Non-small cell lung cancer (NSCLC) accounts for the largest subgroup of lung cancers, resulting in the highest portion of cancer-related mortality worldwide [1
]. Chemo-/radiotherapy represents the standard treatment for locally advanced NSCLC. However, treatment resistance can occur, resulting in 3-year survival rates of 15–20% [5
]. The PI3K/Akt pathway is frequently hyperactivated in NSCLC [7
]. The serine/threonine kinase Akt, also known as protein kinase B (PKB), exists in three isoforms, namely, Akt1/PKBα, Akt2/PKBβ and Akt3/PKBγ [10
]. A high level of activated Akt1 in tumor specimens is a prognostic factor for poor outcomes in NSCLC [12
]. In addition to stimulating tumor cell proliferation, growth and survival [7
], Akt, especially Akt1, promotes DNA double-strand break (DSB) repair and clonogenic survival after irradiation [13
]. Various Akt inhibitors are currently being tested in clinical studies, demonstrating overall tolerable toxicities and promising anti-tumor activities [17
DSBs are the most severe form of irradiation-induced DNA lesions, which can lead to cell death [20
]. γH2AX foci are clusters of histones, which get phosphorylated adjacent to DSBs and, thus, serve as DSB markers [21
]. Cells are capable of DSB repair, a process, which is usually completed within 24 h after DSB induction [23
]. γH2AX histones strongly promote recruitment of DSB repair proteins and are crucial for efficient repair [24
]. The main pathways for the repair of irradiation-induced DSBs are non-homologous end-joining (NHEJ) and homologous recombination (HR) [27
]. Akt1 has been shown to promote NHEJ via interaction with and stimulation of DNA-dependent protein kinase catalytic subunit (DNA-PKcs) in various cell lines including NSCLC cells A549 and H460 [13
HR is a major repair pathway for irradiation-induced DSBs in late S and G2 phases [32
]. In addition, HR is critical for the repair of broken replication forks and DNA interstrand crosslinks [32
]. HR-mediated repair uses a homologous DNA sequence, usually the sister chromatid, as the repair template resulting in high-fidelity repairs. The first step of HR repair is end resection of DSBs to obtain 3′ single-stranded DNA (ssDNA) overhangs. After that, the 3′ ssDNA overhang invades a homologous double-stranded DNA (dsDNA) sequence. The invading 3′ ssDNA is elongated by DNA synthesis, and the triple helix structure is resolved [38
]. DSB end resection is facilitated by a protein complex that includes BRCA1. Invasion of the ssDNA overhang into the homologous dsDNA sequence is dependent on the recombinase Rad51. Accordingly, decreased Rad51/BRCA1 foci formation at the DSB is an indicator for impaired recruitment and decreased HR-dependent repair [40
]. Conversely, an elevated number of residual Rad51/BRCA1 foci suggests reduced foci resolution and impaired HR repair [39
So far, only a few studies have investigated the possible role of Akt1 in the regulation of HR-dependent DSB repair. Akt1 has been reported to impair the nuclear localization and foci formation of BRCA1/Rad51 as well as HR repair in normal tissue and breast cancer cells [46
]. However, Rad51 foci formation has been demonstrated to be independent of Akt activity in HEK cells following etoposide treatment [49
]. Conversely, Akt has been shown to stimulate Rad51 protein expression in several NSCLC cell lines including A549 cells [50
]. Nonetheless, so far it is not known whether the promotive effect of Akt on Rad51 protein level influences HR-mediated DSB repair in NSCLC cells.
In this study, we investigated the effect of Akt1 on HR-dependent DSB repair and the possible underlying mechanism primarily in NSCLC cells A549 as well as in H460 cells. Moreover, we sought to elucidate whether the modulation of HR repair by Akt1 affects clonogenicity after irradiation. We demonstrated that Akt1 stimulates HR-mediated DSB repair in a Rad51-dependent manner. This contributes to the stimulatory effect of Akt1 on DSB repair following irradiation. Thus, our data provide further insights into the role of Akt1 in DSB repair.
The data from the Akt1-KD experiments presented here indicate that Akt1 plays a regulatory role not only for NHEJ repair, as has been reported earlier [14
], but also for HR repair in a Rad51-dependent manner.
To this aim, the results of Rad51 protein expression and foci formation together with the data from the HR-reporter assays strongly suggest that Akt1 stimulates HR-dependent DSB repair. The overall higher number of radiation-induced Rad51 foci in H460 cells as compared to A549 cells is in line with an earlier report describing cell line-dependent differences in the extent of radiation-induced Rad51 foci formation [54
]. Both, A549 and H460 cells are EGFR and TP53 wildtype, but harbor an activating KRAS mutation. In contrast to A549 cells, the H460 cell line shows an activating mutation of PIK3CA coding for PI3K, which acts upstream of Akt1 [55
]. This might underlie the enhanced phosphorylation level of Akt1 (S473, T308) in H460 cells as compared to A549 cells, which has been reported previously [57
]. The higher phosphorylation level, i.e., Akt1 activation, might explain the herein observed more pronounced effect of Akt1-KD on Rad51 foci number in H460 cells compared to A549 cells. Furthermore, Rad51 foci formation is influenced by several other factors such as recruitment of RPA and BRCA2 as well as other repair proteins like Rad52 [58
]. It is possible that these factors are differentially activated in both cell lines, which might also account for the different extent of reduced Rad51 foci number in A549 and H460 cells after Akt1-KD. Another possible explanation might be the strongly increased amount of γH2AX foci in A549-Akt1-KD cells at 8 h after irradiation. In contrast, we reported previously [31
] that in H460 cells Akt1 depletion significantly enhances the number of residual γH2AX foci 24 h after irradiation, but does not affect the number of γH2AX foci at the time point 8 h post-irradiation. It is well known that H2AX phosphorylation is crucial for the recruitment of DSB repair proteins like Rad51 to the damage site [24
]. Thus, it seems plausible that the increased number of γH2AX foci promotes recruitment of Rad51 to the damage site in A549-Akt1-KD cells, while at the same time Akt1-KD reduces Rad51 protein level and consequently foci formation. This interpretation is supported by the observation that Akt1 depletion decreased the fraction of γH2AX foci, colocalized with Rad51 foci, by more than half in A549 cells and approximately half in H460 cells. Contrary to individual data for γH2AX foci and Rad51 foci, looking at colocalization of γH2AX with Rad51 foci allows adjusting for the above-mentioned stimulatory effect of γH2AX foci on recruitment of Rad51. The reason why Akt1-KD affected the number of γH2AX foci more strongly in A549 cells than H460 cells at 8 h after irradiation is speculative, but might indicate that Akt1 exerts a more pronounced effect on early DSB repair in A549 cells compared to H460 cells.
In addition to foci experiments, the reduced clonogenic survival after MMC treatment further indicates that Akt1 promotes HR in A549 and H460 cells. We already tested the stability of Akt1-KD over a period of 7 d in a previous publication showing excellent stability [31
]. Evaluation of Akt1-KD by Western blotting over a period of 10 d (in accordance with duration of clonogenic assay) was not possible due to technical reasons, i.e., cells could not be grown in a monolayer without media change for 10 days. Anyhow, knockdown stability during the first days of clonogenic assay is most important, since the first 24 h after DNA damage induction are most crucial for repair of DSBs, which manifests after several days as clonogenic cell death [21
]. DSBs, which result from MMC treatment, are primarily and accurately repaired via the HR repair pathway [37
]. Misrepair of MMC-mediated DSBs by NHEJ increases MMC-induced cytotoxicity. Consequently, inhibition of the NHEJ pathway has been shown to partially rescue clonogenicity after MMC treatment [37
]. Thus, the observed modest Akt1-KD-mediated sensitization to MMC is in accordance with the stimulatory role of Akt1 in both HR- and NHEJ-dependent repair.
The stimulation of HR repair by Akt1 seems to be in contrast to reports by Plo et al. [47
] and Jia et al. [46
]. These authors have reported an inhibitory effect of Akt1 overexpression on Rad51 foci formation and HR repair using an HR-reporter assay in normal tissue and breast cancer cells [46
]. In line with this report, we observed an increase in the number of Rad51 foci in the breast cancer cell line MCF-7 after Akt1-KD. On the other hand, the data using the two NSCLC cell lines (A549 and H460) presented here are in accordance with the report by Ko et al., who demonstrated that Akt inhibition reduced the clonogenicity of NSCLC cells A549 and H1703 after MMC treatment [51
]. The differential effect of Akt1 reported in earlier studies [46
] and our current study might indicate a cell line-dependent role of Akt1 in HR.
With respect to the mechanism underlying the stimulatory function of Akt1 in HR repair, our data indicate that Akt1 increases Rad51 protein level, especially in the nucleus after irradiation and as a result enhances Rad51 foci formation at the DSB. Lack of a significant difference in HR repair between the cells with single knockdown of Rad51 and those with Akt1/Rad51 double knockdown implies that Akt1 reduces HR repair in a Rad51-dependent manner. These results are supported by results from other studies reporting that inhibition of Akt reduces Rad51 protein levels in different NSCLC cell lines including A549 cells [50
]. Moreover, Ko et al. demonstrated that sensitization of NSCLC cells to MMC after Akt1 inhibition is partially rescued by Rad51 overexpression [51
]. This implies that Akt1 inhibition, which is known to reduce NHEJ [14
], also impairs HR in a direct manner but does not lead to a passive switch from NHEJ to HR in NSCLC cells. Even though Rad51 protein level [62
] and HR repair capacity [32
] are known to be highest during S/G2 phase, interference with the cell cycle does not seem to underlie the effect of Akt1 on Rad51, since cycle distribution of Akt1-KD cells was not significantly affected. In contrast, earlier reports suggest that Akt increases Rad51 protein level by stimulating Rad51 mRNA and protein stability via reduction of Rad51 ubiquitination and proteasomal degradation [50
The data from the γH2AX and BRCA1 foci assays presented here support the conclusion that Akt1 not only promotes NHEJ-mediated repair of radiation-induced DSBs but also exerts a stimulatory role on HR repair. The increased number of γH2AX foci represents non-repaired DSBs [21
], most likely a consequence of both impaired HR- and NHEJ-mediated DSB repair after Akt1-KD. However, Akt1-KD led to an even more pronounced increase in the number of γH2AX foci in A549 cells when it was combined with a DNA-PKcs inhibitor in comparison with that in cells with unhampered NHEJ. Hence, the increase in the number of γH2AX foci after DNA-PKcs inhibition plus Akt1-KD is at least partly due to the inhibition of DSB repair independent of the effect of Akt1 on DNA-PKcs activity, in other words, it is most likely due to the influence of Akt1 on HR. The number of residual Rad51/BRCA1 foci (24 h after irradiation) is a common indicator for the removal of the repair protein after successful repair and thus forms a marker of non-repaired DSBs as a result of hampered HR repair [39
]. In contrast, the number of Rad51/BRCA1 foci at closer time points after DSB induction (~4–12 h after irradiation) indicates the assembly of the protein at the DSB [39
]. Thus, our results suggest that Akt1 stimulates the recruitment of Rad51 to the damage site, whereas BRCA1 recruitment does not seem to be affected. As a consequence of decreased Rad51 protein assembly at the DSB, HR repair cannot be completed and the BRCA1 protein, which was recruited to the damage before Rad51, cannot be removed from the damage leading to the increased number of residual BRCA1 foci. Thus, this further supports the hypothesis that Akt1 enhances repair of radiation-induced DSBs partly by HR. Interestingly, we observed that the Akt1-KD-mediated impairment of DSB repair was weaker after treatment with DMSO as solvent control in A549 cells. The reason for this is speculative but might be attributed to the stimulatory effect of DMSO on NHEJ [65
]. Nonetheless, the data presented here are in accordance with several studies, showing impaired DSB repair after irradiation following AKT1-siRNA or the pharmacological inhibition of Akt [13
]. However, in a previous report, Akt1 depletion did not increase the number of residual γH2AX foci in A549 cells when NHEJ was impaired by DNA-PKcs-KD [14
]. Yet, this experiment was carried out using cells in stationary phase, whereas we used cells in log-phase for the present study. Accordingly, our result is in line with the role of the HR pathway in DSB repair in late S and G2 phases, but not in the G1 phase.
Conversely, the data from colony formation assays indicate that HR plays a minor role in the Akt1-mediated stimulation of clonogenic survival after irradiation. The reduction in clonogenic survival after Akt1-KD and irradiation is in line with previous studies, showing radiosensitization by Akt inhibition or AKT1-siRNA [14
]. The strong radiosensitization of non-synchronized cells but only slight radiosensitization of S/G2 phase synchronized cells following Akt1-KD fits the pattern of NHEJ-deficient but not HR-deficient cells [66
]. Furthermore, the lack of effect of Akt1-KD on clonogenicity after DNA-PKcs inhibition and irradiation is supported by a previous study, showing that Akt1-KD does not induce radiosensitization in DNA-PKcs-deficient cells [15
]. Thus, our results imply that Akt1-KD-mediated reduction of HR decreases clonogenic survival after MMC treatment but does not affect clonogenicity after irradiation. This finding is supported by previous studies, which show that impairment of HR increases sensitivity to MMC to a greater degree than to irradiation [67
]. It is known that HR is a predominant factor for survival after MMC [37
], whereas clonogenic survival after irradiation is influenced by several DNA repair pathways [69
] and additional factors such as autophagy [72
] which might shadow the effect of reduced HR on post-irradiation clonogenicity. Nonetheless, the reason behind why decreased HR repair after Akt1-KD reduces DSB repair following irradiation but does not affect clonogenic survival is speculative. The lack of correlation between DSB repair and clonogenic survival might be explained by other influences, such as autophagy. Several publications have shown that autophagy increases radioresistance [72
]. Moreover, Akt1-KD as well as the inhibition of HR proteins such as BRCA1 is known to induce autophagy and subsequently, increase tumor cell survival [75
]. Thus, Akt1-KD-mediated HR inhibition might lead to enhanced autophagy, which could be sufficient to counteract the radiosensitizing effect of Akt1-KD through impaired HR-mediated DSB repair.
In conclusion, Akt1 promotes HR repair in a Rad51-dependent manner in NSCLC cells A549 and H460. Thus, our data provide further insights into the Akt-mediated resistance of NSCLC to chemo- and radiotherapy. Although further analyses are necessary to investigate the functional interaction between Akt1 and Rad51 in stimulating DSB repair, the present study offers new aspects for the development of novel strategies for selective targeting of NSCLC cells.