Inherited Blood Cancer Predisposition through Altered Transcription Elongation

Despite considerable advances in defining the somatic driver mutations underlying myeloid malignancies, including the myeloproliferative neoplasms (MPNs), a significant heritable component for these diseases remains poorly understood. While common genetic variant association studies have been valuable, they fail to explain the majority of heritable variation. We reasoned that rare variant association studies could provide a valuable complementary approach to identify additional inherited risk factors.

We therefore utilized exome sequencing data from 166,953 UK Biobank participants and performed a gene-based burden analysis for germline genetic variants conferring risk for acquiring a myeloid malignancy. CTR9, which encodes a key component of the PAF1 transcription elongation complex, was among the significant genes identified (SKAT-O p-value = 5.47x10 ^-7). The deleterious variants in CTR9 collectively exhibit a 9.6 (95%CI = 4.86-19.04) increased odds of acquiring a myeloid malignancy and this risk was largely driven by the MPNs. We replicated this association in an independent cohort of 211 MPN patients using external controls. We could show through structural and biochemical analyses that the identified deleterious variants perturbed assembly of the PAF1 complex but did not display dominant negative activity.

Given that increased hematopoietic stem cell (HSC) self-renewal has been shown to predispose to the risk of acquiring MPNs, we sought to define whether CTR9 perturbation could alter HSC self-renewal or function. We achieved predominantly heterozygous loss-of-function in human hematopoietic stem and progenitor cells (HSPCs) by titrating Cas9 ribonucleoprotein delivery with several independent guide RNAs. Partial loss of CTR9 in HSPCs resulted in expansion of phenotypic long-term HSCs (LT-HSCs) and more differentiated short-term HSCs (ST-HSCs). We additionally could show through single cell RNA-sequencing (scRNA-seq) that there was an expansion of molecularly defined HSCs upon partial loss of CTR9. The observed increase in HSCs appeared paradoxical, given that the PAF1 complex has been suggested to be crucial for HSC maintenance. To explore how the observed HSC expansion with CTR9 perturbation may arise, as well as given known interactions between the PAF1 complex and the competing transcriptional super elongation complex (SEC), we examined whether SEC target genes in HSCs, such as mid to posterior HOXA genes, may be activated with partial CTR9 loss. Remarkably, we observed a significant enrichment for hematopoietic SEC target genes upon CTR9 perturbation in HSCs by gene set enrichment analysis (normalized enrichment score = 3.29, p-value < 0.001).

In light of these findings suggesting that SEC activity may be increased with partial CTR9 loss-of-function, as occurs in individuals harboring myeloid malignancy variants, we sought to functionally validate these observations. Using the inhibitors of the SEC, including SR-0813 that targets MLLT3 or with an inhibitor of CDK9, we noted rescue of the CTR9-mediated expansion of phenotypic LT- and ST-HSCs without a significant impact on the bulk HSPC population. To further elucidate underlying mechanisms, we performed immunoprecipitation of PAF1 or SEC component MLLT3 in HSPCs with control or CTR9 editing. While we continued to pull down all PAF1 complex components with PAF1, we also noted pulldown of MLLT3, which increased with CTR9 editing. MLLT3 immunoprecipitation revealed selective pulldowns of PAF1 and CDC73, which also increased with CTR9 editing. These findings show how PAF1 complex components PAF1 and CDC73 interact with and stimulate SEC activity. Our findings reveal how CTR9 usually restricts this activity and constrains transcriptional elongation to limit HSC self-renewal. We functionally validated these findings through selective editing of different PAF1 complex components in HSPCs: we observed reduced HSCs upon editing of PAF1 and CDC73, but increases with editing of other PAF1 complex components.

Our findings collectively demonstrate a mechanism by which a previously undefined myeloid malignancy predisposition occurs. We demonstrate that CTR9 loss-of-function stimulates SEC activity and thereby results in HSC expansion to confer risk for acquiring MPNs and other myeloid malignancies.