DmRecQ4 Depleted S2 Cells Show Decreased Proliferation Due to an S Phase Block
S2 cells were treated with dsRNAi against either the N or the C terminus of DmRecQ4, and the levels of the protein that remained analysed using an antibody raised against the overexpressed N terminal region of DmRecQ4. Fig1A shows that the dsRNA treatment caused at least a 90% decrease in the level of DmRecQ4 by day 3. This was maintained until at least day 7.
Cells which had been treated in this way showed markedly decreased proliferation compared to cells which had been mock treated or were untreated (Fig1B).
In addition FACS analysis of the cell cycle stage of these cells showed that by day 3 a significantly increased percentage of the cells were in S phase (Fig1C). At longer time points many cells remained in the S phase but in addition an increasing percentage of the cells showed a subG1 DNA content.
Despite the fact that a large proportion of the DmRecQ4 cells appeared to be in S phase, these cells did not appear to be actively synthesising DNA, since they showed a significant decrease in BrdU incorporation (). These data suggest that in S2 cells DmRecQ4 plays an important role in the progression of the cells through the S phase.
Reduction of DmRecQ4 in S2 cells interferes with cell proliferation and DNA synthesis.
Both of these dsRNAs showed the same phenotype. The same phenotype was also seen using a third dsRNA against another different region of DmRecQ4 (data not shown). This suggests that the effects observed were specific to DmRecQ4, and not caused by off target depletion of a different protein.
To further clarify the source of the subG1 DNA content, an analysis of the amount of apoptosis occurring was carried out using an Annexin v-FITC kit and flow cytometry. This showed that on day 3 approximately 20% of all cells were annexin positive, rising to about 35% on day 4. This compares with 32% for UV treatment, 53.5% for cycloheximide treatment and 1.8% for untreated cells. Surprisingly, treatment of the RecQ4 depleted cells with the apoptosis inhibitors ALLN and ZVAD (data not shown) had no effect on the production of the subG1 content DNA suggesting that the apoptosis occurred via an apoptotic pathway which did not use caspase.
DmRecQ4 Depleted S2 Cells Show Increased Phospho-H2AvD Staining
To observe if any visible changes occurred to the chromatin on DmRecQ4 depletion, cells at day 4 were fixed and stained with DAPI to visualise the DNA, and antibodies against the phospho-H2AvD protein (the Drosophila homologue of H2Ax). Fig2A shows that more phospho-H2AvD staining was seen in DmRecQ4 depleted cells than those that had been treated with a control dsRNA. This increase could also be seen after chromatin isolation and analysis by western blotting (Fig2B/C). This suggests that DmRecQ4 depletion causes DNA damage. Similar observations have been made in S2 cells on depletion of other proteins that are known to play a role in DNA replication 
S2 cells with reduced DmRecQ4 show increases in phospho-H2AvD.
These observations provide further evidence for an important S phase role for DmRecQ4 in S2 cells.
The N and C Terminal Regions are Both Important for Efficient Rescue of DmRecQ4 Depletion
The above data all support a role for DmRecQ4 in DNA replication in S2 cells, Since the N and C termini of the protein contain homology to different proteins, we were interested to determine whether both parts of the protein were required to support the replication function. It had previously been reported that in Xenopus extracts the isolated N terminus (1–596) 
, at four times higher concentration, could compensate efficiently for depletion of full length DmRecQ4 from the extract, while smaller fragments could partly compensate at much higher concentration 
. DmRecQ4 has a 300 aa insert (228- 610) compared to Xenopus RecQ4. To cover a comparable region we therefore made Drosophila cell lines expressing full length DmRecQ4 and the isolated N (1–707) and C regions (708–1579), all of which were tagged with the V5 epitope ().
Expression of full length, N and C terminal regions in S cells.
All of these proteins were expressed in S2 cells ( and Fig S1A), the N and C terminus at comparable levels while the full length was at a slightly lower level. Analysis of the subcellular location of these proteins by fractionation (fig3B) showed that the full length protein was distributed between the cytoplasm and nucleus. The N terminus is very tightly associated with the nucleus, while the C terminus is predominantly in the cytoplasm. This suggests that different independent elements involved in the nuclear transport of these proteins might be located in the N and C termini.
These cell lines were then assessed for their ability to withstand DmRecQ4 depletion. Initially we attempted to use a dsRNA in the upstream untranslated region of the gene to specifically deplete the endogenous protein. Unfortunately for DmRecQ4 this region is extremely short (42 bp), and therefore we were not able to get efficient DmRecQ4 depletion using only this region (data not shown). We therefore decided to look at the rescue conferred by these constructs on DmRecQ4 depletion by the two individual dsRNAs directed towards the N or C terminus of the DmRecQ4 protein. As shown (Fig S2A) the observed depletion of full length, N and C termini was as expected for these dsRNAs. No endogenous protein was visible in any of the cell lines with either the N or C terminal dsRNA. However it was still possible that small and undetectable levels remained which might influence the results. We reasoned that since the expression of the fusion proteins was comparable between the different cell lines, any titration of the added dsRNA should also be comparable. Therefore any difference in cell survival should accurately reflect the ability of the fusion constructs to compensate for the loss of the endogenous DmRecQ4 protein. Further evidence that supports this was obtained by experiments with additional mutants (see below). In addition, in one of our experiments, (see Fig S2A) the C terminal dsRNA did not fully deplete the C terminal fragment, this would therefore also be expected to be less efficient at depleting the endogenous protein. However the results of this experiment were not significantly different from those where complete depletions were observed.
Cell growth curves for wtS2, and FL, N and C expressing cell lines that have been challenged with dsRNA against the DmRecQ4 N and C termini are shown in . These show that while the expression of full length DmRecQ4 is able to compensate for endogenous DmRecQ4 depletion caused by either the N or the C termini, neither the N or C terminus alone is able to fully compensate for the growth defect. Minor improvements in growth can be seen with the isolated N terminus when the cells are challenged with DmRecQ4 dsDNA from the C terminal region, suggesting a possible small degree of compensation.
Regions in the N and C termini of DmRecQ4 are both required for complete rescue of dsRNA reduction of the endogenous protein.
This is more strongly suggested by the analysis of the cell cycle behaviour of these cells by FACS (). Here on depletion of the endogenous DmRecQ4 with dsRNA located in the C terminus the full length construct appears to revert DmRecQ4 dsRNA challenged S2 cells to a near normal cell cycle pattern, the isolated C terminus showed no rescue, while the isolated N terminus gives some small improvement in the cell cycle. The cells lines expressing the full length and isolated C terminus showed the same result using N terminal dsRNA. In this case however no rescue was seen with the cell line expressing the isolated N terminus, most likely because the levels of expressed N terminus remaining after depletion were too low to be effective.
This suggests that both the N and C termini are important for the cellular function of DmRecQ4.
The SLD2 Domain is Required to Rescue the Replication Effects of DmRecQ4 Depletion but the Drosophila Specific Domain is not
DmRecQ4 contains a region of homology to SLD2, it might therefore be expected that this region is of importance for its replication function. The protein also contains a region from aa 228–610 which is conserved in Drosophila species but not in other organisms. Since Drosophila species lack some of the RecQ helicases, it is possible that this region plays a role in allowing Dm RecQ4 to compensate for their loss. We therefore made V5-tagged constructs of DmRecQ4 missing the SLD2 and the Drosophila specific domain (fig5A). We also made a construct covering the region 1–1234 which had been previously reported in Drosophila embryos to be able to rescue the replication defects as measured by BrdU incorporation – although only rescue viability by 10% 
. All of these constructs were well expressed in Drosophila S2 cells (Fig S1A/B), and showed a similar subcellular distribution to that shown by the full length protein (not shown). These proteins also showed depletion patterns by the N and C terminal dsRNAs as would be expected (Fig S2B/C).
shows that full length protein missing the SLD2 domain has lost most of its ability to rescue the proliferation of cells treated with dsRNA against either the N or C terminus of DmRecQ4, suggesting that the SLD2 domain is important for its cellular function. shows that DmRecQ4 missing the SLD2 domain is also unable to significantly rescue the cell cycle profile of cells treated with dsRNA against the N terminus.
DmRecQ4 proteins without the SLD2 domain cannot rescue either cell cycle or proliferation defects while those without the Drosophila specific domain, or containing only the first 1234aa show differential effects on these parameters.
/E shows the comparable data for the removal of the other domains. The removal of the Drosophila specific domain Δ228–610 produced a protein that could significantly (although not completely) rescue the cell cycle phenotype generated by dsRNA treatment. Surprisingly it has a much smaller compensatory effects on viability. Further removal of the SLD2 domain from this protein removed it's ability to compensate for either. The removal of all amino acids after 1234 produced a protein which showed a slight rescue of the cell cycle phenotype while not compensating at all for the proliferation defect. Unexpectedly the removal of the Drosophila specific domain from the 1234 construct produced a protein that could not compensate for either the cell cycle or proliferation defects (not shown).
DmRecQ4 Depleted Cells Show Abnormal Loading of Key Replication Proteins
In vitro, in Xenopus extracts, the loss of RecQ4 causes a decrease in the loading of RPA and DNA polymerase alpha onto chromatin. To determine whether this was true in vivo, we compared chromatin loaded proteins in DmRecQ4 depleted S2 cells with those in cells that had been treated with low levels of HU. The HU treatment ensures that the cells are largely in the S phase of the cell cycle. These cells also show DNA damage as measured by H2AvD labelling, and therefore should serve as a good control to see which proteins are specifically affected by DmRecQ4 loss. As expected both PCNA and DNA polymerase alpha showed a high level of loading in HU treated cells (). In contrast DmRecQ4 depleted cells on day 3 showed a marked decrease in both of these proteins, suggesting that although an S phase block was apparent from FACS analysis this was not comparable to an HU induced block. Using the same samples Orc2 shows little change in any of the samples. This suggests that the binding of Orc2 is not affected by the loss of DmRecQ4. These decreases were not caused by a decrease in the total amount of RPA or DNA polymerase in the cell (Fig S3) but are specific to chromatin bound proteins.
DmRecQ4 depleted cells show abnormal loading of key replication proteins.
We also looked to see whether the reported interaction with the Mcm protein complex reported in mammalian cells could be seen in Drosophila. Immuno-precipitation of V5-DmRecQ4 from cell lines expressing the full length protein, and analysis of co-immunoprecipitating proteins using Mcm5 antibodies failed to show an interaction between these proteins (data not shown).
Repair Functions of DmRecQ4 in S2 Cells
Although various sources have suggested a role for RecQ4 in 3 types of DNA repair (BER, HR and NER), very little is known about the molecular mechanisms of its involvement in these processes. We therefore set out to determine whether Drosophila S2 cells could provide a convenient system in which to assess this.
DmRecQ4 Accumulates in the Nucleus soon after UV Treatment
The agent selected for investigation of the NER pathway was exposure of the cells to UV light. We first titrated the UV light to determine the level that would be applicable. Fig S4 shows the dose response of S2 cells. We subsequently chose to use a range of UV levels between 80 and 320 J/m2, as others have reported that different effects might be seen under different light intensities.
The most straightforward approach to determine whether DmRecQ4 has a role in the NER pathway would be to analyse whether DmRecQ4 depleted cells were more sensitive to UV light than control S2 cells. Unfortunately the DmRecQ4 depleted cells in the absence of UV were already so compromised that it was not possible to detect any meaningful changes on treatment of these cells with UV.
We then looked to see whether the overexpression of DmRecQ4 in addition to the endogenous protein helped the cells to resist UV treatment. Again no significant advantage in terms of cell growth and viability or recovery of cell cycle profile was observed in cells which were overexpressing the full length DmRecQ4 compared to wt S2 cells.
We also repeated the experiment with the cell lines overexpressing the isolated N and C termini of DmRecQ4. Again there was no advantage in withstanding UV damage to cells expressing either of these. In fact with the isolated N terminus at higher UV concentrations we consistently saw decreased viability of the cells with large numbers of cells appearing in the subG1 region of the FACs profile after 24 hrs (not shown).
Finally we looked to see whether we could detect a change in the DmRecQ4 subcellular distribution in response to the treatment of the cells with UV. We chose to look at the V5 tagged protein as this gave good specific staining on immunofluorescence compared to the antibody raised against the endogenous protein. The observation that it could efficiently substitute for endogenous DmRecQ4 on dsRNA interference suggested that the protein was functional. Analysis of cells fixed directly one hour after UV treatment showed no apparent redistribution of the protein on treatment with UV light at any light intensity. However as the unperturbed distribution of the protein was both nuclear and cytoplasmic we reasoned that this might obscure the visualisation of protein movement. For proteins such as the Mcm proteins it is possible to observe specific behaviour of the proteins after the cells have been extracted with detergent. We therefore subjected the UV treated cells to extraction with a detergent based solution prior to fixation and staining. As can be seen (fig7A) extraction of untreated cells with detergent caused a complete loss of DmRecQ4 staining. However in cells that had been UV treated prior to detergent extraction a strongly bound nuclear fraction could now be detected suggesting that the DmRecQ4 was being selectively retained in the nucleus. When cells were stained 5 h after UV treatment nuclear localisation of DmRecQ4 was no longer observed (not shown).
We also used the same methodology to analyse the N and C terminal fragments of DmRecQ4 (). This analysis showed that while the C terminus was also selectively retained in the nucleus after UV treatment, no retention of the N terminus could be detected. It should be noted that the C terminus can also be seen to be detained in the nucleus without detergent extraction (data not shown), most likely due to the fact that it is usually cytoplasmically located.
DmRecQ4 accumulates in the nucleus soon after UV treatment in a manner dependent on the presence of the C terminus.
This rapid and transient re-localisation of DmRecQ4 on UV treatment suggests that it has a role in the cellular response to UV damage. The fact that this can be seen with the isolated C but not N terminus further suggests that the determinants responsible for this relocation are located in the C terminus of the protein.
DmRecQ4 Localisation does not Change after H2O2 and Etoposide Treatment
Etoposide is often used as an agent to study HR since it generates replication dependent ds DNA breaks. For BER a commonly used agent is H2O2. Dose response curves for both of these reagents are shown in figures S5/S6. From these data we decided to use etoposide at 1 uM which left some percentage of the cells capable of proliferation and 10 uM to provide a more lethal dosage. We chose to use H2O2 at 20 mM. Since UV damage caused nuclear retention of RecQ4, it was possible that the same would be true for other forms of damage we therefore chose to use this as our primary method of analysis for etoposide and H2O2 treatment. Immuno-fluorescence analysis of detergent extracted cells expressing the FL and the N, and C termini one hour or five hours after etoposide or H2O2 treatment did not show nuclear retention of any of these proteins. shows this data for the full length DmRecQ4 treated with etoposide after one hour, but similar results were obtained with the full length protein after H2O2 treatment and the isolated N and C termini after either treatment (not shown) at both time points.
DmRecQ4 localisation does not change after H2O2 or etoposide treatment.
One of the main pieces of evidence for the involvement of RecQ4 in each of these pathways is its interaction with proteins specific to that pathway – XPA for NER, Rad51 for HR, FEN and APE1 for BER. We therefore reasoned that different pathways might be involved in its localisation for each of these pathways. It was therefore possible that nuclear retention in each case might be differentially sensitive to detergent treatment. We therefore repeated the immuno-fluorescence in the absence of prior detergent extraction with the C terminal fragment. For this construct nuclear retention can be seen after UV treatment when the cells are untreated prior to fixation, however neither etoposide or H2O2 produced a similar localisation.