Reply:

Abstract

We appreciate the interest of Drs. Boyault and colleagues in the potential role of KLF6 in hepatocellular cancer (HCC). As we have previously outlined,1 somatic mutation detection in cancer can be highly variable owing to differences in methodology and sample sets. Indeed, reported mutational frequencies of even well-characterized tumor suppressor genes—for example, p53—can range from 13%-67% in HCC.2 Similarly, in prostate cancer, whereas one study failed to identify KLF6 mutations,3 a second one apart from our own4 identified loss of heterozygosity (LOH) in approximately 30% of samples and somatic mutations in 15%.5 In contrast to the negative findings suggested by Drs. Boyault and colleagues, a recent publication6 has identified KLF6 LOH and mutation in HCC at levels almost identical to those we originally reported.7 We would also emphasize that LOH of the KLF6 locus is present much more commonly than mutation in HCC.6, 7 This result parallels the published findings in other cancers, including prostate,4, 5 colorectal,8 and non–small cell lung cancer.9 This feature was not examined by Drs. Boyault and colleagues even though it may lead to a growth advantage through haploinsufficiency. Interestingly, we have also defined a number of tumors that demonstrate only significant degrees of LOH, without somatic mutation. These include ovarian cancer, head and neck squamous cell carcinoma, small cell lung cancer, and glioblastoma, as well as prostate cancer cell lines (manuscripts in preparation), suggesting that the presence of KLF6 mutations is by no means ubiquitous in human cancer. We agree that sequencing artifacts secondary to tissue preservation are an important concern, which is why we have always used similarly prepared and stored paired normal tissue specimens as controls.4, 7, 8 Emerging KLF6 studies are now characterizing functions and features consistent with its role as a tumor suppressor, including interactions with cell cycle regulators,10 suppression of tumorigenicity in vivo,11 induction of proto-oncogene degradation,12 regulation of apoptosis,9 and promoter methylation and/or downregulation in cancer,9, 13 with reduced gene expression levels predicting disease outcome in both pulmonary adenocarcinoma14 and prostate cancer.15 Taken together, we believe these findings highlight the importance of continued independent studies using multiple sample sets to firmly clarify the role and mutation frequency of KLF6, or any candidate tumor suppressor, in both HCC and other human cancers. We welcome interest in this topic and look forward to its continued exploration. Sigal Kremer-Tal M.D.*, Helen Reeves M.D., Ph.D. , Goutham Narla B.S.*, John Martignetti M.D., Ph.D.*, Scott Friedman M.D.*, * Mount Sinai School of Medicine, New York, NY, University of Newcastle upon Tyne, United Kingdom.