Key role for newly identified sensor protein in DNA repair
Researchers in the department have identified a sensor protein for a specific type of DNA damage in the cell.
DNA repair in action: DNA ligase encircles the double helix to repair a broken strand of DNA. (Wikimedia) (Click to enlarge)
Martin Cohn's lab, in collaboration with colleagues at Harvard Medical School, describe the findings in a recent paper in Cell Reports (1).
The discovery gives insight into the Fanconi anemia (FA) pathway, which cells use to repair toxic DNA interstrand crosslinks. In the longer term, it could have clinical relevance by helping to identify molecular targets for personalised cancer treatments.
Interstrand crosslinks (ICLs) are one of the many types of DNA lesions that cells must respond to and repair. The FA pathway carries out repair of ICLs via steps that include nucleotide excision repair, translesion synthesis and homologous recombination.
Researchers have identified 17 genes that work together along the FA pathway. This includes genes such as BRCA1 and BRCA2 that are involved in other DNA repair pathways. A defect in one of the 17 genes can give rise to Fanconi anaemia, a recessive cancer predisposition and developmental syndrome.
As Eric Liang, a third-year DPhil student in the Cohn lab who carried out much of the work explains, a protein called FANCD2 is central to the repair. 'FANCD2 is a core protein that is known to be recruited to the damage site, and other proteins follow afterwards. If FANCD2 is absent, then the pathway is completely shut off.'
The proteins that specifically recognise ICLs, however, are unknown. So Dr Cohn, a Royal Society University Research Fellow, and his former DPhil student Jamie Zhan devised an assay that would help to identify this.
They extracted nuclear protein from cells that has been treated with compounds to activate the ICL repair pathways. Then, adding biotin-labelled DNA substrates with either no ICL or an ICL, they pulled down proteins in the extract that interacted with these two structurally distinct DNA molecules.
They found a number of proteins that were more abundant in the sample with DNA containing an ICL than with control DNA. In collaboration with Steven Gygi's lab at Harvard Medical School, the researchers identified the proteins by mass spectroscopy.
One was a protein called UHRF1 that is known to have a role in epigenetics. The group went on to show, using different assays, that the protein binds directly to DNA ICLs and has a 3-4 fold higher affinity for ICL-containing DNA than for undamaged DNA.
To examine the role of UHRF1 in the cellular response to ICLs in vivo, the group used RNAi to knockdown UHRF1 in human cells. They found that UHRF1 knockdown sensitises cells to the crosslinking agent mitomycin C (MMC). The sensitivity was specific to ICL-inducing agents.
Given the possible involvement of UHRF1 in the FA repair pathway suggested by these results, the group wanted to explore the connection in more detail. They therefore used the UHRF1 knockdown to test whether the recruitment of FANCD2 to nuclear foci might be dependent on UHRF1.
Normal and knockdown cells were treated with MMC and stained for FANCD2 by immunofluorescence at various time-points. Whereas damage increased the chromatin-bound FANCD2 in control cells after just a few hours, in the UHRF1 knockdown, the researchers saw a much slower and modest increase in chromatin-bound FANCD2.
The group turned to live-cell imaging at Micron, the department's advanced bioimaging facility, to follow how these proteins were being recruited onto chromatin. In work carried out by postdoc Yasunaga Yoshikawa, they induced crosslinks at specific sites in the DNA using a drug that intercalates in the DNA. When the site is targeted with UVA, the drug triggers an ICL.
HeLa cells expressing mCherry-tagged UHRF1 (top) or EGFP-tagged FANCD2 (bottom) were pre-treated with an intercalating drug and microirradiated at the indicated sites (white arrows). UHRF1 and FANCD2 are recruited to the induced ICLs sites (Click to enlarge)
In normal cells, they observed very fast recruitment of labelled UHRF1 to the sites of the ICLs. This was followed by recruitment of labelled FANCD2 to same site. Recruitment was specific to the intercalating drug as neither protein was recruited in its absence. When they repeated the experiment in the UHRF1 knockdown, they saw that recruitment of FANCD2 was abolished.
Having provided compelling evidence that UHRF1 is directly recruited to ICLs, and that this is required for subsequent recruitment of FANCD2, the group is ploughing ahead with more detailed mechanistic studies.
'Now that we've identified UHRF1 as a sensor that specifically recognises DNA crosslinks and activates the FA pathway, our next steps are to look at how it is involved in the pathway and exploring its biological and mechanical properties,' says Eric Liang.
Schematic showing the proposed sequence of events leading to repair of ICLs in the genome. The newly identified protein UHRF1 plays a key role (Click to enlarge)
Dr Cohn adds that an additional question is the role of UHRF1 in repair of ICLs during different phases of the cell cycle. Repair of ICLs is crucial during S phase as otherwise these lesions would cause blockage of replication, but the repair process can happen in both a replication-dependent and replication-independent process. The group has found that the UHRF1 protein is recruited to ICLs in both S-phase and G1, but they do not know if its recruitment in both phases is functionally important - something they are currently working on.
With this new role of UHRF1 adding to its previously known role in ensuring maintenance of methylated CpG sequences, the researchers are excited to have uncovered an unexpected connection with a DNA repair pathway. 'Now we have brought UHRF1 into its proper context and found its functional significance,' says Eric Liang.
- Liang C-C, Zhan B, Yoshikawa Y, Haas W, Gygi SP and Cohn MA (2015). UHRF1 is a sensor for DNA interstrand crosslinks and recruits FANCD2 to initiate the Fanconi Anemia pathway. Cell Reports. doi: http://dx.doi.org/10.1016/jcelrep.2015.02.053