Somatic mutations and clonal hematopoiesis: unexpected potential new drivers of age-related cardiovascular disease

Abstract

Increasing evidence shows that conventional cardiovascular risk factors are incompletely predictive of cardiovascular disease, particularly in elderly individuals, suggesting that there may still be unidentified causal risk factors. While the accumulation of somatic DNA mutations is a hallmark of aging, its relevance in cardiovascular disease or other age-related conditions has been, with the exception of cancer, largely unexplored. Here, we review recent clinical and preclinical studies that have identified acquired mutations in hematopoietic stem cells and subsequent clonal hematopoiesis as a new cardiovascular risk factor and a potential major driver of atherosclerosis. Understanding the mechanisms underlying the connection between somatic mutation-driven clonal hematopoiesis and cardiovascular disease will be highly relevant in the context of personalized medicine, as it may provide key information for the design of diagnostic, preventive or therapeutic strategies tailored to the effects of specific somatic mutations.

Aging and cardiovascular disease, an incompletely understood connection

Cardiovascular/cerebrovascular disease (CVD) is the leading cause of morbidity and mortality in both industrialized and low-income to middle-income countries, particularly in the elderly population. Furthermore, this situation is expected to worsen as the world population is aging and CVD incidence steeply increases with age, accounting for more than 40% of total deaths and more than 85% of chronic disease deaths in those older than 70 years.¹ Although epidemiological studies clearly show that advanced age is the most important risk factor for CVD, we have an incomplete understanding of how it promotes disease progression, in particular in the context of atherosclerosis. While the effect of aging on atherosclerotic CVD reflects to a great extent the cumulative exposure to a number of traditional risk factors (e.g. hypercholesterolemia, diabetes, hypertension, smoking),² clinical evidence suggest that additional, still unidentified age-related factors may contribute to atherosclerosis development. Indeed, age remains an independent risk factor for CVD even after statistical adjustment for traditional risk factors,^{3, 4} and most CVD events occur in elderly individuals who are at low or intermediate risk based on traditional risk factors. In one example, an international study that included more than 120,000 atherosclerotic CVD patients revealed that ~65% of patients older than 65 years old had either none or just one conventional CV risk factor.⁵ Consistent with this observation, imaging studies show that most individuals at low CV risk based on conventional risk algorithms exhibit substantial subclinical atherosclerosis.^6–9 Additionally, emerging evidence suggests that the effects of some conventional risk factors on CV risk differ between young and elderly individuals. For example, the impact of hypercholesterolemia on CV risk has been shown to gradually decrease with aging in several studies,^10–12 suggesting the existence of alternative non-conventional risk factors that contribute to the pathogenesis of atherosclerotic CVD in the elderly population. Consistent with these results, recent clinical trials of treatments with high intensity statin regimens or new cholesterol-lowering drugs (e.g. PCSK9 inhibitors) show that, despite massive reductions in blood cholesterol levels, atherosclerotic plaques progress in a significant proportion of individuals^{13, 14} and a substantial level of residual CV risk remains.^{15, 16} Therefore, although the major contribution to atherosclerosis of the cumulative exposure to conventional risk factors –particularly hypercholesterolemia- is undeniable,¹⁷ these aforementioned studies lead inevitably to one question: what else is driving atherosclerotic CVD in the elderly? Emerging evidence from human and mouse studies suggest that the acquisition of somatic mutations in hematopoietic cells may be part of the answer to this question.

Somatic mutations in the hematopoietic system and atherosclerotic CVD: evidence from human studies

Over the past two decades, research on the impact of human genetics on CVD has focused mostly on genetic variation transmitted through the germ line.^{18, 19} While these studies have unveiled a significant contribution of many inherited variants to CVD, recent evidence suggests that non-inherited, acquired mutations that affect somatic cells may also be important players in CVD and potentially other age-related disorders. These somatic mutations occur randomly, beginning in prenatal development.^20–22 In most cases they have no or little phenotypic consequences, as most mutations have no effect on cellular function or in some cases are toxic or even lethal to the mutant cell, leading to its quick disappearance. However, in a few instances, a somatic mutation confers a competitive advantage to the cell, which leads to the progressive expansion of the mutant clone. Consequently, individuals become a mosaic of cells with different genotypes over time.^{21, 23, 24} Accumulating evidence reveals that this somatic genome mosaicism is indeed an inevitable consequence of the normal aging process in many tissues.^{25, 26} This is particularly true in highly proliferative tissues, such as the hematopoietic system, which generates >100 billion cells daily. Recent evidence show that individual hematopoietic stem/progenitor cells (HSPCs) accumulate somatic mutations as a function of age, even in healthy individuals. Human HSPCs have been calculated to develop 0.13 ± 0.02 exonic mutations per year of life, and thus it can be estimated that by age 50 an individual would accumulate an average of five coding gene mutations within each HSPC.²⁴ This sets the stage for a robust Darwinian selection of mutations that provide a competitive advantage to the mutant HSPC by promoting its self-renewal, proliferation or survival. Such a competitive advantage eventually leads to the clonal expansion of the mutant HSPC in the hematopoietic system and, through its cellular progeny, in the blood and other tissues infiltrated by blood cells. Early studies documented the occurrence of age-related clonal hematopoiesis (frequently abbreviated as ARCH) based on the results of X-chromosome inactivation studies,^27–30 and a substantial body of evidence has now demonstrated that this process of somatic mutation-driven clonal hematopoiesis (Figure 1A), is commonly observed in healthy elderly individuals.^31–38 In the case of extreme age, there is sometimes a marked expansion of a few clones and a collapse of HSPC diversity caused by the acquisition of somatic mutations in key hematopoiesis regulators or other mechanisms of clonal expansion.³⁹ For example, deep whole-genome sequencing of the peripheral blood cells from a 115 year old woman revealed that the majority of her white blood cells were derived from just two related HSPC clones.⁴⁰

Figure 1. Somatic mutations in blood cells as a shared mechanism of hematological cancer and cardiovascular disease.

A. The accumulation of somatic mutations in hematopoietic progenitor and stem cells is an inevitable consequence of the process of aging. Some of these random mutations confer a competitive advantage to the mutant cells, leading to its clonal expansion. This phenomenon can be defined as somatic mutation-driven clonal hematopoiesis. B. Most individuals exhibiting somatic mutation-driven clonal hematopoiesis only carry one driver mutation (e.g. in DNMT3A, TET2, ASXL1 or JAK2). While this situation greatly increases the risk of acquiring additional driver mutations and eventually developing a hematologic cancer, this condition is infrequent, even in individuals with clonal hematopoiesis. The main cause of death in individuals exhibiting somatic mutation-driven clonal hematopoiesis is atherosclerotic cardiovascular disease. Grey shade in the figure indicates decreasing frequency.

Clonal hematopoiesis is often driven by a single mutation in a cancer-related gene, and, accordingly, it is associated with increased risk of hematological malignancies.^{35, 36} Despite this, epidemiological studies show that most individuals carrying cancer-related somatic mutations that drive clonal expansion of the mutant cell will never develop blood cancer, since this is a rare condition requiring the accumulation of multiple additional oncogenic gene mutations (Figure 1B).³⁵ Based on these observations, this phenomenon is frequently referred by hematologists as clonal hematopoiesis of indeterminate potential (CHIP).^{42, 43} Although large chromosomal rearrangements^44–47 and point mutations in more than 100 cancer-related genes^{31–37, 41} have been identified as potential drivers of clonal hematopoiesis, most somatic mutations linked to this phenomenon reported to date affect a small subset of genes known to play important roles in hematopoiesis and/or hematological cancers (Table). Notably, most somatic mutations associated with clonal hematopoiesis occur in three genes (DNMT3A, TET2 and ASXL1) that encode for epigenetic regulators that have been implicated in the control of hematopoiesis.^48–55 It can be estimated that at least 10-20% of individuals older than 65 years old exhibit a significant fraction of white blood cells (e.g. >4%) that harbor somatic mutations in these or other known driver genes.^34–36 Furthermore, highly sensitive sequencing techniques that allow the detection of mutations in a smaller percentage of blood cells suggest that low levels of clonal hematopoiesis may be even more prevalent, trending towards inevitability in the very elderly.^{31, 33, 37} Despite these observations, the relevance of clonal hematopoiesis in the context of age-associated disorders other than cancer has been largely neglected. The scarce investigation of the clinical consequences of this phenomenon is somewhat surprising considering that mutations in HSPCs are likely to affect the function of their progeny –immune cells, red blood cells and platelets-, whose dysfunction is at the center of most age-related chronic disorders. This situation is bound to change as evidence has emerged supporting the concept that age-associated somatic mosaicism in the hematopoietic system and clonal hematopoiesis may have major pathophysiological implications in the setting of CVD and other chronic diseases.

Table.

Fifteen most frequently mutated genes identified as candidate drivers of clonal hematopoiesis in cancer-free individuals in several human studies

GENE	REFERENCES
DNMT3A	31–37, 41
TET2	31, 32, 34–38, 41
ASXL1	31, 32, 34–37, 41
JAK2	31–37, 41
TP53	31, 32, 34–37, 41
PPM1D	31, 34, 36, 41
IDH2	31–37, 41
CBLC	31, 32, 34–37, 41
SF3B1	31, 33–37, 41
SRSF2	31, 33, 35–37, 41
GNAS	31, 34, 35, 37, 41
KRAS	32, 33, 35, 37, 41
GNB1	31, 35, 37, 41
NRAS	31–33, 35, 37
MYD88	31, 35–37

Early evidence of a connection between a specific somatic mutation in the hematopoietic system and the development of CVD in the absence of cancer can be found in some small studies that reported the presence of the JAK2 V617F mutation in blood cells of a higher than expected proportion of cancer-free patients with venous thrombosis,^{56, 57} stable coronary heart disease (CHD)⁵⁸ or peripheral artery disease.⁵⁹ While limited by their small sample size and retrospective nature, these studies, among others, opened the door to the possibility of a connection between somatic mutations in blood cells and CVD. This possibility was reinforced when whole-exome sequencing of blood cells of more than 5,000 cancer-free individuals in seven independent cohorts revealed that the presence of single-nucleotide variants and small insertions or deletions in any of 160 analyzed hematopoiesis-related genes is associated with a ~40% increase in all-cause mortality.³⁵ Although the presence of one somatic mutation in any of these genes was associated with an ~11-fold increase in the future risk of hematologic cancer, this only affected 0.5-1% of mutation carriers each year and did not explain the marked increase in all-cause mortality. Unexpectedly, the increased mortality was associated with a substantial increase in deaths due to myocardial infarction or stroke among individuals exhibiting a relative frequency of the mutant allele (variant allele fraction or VAF) greater than 10%, typically indicative of the presence of more than 20% blood cells carrying the mutation in heterozygosis. Consistent with this finding, a significant association was found between the detection of somatic mutations in blood cells and the future incidence of atherosclerotic conditions (CHD and stroke):~45% of mutation carriers developed atherosclerotic CVD during the study, compared to ~20% of non-carriers. Thus, carrying somatic mutations in hematopoiesis-related genes increased >2-fold the risk of developing atherosclerotic CVD, an effect that was comparable or even greater than that of various conventional risk factors.³⁵ While this study reported an association between CVD and a somatic mutation in any of 160 genes, mutations in just four genes, namely DNMT3A, TET2, ASXL1 and JAK2, accounted for more than 70% of the detected mutations. A more recent study has confirmed the existence of independent associations between CHD and mutations in each of these four genes with median VAF ranging from 10% to 20%.⁴¹ However, it is likely that larger studies with longer follow-up periods will identify additional mutated genes linked to CHD, and that significant associations will be observed even with lower VAF values. Indeed, this latter study also reported an association of CHD with any of 56 different somatic mutations in less-frequently mutated genes,⁴¹ supporting the possibility that other mutated genes may be connected to age-related CVD. Somatic mutations may also be linked to the premature development of CVD in young and middle age individuals, as a retrospective case-control analysis of individuals <50 years old revealed a remarkable 4-fold higher risk of early-onset myocardial infarction in individuals exhibiting somatic mutation-driven clonal hematopoiesis.⁴¹ Furthermore, the presence of somatic mutations in blood cells was also found to be associated with a 3-fold higher risk of exhibiting a high degree of coronary calcium calcification, a marker of subclinical atherosclerosis burden. Remarkably, the increased risk of coronary artery calcification was 12-fold in individuals exhibiting a VAF greater than 10%.⁴¹

Overall, these human data suggest a strong and unexpected connection between age-related somatic mutations in hematopoietic cells, clonal hematopoiesis and atherosclerosis. However, it must be noted that the descriptive nature of these types of epidemiological studies does not allow cause-effect relationships to be established. The associations between somatic mutation-driven clonal hematopoiesis and atherosclerotic CVD could simply reflect shared consequences of the aging process or be secondary to confounding factors. Furthermore, these studies do not fully address the issue of directionality; i.e. is clonal hematopoiesis a driver of CVD or are low-grade chronic inflammation and metabolic dysfunction, frequent in subjects at high cardiovascular risk, driving somatic mutagenesis and/or clonal hematopoiesis as suggested by some researchers?^{44, 60} Answering these questions requires the combination of additional human data and carefully designed experimental studies in animal models.

Somatic mutations in the hematopoietic system and atherosclerotic cardiovascular disease: evidence from mouse studies, the case of TET2

As described above, most reported somatic mutations associated with clonal hematopoiesis and CVD occur in a small number of genes, particularly in four genes (DNMT3A, TET2, ASXL1 and JAK2), which are known to modulate hematopoiesis,^{48–55, 61–63} but whose role in CVD remained mostly unexplored until recently. The first of these genes reported to exhibit somatic mutations in blood cells in individuals with clonal hematopoiesis without blood cancer was TET2 (ten eleven translocation 2).³⁸ More than 130 different mutations have been reported in this gene in blood cells of cancer-free individuals,^{31, 32, 34–38, 41} which are considered to result in loss of gene function given that most are small insertions/deletions, non-sense mutations or missense mutations leading to the substitution of key amino acids in the catalytic domain. TET2 is a multifaceted transcriptional regulator that can facilitate both transcriptional activation and repression depending on the molecular/cellular context. It is able to catalyze the conversion of 5-methylcytosine (5mC) into 5-hydroxymethylcytosine (5hmC), a process that facilitates DNA demethylation and transcriptional activation.^64–66 Conversely, TET2 can also mediate transcriptional repression by recruiting histone deacetylases to gene promoters.⁶⁷ Mouse studies have shown that TET2 deletion or haploinsufficiency result in increased HSPC self-renewal and a bias towards differentiation into the myeloid lineage,^{53–55, 68} consistent with human data linking TET2 mutations with clonal hematopoiesis. However, despite the broad expression pattern of TET2,⁶⁹ the relevance of this protein in pathophysiological settings other than stem cell biology or cancer has just recently begun to be explored. In this regard, a recent study from our group demonstrated that TET2 loss of function-driven clonal hematopoiesis accelerates atherosclerosis in hyperlipidemic mice, providing experimental evidence in support of the causal contribution of somatic mutation-driven clonal hematopoiesis to atherosclerotic CVD.⁷⁰ In this study, experimental conditions in mice recapitulated many features of clonal hematopoiesis linked to somatic TET2 mutations in humans. Competitive bone marrow (BM) transplantation experiments were employed to introduce a small number of TET2-deficient cells in atherosclerosis-prone Ldlr−/− mice and mimic the scenario of clonal hematopoiesis, where the mutation is initially present in a small clone that expands over time. In agreement with previous studies,^53–55 TET2-deficient cells expanded progressively in BM, spleen and blood in this experimental setting, and they showed a mild myeloid bias, with a preferential expansion into the Ly-6C-high classical monocyte population.⁷⁰ Importantly, this did not affect total numbers of white blood cells or their distribution in different subsets, consistent with human studies on cancer-free individuals carrying TET2 mutations.^{35, 38} The expansion ofTet2−/− cells in these conditions accelerated atherosclerosis substantially, leading to the formation of ~60% larger plaques. A similar, albeit milder, effect was observed in equivalent transplantation experiments with Tet2+/− cells, which may mimic better the human scenario given that somatic TET2 mutations are likely to be in heterozygosis in most individuals. Additionally, atherosclerotic plaque size was also increased when Tet2 ablation was restricted to myeloid cells.⁷⁰ These findings were recently validated in independent studies in mice exhibiting full hematopoietic ablation of TET2.⁴¹ Collectively, these experimental studies provide strong support to the existence of causal connection between somatic mutations in this gene and the development of atherosclerotic CVD.

From a mechanistic point of view, in vitro and in vivo studies suggest that accelerated atherosclerosis in conditions of TET2 loss of function-driven clonal hematopoiesis is mainly due to the exacerbated expression of pro-inflammatory cytokines and chemokines in TET2-deficient macrophages,^{41, 70, 71} particularly due to the overproduction of the pro-inflammatory cytokine interleukin 1 beta (IL-1β).⁷⁰ Interestingly, this mechanism is consistent with the possibility that a small number of TET2-mutant cells in the plaque may be sufficient to drive exacerbated vascular inflammation and atherosclerosis by promoting pro-inflammatory changes in both mutant and wild-type cells. In this regard, it was found that exacerbated IL-1β production by TET2-deficient plaque cells in mice promotes P-select in expression and endothelial cell activation in the plaque, leading to increased monocyte recruitment regardless of the Tet2 genotype of recruited monocytes.⁷⁰ Furthermore, IL-1β has been reported to stimulate its own expression by an autoregulatory feedback loop^72–76 and therefore it is likely that overproduction of IL-1β by TET2-deficient plaque cells promotes further expression of IL-1β in both TET2-deficient and wild-type cells. Hence, it can be speculated that a few TET2-mutant cells in the plaque act in a catalytic manner, amplifying and perpetuating vascular inflammation and atherosclerosis. Consistent with this “few bad apples spoil the barrel” scenario, the progressive expansion of Tet2−/− cells in a competitive BMT setting seems to have a comparable effect on atherosclerosis than that of the full ablation of TET2 in all BM cells.^{41, 70} This relatively surprising finding likely reflects the complexity of the actions of TET2 in the hematopoietic system and it suggests that the effects of TET2 deficiency on atherosclerosis may be dependent on cell type and/or stage of plaque development. It also provides valuable information on the variety of research strategies available to model clonal hematopoiesis in mice. While conventional full BM transplantation strategies are the most straightforward strategy to study the effects on atherosclerosis of clonal hematopoiesis-related mutations, the use of cell type-specific approaches or competitive BM transplantation strategies with low percentages of mutant cells may be advantageous in some instances. This may be the case when the mutant HSPCs exhibit a clear bias to differentiate into specific lineages or when the ubiquitous expression of the mutated gene in hematopoietic progenitors leads to hematological abnormalities or malignancies not present in most cancer-free individuals exhibiting clonal hematopoiesis (e.g. JAK2 gain-of-function mutations^61–63).

The use of mouse studies tailored to the human scenario of clonal hematopoiesis may provide clinically-relevant information. For instance, the findings in our mouse model of TET2 mutation-mediated clonal hematopoiesis suggest that IL-1β blockade may be particularly effective for the prevention/treatment of CVD in individuals carrying somatic mutations in TET2. Supporting this possibility, pharmacological blockade of NLRP3 inflammasome-mediated IL-1β secretion had greater atheroprotective effects in conditions of TET2 loss of function than in wild-type controls, and completely suppressed differences in atherosclerosis development between genotypes.⁷⁰ These findings are especially relevant in the context of the results of the CANTOS trial, which demonstrated the efficacy of a neutralizing antibody against IL-1β (Canakinumab) in the secondary prevention of atherosclerotic CVD in high-risk patients exhibiting evidence of systemic inflammation.⁷⁷ Even though Canakinumab has shown efficacy against atherosclerotic CVD, its clinical use poses substantive challenges in the setting of chronic disorders, as it also increases the risk of infections.⁷⁷ Therefore, this drug, and probably any anti-inflammatory drug, should probably be administered exclusively to those patients that are predicted to show the greatest response to this therapy and to obtain the most atheroprotection from it. Currently, the analysis of hsCRP, a marker of systemic inflammation, is the best available tool to identify these patients and distinguish between responders and non-responders to anti-inflammatory therapies.⁷⁸ While no conclusive data has yet been provided, it can be speculated that clonal hematopoiesis driven by TET2 loss of function or other pre-leukemic mutations may be associated with increased systemic inflammation in some individuals, and that the genetic analysis of blood cells may be employed in the future to identify responders to anti-inflammatory therapies in the context of precision medicine strategies.⁷⁹ In this regard, our preclinical data suggest that individuals with TET2 mutations may respond more favorably than the general population to IL-1β/NLRP3-targeted therapies, thus creating the basis for a personalized therapy for the prevention of accelerated atherosclerotic CVD in individuals carrying somatic mutations in this gene. This notion could be clinically validated through a genetic screening for somatic TET2 mutations in the CANTOS cohort. Furthermore, given that IL-1β plays a pivotal role in vascular inflammation⁸⁰ and TET2 also affects the expression of other immunomodulatory molecules,^{41, 70, 71} it may be of interest to apply similar strategies in ongoing clinical trials with other broad spectrum anti-inflammatory compounds such as methotrexate⁸¹ or colchicine.⁸²

Somatic mutation-driven clonal hematopoiesis and atherosclerotic cardiovascular disease beyond TET2: other genes and mechanisms?

TET2 was the first gene reported to carry somatic mutations in cancer-free individuals exhibiting clonal hematopoiesis,³⁸ and mutations in this gene have been the first to be causally linked to an age-associated disorder other than cancer, atherosclerotic CVD.^{35, 41, 70} However, human evidence suggest that many other mutated genes may contribute to this condition.^{35, 41} As discussed above, independent associations with atherosclerotic CVD have been reported for three additional genes, DNMT3A, ASXL1 and JAK2, and mutations in other less frequently mutated genes also seem to be associated with increased risk of atherosclerotic CVD.^{35, 41} It is reasonable to speculate that these additional mutated genes will differentially affect CVD due to their unique functions and mechanisms of action. Some mutations may induce qualitative changes in key immune cells, whereas others may affect the differentiation of HSPCs to specific lineages, leading to quantitative changes in specific immune cell subsets, particularly in individuals that exhibit large mutant clones and a reduced HSPC repertoire. Whether somatic mutations in clonal hematopoiesis-related genes other than TET2 contribute to atherosclerosis remains unexplored and additional experimental studies are warranted. However, previous research allows for some speculation on the different mechanisms by which some of these mutated genes may promote CVD.

Accumulating experimental evidence suggest that DNMT3A, the most frequently mutated gene in cancer-free individuals that exhibit clonal hematopoiesis, is an important modulator of both hematopoiesis and inflammatory responses. This gene encodes for a methyl transferase enzyme that catalyzes DNA methylation and modulates gene transcription in multiple scenarios. Supporting a major role of DNMT3A mutations in clonal hematopoiesis in humans, mouse HSPCs exhibiting heterozygotic DNMT3A loss of function develop a competitive advantage and myeloid skewing over time.⁸³ Importantly, DNMT3A deficiency has also been reported to lead to several potentially pro-atherogenic phenotypes in immune cells, including exacerbated pro-inflammatory activation of mast cells,⁸⁴ increased IFNγproduction by T cells^85–87 and restrained immunosuppressive function in myeloid-derived suppressor cells.⁸⁸ In contrast, DNMT3A inhibition has also been shown to increase the expression of interleukin 13 in T cells⁸⁷ and to limit the production of type I interferons in macrophages,⁸⁹ which could potentially protect against atherosclerosis development.^{90, 91} Therefore, given the complexity of the immunomodulatory functions of DNMT3A, carefully designed experimental studies will be required to determine the potential contribution of somatic mutations in this gene to CVD.

Another frequently mutated gene for which previous data support a role in CVD is JAK2, which encodes for an important signaling kinase. Indeed, early evidence linking somatic mutations to CVD was provided by the observation that essential thrombocytopenia and polycythemia vera patients carrying the JAK2 V617F gain-of-function mutation exhibit a higher risk of thrombosis.^92–100 Increased blood counts of leukocytes, red blood cells and/or platelets are potential contributors to the increased CV risk in these patients. However, additional qualitative changes in blood cells are also likely to play a role, as increased CV risk has been observed in individuals carrying this mutation and exhibiting clonal hematopoiesis with little or no changes in blood cell counts.^{35, 101} Some mouse studies suggest that JAK2 V617F-driven clonal hematopoiesis exhibits a clear myeloid, particularly granulocytic, bias¹⁰² and thus it is possible that the effects of this somatic mutation on the CV system are mediated by phenotypic changes in myeloid cells. Given that findings in various experimental models support a role for JAK2 in the modulation of the pro-inflammatory activities of macrophages and neutrophils,^103–105 it will be of interest to evaluate the impact of the JAK2 V617F mutation on CVD in the absence of hematological malignancies.

Beyond the above-mentioned genes, it must be noted that most other mutated genes associated with clonal hematopoiesis have never been evaluated for their role in CVD in any context. Furthermore, controversy exists on whether some of the observed mutations lead to loss or gain of gene function, as is the case of ASXL1 truncation mutations.^{51, 52, 106–108} On the other hand, it is tempting to speculate that some of the somatic mutations that lead to clonal hematopoiesis may be benign or exhibit protective rather than pathogenic effects on the CV system. In other words, some types of clonal hematopoiesis may be uncoupled from CVD or other age-associated chronic diseases, and these forms of clonal hematopoiesis might be overrepresented in individuals that exhibit exceptional longevity.

Somatic mutation-driven clonal hematopoiesis: manyopen questions

Beyond atherosclerosis, a role in other age-related conditions?

The link between acquired somatic mutation-driven clonal hematopoiesis and atherosclerosis development is now supported by both clinical and preclinical data. However, the possibility that mutations driving clonal hematopoiesis also modulate other forms of CVD or other age-related chronic conditions remains largely unexplored and deserves attention. Indeed, recent preclinical studies in mouse models suggest that TET2 loss of function-driven clonal hematopoiesis may accelerate heart failure, a condition that is common in the elderly.¹⁰⁹ Furthermore, human studies have suggested a potential connection between various forms of somatic mutations in blood cells and type 2 diabetes, chronic pulmonary disease or psychiatric conditions.^{31, 32, 35, 44} Future experimental studies are warranted to examine the nature of these connections, especially given that many of the known mutations that drive clonal hematopoiesis affect immune cells and pathways that are at the center of many age-related chronic disorders.

Beyond mutations, what else modulates clonal hematopoiesis?

As indicated above, prior exome sequencing studies have suggested that 10-20% of individuals older than 65 years old exhibit somatic-mutation driven clonal hematopoiesis leading to a substantial percentage (e.g. >4%) of mutant cells in the blood.^34–36 However, more recent studies using highly sensitive sequencing techniques that allow the detection of mutations in a smaller fraction of blood cells found that the prevalence of clonal hematopoiesis is probably at least double than that initially reported.^{31, 33, 37} Furthermore, the seeds for somatic-mutation driven clonal hematopoiesis may be present even in younger individuals, as one recent study employing targeted error-corrected sequencing found that somatic mutations in driver genes (most frequently DNMT3A and TET2) are present in a very small fraction of blood cells (e.g. 1/3000) in 95%of individuals between 50 and 60 years of age.¹¹⁰ Similarly, it has been estimated that almost half of healthy individuals at 50 years of age carry at least one HSPC with a randomly generated mutation in TP53, another potential driver of clonal hematopoiesis.¹¹¹ Thus, the presence of low levels of hematopoietic cells with somatic mutations may be ubiquitous by middle age, but only a fraction of individuals may advance to a condition where the mutant clone undergoes marked expansion. Overall, these observations suggest the existence of additional factors, perhaps genetic or lifestyle factors, that regulate the clonal expansion of mutant HSPCs. For example, some germline variants, most notably in the JAK2 and TERT genes, have been associated with increased prevalence of clonal hematopoiesis associated with the JAK2 V617F mutation.¹¹² Furthermore, smoking has also been found to be associated with clonal hematopoiesis.^{31, 36, 113} Additional investigations of the effect of these and other factors on somatic mutation-driven clonal hematopoiesis may lead to a better understanding of the complex molecular mechanisms that regulate the dynamics of this process, which could contribute to the development of therapies to prevent it.

Somatic mutations in bone marrow cells: a new factor to consider in the application of cell therapies?

Numerous “adult stem cell” therapies are being tested in clinical trials or are offered to the public by commercial entities.^114–116 These therapies often involve the delivery of autologous or allogeneic BM cells or cells from tissues that contain a substantial fraction of BM-derived cells. In addition, some of these procedures employ cells that are selected based upon the expression of cell surface markers, such as CD34, that lead to the enrichment in HSPCs. Given the realization that somatic mutations in HSPCs can promote their clonal expansion and alter the function of their cellular progeny, questions can be raised about the long-term safety of some of these adult stem cell procedures. It is theoretically possible that the delivery of cell fractions from elderly individuals, who may exhibit clonal hematopoiesis, will nullify the therapeutic effect of the transplanted cells or even contribute to adverse outcomes due to the transplantation of hematopoietic cells that harbor somatic mutations. Furthermore, one cannot rule out the possibility that the adoptive transfer of mutant HSPCs could further accelerate their clonal expansion and contribute to disease processes as the patient ages. Future studies are therefore warranted to evaluate whether it might be appropriate to examine transplanted cell populations for their clonal hematopoiesis status prior to use.

Clonal hematopoiesis without known driver mutations: what are we missing?

Most research efforts to date in the context of clonal hematopoiesis have been focused on the identification of mutations in blood cancer-related genes previously known to modulate hematopoiesis. However, it is noteworthy that ~50%, or greater, of cases of clonal hematopoiesis occur without any known driver gene mutation.^{31, 36} While this may be related to technical issues in some cases (e.g. sequencing depth or specific blood cell populations being sequenced), these results strongly suggest the existence of additional, still unidentified somatic mutations that can drive clonal hematopoiesis and perhaps contribute to age-related disorders. The analysis of chromosomal rearrangements or mutations in non-protein coding genes may deserve special attention in this context. Indeed, a significant association between the presence of clonally-expanded blood cells carrying acquired large chromosomal rearrangements and the occurrence of vascular complications of diabetes has been reported.⁴⁴ Alternatively, non-mutational mechanisms may also contribute to clonal hematopoiesis. Computer simulations suggest that, at least in some cases, clonal hematopoiesis can arise in the absence of a driver mutation simply due to stochastic neutral drift in the HSPC population.³¹ Among other non-mutational mechanisms,¹¹⁷ it is possible, for instance, that epigenetic alterations and the associated cell-to-cell variability in gene expression leads to the dominance of specific HSPC populations. Similarly, the heterogeneous distribution of extracellular signals in the BM niche could lead to the preferential expansion of some HSPCs due to their local environment. Interestingly, clonal hematopoiesis without candidate driver gene mutations has also been associated with an increase in all-cause mortality,³¹ although its potential association with CVD remains unexplored. Be it somatic mutations or non-mutational mechanisms, extensive investigation will be required to identify new drivers of clonal hematopoiesis and understand their clinical relevance.

Conclusions

The accumulation of random mutations in the hematopoietic system is an unavoidable consequence of the aging process. While the pathophysiologic relevance of these acquired somatic mutations has been traditionally restricted to the cancer field, emerging evidence suggest that these mutations are important drivers of CVD and potentially many other age-related disorders. A combination of clinical and preclinical studies will be required to understand the pathogenic effects (or lack of) of the various mutated genes, as well as the molecular mechanisms that govern the expansion of mutant hematopoietic stem cells. This information could eventually lead to the design of personalized therapies or preventive strategies for individuals carrying specific somatic mutations in blood cells.

Non-standard Abbreviations and Acronyms

ASXL1

additional sex combs like 1

CHD

coronary heart disease

CHIP

clonal hematopoiesis of indeterminate potential

CVD

cardiovascular disease

DNMT3A

DNA methyltransferase 3 alpha 

HSPC

hematopoietic stem/progenitor cell

JAK2

Janus kinase 2

TET2

tet methylcytosine dioxygenase 2

TP53

tumor protein p53