Digital sequence analysis of dna methylation

#The present invention relates to methods and compositions for determination of and uses of specific methylation patterns indicative of adenoma and carcinoma. In particular, the invention relates to analysis of defined CpG loci that are coordinately methylated in DNAs from cancer and adenoma samples, methods for identifying coordinately methylated loci, and methods of using analysis of coordinately methylated loci in one or more marker regions in the design of assays for adenoma and cancer having improved sensitivity and specificity.

#In higher order eukaryotes, DNA may be methylated at cytosines located 5′ to guanosine in CpG dinucleotides. This modification has important regulatory effects on gene expression, especially when involving CpG rich areas, known as CpG islands, often found in the promoter regions of genes. While approximately 75% of the CpG sites throughout the human genome are methylated, CpG sites within CpG islands are normally unmethylated, and aberrant methylation of CpG islands has been associated with certain diseases, including cancers. For example, CpG island hypermethylation is associated with transcriptional inactivation of defined tumor suppressor genes in human cancers, e.g., colorectal cancer. Therefore, detection of hypermethylated nucleic acid could indicate susceptibility or onset of various forms of cancers.

#Despite indications suggesting a link between CpG island methylator phenotype (CIMP) and cancers (see, e.g., Baylin S B, et al., Adv Cancer Res 1998; 72:141-196 and Jones P A, et al., Nat Rev Genet 2002; 3:415-428), the idea that analysis of methylation status alone could be a useful diagnostic or prognostic tool has been controversial. As discussed by Issa, et al. in an editorial in Gastroenterology 179(3):2005, researchers had mixed results in confirming the link between CI,MP and cancers. Although CIMP was reportedly demonstrated in multiple other malignancies (Shen, I., et al. J Natl Cancer Inst 2002; 94:755-761; Garcia-Manero G, et al., Clin Cancer Res 2002; 8:2217-2224; Toyota M, et al., Blood 2001; 97:2823-2829; Ueki T, et al., Cancer Res 2000; 60:1835-1839; Toyota M, et al., Cancer Res 1999; 59:5438-5442; Strathdee G, et al., Am J Pathol 2001; 158:1121-1127; Abe M, et al., Cancer Res 2005; 65:828-834) and several groups confirmed the original findings using similar markers and technology (Whitehall V L, et al., Cancer Res 2002; 62:6011-6014; van Rijnsoever M, et al., Gut 2002; 51:797-802) other groups were not able to establish such links (Eads C A, et al., Cancer Res 2001; 61:3410-3418; Esteller M, et al., Cancer Res 2000; 60:129-133). As late as 2003, a publication concluded that all methylation events in colorectal cancer were related to aging rather than neoplasia (Yamashita K, et al., Cancer Cell 2003; 4:121-131).

#The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawings will be provided by the Office upon request and payment of the necessary fee.

#FIGS. 1A and 1B provide sequence and CpG information for exemplary marker regions used in the present analysis. For each target gene, the native sequence of the region is shown in the top line. Unmethylated C-residues that would be converted by bisulfite and amplification to Ts are shown as T residues. Candidate methylation positions are shown boxed. Reference numbering for base and CpG positions is shown above each native sequence. Primer locations for amplification are shown as a row of underlined base positions.

#FIGS. 2 A-I provide tables showing the analyses of normal, adenoma, and cancer samples in which average methylation across all of the CpG loci indicated in Table 2 were calculated for each marker in each sample. For the normal samples in FIG. 2A, the average, standard deviation and the mean plus 2 or 3 standard deviations for each marker are indicated. For the adenoma and cancer samples, shaded cells in FIGS. 2B and 2C indicate a positive result, reflected as an average methylation value for that marker that is greater than the mean methylation+3 standard deviations determined for that marker in the normal samples.

#Embodiments of the invention are described in this summary, and in the Summary of the Invention, above, which is incorporated here by reference. Although the invention has been described in connection with specific embodiments, it should be understood that the invention as claimed should not be unduly limited to such specific embodiments.

#The present invention relates to methods and compositions for determination of, and uses of, specific methylation patterns indicative of adenoma and carcinoma. In particular, the invention relates to analysis of defined subsets of CpG loci that are coordinately methylated in DNAs from cancer and adenoma samples, methods for identifying coordinately methylated loci, and methods of using analysis of coordinately methylated loci in one or more markers or regions in the design of assays for adenoma and cancer having improved sensitivity and specificity.

#The present invention relates to the observation that, within marker nucleic acids for which methylation status is indicative of cellular status, e.g., cancerous, pre-cancerous, normal, etc., a subset of the individual methylation loci, e.g., CpG loci, in non-normal cells generally displays a greater degree of methylation relative to the background levels of methylation observed at corresponding loci in normal cells, while other methylation loci in the non-normal cells may exhibit levels of methylation that are closer to background levels. In some embodiments, the degree of methylation observed for a particular locus in plurality of cancerous or pre-cancerous cells relative to normal cells is expressed as a methylation ratio.