0 citations0 references

CREBBP mutation in human cutaneous squamous cell carcinoma

Experimental Dermatology2016Vol. 25(8), pp. 650–651

Citations Over Time

Stephen A. Watt, Karin J. Purdie, Nicoline Y. den Breems, Michelle Dimon, Stefan F. Tucker, Sarah T. Arron, Angela McHugh, Dylan J. Xue, Jasbani H.S. Dayal, Catherine Harwood, Irene M. Leigh, Charlotte M. Proby, Andrew P. South

Abstract

Cutaneous squamous cell carcinoma (cSCC) is the second most frequent skin cancer, with an estimated 400 000 new cases reported yearly in the United States 1. Although metastasis is relatively uncommon, prognosis in patients with lymph node-positive cSCC is poor and 5-year survival rates are low 2. Furthermore, advanced and often inoperable, locoregional disease is frequent with cSCC and such patients would benefit from effective, targeted chemotherapies. The hunt for genetic drivers and thus potential targets for cSCC therapy is hampered by a mutation burden greater than any other known cancer with metastatic potential 3-5, only exceeded by basal cell carcinoma of the skin (S1). Deep sequencing of cSCC has identified key driver mutations in tumor suppressor genes 3-5, which are difficult to target therapeutically as it is easier to inhibit the action of a protein than to replace its function when missing. A pathway-centric approach may be a more appropriate route for therapy development because somatic, cancer-promoting mutations may affect multiple nodes within the same pathway leading to downstream activation 6. We recently identified frequent CARD11 mutations in cSCC, which lead to aberrant NFkB signalling 7. Nuclear factor kappa-b (NFκB) signalling plays a crucial role regulating proliferation and differentiation in the epidermis, and alterations can lead to skin pathologies with significant burden to human health such as psoriasis and cSCC. One gene reported as integral to a number of ubiquitous signalling pathways (including NFκB) is CREBBP which encodes the protein CBP, a transcriptional cofactor with lysine acetyltransferase activity and close homology to p300 (S2). CREBBP loss-of-function (LOF) mutations are frequently identified in a range of different cancers such as lymphomas and carcinomas of the bladder and oesophagus (S3–S5). We previously identified CREBBP mutation in 35% of 20 cSCC using exome sequencing 4. Here, we address whether CREBBP mutation is present in a larger cohort of cSCC samples (n = 91). Please see supporting information (Data S1). To define CREBBP mutation in a larger cohort of human cSCC, we used targeted resequencing of 91 samples previously isolated and used to resequence potential driver genes such as NOTCH1, TP53 and CARD11, as described 4, 7. CREBBP mutation frequency was high (33/91 cSCC tumors, 36%, compared with 7/20, 35%, in the original exome sequencing cohort). Thus, CREBBP is frequently mutated in human cSCC (mutations listed in Table S1). We next compared the spectrum of CREBBP mutation in cSCC with those reported for other tumors in the catalogue of somatic mutations in cancer (COSMIC) (S6) and identified that 50/51 cSCC mutations were missense, in contrast to those reported in COSMIC (n = 758) where >25% of mutations were either nonsense or insertion/deletion mutations leading to frame-shift and a premature termination codon (PTC) (Fig. 1a). The majority of such mutations are predicted to result in expression of a truncated protein that is either greatly reduced or absent as a result of nonsense-mediated decay (NMD). Analysing mutation spectrum by tissue type in COSMIC identified certain tumor groups with greater than 33% mutations leading to PTC such as the haematopoietic and lymphatic system as well as oesophagus. Bladder harboured 75% nonsense or insertion/deletion mutations (Fig. 1b). Two previous studies support our observations that loss-of-function mutation in CREBBP is not prevalent in primary cSCC; Durinck and colleagues identified one missense mutation from eight cSCC 3, while Pickering and colleagues identified 24 mutations in a cohort of 39 aggressive, primary cSCC, 21 of which were missense or silent 5. Given the spectrum of mutation in cSCC, we speculated whether cSCC CREBBP mutation, unlike other tumors with a high CREBBP mutation frequency, does not result in CBP loss of function. To address this immunohistochemistry (IH) was used to interrogate the expression level and distribution of CREBBP in normal skin, and cSCC tumors. Using an antibody raised against the C-terminus of CBP nuclear reactivity was clearly demonstrated throughout basal and suprabasal layers of normal skin (Fig. 2a), and nuclear reactivity in all 21 cSCC samples tested was also observed. These data are in line with two different antibodies published at The Human Protein Atlas 8 (http://www.proteinatlas.org/accessed 17 November 2015). In our study, well-differentiated cSCC tumors (n = 13) exhibited a pattern of strong positive reactivity but loss in central keratinizing areas (Fig. 2b). On the other hand, in eight cSCC samples with poorly differentiated regions, there was a greater variability where regions of weak positivity intermixed with negative tumor cells were observed in three of eight high-grade tumors (Fig. 2c). CREBBP is regularly found to be inactivated through loss-of-function mutations in a range of malignancies (S3–S5). In contrast, we find cSCC harbours missense mutation, and very few mutations that predict loss of function. Furthermore, we identify CBP in all 21 cSCC examined, indicating that protein is expressed and localizes to the correct compartment. It is unclear at this time, however, what role CBP may have in cSCC, but it is interesting to note that CBP/p300 is critical for efficient removal of UV-induced pyrimidine dimers (S7). Furthermore, recent genomic analysis of 29 lymph node metastases from primary cSCC showed frequent (28%) CREBBP mutation 9, but here the majority of alterations (6/8) were truncating or insertion/deletion mutations predicted to lead to LOF. This result may indicate that CBP inactivation is associated with tumor progression in cSCC and agrees with our IH observation, and, also in a limited number of samples, that a degree of CBP loss is observed in high-grade tumors (Fig. 2c). However, these observations will need to be confirmed, and the data presented here together with the recent finding that, unlike NOTCH1, CREBBP mutation is not frequent or positively selected for in normal, sun-exposed skin 10 suggest that CREBBP mutation is unlikely to be an early event in squamous cell carcinogenesis. This work was funded by Cancer Research UK and The British Skin Foundation and the Department of Dermatology and Cutaneous Biology at Thomas Jefferson University. I.M.L is funded by the European Research Council and a strategic grant from the Wellcome Trust. A.P.S., C.M.P., I.M.L., C.A.H., K.J.P. and S.A.W. designed the research and obtained funding. C.A.H. and C.M.P. collected samples. A.P.S., K.J.P., S.A.W., A.M., D.X. and J.H.D. performed the experiments. A.P.S., S.A.W., N.d.B., M.D., S.T. and S.T.A. analysed the data. A.P.S., K.J.P., S.A.W., C.M.P., C.A.H. and I.M.L. wrote the paper. The authors have declared no conflicting interests. Data S1. Materials and Methods. Table S1. CREBBP Mutations. Data S2. Supplementary References. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Citations Over Time

Abstract

Related Papers