Expression Profiling of Graves' Disease Lymphocytes

Dr Simon Pearce, University of Newcastle



Data for eight microarray experiments have been obtained, with lymphocyte samples from four men with "cured" Graves' disease and four age-matched controls. In these studies we have looked at expression levels of 12,500 different genes in the lymphocytes of the Graves' disease people and compared the results to normals.

The first five of these array experiments have been subject to preliminary analysis using proprietary 'Affymetrix' software and this shows that there are 197 genes that have significant (p<0.05) differential expression between lymphocytes from Graves' patients when compared to control subjects. Of these 197 genes, 13 map into chromosomal regions that are known to contain Graves' susceptibility genes from our family linkage studies. There are many interesting candidate genes that appear to be differentially expressed based on this preliminary analysis including the following: IL1OR (P=0.005) TGFßIIR (p=).008), CD4 (p=0.013), Bax (p=0.015) and p50 NF-?B (p=0.025).

We are currently conducting a more detailed analysis of the full data-set using both the Affymetrix software, and a more robust non-parametric scoring algorithm that has been implemented at the Department of Computing Science, Newcastle University, on our request. This will determine which of these genes to take forward into further studies to confirm the findings. We are currently looking to recruit a further 5 men with treated Graves' disease for our next set of experiments.


The blood lymphocytes (known as CD4 cells) are thought to have a key role in causing autoimmune disorders like Graves' disease and we thought that differences in the behaviour of these cells were likely to be important in directing the immune attack on the thyroid gland found in people with overactivity.

From our experiments, we have determined six different genes whose products appear to be present at significantly higher levels in the lymphocytes from Graves' disease people than from unaffected individuals. Only one gene was significantly lower in affected people than in unaffected ones. We are now taking a series of different technical approaches to confirming these microarray findings, including further detailed analysis of gene expression levels in a larger number of people ("real-time RT-PCR") and also we are looking back to the chromosomes to see whether there are DNA changes within any of these 7 genes that might explain the differences in the amount of gene product we found in the lymphocytes. This latter approach seems to have come up trumps for one gene, and further details will be revealed at the American Thyroid Association and the European Thyroid Association scientific meetings this autumn.

All in all, it is still early days for this study, but our initial experiments have provided tantalising results that we are working hard to try and confirm by independent means. I am once again very grateful to the BTF for getting this research area off the ground for us, and to the patients who kindly gave an armful of blood. I feel confident we will have more to report in due course.