UTX, a Histone Demethylase, Is Inactivated through Homozygous Deletion, Mutation, and DNA Methylation in Multiple Myeloma.

Abstract 1798

Poster Board I-824

Histone modifications are known to mediate transcriptional regulation through changes in chromatin condensation and as such can lead to aberrant transcriptional patterns resulting in malignant transformation. Modulation of chromatin structure via histone modification is becoming recognised as an important pathogenic mechanism in myeloma and has been suggested by the over-expression of MMSET, a histone methyltransferase, by the t(4;14) chromosomal rearrangement. More recently inactivation of UTX, a histone demethylase, has also been suggested to have a role in myeloma pathogenesis and both UTX and MMSET are mediators of transcriptional repression. UTX is inactivated in a number of different cancer cell lines but importantly, mutations and deletions have been detected in myeloma cell lines and we wished to follow up on this observation in uniformly treated clinical cases. UTX is a large gene found on the X chromosome covering 240 kb of genomic DNA and consists of 29 exons encoding a protein with both JmjC-domains and tricopeptide repeats responsible for histone demethylation and polycomb protein interactions. Inactivation of UTX occurs through deletions of individual exons through to large whole gene deletions as well as by mutations scattered throughout the 29 exons. A further mechanism of UTX inactivation which has not been looked for to date is via DNA methylation of the CpG island upstream of the transcriptional start site. We set out to determine the status of UTX in our dataset which includes expression, mapping, and methylation array data from presenting myeloma samples entered into the MRC Myeloma IX clinical trial. The gene expression of UTX was measured on 272 samples using Affymetrix U133 Plus 2.0 arrays and showed that 80% of samples do not express UTX transcripts but using expression quartile analysis we could not detect an effect on overall survival. The mechanism underlying the abrogation of expression was investigated further using the Affymetrix 500K SNP mapping array on a subset of 114 samples to detect copy number alterations. UTX was hemizygously deleted in 21 (42%) female samples and was completely deleted in 1 male sample, at the resolution of the arrays. In order to determine if individual exons were deleted, at a resolution below that detectable by mapping arrays, we performed quantitative PCR coupled with high resolution melting (HRM) analysis using the Rotor-gene Q real-time cycler (Qiagen). Exons were amplified, over 40 cycles, to obtain products of ∼200 bp using LC Green Plus mastermix (Idaho Technologies) in a 10 μl reaction on the Rotor-gene Q with a final HRM step from 72-95 °C with increments of 0.1 °C. Amplification plots combined with the HRM step allows us to identify both homozygous deletions and mutations within the exons. We screened all 114 samples for micro-deletions and mutations and found homozygous deletions in ∼7% of samples and identified a significant proportion of mutations using the HRM method which accounted for a total of ∼10% of gene inactivation. In order to determine if methylation could be responsible for inactivation of the remaining allele we used the Illumina Infinium humanmethylation27 array to study the methylation status at the UTX locus. This array interrogates 27,578 highly informative CpG sites per sample at the single-nucleotide resolution using bisulfite converted DNA. The results of this analysis are presented as an average beta-score where 1.0 is fully methylated and 0 is fully unmethylated. Samples were analyzed using Illumina GenomeStudio and the custom differential methylation algorithm. In samples with a diploid copy number of UTX the methylation signals covered 2 ranges: hemi-methylated (0.35-0.55, n=7) and hyper-methylated (0.73-0.89, n=14). In samples with 1 copy of UTX, which includes all males, there were 3 ranges: hypomethylated (0.08-0.21, n=5), hemi-methylated (0.35-0.51, n=3), and hypermethylated (0.66-0.88, n=48). All of the hypomethylated samples with a single copy of UTX were male, and at least 1 of these samples contained an inactivating exonic deletion resulting in complete loss of function. These data indicate that methylation of the residual allele contributes significantly to the inactivation of UTX along with interstitial deletions and mutations. We will go on to present data on the interaction of UTX with variation at the UTY locus and how this modulates behaviour of the myeloma clone.