Clonal hematopoiesis transcending species barriers

In this issue of Blood, Shin et al expand our understanding of clonal hematopoiesis (CH) by showing its natural emergence in aged non-human primates (rhesus macaques), and by demonstrating robust expansion of TET2-mutated clones causing hyperinflammation in a CRISPR-Cas9–based autologous transplantation model in macaques.¹ 

CH describes the overrepresentation of the progeny of a single hematopoietic stem cell (HSC), or clone, in the peripheral blood cell pool. A wealth of data published in recent years has established CH as a universal phenomenon associated with human aging.^2-4 Studies have further demonstrated that larger clones bearing leukemia-associated mutations are associated with an array of adverse outcomes in humans, including development of hematologic malignancies and cardiovascular disease^2,5; this condition has been termed clonal hematopoiesis of indeterminate potential (CHIP) when the variant allele fraction (VAF) exceeds 2%.⁶ Murine models have proven to be important tools in understanding underlying mechanisms, indicating that the association of CHIP with inflammatory diseases is causal and a product of increased inflammation in terminally differentiated myeloid cells.^5,7 Embarking on the study presented here, the authors hypothesized that rhesus macaques might present a faithful model organism for CH because they closely resemble humans in many central attributes of hematopoiesis, while still allowing for experimental engineering of CH that would not be ethical in humans.

Out of 60 aged macaques analyzed with a median age of 25 years, 12 were found to carry naturally occurring somatic mutations that fulfilled the criteria for CH with a known driver mutation present at a VAF > 1% (see figure). This contrasts to aged mice (24 months old) whereby naturally occurring mutations could only be detected in 2% of animals, and at a lower VAF threshold.⁸ As in humans, the most commonly mutated gene was DNMT3A, accounting for 4 cases. The next most frequent alterations were mutations in RUNX1, TP53, NOTCH1, CREBBP, and TET2, a list similar, yet not identical, to the findings in human cohorts. The limited number of cases precludes definitive conclusions about the biological significance of the differences.

Using CRISPR-Cas9–based editing of mobilized peripheral blood CD34⁺ cells, the authors went on to develop an autologous transplantation model of CH in macaques. The authors transplanted autologous CD34⁺ cells engineered to carry loss-of-function mutations in DNMT3A, TET2, and ASXL1 into macaques after myeloablative irradiation (see figure). This model demonstrated robust expansion of TET2-mutated and, to a lesser degree, DNMT3A-mutated clones, consistent with murine models and findings in humans indicating that clone growth rates vary by mutation.^3,4 Moreover, this model produced convincing evidence that TET2-mutated CH in macaques causes interleukin-1β (IL-1β)– and IL-6–driven inflammation. This was shown with several orthogonal assays, including elegant gene expression analysis in genotyped bone-marrow derived colony forming units, functional studies in isolated macrophages, and in the serum. These findings were concordant with prior findings in mice.^5,7 

The authors then harnessed the strengths of this model to conduct a treatment study with the Food and Drug Administration–approved IL-6 receptor antibody tocilizumab, testing the hypothesis that this treatment might impact clone outgrowth. In 1 out of the 3 treated macaques, treatment resulted in a decrease in clone size as assessed by mutant allele fraction. Although the biological interpretation of this part of the study is limited by the small sample size, it does demonstrate that preclinical testing of pharmacologic interventions in this model is feasible.

In sum, this fascinating study advances the field in several ways. It demonstrates the emergence of natural CH in non-human primates, supporting a model in which somatic mutations in specific epigenetic regulators cause CH across species barriers, at least within primates. The experiments add to several recent observations in mice and humans highlighting inflammation as a consequence of CHIP.^5,7 A particularly notable finding in humans was obtained from a subgroup analysis of the CANTOS trial. CANTOS showed that treatment with the IL-1β-targeting antibody canakinumab reduced recurrent cardiovascular events in high-risk patients, albeit at the cost of increasing rates of severe infection.⁹ Subgroup analysis by CHIP status demonstrated that a large part of the reduction in event rates achieved by treatment was attributable to the subgroup of patients with TET2-mutated CHIP.¹⁰ Also, the present study reports the first engineered model of CH in primates that is poised to validate interventions targeting CHIP in the future.

The study also shines renewed light on some of the major unanswered questions in the field. What is special about the recurrent mutations in CH, most prominently those in epigenetic modifiers, that make them so much more common than other mutations? What are the factors, both cell-intrinsic and extrinsic, that determine clonal outgrowth? What are the cellular and molecular underpinnings of hyperinflammation in CHIP? Combining longitudinal studies in humans with perturbation in innovative model systems will be required to fully answer these questions and realize the vast potential of targeting CHIP with therapeutic intention.

Conflict-of-interest disclosure: B.L.E. has received research funding from Celgene, Deerfield, Novartis, and Calico and consulting fees from GRAIL. He is a member of the scientific advisory board and shareholder for Neomorph Therapeutics, TenSixteen Bio, Skyhawk Therapeutics, and Exo Therapeutics. P.J.R. declares no competing financial interests.