Faculty Bibliography

Cancer cells progress through the accumulation of genetic alterations. Familial adenomatous polyposis (FAP) tumors provide an excellent model to unravel the molecular steps underlying malignant transformation. Global genomic damage was assessed in 56 adenomas and 3 carcinomas from six FAP patients and compared with that of sporadic adenomas and carcinomas. Evolutive trees were traced after application of maximum likelihood clustering and split decomposition methods to the analysis of comprehensive genetic profiles generated by diverse molecular approaches: arbitrarily primed PCR, comparative genomic hybridization, and flow cytometry. Overall, genomic damage as assessed by arbitrarily primed PCR was lower in familial adenomas than in sporadic adenomas and carcinomas. Comparative genomic hybridization data also show a low number of alterations in the majority of FAP adenomas. Tumors of the same patient were likely to share specific genetic alterations and may be grouped into one or two clusters. Putative common pathways were also identified, which included tumors of up to three different patients. According to our data, FAP tumors accumulate specific genetic alterations and in a preferred order that is characteristic of each individual. Moreover, the particular genetic background and environmental conditions of a FAP patient restrain the molecular evolution portrait of synchronous tumors.

Hypermethylation associated silencing of the CpG islands of tumor suppressor genes is a common hallmark of human cancer. Here we report a functional search for hypermethylated CpG islands using the colorectal cancer cell line HCT-116, in which two major DNA methyltransferases, DNMT1 and DNMT3b, have been genetically disrupted (DKO cells). Using two molecular screenings for differentially methylated loci [differential methylation hybridization (DMH) and amplification of inter-methylated sites (AIMS)], we found that DKO cells, but not the single DNMT1 or DNMT3b knockouts, have a massive loss of hypermethylated CpG islands that induces the re-activation of the contiguous genes. We have characterized a substantial number of these CpG island associated genes with potentially important roles in tumorigenesis, such as the cadherin member FAT, or the homeobox genes LMX-1 and DUX-4. For other genes whose role in transformation has not been characterized, such as the calcium channel alpha1I or the thromboxane A2 receptor, their re-introduction in DKO cells inhibited colony formation. Thus, our results demonstrate the role of DNMT1 and DNMT3b in CpG island methylation associated silencing and the usefulness of genetic disruption strategies in searching for new hypermethylated loci.

The causes and functional consequences of E-cadherin (E-CD) loss in breast cancer are poorly understood. E-CD loss might act in concert with alterations in the APC/beta-catenin pathway to permit oncogenic beta-catenin signaling. To test this hypothesis, we have analyzed the presence of genetic and epigenetic alterations affecting E-CD (CDH1), APC and beta-catenin (CTNNB1) genes and the immunohistochemical expression of E-CD, beta- and gamma-catenin in a series of 46 infiltrating lobular breast carcinomas (ILCs). Since 80% of ILCs featured complete loss of E-CD expression, we analyzed the molecular alterations responsible for E-CD inactivation in these tumors. We found that 10 of 46 (22%) cases harbored mutations in CDH1, including 1 case with 2 different mutations (1 of which was germline). CDH1 was also inactivated by loss of heterozygosity (LOH; 30/41, 73%) and promoter hypermethylation (19/46, 41%). Interestingly, LOH and mutations were also detected in the corresponding in situ lesions of the ILCs, implying that these alterations are early events in lobular cancer tumorogenesis. Additionally, the presence of a polymorphism in the CDH1 promoter was found to be inversely correlated with CDH1 mutations, but not with E-CD levels. We next examined whether alterations in the APC/beta-catenin pathway also occurred in the same series of ILCs. Although no CTNNB1 or APC mutations were detected, promoter methylation (25/46, 52%) and LOH (7/30, 23%) of APC were found. Moreover, methylation of APC and CDH1 occurred concordantly. However, beta- and gamma-catenin were severely reduced or absent in 90% of these tumors, implying that alterations in CDH1 and APC genes do not promote beta-catenin accumulation in ILC. These molecular alterations were not associated with microsatellite instability. In summary, several different mechanisms (mutations, LOH, methylation) are involved in the frequent CDH1 inactivation in invasive and in situ lobular breast cancer. The same tumors also show genetic and epigenetic alterations of APC gene. However, altered CDH1 and APC genes do not promote beta-catenin accumulation in this tumor type.

Methylation-associated silencing of tumor suppressor genes is recognized as being a molecular hallmark of human cancer. Unlike genetic alterations, changes in DNA methylation are potentially reversible. This possibility has attracted considerable attention from a therapeutics standpoint. Nucleoside-analogue inhibitors of DNA methyltransferases, such as 5-aza-2'-deoxycytidine, are able to demethylate DNA and restore silenced gene expression. Unfortunately, the clinical utility of these compounds has not yet been fully realized, mainly because of their side effects. A few non-nucleoside inhibitors of DNA methyltransferases have been reported, including the anti-arrhythmia drug procainamide. Following this need to find new demethylating agents, we have tested the potential use of procaine, an anesthetic drug related to procainamide. Using the MCF-7 breast cancer cell line, we have found that procaine is a DNA-demethylating agent that produces a 40% reduction in 5-methylcytosine DNA content as determined by high-performance capillary electrophoresis or total DNA enzyme digestion. Procaine can also demethylate densely hypermethylated CpG islands, such as those located in the promoter region of the RAR beta 2 gene, restoring gene expression of epigenetically silenced genes. This property may be explained by our finding that procaine binds to CpG-enriched DNA. Finally, procaine also has growth-inhibitory effects in these cancer cells, causing mitotic arrest. Thus, procaine is a promising candidate agent for future cancer therapies based on epigenetics.

Aberrant methylation of 5'CpG islands is a key epigenetic event in many human cancers. A PCR-based technique of methylated CpG island amplification followed by representational difference analysis was used to identify genes methylated in cancer. Two of the CpG islands identified mapped to the 5' untranslated region of the PAX5 alpha and beta genes. These genes, located on chromosome 9p13, are transcribed from two distinct promoters and form two alternative first exons that are subsequently spliced to the common exons 2-10. The resulting splice variants encode two distinct transcription factors important in cell differentiation and embryonic development. Examination of the methylation status of each gene using methylation-specific PCR revealed that both genes are methylated in approximately 65% of breast and lung tumors. Bisulfite sequencing revealed dense methylation patterns within each 5'CpG island, strongly correlating with transcriptional silencing. Expression in cell lines with dense methylation of either the PAX5 alpha or beta promoter region was restored after treatment with the demethylating agent 5-Aza-2'-deoxycytidine. The PAX5 beta gene encodes for the transcription factor B cell-specific activating protein that, in turn, directly regulates CD19, a gene shown to negatively control cell growth. A strong association was observed between PAX5 beta methylation and loss of expression of the CD19 gene demonstrating that inactivation of the PAX5 beta gene likely contributes to neoplastic development by inhibiting growth regulation through effects on CD19 gene expression. Recent studies have demonstrated the importance of PAX5 gene alterations in human cancer. Our results are the first to identify aberrant promoter methylation as a common mechanism for dysregulation of these genes in solid tumors.

A novel, rapid and simple capillary electrophoretic mobility shift assay (CEMSA) with laser-induced fluorescence (LIF) has been developed for the quantitative study of protein-DNA interactions. This method is particularly useful for the study of basic proteins, the most common of the DNA-interacting proteins. To avoid protein stickiness to the capillary walls we have introduced the use of neutral polyacrylamide that requires the use of reverse polarity. Under these conditions, excellent separation of DNA and protein-DNA complexes was obtained without the requirement of a gel matrix, thereby allowing the easy and reliable quantification of protein-DNA affinities. Analysis of the affinities of histones H2B and H4 for a synthetic oligo have been used to demonstrate the reproducibility and accuracy of this method. We have observed that H4 has a higher affinity for DNA than H2B, with half saturation fractions lying in the micromolar range.

Many genetic and environmental factors contribute to development of cancer, but DNA methylation may provide a link between these influences. Genome stability and normal gene expression are largely maintained by a fixed and predetermined pattern of DNA methylation. In cancer, this idealistic scenario is disrupted by an interesting phenomenon: the hypermethylation of regulatory regions called CpG islands in some tumour suppressor genes--eg, BRCA1, hMLH1, p16INK4a, APC, VHL--which causes their inactivation. Development of new techniques that couple bisulphite modification with PCR has enabled these alterations to be studied in all types of biological fluids and archived tissues. Potentially, there are four types of translational studies that can be used to investigate the aberrant pattern of DNA methylation in cancer. First, CpG island hypermethylation can be used as a marker to identify cancer cells from biological samples, eg, serum and urine. This technique is highly sensitive and informative because profiles of tumour-suppressor-gene inactivation are specific to particular cancers. Second, single and combined genes that are inactivated by promoter hypermethylation, such as p16INK4a and DAPK, can be used as prognostic factors. Third, products of genes that are silenced by DNA methylation can be used as biomarkers of response to chemotherapy or hormone therapy--eg, the DNA repair O6-methylguanine-DNA methyltransferase and the oestrogen receptor. Finally, dormant tumour suppressor genes can be reactivated by DNA demethylating drugs, with the aim of reversing the neoplastic phenotype. These are new avenues worth exploring in the fight against cancer.

The suppressor of cytokine signaling (SOCS) family of proteins has been implicated in the negative regulation of several cytokine pathways, particularly the receptor-associated tyrosine kinase/signal transducer and activator of transcription (Jak/STAT) pathways of transcriptional activation. SOCS-1 (also known as JAB and SSI-1) inhibits signaling by many cytokines. Because of the previously observed hypermethylation-associated inactivation of SOCS-1 in hepatocellular carcinoma and the critical role of interleukin-6 (IL-6) as a survival factor in multiple myeloma (MM), we examined CpG island methylation of the SOCS-1 gene in MM cell lines and primary MM samples. Aberrant SOCS-1 methylation was found in the IL-6-dependent MM cell lines U266 and XG1, which correlated with transcriptional silencing. Treatment of these cell lines with the demethylating agent 5-aza-2'-deoxycytidine (DAC) up-regulated SOCS-1 expression. Methylation-associated inactivation of SOCS-1 in hematopoietic cell lines correlated with greater sensitivity to the chemical JAK inhibitor AG490. Using methylation-specific polymerase chain reaction (MSP), we found that SOCS-1 is hypermethylated in 62.9% (23/35) of MM patient samples. In contrast, methylation analysis of malignant lymphomas of various histologies revealed SOCS-1 hypermethylation in only 3.2% (2/62), and there was no methylation of SOCS-1 in normal peripheral blood leukocytes or bone marrow cells. We conclude that SOCS-1 is frequently inactivated by hypermethylation in MM patients. Silencing of the SOCS-1 gene may impair negative regulation of the Jak/STAT pathway and therefore result in greater responsiveness to cytokines, thus supporting survival and expansion of MM cells.

Cultures of human mammary epithelial cells (HMECs) contain a subpopulation of variant cells with the capacity to propagate beyond an in vitro proliferation barrier. These variant HMECs, which contain hypermethylated and silenced p16(INK4a) (p16) promoters, eventually accumulate multiple chromosomal changes, many of which are similar to those detected in premalignant and malignant lesions of breast cancer. To determine the origin of these variant HMECs in culture, we used Luria-Delbrück fluctuation analysis and found that variant HMECs exist within the population before the proliferation barrier, thereby raising the possibility that variant HMECs exist in vivo before cultivation. To test this hypothesis, we examined mammary tissue from normal women for evidence of p16 promoter hypermethylation. Here we show that epithelial cells with methylation of p16 promoter sequences occur in focal patches of histologically normal mammary tissue of a substantial fraction of healthy, cancer-free women.

Abnormal expression of cadherins and catenins plays a critical role in the initiation and progression of multiple human tumours. This study aimed to evaluate the immunoreactivity of E- and P-cadherin, beta- and gamma-catenin, and p120ctn in premalignant and malignant endometrial lesions and to correlate their membranous expression with clinicopathological features. In addition, we examined whether or not LOH and promoter hypermethylation of the CDH1 gene were associated with E-cadherin expression and clinicopathological variables. Finally, we studied the frequency of beta-catenin mutations in premalignant endometrial lesions. Immunohistochemical staining was performed in 21 atypical endometrial hyperplasias (AEHs), 95 endometrioid carcinomas (EECs), and 33 non-endometrioid carcinomas (NEECs). Reduced E-cadherin expression was observed in 57.8% of the cases, being more frequent in NEECs (87.1%, p = 0.001) and carcinomas of more advanced stage (85.7% of stage III-IV carcinomas, p = 0.01). LOH of CDH1 gene was found in 57.1% of NEECs but only in 22.5% of EECs (p = 0.011) and showed a trend towards association with reduced E-cadherin expression (p = 0.089). CDH1 promoter hypermethylation was found in 21.2% of endometrial carcinomas but was not associated with clinicopathological or immunohistochemical variables. Reduced expression of beta- and gamma-catenin and p120ctn was found in 76.1%, 94.3%, and 63.6% of the cases, respectively, being more frequent in lesions with reduced E-cadherin expression. In addition, beta-catenin, but not gamma-catenin or p120ctn expression, was associated with the histology of the lesion, since it was reduced in 35% of AEHs, 80.3% of EECs, and 96.9% of NEECs (p = 0.000). Mutations in exon 3 of the beta-catenin gene, associated with beta-catenin nuclear expression, were detected in 3 (14.0%) AEH, a frequency similar to that previously reported in this series of ECs. Finally, upregulation of P-cadherin was observed in 28.6% of cases. This alteration was associated with the histology of the lesion, since it was found in 9.5% of AEHs, 27.7% of EECs, and 46.2% of NEECs (p = 0.021).