XPG mRNA Expression Levels Modulate Prognosis in Resected Non-small-cell Lung Cancer in Conjunction With BRCA1 and ERCC1 Expression

Roberta Bartolucci; Jia Wei; Jose Javier Sanchez; Laia Perez-Roca; Imane Chaib; Francesco Puma; Raffaele Farabi; Pedro Mendez; Fausto Roila; Tatsuro Okamoto; Miquel Taron; Rafael Rosell


Clin Lung Cancer. 2009;10(1):47-52. 

In This Article


Several layers of evidence indicate that DNA double-strand breaks and DNA damage response are induced by ionizing radiation,[14] hypoxia,[15] DNA-damaging agents, or activated oncogenes.[16] Recently, it has been found that Bcr-Abl and other fusion tyrosine kinases stimulate single-strand annealing, with enhanced nuclear colocalization of RAD52 and ERCC1.[17] The overexpression of DNA repair genes in melanomas has also been associated with poor prognosis.[18,19]

DNA damage response is a global signaling network that senses different types of damage and coordinates a response that includes activation of transcription, cell cycle control, apoptosis, senescence, and DNA repair processes. At the core of the DNA damage signaling apparatus are a pair of related protein kinases: ataxia telangiectasia mutated (ATM) and ATM and Rad3-related (ATR), which are activated by DNA damage.[20] A large-scale proteomic analysis of proteins phosphorylated in response to DNA damage on consensus sites recognized by ATM and ATR identified > 900 regulated phosphorylation sites, encompassing > 700 proteins.[21] This set of proteins is highly interconnected, with a large number of protein modules and networks not previously linked to DNA damage response (DDR). A module that is central to DDR is BRCA1, which includes BRCA1-associated ring domain protein (BARD1), BRCA2, partner and localizer of BRCA2 (PALPB2), ATM, and E2F transcription factor 1. This module also contains RAD51 and XRCC3 (which are related to homologous recombination repair) and replication protein A and ERCC1 (belonging to the NER pathway).

In chemotherapy-naive patients with early-stage, resected NSCLC, BRCA1 mRNA expression was the only independent prognostic variable.[8] We performed real-time quantitative PCR in frozen lung cancer tissue specimens from 126 patients with early NSCLC who had undergone surgical resection and evaluated the association between survival and expression levels of 9 genes involved in DNA repair pathways and in invasion and metastasis. A strong intergene correlation was observed among expression levels of all 9 genes except nuclear factor of activated T cells (eg, among ERCC1, RRM1, and BRCA1). Along with disease stage (stage I vs. II vs. III), BRCA1 mRNA expression significantly correlated with OS (hazard ratio, 1.98; P = .02). When only patients with stage I disease were examined, median survival was significantly different according to expression levels of ERCC1, myeloid zinc finger 1 (MZF1), TWIST, and BRCA1. Our findings indicate that, although BRCA1 is closely related to ERCC1, RRM1, and other genes like MZF1, it stands out as the most significant prognostic marker of relapse. Patients whose tumors had high BRCA1 expression had significantly poorer survival and should be candidates for adjuvant chemotherapy. Intriguingly, in vitro studies have shown that BRCA1 can regulate differential sensitivity to different classes of chemotherapy agents.[22,23] The absence of BRCA1 results in high sensitivity to cisplatin, whereas its presence increases sensitivity to antimicrotubule agents. Therefore, it is plausible that patients with the highest expression levels would receive more benefit from antimicrotubule, non-platinum-based chemotherapy.

The fact that high levels of ERCC1 or RRM1 transcripts conferred a higher risk of relapse[8] provides further evidence for the role of the loss of let-7 in the upregulation of ERCC1 and RRM1, as well as BRCA1,[24] and for the upregulation of BRCA1 and RRM1 in the SV40 T/t antigen signature.[25] Paradoxically, contradictory findings,[26,27] leading to opposed strategies of customizing adjuvant chemotherapy, have reported that the lack of ERCC1 protein implies a higher risk of relapse and a greater sensitivity to cisplatin-based chemotherapy.[26] Nevertheless, the clinical evidence indicates that the overexpression of ERCC1, RRM1, and especially BRCA1 mRNA confers poor survival in early NSCLC. Against the current standard of cisplatin-based chemotherapy, non-cisplatin-based chemotherapy, including antimicrotubule drugs, might be the proper treatment for the majority of patients with a high risk of relapse.[8] Our findings indicate that a higher risk of relapse is related to high levels of several transcripts, including ERCC1. These patients could be resistant to cisplatin and sensitive to taxanes or other antimicrotubule drugs. No correlation between mRNA and protein levels of ERCC1 was found in cervical carcinoma cell lines.[28] Expression of the ERCC1 gene at the mRNA and protein levels was established by Northern and Western blotting, respectively. There was a significant correlation between ERCC1 mRNA levels and cisplatin resistance; however, there was no relationship between ERCC1 protein levels and cisplatin resistance.

In addition, high mRNA expression of the BRCA1-interacting protein Bach1/Brip1 has been found in aggressive breast cancers.[29] BRCC36 has also been shown to be present in the BRCA1-RAP80 complex[30] and is overexpressed in breast cancer, where it confers radioresistance (resistance to radiation therapy).[31] This highlights the possibility that BRCA1—or several interacting partners—can confer poor prognosis as well as resistance to cisplatin or other DNA-damaging agents.

The present study confirms the poor prognosis conferred by high BRCA1 mRNA levels and shows, for the first time, that low XPG mRNA levels can strongly increase the risk of relapse. For patients with low BRCA1, regardless of XPG mRNA expression levels, DFS was not reached. However, for patients with intermediate/high BRCA1 and low/intermediate XPG, DFS decreased to only 16.3 months (P = .002). Similar differences were observed in OS. There is no facile explanation for this phenomenon. XPG makes the incision on the 3' side of the DNA lesion in the later steps of NER.[10] XPG also plays a role in the assembly of NER complexes because its presence is required for the 5' incision by ERCC1-XPF. Quite a strong physical interaction has been demonstrated between XPG and transcription factor II H (TFIIH), and impairment of this interaction destabilizes the TFIIH complex, dissociating XPD and cyclin-dependent kinase (CDK)-activating protein kinase subcomplex (containing CDK7, cyclin H, and MAT1 subunits). Intriguingly, CDK8 can phosphorylate the cyclin H subunit of mammalian TFIIH, blocking TFIIH activity in transcription.[32] Linking CDK8 to K-ras mutations is the fact that Ras/mitogen-activated protein kinase-dependent CAAT/enhancer-binding protein β phosphorylation could serve as a switch that releases CDK8 from Mediator.[33] CDK8 behaves like an oncogene and was described in acute lymphoblastic leukemia, and recently, it has been reported that CDK8 activates β-catenin activity in colorectal cancer.[34] Therefore, we can speculate that CDK8 can be overexpressed in NSCLCs with K-ras mutations, and this subset of NSCLCs could have β-catenin hyperactivity and increased mammalian target of rapamycin mRNA levels. These data, together with our clinical observations, lead us to speculate that overexpression of CDK8, by hampering TFIIH function, can downregulate XPG and explain the contribution of low XPG expression to the poor prognosis observed in the subgroup of patients with high BRCA1 and low XPG.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.