARP-1 BRCT domain (blue, positive; red, negative; white, neutral).

Figure 5

Comparison of BRCT domains. (a) CLUSTAL-W alignment of the PARP-1 BRCT domain (PARP), the second BRCT domain of XRCC1 (X1BRCTb), and the first BRCA1 BRCT domain (BRCA1). Residues contributing to the BRCA1 phospho-serine binding site are underlined and in red font; dual repeat interacting residues in the tandem BRCA1 BRCT domains are underlined and in green font, and XRCC1 homodimer interface residues are underlined and in cyan font (adapted from [31]). (b) Stereo view of a superposition of the structure of the sixth BRCT domain of TopBP1 (magenta) and PARP-1 BRCT (cyan). (c) Close-up stereo view of the superposition of the BRCA1 BRCT domain phosphoserine binding site (magenta) and the homologous region from the PARP-1 BRCT domain (cyan). Residues interacting with the phosphoserine (orange) in the BRCA1 BRCT domain (magenta), as well as structurally homologous residues from the PARP-1 BRCT domain (cyan) are shown in stick representation.