Absent, small or homeotic discs 1-like (Ash1l) encodes the mammalian homolog of a Trithorax-group protein with a conserved SET domain that harbors histone H3 lysine 36 dimethyltransferase activity. Using mice with constitutive knockdown of Ash1l due to the insertion of a "gene trap" cassette in the first intron (Ash1lGT/GT mice), we previously reported that Ash1l is a critical regulator of quiescence and self-renewal in adult hematopoietic stem cells (HSCs). While wild-type HSCs could readily establish long-term bone marrow reconstitution after transplantation into irradiated recipients, Ash1l-deficient HSCs had markedly decreased quiescence and failed to establish long-term bone marrow reconstitution (Jones, Chase et al., JCI 2015). Here, we addressed three important questions to better understand the role of Ash1l in hematopoiesis: 1) What is the impact of complete, as opposed to partial, Ash1l loss on adult hematopoiesis? 2) Does Ash1l regulate HSC homeostasis in a cell-autonomous manner? 3) Is the catalytic activity of ASH1L required for its function?

To move beyond the limitations of the constitutive knockdown Ash1lGT allele and address the first two questions, we studied newly generated mice carrying conditional knockout (cKO) Ash1l alleles (exon 13 flanked by loxP sites) along with the Mx1-Cre transgene. Induction of Mx1-Cre expression with poly(I:C) in Mx1-Cre+Ash1lf/f mice resulted in nonsense-mediated decay of Ash1l mRNA in hematopoietic cells, thereby completely knocking out Ash1l in adult hematopoietic stem and progenitor cells. Four weeks after inducing Ash1l inactivation in the hematopoietic compartment, we observed a profound depletion of HSCs and multipotent progenitors in Ash1l cKO mice similar to the phenotype of Ash1lGT/GTmice, indicating that conditional Ash1l knockout has a comparable impact on adult hematopoiesis to that of constitutive Ash1l knockdown. Of note, overt hematopoietic failure in steady-state conditions was not observed in either model. To address the second question, we transplanted a mixture of wild-type and Mx1-Cre+Ash1lf/f bone marrow into irradiated wild-type hosts, allowed donor bone marrow to occupy the wild-type niche and establish hematopoiesis, then induced Cre-mediated excision to inactivate Ash1l in Mx1-Cre+Ash1lf/f hematopoietic cells. Upon Cre induction, donor-derived Ash1lΔ/Δ HSCs and myeloid progeny were depleted or outcompeted by wild-type cells, consistent with the model that Ash1l regulates HSC homeostasis in a cell-autonomous manner. Given that Ash1l encodes a SET domain, we next sought to directly examine whether its catalytic activity is required for its role in regulating HSCs. We studied an Ash1l allele with an in-frame deletion of exon 11 and 12, resulting in preserved expression of ASH1L with internally deleted SET domain (ΔSET) (Miyazaki et al., PLOS Genetics 2013). Homozygous ΔSET mice were viable, and phenotypic analysis of adult ΔSET mice revealed normal frequencies of HSCs and multipotent progenitors, in contrast to our observations in Ash1lGT/GTand Ash1l cKO mice. Furthermore, transplanting ΔSET donor bone marrow into irradiated wild-type hosts resulted in sustained long-term reconstitution throughout primary, secondary and tertiary competitive transplantation assays, consistent with preserved HSC function and no alterations in cell cycle regulation. These findings establish that Ash1l regulates HSC homeostasis independently of its SET domain and histone methyltransferase activity. As ASH1L is a very large protein encoding multiple chromatin binding domains, we speculate that ASH1L may serve as a platform to recruit other partners to form novel protein complex(es) that regulate genes critical for HSC homeostasis.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.