Interdisciplinary Training

Robert Kimmel

Gene Amplification and Expression in Thyroid Cancer

We study genomic instability in thyroid cancer. Gain, loss and rearrangement of genomic DNA are major characteristics of cancer, often providing a basis for the altered pattern of gene expression that drives oncogenesis. The least studied aspect of genomic instability in thyroid cancer is the gain of gene copy-number called gene amplification. Since double-stranded DNA breakage by ionizing radiation might enhance gene amplification, we investigated whether the post-Chernobyl papillary thyroid carcinoma (PTC) cohort exhibits specific patterns of gene amplification that can be distinguished from patterns seen in PTC and other thyroid tumors not associated with radiation exposure.

Using a genome-wide, cDNA microarray-based technique, we analyzed genomic DNA from a pilot series of ten post-Chernobyl PTC cases and from several spontaneous thyroid tumors of various types. For reference we analyzed gDNA from normal individuals and from the promyelocytic leukemia cell line, HL60, which harbors a known c-myc amplicon. Changes in copy-number were detected as differences in fluorescence between normal and test genomic DNA that had been random primer labeled with 5-amino-propargyl-2'-deoxyuridine 5'-triphosphate coupled to either Cy3 or Cy5 fluorescent dyes (Cy3-dUTP or Cy5-dUTP), respectively, and co-hybridized to an 18,000 spotted human cDNA glass slide microarray. Fluorescence ratios, obtained after scanning these arrays, were plotted by their radiation hybrid map locations, thus enabling direct visualization of apparent gene amplifications and deletions. Our results showed some evidence of change in DNA copy number in all tumors. However, post-Chernobyl PTC cases exhibited more frequent apparent amplification events than did cases not associated with radiation exposure.

Regression analysis of fluorescence ratios yielded Z-scores and p-values for the most consistent changes in copy-number across the ten post-Chernobyl PTC tumors. Two percent were amplification candidates (p<0.05), while about 1.5% were deletion candidates. Most candidate amplifications occurred as isolated markers. Sixty of over 7000 mapped candidate-amplified genes occurred in 24 clusters over 14 chromosomes, averaging 2.5 (range 2-4) genes/cluster. The clusters averaged 1.4 Mb (range 56-5,520 kb) in length. There was consistent amplification of chorionic gonadotropin-beta (p<0.002), adrenomedullin (p<0.014) and thyroid transcription factor 1 (p<0.008). Numerous other single candidate amplified genes were detected. Amplification of four genes has been confirmed by differential PCR. We concluded that several chromosomal regions may harbor amplicons common to many post-Chernobyl papillary carcinomas and therefore may contain genes of pivotal importance in the oncogenesis of post-Chernobyl thyroid cancer.

We are currently setting up tissue microarray (TMA) resources for spontaneous adult and pediatric thyroid cancers and for post-Chernobyl thyroid cancers. Tissue microarrays provide a potentially useful means of creating thyroid cancer resources from scarce tumor specimens. We will test TMAs with fluorescent labeled probes and antibodies to determine whether genes amplified in thyroid cancer are also over-expressed and/or rearranged. The over-expression of amplified genes might provide a link between changes in gene copy-number and potential effects on oncogenesis, thus providing a basis for the influence of genomic instability in this type of cancer. Evaluation of the gene products of these amplified genes in human tumors is the next critical step to assess their real clinical value. The construction of TMA from well-characterized human tumors is one way to efficiently evaluate the large number of new tumor markers that will be identified through array technologies, and to rapidly translate molecular discoveries into clinical applications.