Histone H1 functions as a stimulatory factor in backup pathways of NHEJ

Abstract

DNA double-strand breaks (DSBs) induced in the genome of higher eukaryotes by ionizing radiation (IR) are predominantly removed by two pathways of non-homologous end-joining (NHEJ) termed D-NHEJ and B-NHEJ. While D-NHEJ depends on the activities of the DNA-dependent protein kinase (DNA-PK) and DNA ligase IV/XRCC4/XLF, B-NHEJ utilizes, at least partly, DNA ligase III/XRCC1 and PARP-1. Using in vitro end-joining assays and protein fractionation protocols similar to those previously applied for the characterization of DNA ligase III as an end-joining factor, we identify here histone H1 as an additional putative NHEJ factor. H1 strongly enhances DNA-end joining and shifts the product spectrum from circles to multimers. While H1 enhances the DNA-end-joining activities of both DNA Ligase IV and DNA Ligase III, the effect on ligase III is significantly stronger. Histone H1 also enhances the activity of PARP-1. Since histone H1 has been shown to counteract D-NHEJ, these observations and the known functions of the protein identify it as a putative alignment factor operating preferentially within B-NHEJ.

Figure 1.

Identification of H1 as a DNA-end-joining factor. (A) Outline of the nuclear extract fractionation scheme utilized. Extracts were first fractionated over a double-stranded DNA cellulose column followed by further fractionation of active fractions over a Mono S column. Shown are the general fractionation schemes, typical chromatograms for each column, as well as SDS–PAGE gels of peak fractions. (B) Effect of fraction IIID on DNA-end joining catalyzed by purified DNA Ligase IIIβ. Reactions were assembled with the indicated amounts of protein. Shown as control are reactions assembled without extract, or with crude nuclear extract. (C) Coomassie blue-stained SDS–PAGE gel (12%) of fractions 86 and 87 (IIID, 20 µl) from a fractionation over a Mono S column. The two prominent bands detected were excised and subjected to mass spectrometry analysis. (D) Sequest search of peptides characterized by LC/MS-MS from the bands shown in C led to the identification of histone H1 variants H1.4 and H1.5 in the upper band and histone H1.2, H1.4 and H1.5 (74) in the lower band (derived by Flicka sequence analysis). The graph shows the corresponding isoforms and the coverage achieved through the peptides analyzed (grey boxes; lines indicate the size of the covered area and the size of the protein isoform itself). (E) Western blot analysis of the indicated fractions with an anti-histone H1 antibody. Ten micrograms of HeLa nuclear extract (NE) and each 10 µl of fractions I/II and III, IIIB1 and IIIB2 as well as 2 µl of fraction IIID2 were separated on a 12% SDS–PAGE gel and blotted onto a nitrocellulose membrane.

Figure 2.

Characterization of H1 function in DNA-end joining. (A) Comparison of end joining between purified histone H1.2 and MonoS fraction IIID. Reactions were assembled with the indicated amount of protein and end-joining activity evaluated. Note that MonoS IIID, when used alone, does not show detectable DNA-end joining activity. (B) MonoS fraction IIID and histone H1.2 enhance DNA-end joining in reactions assembled with 10 ng purified recombinant DNA ligase IIIβ. Ligase alone has only limited and H1.2 alone (50 ng) shows no detectable DNA-end-joining activity. (C) Histone H1.2 titering in DNA-end-joining reactions assembled with 10 ng DNA ligase IIIβ. The DNA substrate, 50 ng pSP65, was linearized with Sal I to generate ends with 4 nt 5′ overhangs and was incubated with histone H1.2 for 10 min at 25° before adding DNA ligase IIIβ and ATP to start end joining, which was carried out at 25°C for 30 min. (D) Titering of Histone H1.2 in DNA-end-joining reactions assembled with 100 ng DNA ligase IV/XRCC4. Other details as in (C). Reactions were assembled without KCl to increase DNA-end-joining activity (see text). (E) DNA-end joining in reactions assembled with 10 ng DNA ligase IIIβ and different amounts of histone H1.2 using pSP65 plasmid digested with Pst I to generate 4 nt 3′ overhangs. (F) As in (E) but for reactions assembled with 100 ng DNA ligase IV/XRCC4. (G) DNA-end joining in reactions assembled with 10 ng DNA ligase IIIβ and different amounts of histone H1.2 using pSP65 plasmid digested with Sma I to generate blunt ends. Other conditions as in (C). (H) As in (G) but for reactions assembled with 100 ng DNA ligase IV/XRCC4.

Figure 3.

The effect of histone H1 on DNA-end joining depends upon the amount and the length of the DNA substrate. (A) Titration of the effect of histone H1 on DNA-end joining for reactions assembled with 10 ng DNA Ligase IIIβ and different amounts of Sal I digested substrate DNA as indicated. Other conditions as in Figure 2C. (B) DNA-end joining in reactions assembled with 20 ng DNA Ligase IIIβ and 30 ng substrate DNA of different lengths as indicated. DNA fragments of 497 and 943 bp were prepared by digesting pUC19 with Alw44I. Other conditions as in Figure 2C. Alw44I recognizes the sequence 5′-G|TGCAC-3′ and generates ends with 3′ 4-bp extensions. (C) The choice of solvent modulates the activity of H1.2 in DNA-end joining. The experiments described earlier were carried out with H1.2 dissolved in reaction buffer (see ‘Materials and methods’). Titration of end-joining reactions with H1.2 dissolved in water or reaction buffer and tested in the presence of 10 ng DNA Ligase IIIβ. Other details as in Figure 2C. Note the shift in the maximum of DNA-end joining from 80 to 20 ng.

Figure 4.

The effect of histone H1 on DNA-end joining cannot be attributed to H1-induced DNA aggregation. Products of DNA-end joining reactions, carried out as described under ‘Materials and methods’ section in the presence of purified DNA Ligase IIIβ and histone H1.2, were analyzed for possible aggregation by gel filtration (Sephacryl S-1000 SF in HR 10/30, void volume 9.7 ml). (A, B) Calibration of the column with pUC19 fragments, pSP65 and λ DNA. (A) shows a typical chromatogram, whereas (B) the DNA fragments found in the different fractions by agarose gel electrophoresis. The DNA mixture contained 0.45 µg λ DNA, 1 µg linearized pSP65 (3005 bp), 0.72 µg 943 bp and 0.4 µg 322 bp DNA fragments. High molecular weight λ DNA is found in early fractions 14–16, pSP65 substrate size DNA in fractions 14–26, and low molecular weight DNA in fractions 22–34, validating thus the separation potential of the column. (C, D, E) Analysis by gel filtration of substrate and products of DNA-end-joining reactions assembled with 20 ng DNA Ligase III and the indicated amounts of histone H1. Gel filtration was carried out in a buffer containing 80 mM NaCl. (F, G, H) Reactions similar to those shown in C, D and E were analyzed by gel filtration in a buffer containing 600 mM NaCl. Other reaction details as in Figure 2C.

Figure 5.

Histone H1 activates PARP-1. (A) End-joining reactions as described in Figure 2C were assembled with 20 ng DNA Ligase IIIβ, in the presence of NAD⁺, and with different amounts of histone H1 and PARP-1, as indicated. PARP-1 activity was measured as autopoly(ADP-ribosylation) by Western blotting using an antibody against poly(ADP-ribose). The lower panel shows the quantification of the results shown in the blot. (B) Results from the same reactions used in the Western blot shown in (A) but analyzed by dot blotting. The lower panel shows the quantification of the results shown in the blot. (C) DNA-end-joining activity measured in the reactions used in A and B to measure PARP-1 activity. Reactions were assembled as described in Figure 2C with different amounts of histone H1 and PARP-1 as indicated. (D) The two pathways of NHEJ. D-NHEJ requires DNA-PK and carries out the final ligation step utilizing the LigIV/XRCC4/XLF complex, whereas B-NHEJ uses histone H1 as an alignment factor and carries out end ligation utilizing LigIII/XRCC1 with some contribution from PARP-1.