Human Oxoguanine Glycosylase 1 Removes Solution Accessible 8

Communication Cite This: Biochemistry XXXX, XXX, XXX−XXX

pubs.acs.org/biochemistry

Human Oxoguanine Glycosylase 1 Removes Solution Accessible 8‑Oxo-7,8-dihydroguanine Lesions from Globally Substituted Nucleosomes Except in the Dyad Region Katharina Bilotti,† Mary E. Tarantino,‡ and Sarah Delaney*,† †

Department of Chemistry and ‡Department of Molecular Biology, Cell Biology, and Biochemistry, Brown University, Providence, Rhode Island 02912, United States S Supporting Information *

axis, while the rotational position of a nucleobase describes its orientation relative to the histone core. Lesions facing outward, rather than in toward the histone core, are more solution accessible. Initial studies suggested that solution accessibility of the damaged base may be predictive of glycosylase activity in an NCP.4 It is also known that increased solution accessibility can derive from transient exposure of lesions via DNA unwrapping, which leads to increased accessibility of nucleobases closer to the ends of the DNA compared to those near the dyad axis.5−9 Several studies have investigated the relationship between solution accessibility and glycosylase activity in NCPs by incorporating single nucleobase lesions with varied translational and rotational positions;10−21 however, the universality of a correlation between increased glycosylase activity and increased solution accessibility has yet to be demonstrated. Notably, an examination of uracil (U) excision throughout an NCP demonstrated that glycosylase activity does not necessarily correlate with solution accessibility.11,21 Steric interactions with the histone core, local distortions of the DNA, and transient unwrapping of the DNA ends may play a role.4,11 In this study, we evaluated BER in the context of packaged DNA and in a global fashion, namely using 8-oxoG lesions with a variety of translational and rotational positions. The substrates are based on the Widom 601 DNA positioning sequence, which binds in a single translational and rotational position around the histone core.22 While previous studies relied on polymerase chain reaction for incorporation of U, similar techniques are incompatible with 8-oxoG because of polymerase infidelity. To overcome this limitation, we used synthetic methods to create a population of DNA that contains global 8-oxoG substitutions (Figure 1). To prepare this DNA, we used a mixture of G/8-oxoG phosphoramidites as a reagent during chemical synthesis. The molar ratio of this mixture is defined by a Poisson distribution to minimize the number of strands with two or more lesions; ultimately, 95% of oligomers have at most one 8-oxoG. This method of 8-oxoG incorporation yields a population of oligomers with an unbiased distribution of G to 8-oxoG substitutions throughout the length of the “I strand” of the Widom 601 DNA duplex (Scheme S1).23 This substrate allows us to evaluate the global profile of repair in packaged DNA and examine the influence of solution

ABSTRACT: Persistent DNA damage is responsible for mutagenesis, aging, and disease. Repair of the prototypic oxidatively damaged guanine lesion 8-oxo-7,8-dihydroguanine (8-oxoG) is initiated by oxoguanine glycosylase (hOGG1 in humans). In this work, we examine hOGG1 activity on DNA packaged as it is in chromatin, in a nucleosome core particle (NCP). We use synthetic methods to generate a population of NCPs with G to 8oxoG substitutions and evaluate the global profile of hOGG1 repair in packaged DNA. For several turns of the helix, we observe that solution accessible 8-oxoGs are sites of activity for hOGG1. At the dyad axis, however, hOGG1 activity is suppressed, even at lesions predicted to be solution accessible by hydroxyl radical footprinting (HRF). We predict this diminished activity is due to the properties of the DNA unique to the dyad axis and/or the local histone environment. In contrast to the dyad axis, the DNA ends reveal hOGG1 activity at sites predicted by HRF to be both solution accessible and inaccessible. We attribute the lack of correlation between hOGG1 activity and solution accessibility at the ends of the DNA to transient unwrapping of the DNA from the protein core, thus exposing the inward-facing lesions.

T

he DNA in our cells is damaged every day by both exogenous and endogenous agents, including free radicals, environmental toxins, and ultraviolet radiation.1 If left unrepaired, these lesions are mutagenic and have been implicated in the aging phenotype and carcinogenesis.1 Nucleobase lesions are repaired by the base excision repair (BER) pathway. BER is initiated by recognition and excision of the damaged nucleobase by a glycosylase specific to the lesion, leaving an abasic site that will continue through the repair process.2,3 For example, the prototypic oxidatively damaged guanine lesion 8-oxo-7,8-dihydroguanine (8-oxoG) is removed by oxoguanine glycosylase (hOGG1 in humans). Though the activity of glycosylases has been examined extensively in free duplex substrates, genomic DNA is packaged into chromatin in vivo. The basic repeating unit of chromatin is the nucleosome core particle (NCP), consisting of 145−147 bp wrapped ∼1.7 times in superhelical coils around an octameric histone core (two copies each of H2A, H2B, H3, and H4) with a central dyad axis of pseudosymmetry. The translational position of a nucleobase describes the distance from this dyad © XXXX American Chemical Society

Received: November 7, 2017 Revised: January 15, 2018 Published: January 17, 2018 A

DOI: 10.1021/acs.biochem.7b01125 Biochemistry XXXX, XXX, XXX−XXX

Communication

Biochemistry

Figure 1. Representation of an NCP. Merged crystal structure of an NCP containing Widom 601 DNA with a histone octamer containing tails (Protein Data Bank entries 3lz0 and 1kx5, respectively). Guanine sites are colored red, and the dyad axis is indicated by a dashed line.

accessibility, DNA dynamics, and the local histone environment on hOGG1 activity in an NCP. Incubating free duplex DNA with hOGG1 confirmed that the synthesis strategy provides a population in which collectively each G is replaced with 8-oxoG (Figure 2, duplex lane). The integrated band area at each base position is shown in Figure 3A. The G sites in the original 601 DNA are colored red; activity of hOGG1 is observed at each site. The signal intensity is generally greater at the 3′-end of the sequence than at the 5′end; this difference has been documented previously24 and is due to the weakened propensity of small DNA fragments to precipitate during sample workup. We attribute other minor differences in reactivity at individual 8-oxoG to modest flanking sequence preferences, as reported previously for hOGG1 activity in duplex DNA substrates.25 We next formed NCPs using this DNA with globally incorporated 8-oxoG. Xenopus laevis histones were expressed in Escherichia coli, purified, assembled into an octamer, and used to form NCPs with the DNA following the salt dialysis method.26 The success of NCP reconstitution was evaluated by nondenaturing polyacrylamide gel electrophoresis, and only >95% pure preparations of NCPs were used in the experiments (Figure S1). Hydroxyl radical footprinting (HRF) was used to evaluate the structure of NCPs and establish the solution accessibility of each lesion position. As reported previously, DNA in an NCP shows an oscillating pattern of protection and damage by hydroxyl radicals, reflecting the DNA wrapping in toward the histone core and out toward solution, respectively.27 This distinct pattern is observed for our population of NCPs

Figure 2. Global reactivity of hOGG1 in free duplex and NCP-bound DNA. Denaturing polyacrylamide gel for examining cleavage patterns in Widom 601 DNA. The DNA is numbered such that 0 corresponds to the center of the Widom 601 “I strand”, the 5′-end is −72, and the 3′-end is +72. The A/G lanes display a sequence ladder created using Maxam−Gilbert reactions. A negative control (QC lane) shows any pre-existing and incidental damage formed by heat/NaOH treatment of the free duplex DNA. Base-catalyzed strand breaks after hOGG1 incubation reveal excision of 8-oxoG from the free duplex DNA substrate (duplex lane) and NCP (NCP lane). Hydroxyl radical footprinting of NCP reveals an oscillating pattern of damage (HRF lane).

containing 8-oxoG (Figure 2, HRF lane). Notably, as seen in Figure 3B, the lesions span regions from low to high HRF reactivity, reflecting varying levels of solution accessibility. After the characterization of the solution accessibility of the 8-oxoG lesions in NCPs, the substrates were incubated with hOGG1. It is reasonable to predict that lesions with an outward-facing rotational position will be more easily excised by hOGG1 than inward-facing lesions, in accordance with their B

DOI: 10.1021/acs.biochem.7b01125 Biochemistry XXXX, XXX, XXX−XXX

Communication

Biochemistry

suggestions that BER cannot be efficiently initiated in NCP without the assistance of chromatin remodelers.20,30 Inhibition at the dyad axis has also been observed for BER enzymes that act downstream of the glycosylase, such as polymerase β, DNA ligase I, and DNA ligase IIIα/XRCC1.15,17,31 Our data suggest, however, that for some enzymes the conclusions made from studies at the dyad axis may not be representative of the global picture of repair in packaged DNA. In the case of hOGG1, we demonstrate that activity at the dyad axis of the Widom 601 NCP is an exception rather than the rule. We also observe a disagreement between the solution accessibility and hOGG1 activity at some lesion sites closer to the 5′-end of the sequence. Of particular interest is site −54, which is predicted by HRF to be inward-facing but is a site of hOGG1 activity. We attribute this divergence to the welldocumented phenomenon of transient unwrapping of the DNA ends from the histone core to expose sites that are normally occluded.5−9,32,33 Interestingly, site −49 is the most solution accessible site within the helical turn that also contains site −54, yet this site is least reactive toward hOGG1. We suggest that additional determining factors, including the presence of local interactions with the histone core and/or tails, can influence excision of 8-oxoG. Indeed, we previously showed that hOGG1 removal of 8-oxoG at site −49 was modulated by the tail of histone H2B.18 It is notable that not all sites at the most 5′-end of the sequence are efficiently repaired, suggesting that transient unwrapping alone is insufficient to explain hOGG1 activity and additional factors; likely steric obstruction by the histone tails and/or superhelical coils play a role in dictating lesion excision. In summary, we observe that both the translational position and the rotational orientation of an 8-oxoG lesion are factors affecting hOGG1 activity in an NCP. While we observe a correlation between solution accessibility and hOGG1 activity in some regions of the NCP, there are exceptions in the dyad region and the ends of the DNA. At the dyad axis, there is a significant drop in hOGG1 activity regardless of rotational orientation, while closer to the ends of the DNA, some sterically occluded lesions are excised, suggesting that transient unwrapping is the dominating force in accessibility of these lesions.6 Our globally substituted NCP substrate demonstrates the complexity of DNA repair in the context of packaged DNA and provides context for further studies of cellular DNA chemistry.

Figure 3. (A) Global activity of hOGG1 on free duplex Widom 601 DNA, (B) HRF on NCP, and (C) activity of hOGG1 on NCPs. In all cases, the DNA contains globally incorporated 8-oxoG. Sites of 8oxoG substitution are colored red, and other bases are colored gray. In panel C, data are represented as NCP/duplex (see the Supporting Information).

solution accessibility and potential steric block of the histones. Interestingly, we do observe a correlation between hOGG1 activity and the rotational orientation of the lesion site in some regions of the NCP (Figure 3C). For example, sites −19, −30, and −40 are outward-facing as determined by HRF and are sites of high hOGG1 reactivity. In a similar manner, sites −24, −34, and −55 are inward-facing as determined by HRF and are sites of low hOGG1 activity. On the other hand, there is significantly less hOGG1 activity on the ∼20 bp centered at the dyad axis regardless of the rotational orientation of the lesion. For example, site +02 is outward-facing but shows much less hOGG1 reactivity than other outward-facing sites in the NCP. DNA at the dyad axis is known to be relatively straight compared to other regions of an NCP and to adopt an unusual helical periodicity;27−29 these factors may modulate hOGG1 activity. Another intriguing observation in Figure 3C is that sites −11 and −07 show nearly identical HRF reactivity and are separated by only 3 bp, but hOGG1 has dramatically different activity at these two sites. We speculate this difference arises from the properties of DNA near the dyad axis, the local histone environment, and/or the presence of superhelical coils. It is of note that many of the initial studies that examined glycosylase activity in NCPs used lesions positioned at the dyad axis.11−13,15−17,19,21 In particular, hOGG1 exhibits limited glycosylase reactivity at the dyad axis,19,20,30 prompting earlier

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.biochem.7b01125.

■

Experimental procedures and a supplemental figure (PDF)

AUTHOR INFORMATION

Corresponding Author

*Department of Chemistry, Brown University, Providence, RI 02912. E-mail: [email protected]. Phone: +1-401-863-3590. Fax: +1-401-863-2594. ORCID

Sarah Delaney: 0000-0002-8366-3808 C

DOI: 10.1021/acs.biochem.7b01125 Biochemistry XXXX, XXX, XXX−XXX

Communication

Biochemistry Author Contributions

(26) Luger, K., Rechsteiner, T. J., and Richmond, T. J. (1999) Methods Enzymol. 304, 3−19. (27) Hayes, J. J., Tullius, T. D., and Wolffe, A. P. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 7405−7409. (28) Hayes, J. J., Clark, D. J., and Wolffe, A. P. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 6829−6833. (29) Arents, G., and Moudrianakis, E. N. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 10489−10493. (30) Menoni, H., Shukla, M. S., Gerson, V., Dimitrov, S., and Angelov, D. (2012) Nucleic Acids Res. 40, 692−700. (31) Chafin, D. R., Vitolo, J. M., Henricksen, L. A., Bambara, R. A., and Hayes, J. J. (2000) EMBO J. 19, 5492−5501. (32) Linxweiler, W., and Hörz, W. (1984) Nucleic Acids Res. 12, 9395−9413. (33) Polach, K., and Widom, J. (1995) J. Mol. Biol. 254, 130−149.

K.B., M.E.T., and S.D. designed the research. K.B. and M.E.T. performed the experiments. All authors contributed to the writing of the manuscript. K.B. and M.E.T. contributed equally to this work. Funding

This work was supported by Brown University. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors thank Dr. Eric Olmon for conducting preliminary experiments and Paul Caffrey, Erin Kennedy, and Chuxuan Li for careful reading of the manuscript.

■

REFERENCES

(1) Dizdaroglu, M. (2015) Mutat. Res., Rev. Mutat. Res. 763, 212− 245. (2) Schermerhorn, K. M., and Delaney, S. (2014) Acc. Chem. Res. 47, 1238−1246. (3) Wilson, S. H., and Kunkel, T. A. (2000) Nat. Struct. Biol. 7, 176− 178. (4) Odell, I. D., Wallace, S. S., and Pederson, D. S. (2013) J. Cell. Physiol. 228, 258−266. (5) Li, G., and Widom, J. (2004) Nat. Struct. Mol. Biol. 11, 763−769. (6) Li, G., Levitus, M., Bustamante, C., and Widom, J. (2005) Nat. Struct. Mol. Biol. 12, 46−53. (7) Tims, H. S., Gurunathan, K., Levitus, M., and Widom, J. (2011) J. Mol. Biol. 411, 430−448. (8) Tomschik, M., Zheng, H., van Holde, K., Zlatanova, J., and Leuba, S. H. (2005) Proc. Natl. Acad. Sci. U. S. A. 102, 3278−3283. (9) Wei, S., Falk, S. J., Black, B. E., and Lee, T.-H. (2015) Nucleic Acids Res. 43, e111−e111. (10) Prasad, A., Wallace, S. S., and Pederson, D. S. (2007) Mol. Cell. Biol. 27, 8442−8453. (11) Ye, Y., Stahley, M. R., Xu, J., Friedman, J. I., Sun, Y., McKnight, J. N., Gray, J. J., Bowman, G. D., and Stivers, J. T. (2012) Biochemistry 51, 6028−6038. (12) Hinz, J. M., Rodriguez, Y., and Smerdon, M. J. (2010) Proc. Natl. Acad. Sci. U. S. A. 107, 4646−4651. (13) Nilsen, H., Lindahl, T., and Verreault, A. (2002) EMBO J. 21, 5943−5952. (14) Beard, B. C., Wilson, S. H., and Smerdon, M. J. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 7465−7470. (15) Odell, I. D., Barbour, J.-E., Murphy, D. L., Della-Maria, J. A., Sweasy, J. B., Tomkinson, A. E., Wallace, S. S., and Pederson, D. S. (2011) Mol. Cell. Biol. 31, 4623−4632. (16) Odell, I. D., Newick, K., Heintz, N. H., Wallace, S. S., and Pederson, D. S. (2010) DNA Repair 9, 134−143. (17) Rodriguez, Y., and Smerdon, M. J. (2013) J. Biol. Chem. 288, 13863−13875. (18) Bilotti, K., Kennedy, E. E., Li, C., and Delaney, S. (2017) DNA Repair 59, 1−8. (19) Olmon, E. D., and Delaney, S. (2017) ACS Chem. Biol. 12, 692− 701. (20) Menoni, H., Gasparutto, D., Hamiche, A., Cadet, J., Dimitrov, S., Bouvet, P., and Angelov, D. (2007) Mol. Cell. Biol. 27, 5949−5956. (21) Cole, H. A., Tabor-Godwin, J. M., and Hayes, J. J. (2010) J. Biol. Chem. 285, 2876−2885. (22) Lowary, P. T., and Widom, J. (1998) J. Mol. Biol. 276, 19−42. (23) Vasudevan, D., Chua, E. Y. D., and Davey, C. A. (2010) J. Mol. Biol. 403, 1−10. (24) Shaytan, A. K., Xiao, H., Armeev, G. A., Wu, C., Landsman, D., and Panchenko, A. R. (2017) Nucleic Acids Res. 45, 9229−9243. (25) Sassa, A., Beard, W. A., Prasad, R., and Wilson, S. H. (2012) J. Biol. Chem. 287, 36702−36710. D

DOI: 10.1021/acs.biochem.7b01125 Biochemistry XXXX, XXX, XXX−XXX