NOTE: Received NCE to 9/30/2010, per J. Dardano/JSC (7/2009)

NOTE: End date changed to 9/30/2009, per K. Willison (3/07)

NOTE: Question re end date, per JSC; changed back to 9/30/2006 (1/07)

NOTE: project extended for full length of proposal, per J. Dardano (2/06)

To test the contribution of homologous recombinational repair (HR) in repairing DNA damaged sites induced by high-energy Fe ions, we used: 1) HR-deficient rodent cells carrying a deletion in the RAD51D gene, 2) syngeneic human cells impaired for HR by RAD51D or RAD51 knockdown using RNA interference. We show that in response to Fe ions, HR is essential for cell survival in rodent cells, and that HR-deficiency abrogates RAD51 foci formation. Complementation of the HR defect by human RAD51D rescues both enhanced cytotoxicity and RAD51 foci formation. For human cells irradiated with Fe ions, cell survival is decreased, and, in p53 mutant cells, the levels of mutagenesis are increased when HR is impaired. Human cells synchronized in S phase exhibit more pronounced resistance to Fe ions as compared with cells in G1 phase, and this increase in radioresistance is diminished by RAD51 knockdown. These results implicate a role for RAD51-mediated DNA repair (i.e. HR by strand invasion) in removing a fraction of clustered DNA lesions induced by charged particle irradiation. Our results are the first to directly show the requirement for an intact HR pathway in human cells in ensuring DNA repair and cell survival in response to high-energy high LET radiation.

To test the contribution of homologous recombinational repair (HR) in repairing DNA damaged sites induced by high-energy Fe ions, we used: 1) HR-deficient rodent cells carrying a deletion in both alleles of the RAD51D gene, and 2) syngeneic human cells impaired in HR by RAD51D or RAD51 knockdown using RNA interference. We show that, in response to Fe ions, HR is essential for cell survival in rodent cells, and that HR (i.e. RAD51D)-deficiency abrogates RAD51 focus formation in response to Fe ions. Complementation of the HR defect by human RAD51D in rad51d-deficient CHO cells rescues both enhanced cytotoxicity and loss of RAD51 focus formation. In these experiments we chose a RAD51D-complemented CHO clone whose protein expression level for human RAD51D was very similar to that of endogenous hamster RAD51D expressed in AA8 wild-type CHO cells. Our results showing full complementation for cell survival in response to graded dosed of X-rays also support a role for human RAD51D in the response to low LET ionizing radiation. We obtained RBE values for Fe ions of 2.4 and 2.0 for wild-type and rad51d-deficient cells, respectively. From our data we infer that HR is active in rodent cells after exposure to 1 GeV/n Fe ions and that proper function of this high-fidelity DNA DSB repair pathway is required to fully protect CHO cells from the cytotoxic effects of Fe ions.

For human cells irradiated with Fe ions, cell survival is decreased, and, in p53 mutant cells, the levels of mutagenesis are increased when HR is impaired. The effects of RAD51D protein knockdown on cell survival are more pronounced in p53 wild-type TK6 cells than in p53 mutant WTK1 cells. The reason for this difference is unclear at this point, but it is possible that mutagenic DNA DSB repair pathways (i.e. NHEJ, single-strand annealing (SSA) or microhomology-mediated end-joining (MMEJ)) may substitute at higher levels in p53 mutant cells, rescuing cytotoxicity in the absence of fully functional HR. Our results also suggest that, after exposure to 1 GeV/n Fe ions, p53 wild-type human cells rely more heavily on faithful HR and do not tolerate unfaithful DSB repair mechanisms in substitution for HR. Our results for high and low LET radiation-induced mutagenesis are in support of this, because very different effects of RAD51D knockdown on Fe ion-induced mutagenesis are observed in TK6 and WTK1 cells. When we reduced the expression of RAD51D protein by RNA interference, TK mutant fractions were increased significantly in Fe ion- and X-irradiated WTK1 cells which express mutant p53. In these cells, approximately twice as many TK mutants were recovered from RAD51D-depleted cells than from control-transfected cells at all Fe ion and X-ray doses investigated. These results suggest that in the absence of faithful HR, unfaithful DNA repair pathways take over, leading to elevated levels of mutagenesis in permissive (i.e. p53 mutant) human cells. Conversely, no such increase in mutagenesis was observed in HR-compromised p53 wild-type cells, suggesting that precise HR is tightly regulated in TK6 cells, does not lead to the induction of TK mutations, and cannot be substituted for by error-prone DNA damage repair pathways. However, downregulation of the NHEJ gene XRCC4 in p53 wild-type cells limited mutagenesis after both Fe ion- and X-irradiation, demonstrating that NHEJ is a mutagenic DNA DSB repair pathway that is active in p53 wild-type cells after exposure to graded does of Fe ions and X-rays. Unexpectedly, this was not observed for XRCC4-depleted WTK1 cells, for which the levels of Fe ion- and X-ray-induced TK mutant fractions remained at those of control-transfected cells. These results indicate that NHEJ does not contribute to mutagenesis in WTK1 cells under wild-type repair conditions. These findings also suggest that error-prone homology-mediated events such as SSA, MMEJ or PARP-dependent alternative end-joining may account for radiation-induced mutagenesis in these p53 mutant cells.

Furthermore, we find that human U2OS cells synchronized in S phase exhibit more pronounced resistance to Fe ions as compared to cells irradiated in G1 phase, and that this increase in radioresistance is diminished by RAD51 knockdown. To the best of our knowledge, our results for U2OS cells are the first to show S-phase-dependent radioresistance for human cells after exposure to high LET Fe ions. To our surprise, the difference in radioresistance between U2OS cells irradiated in G1 or S phase was almost as pronounced after exposure to Fe ions as after X-irradiation (sensitivity factors (D10) are 1.6 and 1.8, respectively), suggesting that S-phase-specific DNA repair pathways (i.e. HR and SSA) are as important after Fe ion exposure as after X-irradiation. Notably, whereas knockdown of RAD51 in S-phase U2OS cells reduced X-ray resistance to the levels of G1 phase cells, a significantly smaller decrease in radioresistance was observed for Fe ion-exposed RAD51-depleted cells in S-phase. We propose that SSA, a mutagenic sub-pathway of HR, that is restricted to S/G2 cells and that requires DSB end resection and is independent of RAD51, operates at higher levels after high LET radiation than after low LET radiation. It is also possible, that in Fe ion-irradiated S-phase cells a higher fraction of the induced DSBs are repaired by NHEJ. Compared to X-rays, HR in S-phase would than be less important after exposure to 1 GeV/n Fe ions. In summary, proper HR is required to fully protect human and rodent cells from the cytotoxic effects of Fe ions. Furthermore, in permissive human cells (i.e. cells with a gain-of-function mutant p53), a reduced level of HR leads to a significant increase in Fe ion-induced mutant fractions. No such increase in mutagenesis is observed in HR-compromised p53 wild type human cells. While HR is important for some lesion repair (i.e. helps limit cytotoxicity after Fe ions), recombinational repair events in p53 wild-type cells appear to be highly controlled, are mostly faithful, and cannot easily be substituted for by other DSB repair pathways. In addition to HR, our results support a role for other S-phase-specific DSB repair pathways after high LET radiation. These pathways are currently the subject of intense investigation in rodent cells, and should also be investigated in human cells to better understand the cancer risks for astronauts after exposure to space radiation.

NOTE: Received NCE to 9/30/2010, per J. Dardano/JSC (7/2009)

NOTE: End date changed to 9/30/2009, per K. Willison (3/07)

NOTE: Question re end date, per JSC; changed back to 9/30/2006 (1/07)

NOTE: project extended for full length of proposal, per J. Dardano (2/06)

We have shown that CHO cells deleted for the RAD51D gene (i.e. deficient in HR) show increased sensitivity to the cytotoxic effects of iron ions compared to wild type cells, and that ectoptic expression of human RAD51D in rad51d-deleted cells reverts their sensitivity to iron ions to close to wild type levels. In these experiments, we have chosen a RAD51D-complemented CHO clone whose protein expression level for human RAD51D is very similar to that of endogenous hamster RAD51D expressed in AA8 wild type CHO cells, as demonstrated previously (1). Our results showing full complementation for cell survival in response to graded doses of X-rays also support a role for human RAD51D in response to ionizing radiation. We obtain RBE values for iron ions (D10) of 2.4 and 2 for wild type and rad51d-deficient cells, respectively, that are virtually identical to the RBE values reported by Wang and collaborators (2) for the same high LET radiation type and AA8 and irs1SF cells. Irs1SF cells also are deficient in HR due to inactivation of the XRCC3 gene (3). Although NHEJ-deficient rodent cells were not included in our study, they were part of the investigation by Wang and collaborators (2), who obtained RBE values (D10) ~1 for ku80-deficient mouse embryonic fibroblasts (MEFs) exposed to 1 GeV/n iron ions, supporting their additional results for the specific inhibition of the NHEJ pathway after high LET radiation in rodent cells. From our data we infer that HR is active in rodent cells after exposure to 1 GeV/n iron ions, and that proper function of this high-fidelity DNA repair pathway is required to fully protect CHO cells against the cytotoxic effects of iron ions.

RAD51 is the key protein in HR and forms a filament on single-stranded DNA, a filament that is essential for homology search and strand invasion. Ongoing HR can be detected as RAD51 focus formation by immunostaining and fluorescence microscopy, and rodent and human cells impaired in HR due to loss of a RAD51 paralog are impaired in RAD51 foci formation after low LET IR (4, 5). As reported here, in response to high LET iron ions rad51d-deficient CHO cells also were unable to form RAD51 foci, but ectopic expression of human RAD51D in rad51d-/- cells reverted RAD51 foci formation back to wild type levels, indicative for 1) the causal relationship between cellular sensitivity and ongoing HR, and 2) the full functionality of the heterologous human RAD51D protein in CHO cells deleted for the hamster RAD51D gene. In near-normal human cells exposed to iron ions, RAD51 foci formation is restricted to S/G2 phase and absent from G1 phase cells, further pointing to the biological relevance of the HR pathway after high LET radiation. Since RAD51 directly binds to single-stranded DNA, it also accumulates into micro-foci after high LET radiation, as demonstrated earlier after laser micro-irradiation (6).

Human lymphoblast cells depleted for RAD51D also show enhanced sensitivity to iron ions. However, the effects of RAD51D protein knockdown on cell survival are more pronounced in p53 wild type TK6 cells than in p53 mutant WTK1 cells. The reason for this difference it unclear at this point, but it could be speculated that mutagenic DNA repair pathways (i.e. NHEJ, single-strand annealing (SSA) or microhomology-mediated end-joining (MMEJ)) may substitute at higher levels in p53 mutant cells, rescuing cytotoxicity in the absence of properly functional HR. Conversely, since we observe that RAD51D knockdown is more deleterious for iron ion-induced cytotoxicity in TK6 cells, these p53 wild type cells may more heavily rely on faithful HR, and may not tolerate unfaithful DNA repair events in substitution for HR. Our results obtained for high and low LET radiation-induced mutagenesis are in support of this speculation, as very different effects of RAD51D knockdown on iron ion-induced mutagenesis are observed when TK6 and WTK1 cells are assessed (see below). Although the RBE values (D10) obtained for cell death of control and DNA repair protein-depleted cell lines (both TK6 and WTK1 cells) ranged from ~1-1.5 only, it is interesting to note that for both TK6 cells and WTK1 cells the smallest RBE values (i.e. ~1) were obtained for their respective XRCC4-depleted cell populations. This observation is in accord with the recently published report on repair-deficient rodent cells, demonstrating smallest RBEs for NHEJ-deficiency due to the specific inhibition of this DNA repair pathway by 1 GeV/n iron ions (2).

The inhibition of the HR pathway via knockdown of RAD51D has different effects on the outcome of mutagenesis at the thymidine kinase (TK) locus in Fe ion-exposed TK6 and WTK1 cells. Notably, when expression of RAD51D is reduced, TK mutant fractions are enhanced significantly in iron ion- and X-irradiated WTK1 cells. In these cells approximately twice as many TK mutants were recovered from RAD51D-depleted cells than from control-transfected cells at all Fe ion and X-ray doses investigated. These results suggest that in the absence of faithful HR, unfaithful DNA repair pathways take over, leading to elevated levels of mutagenesis in permissive (i.e. p53 mutant) human cells. Conversely, no such increase in mutagenesis was observed in HR-compromised p53 wild type cells, suggesting that non-mutagenic HR is tightly regulated in TK6 cells, and cannot be substituted for by error-prone DNA damage repair pathways. However, down-regulation of NHEJ (i.e. XRCC4) in p53 wild type cells limited mutagenesis after both Fe ions and X-rays, demonstrating that NHEJ is a mutagenic DNA repair pathway active in TK6 cells after both high and low LET radiation. Unexpectedly, this was not observed for XRCC4-depleted WTK1 cells, for which the levels of iron ion- and X-ray-induced TK mutant fractions remained at those of control-transfected cells. These results indicate that NHEJ does not contribute to mutagenesis in WTK1 cells under wild type repair conditions. This finding suggests that error-prone homology-mediated events, such as SSA, MMEJ (for review see (7)), or PARP-dependent alternative end-joining (8, 9) may account for radiation-induced mutagenesis in these p53 mutant cells. It is unlikely that the levels of XRCC4-depletion were insufficient to uncover a phenotype for reduced radiation-induced mutagenesis in WTK1 cells under conditions when NHEJ was impaired, since the same cells demonstrated reduced clonogenic potential after irradiation.

HR is restricted to late S and G2 phase of the cell cycle when the sister chromatid is present, and HR-defective cells show S phase-dependent radiosensitivity (for review see (10)). In repair-proficient human cells, S phase-dependent radioresistance to high-energy Fe ions had not been investigated. To the best of our knowledge, our results for a derivative of U2OS cells are the first to show S phase-dependent radioresistance for human cells in response to high LET iron-ion irradiation, and our findings are in accord with findings from Blakeley and co-workers (11) studying synchronized human T-1 cells and 425 MeV/n neon ions. To our surprise, the difference in radioresistance between U2OS cells irradiated in G1 or S phase was almost as pronounced after iron ions as after X-rays (sensitivity factors (D10) were 1.6 and 1.8, respectively), suggesting that S phase-specific DNA repair pathways (i.e. HR and SSA) are as important after iron ion exposure as they are after X-rays. Notably, whereas knockdown of RAD51 in S phase U2OS cells reduced X-ray radioresistance to the levels of G1 phase cells, significantly less decrease in radioresistance was observed for iron-ion exposed RAD51-depleted cells in S phase. As recently discovered in rodent cells (12), we propose that SSA, a mutagenic sub-pathway of HR that is restricted to S/G2 cells, and that requires DSB end resection but is RAD51 independent, operates at higher levels after high LET radiation damage than after low LET exposure.

In summary, proper HR is required to fully protect both human and rodent cells from the cytotoxic effects of iron ions. Furthermore, in permissive human cells (i.e. cells with gain-of-function mutant p53) reduced level of HR leads to a significant increase in iron ion-induced mutant fractions. No such increase in mutagenesis is observed in HR-compromised p53 wild type cells. While HR is important for some lesion repair (i.e. helps limit cytotoxicity), recombinational repair events in p53 wild type cells are highly controlled, they are mostly faithful, and cannot easily be substituted for by other DSB repair pathways. In addition to HR, our results support a role for other S phase-specific DNA repair pathways after high LET radiation. These pathways are currently the subject of intense investigation in rodent cells (12), and, in the future, should be investigated in human cells to better understand the astronauts’ risk for cancer from space radiation.

References:

1. C. Wiese, J. M. Hinz, R. S. Tebbs, P. B. Nham, S. S. Urbin, D. W. Collins, L. H. Thompson and D. Schild, Disparate requirements for the Walker A and B ATPase motifs of human RAD51D in homologous recombination. Nucleic Acids Research 34, 2833-2843 (2006).

2. H. Wang, X. Wang, P. Zhang and Y. Wang, The Ku-dependent non-homologous end-joining but not other repair pathway is inhibited by high linear energy transfer ionizing radiation. DNA Repair 7, 725-733 (2008).

3. N. Liu, J. E. Lamerdin, R. S. Tebbs, D. Schild, J. D. Tucker, M. R. Shen, K. W. Brookman, M. J. Siciliano, C. A. Walter, et al., XRCC2 and XRCC3, new human Rad51-family members, promote chromosome stability and protect against DNA cross-links and other damages. Molecular Cell 1, 783-793 (1998).

4. J. M. Hinz, R. S. Tebbs, P. F. Wilson, P. B. Nham, E. P. Salazar, H. Nagasawa, S. S. Urbin, J. S. Bedford and L. H. Thompson, Repression of mutagenesis by Rad51D-mediated homologous recombination. Nucleic Acids Research 34, 1358-1368 (2006).

5. C. Wiese, E. Dray, T. Groesser, J. San Filippo, I. Shi, D. W. Collins, M. S. Tsai, G. J. Williams, B. Rydberg, et al., Promotion of homologous recombination and genomic stability by RAD51AP1 via RAD51 recombinase enhancement. Molecular Cell 28, 482-490 (2007).

6. S. Bekker-Jensen, C. Lukas, R. Kitagawa, F. Melander, M. B. Kastan, J. Bartek and J. Lukas, Spatial organization of the mammalian genome surveillance machinery in response to DNA strand breaks. The Journal of Cell Biology 173, 195-206 (2006).

7. M. McVey and S. E. Lee, MMEJ repair of double-strand breaks (director's cut): deleted sequences and alternative endings. Trends Genet 24, 529-538 (2008).

8. S. J. DiBiase, Z. C. Zeng, R. Chen, T. Hyslop, W. J. Curran, Jr. and G. Iliakis, DNA-dependent protein kinase stimulates an independently active, nonhomologous, end-joining apparatus. Cancer Research 60, 1245-1253 (2000).

9. R. Perrault, H. Wang, M. Wang, B. Rosidi and G. Iliakis, Backup pathways of NHEJ are suppressed by DNA-PK. Journal of Cellular Biochemistry 92, 781-794 (2004).

10. P. Tamulevicius, M. Wang and G. Iliakis, Homology-directed repair is required for the development of radioresistance during S phase: interplay between double-strand break repair and checkpoint response. Radiation Research 167, 1-11 (2007).

11. E. A. Blakely, P. Y. Chang and L. Lommel, Cell-cycle-dependent recovery from heavy-ion damage in G1-phase cells. Radiat Res Suppl 8, S145-157 (1985).

12. M. Frankenberg-Schwager, A. Gebauer, C. Koppe, H. Wolf, E. Pralle and D. Frankenberg, Single-strand annealing, conservative homologous recombination, nonhomologous DNA end joining, and the cell cycle-dependent repair of DNA double-strand breaks induced by sparsely or densely ionizing radiation. Radiation Research 171, 265-273 (2009).

NOTE: End date changed to 9/30/2009, per K. Willison (3/07)

NOTE: Question re end date, per JSC; changed back to 9/30/2006 (1/07)

NOTE: project extended for full length of proposal, per J. Dardano (2/06)

In the last year we have conducted two experimental runs at BNL-NSRL using the 1 GeV/n Fe beam to test whether HR contributes to the repair of complex DNA DSBs induced by Fe ions. For comparison purposes, we have also conducted several experiments using X-rays. We have further refined our experimental approach in which we use RNAi transfection or conditional expression of shRNA to induce gene-specific knockdown of essential proteins functioning in HR. Furthermore, we now have included one NHEJ protein, XRCC4, a stimulator of DNA ligase IV, into our mutation analyses to better understand the complexity of our results under conditions when HR is compromised. We would like to point out that we have conducted our mutation experiments using comparatively low radiation doses (i.e. from 0.3 to 1.2 Gy) for both X-rays and Fe ions.

In summary, our results using both human and hamster cell lines show that, after exposure to Fe ions, HR deficiency mildly but reproducibly decreases cellular survival and abrogates RAD51 foci formation. In addition, RAD51 foci formation in response to Fe ions is cell cycle regulated and occurs in S/G2 phase cells to facilitate strand exchange with the sister chromatid, but not in G1 cells. In rad51d-knockout hamster cells, the inability to recruit RAD51 into RAD51 foci after exposure to Fe ions and the increased cytotoxicity after Fe ion exposure are very likely to be intertwined, although this has not directly been investigated here. Human fibroblasts and epithelial cells depleted for RAD51D are impaired in RAD51 foci formation both after X-rays and after Fe ions.

Our data on Fe ion-induced mutagenesis in HR-impaired human cells are complex and show that functional HR is a prerequisite for ensuring genome maintenance and proper repair of DNA damage induced by 1 GeV/n Fe ions in “permissive” human lymphoblastoid cells (i.e. p53 mutant WTK1 cells). Furthermore, our results clearly demonstrate that the susceptibility to mutagenesis is dependent on both the locus investigated (i.e. HPRT vs. TK) and the genetic background (i.e. wild type vs. defective HR). Compared to control transfected cells, we recover more TK mutants from HR defective WTK1 cells after exposure to Fe ions, suggesting that at this locus DNA repair fidelity is reduced significantly when decreased levels of HR proteins are expressed. We conclude that ability to properly perform HR after exposure to Fe ions is essential for limiting mutagenesis at an autosomal locus. Interestingly, simultaneous depletion of the NHEJ protein XRCC4 reduces TK mutation levels significantly when HR is impaired. Our results show that ablation of HR leads to an increase in NHEJ-driven mutagenesis at TK. Furthermore, in WTK1 cells XRCC4-dependent NHEJ also contributes to Fe ion-induced mutagenesis at the hemizygous HPRT locus, but defects in HR do not affect HPRT mutation levels after Fe ions at this hemizyous locus, indicating either that the contribution of HR-mediated mechanisms to mutagenesis at HPRT is minor, or that, under conditions of impaired HR, HPRT mutants derived from NHEJ events cannot be recovered. In wild type p53 TK6 cells, NHEJ contributes to mutagenesis at TK both after Fe ions and after X-rays, but alterations in the ability to perform HR apparently have no effect, suggesting that HR does not contribute to mutagenesis at this heterozygous locus (as expected) and may only play a minor role in repairing radiation-induced DSBs within TK or within the genomic region flanking TK. We consider the possibility that p53 wild type cells impaired in HR by depletion of RAD51D are unable to process endogenous replication damage and exogenous DNA damage due to Fe ion or X-ray exposure, and may escape the mutation analysis due to increased cell death. Taken together, ablation of HR greatly affects Fe ion-induced TK mutagenesis in p53 mutant but not in p53 wild type human lymphoblastoid cell lines. Since the inactivating p53 mutation in WTK1 cells is in the sequence-specific DNA binding domain and since the same mutation is observed in some human tumors, the phenotype of WTK1 cells is somewhat typical of many p53 mutant cells. It has been estimated that a certain fraction of cells within a healthy individual’s body spontaneously acquire p53 mutations that then allow oncogenic transformation to occur. For this, mutation analyses in genetically predisposed cell lines can provide a useful means to derive radiation risk estimates for carcinogenesis in humans traveling to space, since these humans presumably already contain precancerous cells with p53 mutations.

NOTE: Question re end date, per JSC; changed back to 9/30/2006 (1/07)

NOTE: project extended for full length of proposal, per J. Dardano (2/06)

Homologous recombination (HR) is conserved in all organisms and is a prerequisite for both DNA repair and the resumption of stalled replication forks. As replication fork stalling occurs during most cycles of replication, proteins required for HRR generally are essential and cannot easily be deleted in human cells. Therefore, RNA interference is the method of choice for attenuating HRR in near-normal human cells.

Importantly, HRR is a DNA repair pathway with close ties to cancer biology and crucial for maintaining genomic stability, limiting mutagenesis and preventing carcinogenesis. It is well established that conditions promoting reduced levels of HRR compromise the fidelity of DSB repair and correlate with an elevated cancer risk. The risk of developing cancer is increased in individuals exposed to space radiation, and defects in HRR, leading to the increased utilization of error-prone DNA repair pathways, are likely to contribute to this process. Therefore, it is a necessity to establish the relevance of HRR to HZE radiation, both for better prediction of the astronaut’s sensitivity to radiation carcinogenesis and for proper radiation risk assessment.

Using syngeneic human cells, this Research Project is assessing whether defects in HRR affect the extent of cell killing (Aim 1) and the levels and mechanisms of mutagenesis (Aim 2) by densely ionizing Fe ions. RNA interference is used to impair HRR (targeting proteins essential for HRR: XRCC3 and RAD51AP1) in two human lymphoblastoid cell lines of different p53 status (Aims 1&2). The Fe particle-induced mutation frequencies were determined at the autosomal TK locus and at the X-linked HPRT locus for one representative derivative of each cell line with a ‘loss-of-function’ phenotype for HRR (Aim 2), and these mutation frequencies are compared to control cells transfected with a non-depleting hairpin. In future NSRL runs, sets of TK mutants will be collected and the Fe ion-induced, X-ray-induced and spontaneous mutation spectra will be assessed to discriminate between recombinational and deletional events (Aim 2). As part of Aim 3 synchronized hTERT-immortalized normal human dermal fibroblasts (NHDF) were tested for their ability to localize RAD51 into RAD51 foci in response to Fe ions. HRR-defective Chinese Hamster Ovary (CHO) cells have also been tested for their response to 1 GeV/n Fe ions, investigating both RAD51 foci formation and cell survival.

We have conducted two experimental runs at NSRL using 1 GeV/n Fe ions during the last year. We have found that human lymphoblastoid cells depleted for either XRCC3 or RAD51AP1, both proteins essential for HRR, are mildly sensitized (~1.2- to 1.4-fold) to the cytotoxic effects of 1 GeV/n Fe ions. This sensitization is more pronounced for TK6 cells expressing wild type p53 than for WTK1 cells expressing mutant p53. Furthermore, we have determined the levels of spontaneous and radiation-induced mutagenesis in human lymphoblastoid cell lines with silenced XRCC3. As expected, more spontaneous TK and HPRT mutants were detected in TK6 and WTK1 cells expressing low levels of XRCC3, implicating that under conditions that limit DSB repair by HRR, error-prone DSB repair mechanisms (i.e. NHEJ) may elevate spontaneous mutagenesis. Following exposure to 1.2 Gy of 1 GeV/n Fe ions, significantly more HPRT-deficient mutants were recovered from cells with silenced XRCC3 compared to control-depleted cells, suggesting that HRR-mediated repair events or wild-type levels of XRCC3 protein limit Fe ion-induced mutagenesis at HPRT. However, the levels of Fe ion-induced TK mutants did not vary significantly between XRCC3-deficient and control cells, suggesting that the contribution of recombinational events to mutagenesis at TK may be minor or that, due to the impaired levels of HRR in XRCC3-depleted cells, viable TK mutants arising from NHEJ events can not be recovered. Our findings require replication in future NSRL runs.

Unlike for human cells, HRR-knockouts of CHO cells are viable and were tested by us for their response to 1 GeV/n Fe ions. Using colony formation assays, we have found that, compared to CHO wild type cells, RAD51D-deficient CHO cells show enhanced sensitivity (~ 1.5-fold) to the cytotoxic effects of 1 GeV/n Fe ions. Expression of human RAD51D in the RAD51D-knockout reverted cell survival back to wild type levels, supporting the notion that HRR contributes to cell survival after exposure to Fe ions. In wild type, RAD51D-deficent and complemented CHO cell lines we have investigated RAD51 and 53BP1 foci formation in response to 1 Gy of 1 GeV/n Fe ions. In CHO wild type and RAD51D-complemented RAD51D-knockout cells, both 53BP1, a DSB sensor and checkpoint protein, and RAD51, the HRR strand transferase, form microscopically discernible accumulations as early as 20 min after exposure to 1 Gy of 1 GeV/n Fe ions. In HRR-defective RAD51D-knockout cells 53BP1 foci form normally, but RAD51 foci formation is abrogated after exposure to Fe ions, and this defect in localization of RAD51 into focal accumulations is likely to contribute to the enhanced levels of Fe ion-induced cytotoxicity observed for RAD51D-knockout cells. As mentioned above, ectopic expression of human RAD51D in RAD51D knockout cells restores RAD51 foci formation in response to Fe ions.

HR largely relies on the presence of a sister chromatid that serves to provide the template for lesion repair. As a consequence, HRR predominantly occurs in late S- and G2-phases when the sister chromatid is present. We have used hTERT-immortalized human fibroblasts synchronized in G1- or S-phase and investigated whether the formation of RAD51 foci after exposure to Fe ions is cell cycle regulated. We have found that hTERT-immortalized human fibroblasts in G1 are able to form both 53BP1 and g-H2AX foci, but unable to form RAD51 foci in response to Fe ion irradiation. However, when synchronized in S-phase, hTERT-immortalized human fibroblasts show RAD51, 53BP1 and g-H2AX foci after exposure to Fe ions, all of which co-localize along the ion trajectory where the radiation damage induced is expected to be most severe. Our results show that the formation of RAD51 foci in response to Fe ions is cell cycle regulated.

During the course of the last year we have collected evidence showing that RAD51, the strand transferase that is essential for cell viability and the key player in HRR, readily is recruited to DNA lesions induced by 1 GeV/n Fe ions in a cell cycle dependent manner. Compared to X-rays, RAD51 foci following exposure to Fe ions are observed at lower doses and at earlier times. In addition, our cell survival studies and mutation analyses point to the importance of a fully functional HRR pathway in the repair of radiation damage induced by 1 GeV/n Fe ions.

NOTE: Question re end date, per JSC; changed back to 9/30/2006 (1/07)

NOTE: project extended for full length of proposal, per J. Dardano (2/06)

Homologous recombination (HR) is conserved in all organisms and a prerequisite for both DNA repair and the resumption of stalled replication forks. As replication fork stalling occurs during most cycles of replication, proteins required for HRR generally are essential and cannot easily be deleted in human cells. Therefore, RNA interference is the method of choice for attenuating HRR in near-normal human cells.

Importantly, HRR is a DNA repair pathway with close ties to cancer biology and crucial for maintaining genomic stability, limiting mutagenesis and preventing carcinogenesis. It is well established that conditions promoting reduced levels of HRR compromise the fidelity of DSB repair and correlate with an elevated cancer risk. The risk of developing cancer is increased in individuals exposed to space radiation, and defects in HRR, leading to the stimulation of error-prone DNA repair pathways, are likely to contribute to this process. Therefore, it is a necessity to establish the relevance of HRR to HZE radiation, both for better prediction of the astronaut’s sensitivity to radiation carcinogenesis and for proper radiation risk assessment. Using syngeneic human cells, this Research Project is assessing whether defects in HRR affect the extent of cell killing (Aim 1) and the levels and mechanisms of mutagenesis (Aim 2) by densely ionizing Fe ions. RNA interference is used to impair HRR (targeting proteins essential for HRR: XRCC3, RAD51, RAD51D, RAD51AP1) in the human lymphoblastoid cell line TK6 (Aims 1&2). TK6 cells were not part of the original grant application due to their inherently lower ability to perform HRR. However, the reviewers strongly recommended also including a wild type p53-expressing cell appropriate for mutation studies (i.e. TK6 cells) into this Research Project in order to be able to obtain results relevant to the risk estimate for normal human tissues. The Fe particle-induced mutation frequencies will be determined at the autosomal TK1 locus and at the X-linked HPRT locus for one representative derivative of TK6 cells with a ‘loss-of-function’ phenotype for HRR (Aim 2) and these mutation frequencies will be compared to parental TK6 cells studied in our earlier investigation. Sets of TK1 mutants will be collected and the Fe ion-induced, X-ray-induced and spontaneous mutation spectra will be compared to discriminate between recombinational and deletional events (Aim 2).

As previously mentioned and in response to the reviewers’ critiques, we have decided to redirect one major part of the proposed research within this Research Project to be able to employ the wild type p53 human TK6 cell line instead of its related WTK1 cell line that expresses mutant p53. We now have proceeded to 1) develop retroviral vectors highly competent for the expression of shRNAs targeting HRR genes in near-normal (i.e. TK6 cells and hTERT-immortalized human cells) and finite lifespan (i.e. normal) human cells, and 2) accumulate data using TK6 cells depleted for several HRR proteins individually using the plasmid-based system for the expression of shRNAs. Since the retroviral vectors are reported to yield higher transduction efficiencies and therefore more extensive knockdown of the proteins of interest, we are anticipating only using the retroviral system in future experiments. This approach also allows us to make progress independent of the proposed but difficult generation of the XRCC3-/- knockout from hTERT-immortalized human fibroblasts (currently carried out by our collaborator Dr. L.H. Thompson, LLNL).

Among the several endpoints that we are planning to investigate, we have concentrated on cell cytotoxicity and colony formation assays to date. We have successfully developed an approach to deplete essential HRR proteins in the human lymphoblastoid TK6 cell line that expresses wild type p53. Moreover, we have assessed these cells for their ability to carry out homologous recombinational repair (i.e. gene conversion). Using the plasmid-based system for expression of shRNAs we now are able to significantly reduce the levels of XRCC3, RAD51 or RAD51AP1 (three different proteins required for HRR) in p53 wild type TK6 cells. In TK6 cells, depletion of either XRCC3 or RAD51 results in suppression of HRR (3- to 4-fold; determined by monitoring the frequency of gene conversion at a site-specific DSB within an integrated recombinational reporter construct). These findings are directly relevant to our Research Project since the demonstration that the expression of shRNAs targeting either XRCC3 or RAD51 indeed confers a defect in HRR is a prerequisite for the proposed study. We have assessed the radiosensitivity of TK6 cells depleted for XRCC3 or RAD51AP1 by exposing these cells to increasing doses of sparsely ionizing radiation and determining the fractions of cells surviving. For control purposes, we have used TK6 cells transfected with the plasmid encoding the mutated hairpin, expression of which does not result in protein depletion. As expected and compared to control cells (transfected with plasmid encoding a non-depleting hairpin), asynchronous cultures of TK6 cells depleted for either XRCC3 or RAD51AP1 are not sensitized to the cytotoxic effects of low LET ionizing radiation. Studies are underway to also perform similar experiments with RAD51-depleted TK6 cells, but more importantly, in the attempt to synchronize TK6 cells, we have initiated experiments to perform centrifugal elutriation with XRCC3-depleted TK6 cells. In preliminary studies, but using HeLa cells, we have synchronized RAD51AP1-depleted cells and found that RAD51AP1-depleted S-phase cells are more sensitive to X-rays than are control cells.

We have generated lentiviral vectors that encode shRNAs to deplete XRCC3, RAD51 or RAD51AP1 for our analyses in near-normal (i.e. hTERT-immortalized) and finite lifespan (i.e. normal) human cells. These lentiviral vectors have considerable advantages over normal plasmid expression vectors, in that they allow for higher efficiency of transduction, as compared to plasmid transfection. In addition, the lentiviral constructs are very efficiently integrated into the genome, enabling studies in normal human cells directed towards obtaining results relevant to the radiation risk estimate for normal human tissues. Although our progress on the original three aims is slower than anticipated, this is largely due to the suggested (i.e. by the reviewers) and acknowledged (i.e. by us) necessity to perform our Research Project in p53 wild type human cells. We have presented evidence that we are now fully capable of performing the proposed study in TK6 cells, using the plasmid-based shRNA expression system. Moreover, we have spent considerable time and effort in generating retroviral shRNA-expressing constructs that are likely to work for many different human cell types, making our investigation more practical for human space travel risk assessment.