Dangerous Liaisons between Tet2 Mutation, Inflammatory Monocytes, and Leukemogenesis

Studies over the last decade have revealed the presence of hematopoietic clonal expansions, often associated with mutations in known leukemia driver genes, that are particularly evident in the blood of the elderly. This condition, referred to as clonal hematopoiesis of indeterminate potential (CHIP), is associated with increased risks for leukemias, other cancers, cardiovascular diseases, other diseases of old age, and overall mortality (1).

In a report in this issue, Yeaton and colleagues generate a novel mouse model of a common TET2 mutation found in CHIP to understand the conditions that can lead from CHIP to leukemia with advancing age (2). First, using molecular dynamics modeling and mass spectrometry, they provide evidence that this point mutation, H1881R, compromises the catalytic activity of TET2, indicating the importance of H1881 in the coordination of the divalent cation and providing a potential mechanism for how the H1881R mutation prevents oxidation of the DNA substrate. They then generated a mouse model that carries this mutation and show that aged heterozygous Tet2-mutant (Tet2^HR) mice develop acute myeloid leukemia (AML)–like neoplasms associated with an increased inflammatory gene signature. Combining Tet2^HR with other co-occurring mutations such as FLT3-ITD and AML-ETO9a led to faster development of leukemias in line with prior studies using Tet2 knockout cells (3). Notably, cooperating mutations were frequently not evident in the leukemias that developed in older mice bearing only the Tet2^HR mutations. Together with their observation that Tet2^HR mice are relatively normal when young but develop myelodysplastic sydrome (MDS)/myeloproliferative disease– to AML-like phenotypes when old, secondary (and apparently nonmutational) factors that are altered with age appear to contribute to leukemia development.

Single-cell RNA sequencing analyses revealed a relatively normal hematopoietic profile of the young Tet2^HR mice, but severe monocytic skewing and abnormal gene expression profiles in old Tet2^HR mice (with the complication that these mice exhibit leukemic disease). Gene expression perturbations, with a notable upregulation of interferon pathways, were evident across differentiation stages in the old but not young Tet2^HR mice, coinciding with levels of inflammatory cytokines in the bone marrow. Of note, they identify a population of MHC-II^hi monocytes with a specific inflammatory gene expression signature in the Tet2^HR mice (Fig. 1). The presence of these monocytes coincides with leukemic progression in the mice, and the MHC-II^hi monocytic gene signature is associated with worse prognosis in patients with AML. TET2-mutant human AMLs are more likely to exhibit this signature, and the MHC-II^hi monocytes were not increased in frequency in engineered non–Tet2-mutated mouse myeloid leukemias. Given that the Tet2 mutation is present in all cells of the Tet2^HR mice, the authors’ demonstration that this inflammatory monocytic population also arises with age in mice that received transplanted Tet2^HR bone marrow is important. In all, these observations of the aging-dependent manifestation of TET2 mutations are quite interesting and provide insight into the specific prevalence of AMLs driven by TET2 mutations in the elderly.

An interesting facet of Yeaton and colleagues’ results lies in the ability of inflammation to reconfigure the transcriptional programs of Tet2^HR hematopoietic cells to promote differentiation toward an inflammatory monocyte type expressing high levels of MHC-II. Tet2^HR hematopoietic cells showed no significant differences in gene expression in the young mice, and yet somehow Tet2 loss of function interacts with an inflammatory and/or aged environment to lead to altered gene expression programs and altered hematopoietic output. Understanding this “somehow” will require further mechanistic studies. In particular, the extent to which the MHC-II^hi cells represent an epigenetically encoded differentiation pathway triggered in Tet2-deficient cells by inflammatory activity remains to be addressed.

Is the interplay between this monocytic population, inflammation, and Tet2^HR hematopoiesis important for leukemia progression from Tet2-mutant hematopoiesis? The authors provide some data in support of such a model (Fig. 1). First, they show that inhibiting Toll-like receptor (TLR) signaling using either antibiotics or inhibitors that block downstream IRAK1/4 signaling reduces the accumulation of these MHC-II^hi monocytes in the bone marrow. Second, as a complementary approach, they show that triggering inflammation via the TLR/IRAK pathway with the TLR4 ligand lipopolysaccharide (LPS) is sufficient to cause the expansion of the MHC-II^hi monocyte population, with some indications of malignant progression. Further studies will be needed to uncover direct functional links between the monocyte population and leukemia development, such as by testing whether leukemogenesis is blocked by the inhibition of inflammatory signaling or depletion of this monocytic population. In particular, a direct demonstration that the identified MHC-II^hi monocytes promote Tet2^HR-initiated leukemogenesis, while not trivial, will be key to establishing that these monocytes represent a cause versus an effect of leukemic transformation. This article provides an important rationale for such downstream studies.

Altogether, this work raises several important questions for the field. Previous studies have shown how inflammation (such as from infections) promotes Tet2-mutant clonal expansions in mouse models (1). This current study shows how inflammation can (i) stimulate Tet2-mutant hematopoietic stem and progenitor cell (HSPC) differentiation into inflammatory MHC-II^hi monocytes and (ii) promote malignant progression of the Tet2-mutant clone to leukemia. These studies together suggest a feed-forward loop between inflammation and Tet2-mutant clonal expansion. These findings also highlight the need to consider that the presence of a CHIP clone may reflect an overall systemic decline, whether mediated by aging or lifestyle, with influences from genetics. Thus, CHIP and other risks such as for cancers and heart disease could reflect common manifestations of this overall decline in the soma (4). The results from this study showing how modulating inflammation and the microbiome can prevent the emergence of MHC-II^hi monocytes suggest that interventions that limit the development of these monocytes could have wide-reaching implications for health in individuals with TET2-mutant CHIP. Hence, it appears that a high TET2-mutant clonal burden can be disconnected from pathology, being apparently asymptomatic in youth but pathogenic in old age or with inflammation (reversibly so?), likely in part through modulation of myeloid development to generate MHC-II^hi monocytes. These findings are cause for reflection as to what specific parameters of CHIP possess the highest likelihood of triggering morbidity and/or mortality and should be targeted in the clinic versus which features may be either adaptive or simply neutral phenotypes in the setting of aging. A study of oldest-old populations with CHIP shows that risk for the spectrum of nonmalignant comorbidities such as cardiovascular disease drops off in these individuals (5), implying that in the absence of pathologic triggers due to lifestyle, underlying disease, and/or other conditions, CHIP itself may represent a relatively benign “fitness peak” that maintains blood system function in the setting of aging. Along these lines, individuals receiving hematopoietic stem cell transplants containing non–TP53-mutant CHIP mutations have a lower risk of relapse, perhaps owing to the increased regenerative and/or competitive potential of the CHIP clone(s) in myeloablated bone marrow (6). Hence, therapeutic approaches aimed at mitigating the emergence of MHC-II^hi cells by modulating inflammation could represent a straightforward approach to mitigating CHIP-related morbidity. Whether treatment with antibiotics or IRAK1/4 inhibitors reduces the spectrum of diseases, including myeloid malignancy, in Tet2^HR mice remains an enticing question.

Along these same lines, it is worth noting that in humans, although the presence of TET2 mutations at any frequency carries an intrinsically higher risk for leukemic transformation, most individuals with TET2-mutant CHIP (and CHIP in general) will never progress to malignancy. Instead, key parameters indicative of low overall bone marrow fitness, including high mutant variant allele frequency, increased red cell distribution width, and the presence of cytopenia, strongly predict for progression to MDS and/or AML (7). Chronic inflammation is a likely partner in this process, though in both humans and mice its precise role and the underlying mechanism remain to be demonstrated (1). Although not shown explicitly here, the authors’ work would suggest that TET2 mutations likely alter the bone marrow microenvironment to favor leukemia development by promoting the expansion of inflammatory MHC-II^hi monocytes. The contributory role of the microenvironment is likewise implied by a recent study showing that although impairment of the PU.1 network collaborates with Tet2 deficiency to initiate myeloid leukemogenesis (8), AML-like disease occurs only in aged animals where an increased inflammatory burden is present. Furthermore, the link between the presence of MHC-II^hi monocytes with poor AML patient outcomes raises the important question as to whether the monocytes are triggering conditions that increase the fitness of the leukemic cells themselves (by promoting their survival and proliferation) versus impairing the function of nonleukemic HSPC clones, thereby creating a fitness differential via induction of apoptosis, metabolic suppression, and/or disruption of the bone marrow niche. Indeed, although people with CHIP are at substantially increased risk of leukemia development, the CHIP mutation in question is not always present in subsequent leukemia (4). The authors’ clinical data set suggests that CHIP clones could affect disease progression in trans, by engendering a microenvironment that favors leukemia evolution. Using newly devised models of cell competition and sequential oncogenesis to address the extent to which TET2-deficient (and other CHIP-mutant) HSPC clones and their myeloid progeny accelerate myeloid oncogenesis, and even favor certain mutant clones more than others, remains a critical next step in defining approaches for early intervention.

Lastly, it is worth considering the broader immunologic importance of MHC-II^hi monocyte populations in the setting of CHIP, as the authors’ data indicate that this population may represent a durable response to inflammatory insults. Aging and lifestyle factors that trigger dysbiosis and inflammation can induce changes in gut permeability, leading to the release of bacterial components into the bloodstream. Furthermore, the proinflammatory features of Tet2-deficient clonal hematopoiesis are also sufficient to increase gut permeability and induce subclinical sepsis in mouse models (9). Understanding the extent to which CHIP-mutant MHC-II^hi myeloid cells represent a maladaptive immune response to changes in gut permeability may be important for understanding the etiology of cardiovascular disease, autoimmune diseases, and other inflammatory conditions such as gout, which may be linked to CHIP. Furthermore, it may be important to identify whether these MHC-II^hi myeloid cells could have an adaptive role in bolstering adaptive immune function, and thus survival, in aged individuals. It is well known that aging is associated with loss of thymic activity and immunosenescence, with greatly reduced generation of new T and B lymphocytes and reduced functionality of existing clones. Lymphocytes are commonly activated by the presentation of antigens engulfed by macrophages and dendritic cells on MHC-II. Higher levels of MHC-II expression may thus provide a stronger or more frequent activating signal for lymphocytes, potentiating host immune surveillance. On the other hand, such signals could also trigger autoimmune activity, particularly if key players that normally suppress spurious immune activation, like regulatory T cells, are dysfunctional. Of note, the bone marrow is often considered an immune-privileged tissue (10). Whether the expansion of MHC-II^hi monocytes in the bone marrow compromises that privilege, leading to activation of immune processes that trigger myelodysplasia, remains an open and important question.

In all, although key questions remain in terms of the direct relationship between inflammatory monocytes and oncogenesis, and the molecular mechanisms driving the activity of MHC-II^hi monocytes, this study opens important avenues for investigation. Hence, we can better appreciate how aging (influenced by other factors like lifestyle) can generate a feed-forward loop through CHIP, inflammation, and tissue decline (each self-reinforcing the other) to lead to an exponential increase in the risks of diseases such as leukemias, other cancers, and cardiovascular disease as we age.

J. DeGregori reports grants from the NIH during the conduct of the study. No disclosures were reported by the other author.

We thank Daniela Ortiz Chavez for creating the figure. This work was supported by grants from the NIH (R01-AG067584 to J. DeGregori and R01-DK119394 to E.M. Pietras) and the Leukemia & Lymphoma Society SCOR grant 7020-19 to J. DeGregori.