Asxl1 mutant cells suppress Dnmt3a mutant clonal expansion. Free

Clonal hematopoiesis (CH) is ubiquitous among people over age 55, and 25% of aged individuals have multiple mutant clones (Fabre, Nature 2022). Most studies of CH have investigated mechanisms of clonal expansion one clone at a time; little is known about how emerging clones impact the expansion of neighboring clones.

Prior work showed that HSCs with mutations in DNMT3A, ASXL1 or TET2 (D-A-T), three commonly mutated CH genes, have mutation-specific selective advantage in inflammatory conditions (Florez, Cell Stem Cell 2022). Selective advantage is achieved via enhanced cytokine production paired with mutation-specific mechanisms to survive inflammation, creating a feed-forward loop. We therefore hypothesized that D-A-T mutant clones cooperate to promote each other's expansion.

Results Human Genetic Analyses Reveal ASXL1-DNMT3A Antagonism We analyzed co-mutation patterns and clonal dynamics across 4 CH/MDS human cohorts. Contrary to our expectations, across these datasets, we observed a consistent and pronounced mutual exclusivity between DNMT3A and ASXL1 mutations, with the strongest signal in All of Us (observed = 48 vs. expected = 123, p = 6.3×10⁻²⁹, OR = 0.16). DNMT3A-ASXL1 mutual exclusivity ranked among the strongest of all tested driver pairs, including known enriched co-mutations like TET2-JAK2, used to validate our analytic pipeline.

Using logistic regression models, we tested whether clone size influenced these dynamics. Increasing ASXL1 VAF was associated with reduced odds of a DNMT3A co-mutation, while increasing DNMT3A VAF had no significant effect on ASXL1 detection. This indicates a directional and dose-dependent suppressive effect. Increased TET2 VAF was positively associated with DNMT3A detection (OR > 1), though these genes rarely co-occurred, hinting at context-dependent interactions. These findings suggest ASXL1 mutant clones suppress DNMT3A clonal expansion, while TET2 may promote it under certain conditions.

Murine Models Support Directional Suppression To model these interactions and investigate mechanisms, we generated a 1% mosaic competitive transplant murine model where Dnmt3a^-/-, Tet2^-/-, and Asxl1-truncation transgene (D-A-T) cells were transplanted alone, pairwise (2x) and all combined (3x). Given most CH murine models use a 10% starting engraftment, we first ascertained whether dynamics differ at a starting engraftment of 1%. We found no difference in the growth curves of D-A-T mutants with a starting clone size of 1% versus 10%, suggesting that mutant clones have a set expansion rate under steady state conditions. In contrast, with chronic inflammatory stress induced by Mycobacterium avium or intermittent low dose LPS, Dnmt3a^-/- and Tet2^-/-clones, which are reported to have accelerated expansion at a starting engraftment of 10%, did not expand at a 1% starting engraftment. These findings suggest that the power of external inflammatory stimuli to promote clonal expansion is dependent on clone size.

Turning to the 2x and 3x transplants, Dnmt3a^-/- and Tet2^-/- clones promoted each other's expansion when co-occurring as observed in the human data. Notably, Dnmt3a^-/- clonal expansion was suppressed in the presence of Asxl1 mutant cells in both the 2x and 3x mosaic models. Statistical modeling revealed a 4% decrease in weekly growth rate for Dnmt3a^-/- cells when in competition with Asxl1 mutants. Interestingly, ex vivo competition experiments show Dnmt3a^-/- progenitor cells are not directly suppressed by Asxl1 mutant progenitor cells, suggesting terminally differentiated mutant cells likely mediate the suppressive phenotype. In ongoing work, scRNA sequencing data from 1x versus 2x models will identify mechanisms of Asxl1-mediated Dnmt3a suppression. Given known transcriptional differences between these clones (for example, Asxl1-mutant cells survive inflammation by upregulating Socs3a, whereas Dnmt3a mutant cells have lower Socs3a but enhanced p21 to promote survival), we expect to find transcriptional differences that underly this suppressive effect (King, Exp Hem 2021).

Conclusion We found that Asxl1 mutant cells suppress Dnmt3a mutant clonal expansion. To our knowledge, this is the first epidemiologic and mechanistic report of intraclonal dynamics that affect early clonal expansion. Given that mutations in DNMT3A account for ~60% of CH, these findings could reveal therapeutic avenues for slowing DNMT3A mutant clonal expansion and reduce overall risk of malignant transformation.