Loss of the cyclin kinase inhibitor p27kip1 opens a
therapeutic window by deregulating pathway choice
after DNA double strand breaks

doi:10.21203/rs.3.rs-2889500/v1

Download PDF

Article

Loss of the cyclin kinase inhibitor p27kip1 opens a therapeutic window by deregulating pathway choice after DNA double strand breaks

https://doi.org/10.21203/rs.3.rs-2889500/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Reduced expression of the cyclin kinase inhibitor p27kip1 is found in many human tumors and correlates with inferior prognosis. We tested the importance of p27 loss for the induction of genetic instability in Notch driven cholangiocarcinoma (CCA). Specifically, we asked how loss of p27 interferes with DNA repair pathway choice i.e. non-homologous-end-joining (NHEJ) or homologous recombination (HR). We detected all hallmarks of HR, in p27 deficient cells even though these cells had not undergone DNA replication. This defect in pathway choice depends on a previously unknown interaction of p27 with the RAD17 protein. In a cholangiocarcinoma mouse model loss of p27 greatly accelerated tumor formation but also resulted in a strikingly increased sensitivity against DNA damage response (DDR) targeting agents. We find that 30% of human CCC patients cluster in a group which corresponds to the DDR sensitive phenotype we have identified. In this work we show that the cyclin kinase inhibitor p27kip1 has an essential role in regulating the DNA damage response. This new activity of p27 is necessary to suppress homologous recombination-based DNA repair during the G1 phase. These findings point to a role of DDR targeting drugs in CCA with low levels of p27.

Biological sciences/Cancer/Gastrointestinal cancer/Biliary tract cancer

Health sciences/Biomarkers/Predictive markers

DNA double strand breaks

homologous recombination

cyclin-dependent kinase inhibitor p27kip1

Intrahepatic cholangiocarcinoma (iCCA) has shown a global trend of rising incidence and mortality [1]. Systemic chemotherapy with gemcitabine and cisplatin is still the main option for standard of care [2]. One frequently altered signal transduction cascade in iCCA is the Notch pathway. We recently showed that alteration of this pathway can lead to the transformation of undifferentiated hepatic progenitor cells and the formation of iCCA[3]. We showed that the cyclin E gene which is essential for this process is transcriptionally controlled by Notch-ICD. The cyclin kinase inhibitor (CKI) p27^kip1 primary function is to bind cyclin E/CDK2 complexes and regulate their function at the G1/S boundary[4]. In line with its function as a CKI, loss of p27 results in reduced dependency on mitogens for cells which re-enter the cell cycle from quiescence.

In agreement with these studies, decreased levels of p27 have been shown to associate with unfavorable tumor stage and grade in a number of human cancers [5]. In biliary tumors p27 expression correlates strongly with the malignancy of the tumor and overall survival of the patient [6]. p27 protein levels are primarily regulated by ubiquitin dependent proteasomal degradation which restricts p27 expression to the G0 and G1 phases of the cell cycle. Meanwhile, several additional binding partners of p27 have been discovered, encompassing processes like cell adhesion, cell migration and cell death. These additional functions are often mediated by direct interactions with proteins like RhoA, Rac or stathmin [7–9].

In this work we show that cells which have lost p27 cannot sufficiently repair DNA double strand breaks. This function of p27 is independent of its cyclin kinase inhibitory activity but depends on a so far unknown interaction of p27 with the RAD17 protein. Loss of p27 promotes phosphorylation of RAD17 during G1 by the ATM kinase in response to DNA damage which in turn initiates homologous recombination dependent DNA repair even though the cell has not entered S-phase.

We tested the physiological importance of this mechanism using a Notch/Akt driven cholangiocarcinoma mouse model and histological analyses of primary human CCA tissues. Our analyses revealed a synthetic lethal interaction between loss of p27 and DDR targeting drugs pointing towards new treatment options in p27 deficient tumors.

Detailed description of material and methods in supplementary

Loss of p27 accelerates CCA formation and predisposes tumor cells to DNA damage

To study the impact of p27 in cholangiocarcinoma we used a previously established mouse model in which Notch and AKT expressing vectors (Notch-ICD; myr-AKT) are stably delivered into the liver by hydrodynamic tail-vein injection (HDTVi) [10]. We had previously shown that aberrant expression of NOTCH receptors is often found in human CCA and that mice which overexpress NOTCH-ICD develop CCA [3]. Importantly NOTCH-ICD will induce transcription of the Cyclin E gene which in turn results in the formation of genetically unstable cells and cellular transformation [3].

To ablate p27 expression in combination with Notch/Akt expression we generated an additional vector system co-expressing a p27 specific sgRNA together with the CAS9 endonuclease (Supplemental Fig. S1A) (further denoted as Notch/AKT/sgp27 mice). Transfection efficacy, evaluated seven days post HDTVi, was unaltered independent of the p27 genotype (Supplemental Fig. S1 B and C). After injection of these plasmids into C57BL/6J mice, we performed in vivo imaging by MRI and observed formation of CCA with a latency of 4 to 5 weeks. As expected, loss of p27 resulted in a higher incidence of hepatic lesions 4 weeks after HDTVi with 68% of the Notch/AKT/sgp27 showing MRI-detectable tumors at 4 weeks, while only 29% of Notch/AKT mice showed tumors (Fig. 1A). No difference in liver weight, body weight, liver/body weight ratio or serum liver enzymes was found (Supplemental Fig. S2). Chemotherapy involving gemcitabine constitutes the current standard of care for advanced CCA. To test the role of p27 in the response to treatment, we initiated treatment by intraperitoneal injection of gemcitabine (0.1 mg/ gram mouse body weight) twice per week. To ensure that the tumor incidence in both groups was comparable, we only started the treatment when the mice were tumor-positive by MRI. No difference in survival of vehicle treated mice caring wtp27 or p27 deficient tumors were indicated (Fig. 1B). Surprisingly, mice with p27 deficient tumors survived significantly longer compared to p27 wildtype tumor mice (52 vs. 24 days CI: 0.15–1.43, p = 0.002) (Fig. 1C) after gemcitabine treatment. Histological analyses (H&E) and immune histochemistry for Ki67 revealed no significant differences in proliferation rate between tumors in both groups after Gemcitabine treatment (Fig. 1D and E). Also, no differences in liver weight, body weight, liver/body weight ratio or serum liver enzymes were found (Supplemental Fig. S3).

However, expression of the apoptosis marker cleaved caspase 3 and the DNA damage marker γH2AX were significantly higher in the p27 deficient compared to the wildtype mouse tumors. After treatment of both groups with gemcitabine this difference became even more apparent. To ensure that these differences were due to changes in the tumor cells we corroborated the results using immunohistochemical staining (Fig. 1F and G). Together these results pointed towards a defect in DNA repair in tumor cells which had lost p27.

Loss of p27 promotes an aberrant homologous recombination-dependent DNA damage response during G1

To understand the molecular mechanism by which loss of p27 results in increased levels of DNA damage we turned to a mouse embryonic fibroblast cell line with targeted deletions of the p27 gene. The use of MEF cells also allowed us to assess the function of p27 as a cyclin kinase inhibitor by using a cell line derived from a knock-in mouse strain which expresses a form of p27 which cannot bind to cyclin-cdk complexes

(p27^{c − k−})[11].

These three cell lines were synchronized and irradiated in different stages of the cell cycle (G1, S-phase). We found a significant increase in the number of γH2AX foci 24h after irradiation in p27 knockout cell which were irradiated during the G1-phase as compared to WT cells. No difference in γH2AX foci formation was found in cell lines which were irradiated during S-phase (Fig. 2A). Importantly, cells expressing the non-cyclin/cdk binding form of p27 (p27^{c − k−}) behaved like wildtype cells after irradiation pointing towards a cdk independent mechanism by which p27 is involved in the DNA damage response. To ensure that this phenotype was not restricted to fibroblasts, cholangiocarcinoma cells (TFK1) in which p27 expression was ablated by siRNA were irradiated after synchronization. As in fibroblasts we detected a significant difference in the levels of γH2AX comparing p27siRNA to p27 wildtype cells when cells were irradiated in G1 but not after irradiation in S phase (Fig. 2B).

As p27 deficient cells accumulated γH2AX foci specifically when the damage occurred during G1 we first analyzed whether non-homologous end joining (NHEJ) dependent DNA repair was altered after loss of p27. The protein p53BP1 restricts end resection of DNA double strand breaks and promotes NHEJ pathway choice [12]. When cholangiocarcinoma cells (TFK1) cells were irradiated in G1 or S-phase no differences in the number of p53BP1 foci were detected in p27 deficient or control cells (Fig. 2C) indicating, that NHEJ was not affected.

Next, we tested if homologous recombination (HR) dependent DNA repair was affected by loss of p27. BRCA1 has been shown to promote HR and should therefore be largely restricted to cells which have undergone damage during S or G2-phases [13]. Surprisingly, we found that p27 depleted CCA cells in which DNA damage was induced during G1 displayed significantly more pBRCA1 positive foci than p27 expressing cells. However, no difference was detected in cells which were damaged in S phase (Fig. 2D). In agreement with an aberrant activation of HR, we also detected a significant increase in phosphorylation of the BRCA1 protein in p27 deficient TFK1 cells which were irradiated in G1 as compared to wildtype cells (Fig. 2E) [14].

To further corroborate the finding that HR dependent DNA double strand repair is activated during the G1 phase in p27 deficient cell we used pRAD51 immunofluorescence staining. RAD51 is often used as an indicator of HR as its loading on DBS has been shown to be essential for this process [15]. As shown in Fig. 2F p27 depleted TFK1cells showed significantly more pRAD51 foci compared to WT cell lines when damaged during G1. No difference in pRAD51 foci formation was observed, when the damage occurred during S phase (Fig. 2F and Supplemental Fig. S4A). This finding was also not dependent on the ability of p27 to act a as cyclin kinase inhibitor as p27^{C − K−} MEF behaved like p27 wildtype cells while p27 deficient cells showed a significant increase in RAD51 foci formation (Supplemental Fig. S4B)

To directly test whether the HR-dependent DNA repair mechanism is indeed active in p27^−/− cells and independent of cyclin kinase inhibitory activity, we created p27^+/+, p27^−/−, and p27^{c − k−} cell lines with a stably integrated plasmid expressing a pDR-GFP recombination reporter which allows the detection of cells undergoing HR in cells [16] (Supplemental Fig. S5A). pDR-GFP expressing cells were transfected with the pCβA-SceI plasmid to allow the transient expression of the SceI endonuclease and synchronized in G1 to measure HR during this stage in the cell cycle (Supplemental Fig.S5B). p27^−/− cells showed significantly increased GFP expression as compared to p27^+/+ and p27^{c − k−} cells indicating that these cells were able to activate a HR response after damage (Fig. 3A). HR is initiated with the resection of the 5´ strands to generate 3´ single-stranded DNA (ssDNA), which is required for RAD51 binding and strand invasion. We therefore compared 5´strand resection in p27^−/− cells to p27^+/+ cells using a previously described method [17] which uses the ER-AsiSI restriction enzyme to induce DSB at sequence specific sites. A higher percentage of ssDNA was found in p27^−/− than WT HeLa cells synchronized in G1 upon expression of ER-AsiSI (Fig. 3B)

These data suggests that cells which have lost p27 are undergoing an aberrant attempt to repair DNA double strand breaks which occur during the G1 phase by homologous recombination. Given that during G1 no sister chromatids are available for recombination, we wondered if the increased propensity to initiate HR dependent repair might also increase the rate of interhomolog recombination. To test this, we used the previously described S/P cell line which allows for the detection of interhomolog recombination events [18]. To test the effect of p27 on homologous recombination we deleted p27 using CRISPR/Cas and derived two independent p27^−/− S/P cell lines (Supplemental Fig.S6). Upon transient expression of the I-SceI endonuclease and selection of the surviving cells we found that loss of p27 led to a significant increase in interhomolog recombination (Fig. 3C).

Taken together, these findings indicate that p27 expression is necessary to block the initiation of HR during the G1 stage of the cell cycle. Aberrant activation of HR during G1 results in persistent DNA damage and cell death. However, loss of p27 also promotes recombination between homologous chromosomes. Surprisingly, the HR suppressive activity is independent of the ability of p27 to bind and inhibit cyclin-CDK complexes. We therefore tried to identify the underlying molecular mechanism which requires p27 to suppress HR before S-phase.

The interaction of p27 and RAD17 is necessary to prevent HR-dependent DNA double strand repair during G1

In an independent shRNA screen, which we conducted to identify resistance mechanisms to p27 stabilizing compounds [19], we found an interaction between p27 and the RAD17 protein [20, 21]. RAD17 is an essential component of the DDR machinery required to allow binding of the 9-1-1 (RAD9-HUS1-RAD1) clamp loading complex to damaged DNA [22–24]. Recently, RAD17 was shown to be also involved in the recruitment of the MRN (MRE11-RAD50-NBS1) complex to sites of DNA damage [25, 26]. Upon irradiation, RAD17 is phosphorylated by the ATM kinase at two conserved serines (S645 and S635) [27, 28] which is believed to be essential for the interaction of RAD17 with the MRN complex.

To analyze p27, RAD17 and ATM binding we synchronized CCA cells and measured complex formation of the endogenous proteins at different stages of the cell cycle

in irradiated and control cells. Immunoprecipitation experiments showed that p27 and RAD17 form a complex during the G1 phase which is not disrupted when cells are irradiated. However, RAD17 was no longer in complex with p27 when irradiated cells reached the G1/S transition and p27 levels dropped (Fig. 4A). Mutation of RAD17 phosphorylation sites (S635, S645) prevented dissociation of p27 and RAD17 after irradiation (Supplemental Fig. S7) indicating that the cell cycle regulated decrease in p27 levels at the G1/S transition and the phosphorylation of RAD17 regulate the release of RAD17 from p27. During S phase p27 is not expressed and hence no complex formation was detectable. As expected, ATM became phosphorylated at S1981 in response to irradiation irrespective of the cell cycle stage (Fig. 4A). A complex of RAD17 and ATM or phosphorylated ATM in irradiated p27 expressing cells was not detectable before the cells had reached the G1/S transition (Fig. 4A). In line with this observation, RAD17 phosphorylation (S635/S645) was only weakly detectable in irradiated G0 or G1 p27 WT cells while RAD17 was strongly phosphorylated when it was bound to ATM at the G1/S transition (Supplemental Fig. S8). Importantly knockdown of p27 expression promoted binding of RAD17 to ATM and allowed phosphorylation of RAD17 at S635 and S645 before cells had reached the G1/S transition (Supplemental Fig. S8). This data suggested that p27 expression during G1 is necessary to prevent RAD17 binding to and phosphorylation by ATM.

The initiation of double strand repair depends on the recruitment of the MRN (MRE11-RAD50, NBS1) complex onto damaged DNA. RAD17 was shown to possess an additional function in allowing MRN recruitment [25].

We expressed RAD17 wildtype or a non-phosphorylatable mutant (RAD17AA) in Hela cells in which the endogenous RAD17 was knocked down using siRNAs. At the same time, we depleted p27 by siRNA treatment to obtain cells which differ in their p27 status (wildtype or knockdown) and in their RAD17 (wildtype or phospho-mutant) status. Upon synchronization of these cells in G1 and irradiation with 4Gy we again detected complexes between ATM and RAD17 in cells in which p27 was depleted but not in wildtype expressing cells during the G1 phase (Supplemental Fig. S8).

To analyze complex formation between RAD17 and MRN components we immunoprecipitated RAD17 from damaged chromatin and measured the amount of bound MRE11, Rad50 and NBS1. In agreement with the unscheduled formation of ATM and RAD17 complexes we also detected significantly increased binding of MRN components to RAD17 on damaged chromatin (Fig. 4B). Importantly loss of RAD17 phosphorylation did not prevent ATM binding but prevented binding to MRN components almost completely. Moreover, loss of RAD17 phosphorylation prevented yH2AX appearance completely (Supplemental Fig. S9). This role of p27 in preventing premature MRN interaction with RAD17 on chromatin during G0 and G1 phase was independent of p27 ability to inhibit cyclin cdk complexes (Supplemental Fig. S10).

This data suggests that phosphorylation of RAD17 by the ATM kinase is critically important for the observed phenotype of unscheduled HR in p27 deficient cells. However, binding of ATM to RAD17 was not affected by RAD17 phosphorylation. We therefore tested if ATM kinase activity itself is important for complex formation. For this we synchronized irradiated CCA cells in G1 and measured the effect of caffeine (nonspecific kinase inhibitor), ATM inhibitor, CHK2, ATR and CHK1 inhibitor in p27 depleted or wildtype expressing cells. While loss of p27 resulted in increased binding of ATM and RAD17, caffeine and ATM inhibition significantly reduced complex formation. All other inhibitors had no effect (Fig. 4C).

Finally, we measured the levels of RAD17 phosphorylation in primary mouse cholangiocarcinoma tissues derived from our treatment study in p27 WT and deficient mice. In agreement with the results from cell lines loss of p27 resulted in a striking increase in RAD17 and ATM activation in tumors from p27 deficient mice which had been treated with gemicitabine as compared to wildtype p27 mice (Fig. 4D).

Using embryonic fibroblast and CCA cell lines as well as primary mouse tumor tissues this data shows that loss of p27 promotes premature binding of ATM to RAD17 during the G1 phase which results in phosphorylation of RAD17 which in turn leads to premature binding to the MRN complex to Rad17 onto damaged chromatin.

Loss of p27 sensitizes tumor cells against drugs that target the DDR response

As our data suggested a critical role for ATM inhibition, we tested a series of pharmacological inhibitors of the DDR for their ability to synergize with loss of p27 to affect the viability of cholangiocarcinoma derived mouse organoids. We tested ATM, ATR, DNA-PK, Chk1, Chk2 and PARP inhibitors in combination with either gemcitabine or irradiation on CCA organoids derived from p27 deficient or wildtype mice. We found that already 48hr after treatment with ATM or PARP inhibitors organoids derived from p27 deficient mice started to undergo apoptosis as measured by caspase 3 cleavage when combined with irradiation or gemcitabine treatment (Supplemental Fig. S11). Organoids derived from p27 wildtype mice were much more resistant to induction of cell death under these conditions (Fig. 5A, 5B). We then quantitated cell viability of the different organoid cell lines in response to different amounts of ATM or PARP inhibitor. While both inhibitors alone lead only to small changes in viability (Supplemental Fig. S12), combination with irradiation or gemcitabine resulted in significantly increased levels of cell death in the p27 deficient cholangiocarcinoma organoids (Fig. 4C,D) as compared to wildtype organoids. Moreover, organoids derived from p27 wild type mice were much more resistant to the combination of DNA damage and inhibition of PARP activity as determined via morphological examination (Fig. 4E). To ensure that the selective sensitivity to combination treatment was not caused by a defect in cdk binding we treated MEFs derived from p27^+/+, p27^−/− and p27^{c − k−} mice with the PARP inhibitor Olaparib alone or in combination with irradiation. Only p27 knockout cells respond to combination treatment with apoptotic cell death while the other cell lines are largely resistant, reinforcing the notion that the DDR phenotype observed in p27 knockout cells is independent of cyclin kinase complex inhibition (Supplemental Fig. S13).

These results show that the aberrant induction of HR-dependent DNA damage repair in cells which have lost p27, sensitizes both cell lines and CCA organoids to drugs that target the DDR in combination with irradiation or gemcitabine. Moreover, our results also point towards a treatment option for patients suffering from cholangiocarcinoma. Therefore, we finally asked if a iCCA patient subgroup exists that corresponds to our murine iCCA model being characterized by a highly aggressive/poorly differentiated phenotype with low levels of p27 and high appearance of DNA damage markers such as γH2AX. To this end, we examined the presence of both markers by immunohistochemistry (IHC) in 178 iCCA patients using TissueMicroArrays (TMAs) containing well (G1), moderately (G2), and poorly differentiated (G3) tumors. We identified three distinct subgroups by unsupervised model-based clustering using p27 and γH2AX IHC scores as numerical variables and the histologic grade as categorical variable. Among these, cluster 3 corresponds very well to the murine CCA model exclusively consisting of poorly differentiated tumors (G3) showing low p27 and higher γH2AX IHC scores (Fig. 6A,B Supplemental Table S1) for detailed information regarding cluster characteristics). Notably, ~ 30% (53/178) of all analyzed patients fall into cluster 3 and could, based on our data, benefit from a PARP inhibitor-containing combinatorial treatment. Further in line with our findings is the fact that none of the analyzed iCCAs showed both, high p27 and high γH2AX levels (Fig. 6C red box).

In this work we uncover a new function of the p27kip1 protein in preventing unscheduled induction of homologous recombination-based DNA double-strand repair before S-phase. In contrast to other defects in the DDR i.e. BRCA1, BRCA2, Rad51, ATM loss of p27 does not result in a defect in the repair machinery itself [29, 30]. Instead loss of p27 results in the initiation of homologous recombination during G1 a time in the cell cycle where non-homologous end joining (NHEJ) [31] is the pivotal pathway of DNA double strand repair [32]. This new activity of p27 is based on its ability to bind to and prevent the phosphorylation and thereby activation of the RAD17 protein. Loss of p27 leads to the premature phosphorylation of RAD17 which in turn promotes loading of the MRN complex onto damaged DNA (Fig. 6D). The mechanisms which couple cell cycle progression with DDR pathway choice have been extensively studied. It has meanwhile been established that several decision points are required to allow DSB repair by HR. In addition to MRN association with the DSB, DNA end resection and RAD51 loading are essential for HR to occur. Activation of end resection by the MRN complex depends on the co-localization of CtIP with MRN on chromatin a process which is facilitated by ATM and CDK phosphorylation and ubiquitylation by BRCA1[33–35]

The cell cycle regulated activation of cdk therefore plays a pivotal role in ensuring the coupling of cell cycle phases to pathway choice. While loss of p27 increases the levels of cdk activity during G1, a mutant version of p27 which cannot bind and inhibit CDK behaved like wildtype p27 i.e. suppressed the activation of HR before S-phase. This result argues against a primary role of CDK de-inhibition through loss of p27 in this process and favours a model in which RAD17 sequestration by p27 is critically important to prevent HR induction during G1. Loss of RAD17 sequestration and de-inhibition of cdk activity might however collaborate to create a HR permissive environment during the G1 phase. In support of this idea cell lines engineered to express a GFP based HR reporter system revealed a significantly increased propensity to undergo HR based repair during G1 when p27 was lost as compared to wildtype cells.

The attempt to undergo HR during a time in the cell cycle when no sister chromatids are available results in an aberrant or incomplete repair of DNA double strand breaks as shown by a persistent presence of γH2AX in cells which have lost p27 as compared to wildtype cells. While this type of DNA damage led to an increased rate of cell death in p27 deficient cells after DSB we also detected an increase in the rate of interhomolog recombination a type of homologous recombination which is prone to induce additional DNA rearrangements and LOH [18].

In an interesting parallel to our findings it had previously been shown, that the CKI p21 also has a dual role in cell proliferation and replication control [36]. In addition to its function as an CDK inhibitor, it was shown to be able to bind the PCNA protein thereby inhibiting its function in DNA replication upon induction by p53. p27 therefore affects tumor formation at least at two critical decision points: first it controls entry of cells into S-phase by inhibiting cyclin E and A dependent CDK-activity [37] and second it prevents activation of HR-dependent DNA repair during the G1 phase. Moreover, both functions play a combined role as suppression of cdk activation during G1 is also important for suppression of HR even though with respect to p27 the sequestration of RAD17 seems to be the critical activity within this process. Cells which have lost p27 are therefore less dependent on mitogenic stimuli and more prone to undergo DNA damage which cannot be sufficiently repaired. We tested the impact of this defect in DNA repair in a cholangicocarcinoma (CCA) mouse model. Cholangiocarcinoma is a devastating disease with very limited therapeutic options in advanced stages. One frequently altered signal transduction cascade in iCCA is the Notch pathway. We and others showed that alteration of this pathway can lead to the transformation of undifferentiated hepatic progenitor cells and the formation of iCCA. [3, 38, 39]. We showed that the cyclin E gene is under direct transcriptional control of Notch-intra cellular domain (NOTCH-ICD) and that overexpression of cyclin E is required for the transformation of hepatic progenitors. Moreover the F-Box protein Fbw7 which is frequently mutated in cholangiocarcinoma [40] has been shown to lead to increased levels of Notch and Cyclin E [41], which in turn results in degradation of p27 after phosphorylation and ubiquitylation by SCF^skp2 dependent E3 ligase [42]. It is interesting to note that CCA which express low levels of Fbw7 show a more aggressive phenotype but respond better to chemotherapy [43]. Notch driven CCA therefore comprise a type of tumor in which the dysregulation of p27 is directly connected to the pathogenesis of the cancer.

In line with this conclusion within a multitude of prognostic markers which have been evaluated in CCA low levels of p27 expression correlate strongly with the aggressiveness of the tumor and overall survival of the patient [6]. The importance of p27 dysregulation in CCA and our finding of a specific defect in the DNA damage response pathway connected to loss of p27 provides the rationale for exploring DNA damage response targeted drugs (DDR) in these types of tumors. Indeed, our analysis of CCA bearing mice which were treated with gemcitabine, revealed a significantly increased sensitivity of p27 deficient tumors towards this drug accompanied with high levels of RAD17 and ATM phosphorylation. Based on our observations we propose molecular model of the interactions between p27-RAD17 and ATM in response to DNA damage occurring during G0 and G1 phase in p27 wildtype and knockout cells (Fig. 6D). Using organoid cultures from Notch/Akt driven cholangiocarcinomas we observed a striking difference in the sensitivity against DDR targeted drugs in combination with DNA damaging treatments in p27 deficient organoids. This was especially pronounced when gemcitabine was combined with the PARP inhibitor olaparib. Loss of p27 therefore results in accelerated tumor development but at the same time opens a therapeutic window as it sensitizes cells which express low levels of p27 to DDR inhibitors which interfere with NHEJ [44, 45].

Our analysis of primary iCCA also revealed the existence of a substantial number of patients which show an aggressive undifferentiated phenotype which was characterized by low levels of p27 and significantly increased presence of γH2AX.

In addition to the well-known dysregulation of p27 expression by ubiquitin dependent proteolysis a number of cdkn1b gene inactivating mutations in different tumor types either affecting the germline or somatic mutations in tumor tissues have been identified recently. Interestingly multiple endocrine neoplasias (MEN) a group of endocrine disorders which are characterized by mutations in the Ret gene can be caused by germline mutations in cdkn1b independent of Ret a type of syndrome which was named MEN 4 [46]. In hairy cell leukemias cdkn1b mutations account for 16% [47] of alterations making it the second most frequently mutated gene in this type of leukemia. Finally, small intestine neuroendocrine cancers show frameshift mutations and hemizygous deletions in a substantial number of patients [48]. These genetic p27 mutations could also provide a biomarker which guides the use of DDR targeted drugs like PARP inhibitors [49] in these groups of patients.

In this work we show that loss of the cyclin kinase inhibitor (cki) p27kip1 does possess a central role in suppressing HR independent of its role as a cki. Moreover, the importance of this mechanism in a Notch/Akt driven cholangiocarcinoma mouse model in vitro and vivo provides better therapeutic strategies of CCC.

All authors read and approve the manuscript

Ethical Approval

All animal experiments have been performed in accordance to the approval of the Regierungspräsidium Karlsruhe, Germany (G237/18).

Human samples were used for retrospective analysis following the institutional rules of the University Hospital Tübingen

Competing interests

All authors declare that there is no conflict of interest.

Authors' contributions

Przemyslaw Bozko design and performed experiments

Khac Cuong Bu, Mohammad Rahbari, Sven Mattern, Mihály Sulyok, Stephanie Rössler, Benjamin Goeppert, Tim Scholta, Thi Mai Ly Nguyen, Mahmoud Toulany, Ying Shi, Ramona Rudalska, Daniel Dauch, Stephan Singer performed experiments.

Mirian Fernandez, Pavlos Missios, Maria Garcia-Beccaria, Julian Götze, Mathias Riebold, Lars Zender, Daniel Zips, Sebastian Reuter, Peter Schirmacher provided expertise and feedback.

Mathias Heikenwälder design experiments, provided expertise and feedback

Nisar P. Malek conceived and design experiments, wrote the manuscript and secured funding

Funding

German Research Fundation (Deutsche Forschungsgemeinschaft) SFB/TR 209

Availability of data and materials

All original data available on request

Banales JM, Marin JJG, Lamarca A, Rodrigues PM, Khan SA, Roberts LR et al. Cholangiocarcinoma 2020: the next horizon in mechanisms and management. Nat Rev Gastroenterol Hepatol 2020; 17: 557-588.
Valle J, Wasan H, Palmer DH, Cunningham D, Anthoney A, Maraveyas A et al. Cisplatin plus gemcitabine versus gemcitabine for biliary tract cancer. N Engl J Med 2010; 362: 1273-1281.
Zender S, Nickeleit I, Wuestefeld T, Sorensen I, Dauch D, Bozko P et al. A critical role for notch signaling in the formation of cholangiocellular carcinomas. Cancer cell 2013; 23: 784-795.
Polyak K, Kato JY, Solomon MJ, Sherr CJ, Massague J, Roberts JM et al. p27Kip1, a cyclin-Cdk inhibitor, links transforming growth factor-beta and contact inhibition to cell cycle arrest. Genes & development 1994; 8: 9-22.
Wander SA, Zhao D, Slingerland JM. p27: a barometer of signaling deregulation and potential predictor of response to targeted therapies. Clin Cancer Res 2011; 17: 12-18.
Ruys AT, Groot Koerkamp B, Wiggers JK, Klümpen HJ, ten Kate FJ, van Gulik TM. Prognostic biomarkers in patients with resected cholangiocarcinoma: a systematic review and meta-analysis. Ann Surg Oncol 2014; 21: 487-500.
Besson A, Gurian-West M, Schmidt A, Hall A, Roberts JM. p27Kip1 modulates cell migration through the regulation of RhoA activation. Genes & development 2004; 18: 862-876.
Baldassarre G, Belletti B, Nicoloso MS, Schiappacassi M, Vecchione A, Spessotto P et al. p27(Kip1)-stathmin interaction influences sarcoma cell migration and invasion. Cancer cell 2005; 7: 51-63.
Connor MK, Kotchetkov R, Cariou S, Resch A, Lupetti R, Beniston RG et al. CRM1/Ran-mediated nuclear export of p27(Kip1) involves a nuclear export signal and links p27 export and proteolysis. Mol Biol Cell 2003; 14: 201-213.
O'Rourke CJ, Matter MS, Nepal C, Caetano-Oliveira R, Ton PT, Factor VM et al. Identification of a Pan-Gamma-Secretase Inhibitor Response Signature for Notch-Driven Cholangiocarcinoma. Hepatology 2020; 71: 196-213.
Besson A, Hwang HC, Cicero S, Donovan SL, Gurian-West M, Johnson D et al. Discovery of an oncogenic activity in p27Kip1 that causes stem cell expansion and a multiple tumor phenotype. Genes & development 2007; 21: 1731-1746.
Bunting SF, Callen E, Wong N, Chen HT, Polato F, Gunn A et al. 53BP1 inhibits homologous recombination in Brca1-deficient cells by blocking resection of DNA breaks. Cell 2010; 141: 243-254.
Chapman JR, Sossick AJ, Boulton SJ, Jackson SP. BRCA1-associated exclusion of 53BP1 from DNA damage sites underlies temporal control of DNA repair. J Cell Sci 2012; 125: 3529-3534.
Cortez D, Wang Y, Qin J, Elledge SJ. Requirement of ATM-dependent phosphorylation of brca1 in the DNA damage response to double-strand breaks. Science 1999; 286: 1162-1166.
Bhat KP, Cortez D. RPA and RAD51: fork reversal, fork protection, and genome stability. Nat Struct Mol Biol 2018; 25: 446-453.
Pierce AJ, Johnson RD, Thompson LH, Jasin M. XRCC3 promotes homology-directed repair of DNA damage in mammalian cells. Genes & development 1999; 13: 2633-2638.
Zhou Y, Caron P, Legube G, Paull TT. Quantitation of DNA double-strand break resection intermediates in human cells. Nucleic Acids Res 2014; 42: e19.
Stark JM, Jasin M. Extensive loss of heterozygosity is suppressed during homologous repair of chromosomal breaks. Molecular and cellular biology 2003; 23: 733-743.
Nickeleit I, Zender S, Sasse F, Geffers R, Brandes G, Sorensen I et al. Argyrin a reveals a critical role for the tumor suppressor protein p27(kip1) in mediating antitumor activities in response to proteasome inhibition. Cancer cell 2008; 14: 23-35.
Wang X, Zou L, Zheng H, Wei Q, Elledge SJ, Li L. Genomic instability and endoreduplication triggered by RAD17 deletion. Genes & development 2003; 17: 965-970.
Wang X, Zou L, Lu T, Bao S, Hurov KE, Hittelman WN et al. Rad17 phosphorylation is required for claspin recruitment and Chk1 activation in response to replication stress. Mol Cell 2006; 23: 331-341.
Bermudez VP, Lindsey-Boltz LA, Cesare AJ, Maniwa Y, Griffith JD, Hurwitz J et al. Loading of the human 9-1-1 checkpoint complex onto DNA by the checkpoint clamp loader hRad17-replication factor C complex in vitro. Proceedings of the National Academy of Sciences of the United States of America 2003; 100: 1633-1638.
Medhurst AL, Warmerdam DO, Akerman I, Verwayen EH, Kanaar R, Smits VA et al. ATR and Rad17 collaborate in modulating Rad9 localisation at sites of DNA damage. Journal of cell science 2008; 121: 3933-3940.
Zou L, Cortez D, Elledge SJ. Regulation of ATR substrate selection by Rad17-dependent loading of Rad9 complexes onto chromatin. Genes & development 2002; 16: 198-208.
Wang Q, Goldstein M, Alexander P, Wakeman TP, Sun T, Feng J et al. Rad17 recruits the MRE11-RAD50-NBS1 complex to regulate the cellular response to DNA double-strand breaks. Embo j 2014; 33: 862-877.
Zeng Y, Jie X, Wu B, Wu G, Liu L, Xu S. IQGAP3 interacts with Rad17 to recruit the Mre11-Rad50-Nbs1 complex and contributes to radioresistance in lung cancer. Cancer Lett 2020; 493: 254-265.
Post S, Weng YC, Cimprich K, Chen LB, Xu Y, Lee EY. Phosphorylation of serines 635 and 645 of human Rad17 is cell cycle regulated and is required for G(1)/S checkpoint activation in response to DNA damage. Proc Natl Acad Sci U S A 2001; 98: 13102-13107.
Bao S, Tibbetts RS, Brumbaugh KM, Fang Y, Richardson DA, Ali A et al. ATR/ATM-mediated phosphorylation of human Rad17 is required for genotoxic stress responses. Nature 2001; 411: 969-974.
Jiricny J. The multifaceted mismatch-repair system. Nat Rev Mol Cell Biol 2006; 7: 335-346.
Jackson SP, Bartek J. The DNA-damage response in human biology and disease. Nature 2009; 461: 1071-1078.
Chang HHY, Pannunzio NR, Adachi N, Lieber MR. Non-homologous DNA end joining and alternative pathways to double-strand break repair. Nat Rev Mol Cell Biol 2017; 18: 495-506.
Hustedt N, Durocher D. The control of DNA repair by the cell cycle. Nat Cell Biol 2016; 19: 1-9.
Yu X, Chen J. DNA damage-induced cell cycle checkpoint control requires CtIP, a phosphorylation-dependent binding partner of BRCA1 C-terminal domains. Molecular and cellular biology 2004; 24: 9478-9486.
Yue X, Bai C, Xie D, Ma T, Zhou PK. DNA-PKcs: A Multi-Faceted Player in DNA Damage Response. Front Genet 2020; 11: 607428.
Yun MH, Hiom K. CtIP-BRCA1 modulates the choice of DNA double-strand-break repair pathway throughout the cell cycle. Nature 2009; 459: 460-463.
Waga S, Hannon GJ, Beach D, Stillman B. The p21 inhibitor of cyclin-dependent kinases controls DNA replication by interaction with PCNA. Nature 1994; 369: 574-578.
Coats S, Flanagan WM, Nourse J, Roberts JM. Requirement of p27Kip1 for restriction point control of the fibroblast cell cycle. Science 1996; 272: 877-880.
Sekiya S, Suzuki A. Intrahepatic cholangiocarcinoma can arise from Notch-mediated conversion of hepatocytes. J Clin Invest 2012; 122: 3914-3918.
Cigliano A, Wang J, Chen X, Calvisi DF. Role of the Notch signaling in cholangiocarcinoma. Expert Opin Ther Targets 2017; 21: 471-483.
Akhoondi S, Sun D, von der Lehr N, Apostolidou S, Klotz K, Maljukova A et al. FBXW7/hCDC4 is a general tumor suppressor in human cancer. Cancer Res 2007; 67: 9006-9012.
Davis RJ, Welcker M, Clurman BE. Tumor suppression by the Fbw7 ubiquitin ligase: mechanisms and opportunities. Cancer cell 2014; 26: 455-464.
Kossatz U, Dietrich N, Zender L, Buer J, Manns MP, Malek NP. Skp2-dependent degradation of p27kip1 is essential for cell cycle progression. Genes & development 2004; 18: 2602-2607.
Ishii N, Araki K, Yokobori T, Watanabe A, Tsukagoshi M, Kubo N et al. Poor prognosis in cholangiocarcinoma patients with low FBXW7 expression is improved by chemotherapy. Oncol Lett 2017; 13: 3653-3661.
Farmer H, McCabe N, Lord CJ, Tutt AN, Johnson DA, Richardson TB et al. Targeting the DNA repair defect in BRCA mutant cells as a therapeutic strategy. Nature 2005; 434: 917-921.
Bryant HE, Schultz N, Thomas HD, Parker KM, Flower D, Lopez E et al. Specific killing of BRCA2-deficient tumours with inhibitors of poly(ADP-ribose) polymerase. Nature 2005; 434: 913-917.
Lee M, Pellegata NS. Multiple endocrine neoplasia type 4. Front Horm Res 2013; 41: 63-78.
Dietrich S, Hüllein J, Lee SC, Hutter B, Gonzalez D, Jayne S et al. Recurrent CDKN1B (p27) mutations in hairy cell leukemia. Blood 2015; 126: 1005-1008.
Francis JM, Kiezun A, Ramos AH, Serra S, Pedamallu CS, Qian ZR et al. Somatic mutation of CDKN1B in small intestine neuroendocrine tumors. Nat Genet 2013; 45: 1483-1486.
Cleary JM, Aguirre AJ, Shapiro GI, D'Andrea AD. Biomarker-Guided Development of DNA Repair Inhibitors. Mol Cell 2020; 78: 1070-1085.

There is NO conflict of interest to disclose.

Download PDF

Unknown event
21 May, 2024
Editorial decision: Reject after peer review
12 Sep, 2023
Review #2 received at journal
17 Aug, 2023
Review #1 received at journal
22 Jul, 2023
Reviewer #2 agreed at journal
17 Jul, 2023
Reviewer #1 agreed at journal
12 Jul, 2023
Reviewers invited by journal
11 May, 2023
Submission checks completed at journal
04 May, 2023
First submitted to journal
04 May, 2023
Unknown event
04 May, 2023
Editor assigned by journal
03 May, 2023

You are reading this latest preprint version

Loss of the cyclin kinase inhibitor p27kip1 opens a therapeutic window by deregulating pathway choice after DNA double strand breaks

Status:

Version 1

Abstract

Figures

Background

Methods

Results

Loss of p27 accelerates CCA formation and predisposes tumor cells to DNA damage

Loss of p27 promotes an aberrant homologous recombination-dependent DNA damage response during G1

Loss of p27 sensitizes tumor cells against drugs that target the DDR response

Discussion

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1