Abstract

In multiple myeloma (MM), inactivating mutations and deletions affecting the histone demethylase KDM6A locus are found in up to 10% of newly diagnosed patients and associated with poor prognosis. KDM6A (also named UTX, Ubiquitously transcribed Tetratricopeptide repeat, X chromosome) belongs to a family of Jumonji-C (Jmj-C)-containing demethylases that work as a scaffold for a multiprotein complex containing H3K4 specific methyltransferases KMT2D and/or KMT2C (MLL2/3), the histone acetyltransferase CBP/p300 and members of the SWI/SNF chromatin-remodeling complex. In a concerted manner this complex appears to add activation marks on histones and remove methylation of lysine 27 on histone H3 (H3K27me) associated with gene repression. Importantly, all these coregulators are found significantly mutated in MM and their function may converge into a tumor suppressive pathway. Our goal is to understand how KDM6A loss contributes to the development of MM.

We modeled the loss of KDM6A in MM cell lines using CRISPR-Cas9 ribonucleotide protein (RNP) technology. Mutant allele frequency over time post electroporation of RNP revealed a growth advantage of KDM6A mutant alleles. By 2 weeks of growth most of the cells in culture harbored KDM6A gene disruption and exhibit elimination of KDM6A protein confirming the tumor suppressive role of KDM6A in MM. We used these isogenic polyclonal edited cell lines with KDM6A wild type or mutated to identify KDM6A binding sites and enhancers affected by the loss of KDM6A. As well, we knock-in an HA tag on endogenous KDM6A to identify DNA regions occupied by KDM6A. To understand the importance of KDM6A demethylase activity in the tumor suppressive effect of KDM6A, we developed stable cell lines with a doxycycline-inducible form of KDM6A wild-type (WT) or lacking demethylase activity (JmjC-dead). We found that about 20% of the genes deregulated by re-expression of WT and jmjC-dead KDM6A overlap suggesting that demethylase activity is not essential for all KDM6A functions in MM.

Importantly, we confirmed the tumor suppressive role of KDM6A in a novel mouse model of MM in which KDM6A is deleted specifically in the B cell compartment. Briefly, we isolated CD19cre-/+ (control) or CD19cre-/+ Kdm6afl/fl fetal liver cells and transduced these cells with a constitutively activated form of the IL-6 coreceptor (L-GP130) that activates the JAK/STAT pathway. Mice transplanted with CD19cre-/+ Kdm6afl/fl fetal liver cells developed MM by 7 weeks post transplantation while mice transplanted with CD19cre+/- fetal liver cells did not developed MM by20 weeks. Necroscopy and flow cytometry analysis demonstrated infiltration of CD138+ cells in bone marrow, spleen, liver and kidney of mice that developed MM. In the future we will use this model to explore how loss of KDM6A affects chromatin structure in vivo and how it changes the characteristics of MM. These studies are expected to provide new insights that lead to the development of more effective MM therapies which directly target mechanisms of chromatin structure regulation.

Disclosures

Licht:Celgene: Research Funding.

Author notes

*

Asterisk with author names denotes non-ASH members.