Clonal hematopoiesis associated with TET2 deficiency
accelerates atherosclerosis development in mice
José J. Fuster
1,*
, Susan MacLauchlan
1
, María A. Zuriaga
1
, Maya N. Polackal
1
, Allison C.
Ostriker
2
, Raja Chakraborty
2
, Chia-Ling Wu
1
, Soichi Sano
1
, Sujatha Muralidharan
1
, Cristina
Rius
3
, Jacqueline Vuong
1
, Sophia Jacob
1
, Varsha Muralidhar
1
, Avril A. B. Robertson
4
,
Matthew A. Cooper
4
, Vicente Andrés
3
, Karen K. Hirschi
5
, Kathleen A. Martin
2
, and Kenneth
Walsh
1,*
1
Molecular Cardiology, Whitaker Cardiovascular Institute, Boston University School of Medicine,
Boston, MA 02118, USA.
2
Yale Cardiovascular Research Center, Vascular Biology and Therapeutics Program, and
Departments of Medicine and Pharmacology, Yale University School of Medicine, New Haven, CT
06511, USA.
3
Centro Nacional de Investigaciones Cardiovasculares Carlos III (CNIC) and CIBER de
Enfermedades Cardiovasculares, Madrid, Spain.
4
Institute for Molecular Bioscience, The University of Queensland, Brisbane, Queensland,
Australia.
5
Yale Cardiovascular Research Center and Yale Stem Cell Center, Yale University School of
Medicine, New Haven, CT 06511, USA.
Abstract
Human aging is associated with an increased frequency of somatic mutations in hematopoietic
cells. Several of these recurrent mutations, including those in the gene encoding the epigenetic
modifier enzyme TET2, promote expansion of the mutant blood cells. This clonal hematopoiesis
correlates with an increased risk of atherosclerotic cardiovascular disease. We studied the effects
of the expansion of
Tet2
-mutant cells in atherosclerosis-prone, low-density lipoprotein receptor–
deficient (
Ldlr
−/−
) mice. We found that partial bone marrow reconstitution with TET2-deficient
cells was sufficient for their clonal expansion and led to a marked increase in atherosclerotic
plaque size. TET2-deficient macrophages exhibited an increase in NLRP3 inflammasome–
mediated interleukin-1β secretion. An NLRP3 inhibitor showed greater atheroprotective activity in
chimeric mice reconstituted with TET2-deficient cells than in nonchimeric mice. These results
support the hypothesis that somatic
TET2
mutations in blood cells play a causal role in
atherosclerosis.
*
Corresponding author. jjfuster@bu.edu (J.J.F.); kxwalsh@bu.edu (K.W.).
SUPPLEMENTARY MATERIALS
www.sciencemag.org/content/355/6327/842/suppl/DC1
HHS Public Access
Author manuscript
Science
. Author manuscript; available in PMC 2017 August 03.
Published in final edited form as:
Science
. 2017 February 24; 355(6327): 842–847. doi:10.1126/science.aag1381.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Cardiovascular disease (CVD) is the leading cause of death in the elderly, but almost 60% of
elderly patients with atherosclerotic CVD have either no conventional risk factors (e.g.,
hypertension, hypercholesterolemia, etc.) or just one risk factor (1). Furthermore, increasing
evidence suggests that most middle-aged individuals at low risk of CVD, based on
conventional risk factors, exhibit subclinical atherosclerosis (2, 3). These clinical data
suggest that unidentified age-dependent risk factors contribute to the development of CVD.
The accumulation of somatic DNA mutations is a hallmark of aging, particularly in
proliferating tissues, which over time may become a mosaic of cells with different genotypes
due to the clonal expansion of single de novo mutations (4). However, though human studies
suggest that somatic mutations may be associated with a broad spectrum of human disease
(5–7), there is little information on the potential causal role of somatic mutations in age-
associated disorders other than cancer. Recent human studies have shown that normal aging
is associated with an increased frequency of somatic mutations in the hematopoietic system,
which provide a competitive growth advantage to the mutant cell and allow its progressive
clonal expansion (clonal hematopoiesis) (7–11). This acquired clonal mosaicism in the
hematopoietic system of healthy individuals correlates with an increased risk of subsequent
hematologic cancer (7–9), but it has also been associated with higher prevalence of vascular
complications of diabetes, greater incidence of atherosclerotic conditions (i.e., coronary
heart disease, stroke), and increased frequency of CVD-related deaths (6, 7). Although these
human studies suggest an unexpected connection between somatic mutations in
hematopoietic cells, clonal hematopoiesis, and atherosclerosis, their descriptive nature does
not allow cause-effect relationships, or even directionality, to be established.
Most of the reported somatic mutations associated with age-related clonal hematopoiesis
occur in a small number of genes encoding epigenetic regulators (7–10). The present study
focuses on one of these genes,
TET2
(ten-eleven translocation 2), the first gene reported to
exhibit somatic mutations in blood cells in individuals with clonal hematopoiesis without
hematological malignancies (10). More than 70 different mutations have been reported in
this gene (7–10, 12). The protein encoded by
TET2
is an epigenetic regulatory enzyme that
catalyzes the oxidation of 5-methylcytosine (5mc) in DNA to 5-hydroxymethylcytosine
(5hmc) and also exerts noncatalytic actions. TET2 modulates hematopoietic stem and
progenitor cell (HSPC) self-renewal (13–16), but its role in CVD remains largely
unexplored.
To mimic the human scenario of clonal hematopoiesis and test whether clonal expansion of
TET2-deficient hematopoietic cells contributes to atherosclerosis, we used a competitive
bone marrow transplantation (BMT) strategy to generate atherosclerosis-prone, low-density
lipoprotein receptor–deficient (
Ldlr
−/−
) chimeric mice with a small proportion of TET2-
deficient HSPCs. Lethally irradiated
Ldlr
−/−
recipients were transplanted with suspensions
of bone marrow (BM) cells containing 10%
Tet2
−/−
cells and 90%
Tet2
+/+
cells [10%
knockout (KO)– BMT mice] and then fed a normal diet (ND) or a high-fat/high-cholesterol
(HFHC) diet for 9 weeks to induce atherosclerosis development (fig. S1). To distinguish
donor
Tet2
−/−
and
Tet2
+/+
cells in this experimental setting,
Tet2
+/+
cells were obtained from
mice carrying the CD45.1 variant of the CD45 hematopoietic antigen, whereas
Tet2
−/−
cells
were obtained from mice carrying the CD45.2 variant of this protein. Control mice [10%
Fuster et al. Page 2
Science
. Author manuscript; available in PMC 2017 August 03.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
wild-type (WT)–BMT] were transplanted with 10% CD45.2
+
Tet2
+/+
cells and 90%
CD45.1
+
Tet2
+/+
cells. Flow cytometry analysis of CD45.2
+
blood cells established that this
BMT strategy led to the clonal expansion of
Tet2
−/−
hematopoietic cells to an extent
consistent with variant allelic fractions for somatic
TET2
mutations observed in human
studies linking clonal hematopoiesis to accelerated CVD (7). At the start of ND or HFHC
diet feeding (4 weeks after BMT), CD45.2
+
cells represented ~28% of blood cells in 10%
KO-BMT mice, and they expanded further over time, reaching 42% of blood cells 6 weeks
after BMT and 56% 12 weeks after BMT (Fig. 1, A and B). This clonal expansion of TET2-
deficient hematopoietic cells is similar to that observed when human cells carrying somatic
TET2
mutations are transplanted into immune-deficient mice (17).
TET2 ablation in CD45.2
+
cells of HFHC-fed 10% KO-BMT mice was confirmed by
quantitative real-time polymerase chain reaction (qRT-PCR) analysis of CD45.2
+
white
blood cell (WBC) fractions (Fig. 1C and fig. S2A). No changes were observed in the
expression of TET1 or TET3, two related epigenetic modulators. Consistent with the
enzymatic activity of TET2, ablation of the gene was paralleled by a decrease in 5hmC
levels in WBCs and macrophages (fig. S2, B and C). Although the absolute number of
HSPCs [defined as lineage
−
, Sca1
+
, c-Kit
+
(LSK) cells] was comparable between genotypes
(fig. S3A), CD45.2
+
cells represented 69% of LSK cells in the BM (fig. S3B) and 61% in
the spleen (fig. S3C) of 10% KO-BMT mice at 13 weeks post-BMT, consistent with
previous studies reporting that TET2 inactivation enhances HSPC self-renewal (13–16).
Transplanted
Tet2
−/−
BM cells expanded into all blood cell lineages, regardless of type of
diet, although with a slight myeloid bias and a reduced expansion into the T-lymphoid
lineage in the BM, spleen, and blood (Fig. 1D and fig. S3, D to F), in agreement with
previous studies with TET2-deficient mice (13–16). The expansion of
Tet2
−/−
HSPCs did
not affect blood cell counts (fig. S3G), consistent with findings in cancer-free individuals
carrying
TET2
mutations in blood cells (7, 10).
Having demonstrated that the competitive BMT strategy leads to the clonal expansion of
TET2-deficient HSPCs and mimics the human scenario of clonal hematopoiesis associated
with
TET2
mutations, we next evaluated whether the clonal expansion of TET2-deficient
HSPCs affects atherogenesis and related metabolic abnormalities. We observed no effects on
body weight (fig. S4A), spleen weight (fig. S4B), blood glucose levels (fig. S4C), systemic
insulin sensitivity (fig. S4D), or plasma cholesterol levels (fig. S4E). ND-fed mice developed
no aortic atherosclerosis, regardless of BM genotype (fig. S4F). In contrast, clonal expansion
of TET2-deficient BM cells had a profound effect on HFHC-induced atherosclerosis, as 10%
KO-BMT mice exhibited 60% larger plaques in the aortic root than did WT controls (Fig.
1E). Competitive BMT experiments with
Tet2
+/−
cells revealed that TET2 heterozygosity is
sufficient to accelerate atherosclerosis, despite the slower kinetics of TET2-heterozygous
cell expansion (fig. S5). Increased atherogenesis in 10% KO-BMT mice was paralleled by
an increase in total macrophage content in the intima, although this parameter was not
statistically significant when normalized to plaque size (fig. S6). BM genotype did not affect
lesional content of collagen or vascular smooth muscle cells (fig. S6), apoptosis (fig. S7A),
necrotic core extension (fig. S7B), or proliferation rates of total plaque cells or lesional
macrophages (fig. S7C). Overall, these data demonstrate that clonal expansion of TET2-
deficient hematopoietic cells accelerates atherogenesis in a manner independent of
Fuster et al. Page 3
Science
. Author manuscript; available in PMC 2017 August 03.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
alterations in systemic metabolism, changes in blood cell counts, or macrophage
proliferation or apoptosis in the plaque.
Consistent with their above-mentioned preferential differentiation into myeloid cells, TET2-
deficient HSPCs expanded preferentially into the macrophage population in the
atherosclerotic vascular wall. CD45.2
+
cells represented 58% of total immune cells, 62% of
macrophages, and 35% of T cells present in the aortic wall of 10% KO-BMT mice (Fig. 1, F
and G, and fig. S8). On the basis of these findings, we hypothesized that TET2-deficient
hematopoietic cells accelerate atherosclerosis mainly by generating a pool of macrophages
with enhanced proatherogenic activities. To test this possibility, we used BMT and LysM-
Cre/LoxP strategies to generate atherosclerosis-prone mice exhibiting TET2 deficiency
restricted to myeloid cells (Mye-
Tet2
-KO mice). Although this strategy led to a partial
(~80%) inactivation of
Tet2
in BM-derived macrophages (Fig. 2A), it was sufficient to
increase plaque size in the aortic root of HFHC-fed mice (Fig. 2B), with no differences in
body or spleen weight, blood monocyte counts, or glucose and cholesterol levels (fig. S9).
These results demonstrate that TET2 deficiency in myeloid cells is sufficient to promote
atherogenesis and suggest that macrophages play a major role in the accelerated
atherosclerosis associated with expansion of TET2-deficient HSPCs. However, they do not
rule out a potential contribution from other BM-derived cells. Analysis of aorta and aorta-
draining mediastinal lymph nodes showed that the expansion of TET2-deficient HSPCs does
not affect T cell numbers or aortic expression of T cell activation markers (fig. S10, A to C),
although it leads to modest changes in the frequency of various T cell subsets (fig. S10, D to
G), consistent with recent studies (18, 19). Such changes were not observed in Mye-
Tet2
-
KO mice (fig. S10, H to J). Therefore, although a contribution of TET2-deficient T cells to
the atherogenic effects of the expansion of TET2-deficient HSPCs cannot be excluded, these
data demonstrate that changes in T cells are not essential for the accelerated atherosclerosis
associated with TET2 loss of function and suggest instead a predominant role of
macrophages in this context.
We next evaluated the effects of TET2 deficiency on the function of macrophages in culture.
Consistent with the in vivo observations, TET2 deficiency did not affect macrophage
proliferation (fig. S11A), apoptosis (fig. S11B), oxidized low-density lipoprotein (oxLDL)
uptake (fig. S11C), or the expression of cholesterol trafficking regulators (fig. S11, D and E).
To evaluate whether TET2 deficiency affects proinflammatory macrophage activation, we
performed Affymetrix microarray analysis on
Tet2
−/−
macrophages and WT controls in
resting conditions and after treatment with a combination of lipopolysaccharide (LPS) and
interferon-γ (IFN-γ). Whereas no genes were differentially expressed in unstimulated
macrophages (
q
value < 0.05; fig. S12), a widespread alteration in gene expression was
found in
Tet2
−/−
macrophages after a 10-hour treatment with LPS/IFN-γ. Expression of 475
genes was altered by more than a factor of 1.5 when compared with WT macrophages (
q
<
0.05; Fig. 2C). PANTHER functional annotation software revealed that transcripts encoding
cytokine, chemokine, and signaling molecules were the top overrepresented classes altered
in the transcriptome of LPS/IFN-γ–treated TET2-deficient macrophages (Fig. 2D). Genes in
these classes with known proinflammatory actions were mostly up-regulated in TET2-
deficient macrophages (Fig. 2E). Consistent with this observation, qRT-PCR analysis
revealed that TET2-deficient macrophages exhibit markedly increased expression of
Fuster et al. Page 4
Science
. Author manuscript; available in PMC 2017 August 03.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
proinflammatory cytokines (Fig. 2F and fig. S13, A and B), chemokines (fig. S13C), and
enzymes (fig. S13D). This pattern of gene expression was also evident in
Tet2
+/−
macrophages and macrophages isolated from Mye-
Tet2
-KO mice (fig. S13, E and F). TET2
deficiency also resulted in increased interleukin-6 (IL-6) protein levels in macrophage
culture supernatants (fig. S13G). These data suggest that TET2 acts as a negative
transcriptional regulator of proinflammatory responses and are consistent with a previous
study reporting that TET2 represses LPS-induced IL-6 expression (20).
However, the situation in vivo in the atherosclerotic plaque is particularly complex, as
lesional macrophages are exposed to multiple signals simultaneously. Therefore, the anti-
inflammatory actions of TET2 in cultured macrophages were further evaluated by testing
their effect on macrophage response to a cocktail of low doses of oxLDL, tumor necrosis
factor (TNF), and IFN-γ, three stimuli present in atherosclerotic plaques. These conditions
minimized the impact of TET2 deficiency on cytokine and chemokine expression (fig. S14),
with the exception of IL-1β, which was markedly up-regulated in TET2-deficient
macrophages at all time points of oxLDL/TNF/IFN-γ stimulation (Fig. 3A). Supporting a
predominant role for IL-1β in the exacerbated atherosclerosis associated with TET2 loss of
function, gene expression analysis of the aortic arch of HFHC-fed mice revealed a
significant >twofold increase in transcript levels of IL-1β in 10% KO-BMT versus 10% WT-
BMT mice (Fig. 3B), whereas few significant differences were observed in the aortic
expression of other cytokines and chemokines or other macrophage-enriched genes (fig.
S15). 10% KO-BMT also exhibited increased IL-1β protein levels in atherosclerotic plaques
and plaque macrophages, as revealed by immunofluorescence staining coupled to confocal
microscopy (Fig. 3, C and D).
To examine the molecular mechanisms underlying the effects of TET2 deficiency on IL-1β
expression, we performed cell culture studies with LPS/IFN-γ–treated macrophages
transiently overexpressing either WT TET2 or a TET2 mutant unable to catalyze the
oxidation of 5mC to 5hmC. Both WT-TET2 and mutant-TET2 overexpression led to a ~90%
reduction in IL-1β expression and blunted differences between
Tet2
+/+
and
Tet2
−/−
macrophages (Fig. 3E), suggesting that TET2 modulates IL-1β expression independent of its
catalytic activity. Therefore, we investigated noncatalytic mechanisms of TET2-mediated
transcriptional repression. Consistent with previous studies showing that TET2 inhibits gene
transcription via histone deacetylase (HDAC)–mediated histone deacetylation (20),
treatment with the HDAC inhibitor trichostatin A (TSA) increased IL-1β expression in LPS/
IFN-γ–treated macrophages and abolished expression differences between TET2-deficient
and WT genotypes (Fig. 3F). Further supporting a role for histone deacetylation in TET2-
mediated repression of IL-1β, chromatin immunoprecipitation (ChIP)–qPCR analysis
revealed greater histone H3 acetylation at the
Il1b
gene promoter in TET2-deficient
macrophages (Fig. 3G). In contrast, treatment with TSA reduced the expression of
transcripts corresponding to other proinflammatory cytokines (such as IL-6), in agreement
with previous reports (21), and did not affect differences between genotypes (fig. S16A).
Furthermore, H3 acetylation at the
Il6
gene promoter was not affected by TET2 ablation
(fig. S16B). Overall, these data suggest that a reduction in HDAC-mediated histone
deacetylation accounts for the effects of TET2 loss of function on IL-1β expression in
macrophages, whereas alternative mechanisms contribute to its effects on other genes.
Fuster et al. Page 5
Science
. Author manuscript; available in PMC 2017 August 03.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript