Advertisement
Research LetterHematologyInflammationMetabolism
Open Access | 
10.1172/JCI197100
Clonal hematopoiesis driven by Dnmt3a mutations promotes metabolic disease development in mice
Bowen Yan, Qingchen Yuan, Marco M. Buttigieg, Prabhjot Kaur, Annalisse R. McKee, Daniil E. Shabashvili, Caitlyn Vlasschaert, Alexander G. Bick, Michael J. Rauh, and Olga A. Guryanova
1University of Florida College of Medicine, Gainesville, Florida, USA.
2University of Florida Health Cancer Center, Gainesville, Florida, USA.
3Queen’s University, Kingston, Ontario, Canada.
4Vanderbilt University Medical Center, Nashville, Tennessee, USA.
Address correspondence to: Olga A. Guryanova, 1200 Newell Drive, PO Box 100267, Gainesville, Florida, 32610, USA. Phone: 352.294.8590; Email: oguryanova@ufl.edu.
Find articles by Yan, B. in: PubMed | Google Scholar
1University of Florida College of Medicine, Gainesville, Florida, USA.
2University of Florida Health Cancer Center, Gainesville, Florida, USA.
3Queen’s University, Kingston, Ontario, Canada.
4Vanderbilt University Medical Center, Nashville, Tennessee, USA.
Address correspondence to: Olga A. Guryanova, 1200 Newell Drive, PO Box 100267, Gainesville, Florida, 32610, USA. Phone: 352.294.8590; Email: oguryanova@ufl.edu.
Find articles by Yuan, Q. in: PubMed | Google Scholar
1University of Florida College of Medicine, Gainesville, Florida, USA.
2University of Florida Health Cancer Center, Gainesville, Florida, USA.
3Queen’s University, Kingston, Ontario, Canada.
4Vanderbilt University Medical Center, Nashville, Tennessee, USA.
Address correspondence to: Olga A. Guryanova, 1200 Newell Drive, PO Box 100267, Gainesville, Florida, 32610, USA. Phone: 352.294.8590; Email: oguryanova@ufl.edu.
Find articles by Buttigieg, M. in: PubMed | Google Scholar
1University of Florida College of Medicine, Gainesville, Florida, USA.
2University of Florida Health Cancer Center, Gainesville, Florida, USA.
3Queen’s University, Kingston, Ontario, Canada.
4Vanderbilt University Medical Center, Nashville, Tennessee, USA.
Address correspondence to: Olga A. Guryanova, 1200 Newell Drive, PO Box 100267, Gainesville, Florida, 32610, USA. Phone: 352.294.8590; Email: oguryanova@ufl.edu.
Find articles by Kaur, P. in: PubMed | Google Scholar
1University of Florida College of Medicine, Gainesville, Florida, USA.
2University of Florida Health Cancer Center, Gainesville, Florida, USA.
3Queen’s University, Kingston, Ontario, Canada.
4Vanderbilt University Medical Center, Nashville, Tennessee, USA.
Address correspondence to: Olga A. Guryanova, 1200 Newell Drive, PO Box 100267, Gainesville, Florida, 32610, USA. Phone: 352.294.8590; Email: oguryanova@ufl.edu.
Find articles by McKee, A. in: PubMed | Google Scholar
1University of Florida College of Medicine, Gainesville, Florida, USA.
2University of Florida Health Cancer Center, Gainesville, Florida, USA.
3Queen’s University, Kingston, Ontario, Canada.
4Vanderbilt University Medical Center, Nashville, Tennessee, USA.
Address correspondence to: Olga A. Guryanova, 1200 Newell Drive, PO Box 100267, Gainesville, Florida, 32610, USA. Phone: 352.294.8590; Email: oguryanova@ufl.edu.
Find articles by Shabashvili, D. in: PubMed | Google Scholar
1University of Florida College of Medicine, Gainesville, Florida, USA.
2University of Florida Health Cancer Center, Gainesville, Florida, USA.
3Queen’s University, Kingston, Ontario, Canada.
4Vanderbilt University Medical Center, Nashville, Tennessee, USA.
Address correspondence to: Olga A. Guryanova, 1200 Newell Drive, PO Box 100267, Gainesville, Florida, 32610, USA. Phone: 352.294.8590; Email: oguryanova@ufl.edu.
Find articles by Vlasschaert, C. in: PubMed | Google Scholar
1University of Florida College of Medicine, Gainesville, Florida, USA.
2University of Florida Health Cancer Center, Gainesville, Florida, USA.
3Queen’s University, Kingston, Ontario, Canada.
4Vanderbilt University Medical Center, Nashville, Tennessee, USA.
Address correspondence to: Olga A. Guryanova, 1200 Newell Drive, PO Box 100267, Gainesville, Florida, 32610, USA. Phone: 352.294.8590; Email: oguryanova@ufl.edu.
Find articles by
Bick, A.
in:
PubMed
|
Google Scholar
|
1University of Florida College of Medicine, Gainesville, Florida, USA.
2University of Florida Health Cancer Center, Gainesville, Florida, USA.
3Queen’s University, Kingston, Ontario, Canada.
4Vanderbilt University Medical Center, Nashville, Tennessee, USA.
Address correspondence to: Olga A. Guryanova, 1200 Newell Drive, PO Box 100267, Gainesville, Florida, 32610, USA. Phone: 352.294.8590; Email: oguryanova@ufl.edu.
Find articles by Rauh, M. in: PubMed | Google Scholar
1University of Florida College of Medicine, Gainesville, Florida, USA.
2University of Florida Health Cancer Center, Gainesville, Florida, USA.
3Queen’s University, Kingston, Ontario, Canada.
4Vanderbilt University Medical Center, Nashville, Tennessee, USA.
Address correspondence to: Olga A. Guryanova, 1200 Newell Drive, PO Box 100267, Gainesville, Florida, 32610, USA. Phone: 352.294.8590; Email: oguryanova@ufl.edu.
Find articles by Guryanova, O. in: PubMed | Google Scholar
Published September 30, 2025 - More info
J Clin Invest. 2025;135(23):e197100. https://doi.org/10.1172/JCI197100.
© 2025 Yan et al. This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
To the Editor: Clonal hematopoiesis (CH) is associated with an increased risk of nonhematologic chronic diseases, including metabolic disorders, yet causality remains poorly defined. DNA methyltransferase 3A (DNMT3A) is the most altered gene in CH, commonly through monoallelic loss-of-function (LOF) and Arg882His (RH) mutations. Here, we demonstrate in a mouse model that CH driven by Dnmt3a RH and especially LOF promotes obesity, diabetes, and chronic liver disease, effects that are further exacerbated by high-fat diet (HFD).
CH is defined as expansion of a blood cell clone marked with somatic mutations absent diagnosis of hematologic malignancies. In addition to a risk of future leukemia, CH is notably associated with chronic nonhematologic diseases, such as inflammatory disorders and cardiovascular disease (1). Epidemiological studies found that CH, especially when driven by TET2 loss, was enriched in individuals who were overweight and those with type 2 diabetes (T2D) and chronic liver disease (2, 3). Whether CH is a cause or a consequence of these comorbidities — a question of high translational significance — is incompletely understood.
DNMT3A is the most mutated gene in CH (4). While the RH variant is enriched in acute myeloid leukemia, most CH-related DNMT3A alterations are LOF (5), indicating they may not be fully mechanistically equivalent. The specific impact of CH driven by DNMT3A(LOF) versus DNMT3A(RH) on chronic disease development has not been investigated.
In meta-analysis across nonobese individuals in the UK Biobank and All of Us cohorts, we found that CH, including DNMT3A CH, was associated with incident obesity risk (DNMT3A CH HR = 1.14 [95% CI, 1.01–1.27], Figure 1A). To elucidate the functional impact of CH with DNMT3A LOF and RH mutations on metabolic disease development, we created a mouse bone marrow transplantation–based chimeric model with 20% of Dnmt3a+/– (representing LOF), Dnmt3a+/R878H (corresponding to human RH), or WT control cells (marked by pan-leukocytic CD45.2) mixed with 80% WT support (CD45.1) to mimic a clinically meaningful large CH clone (Figure 1B and Supplemental Figure 1, A and B; supplemental material available online with this article; https://doi.org/10.1172/JCI197100DS1). Eight weeks after transplantation, to allow hematopoietic reconstitution, and after confirming engraftment, mice were randomized to high-fat high-glucose Western diet (HFD) or normal chow groups. Animals on control chow harboring Dnmt3a+/– hematopoietic cells exhibited faster body weight gain (1.7-fold, P < 0.001) and a moderately higher food intake compared with Dnmt3aWT-engrafted controls (Figure 1, C and D, and Supplemental Figure 1, C–E), accompanied by subcutaneous white adipocyte hypertrophy (Supplemental Figure 1F and Figure 1E). HFD accentuated bodyweight increase and adipocyte hypertrophy, particularly in the Dnmt3a+/–-CH group (Figure 1, C and E, and Supplemental Figure 1, D and F), which became overweight and obese earlier than Dnmt3aWT controls (Supplemental Figure 1E). Furthermore, Dnmt3a+/–-CH animals exhibited higher plasma leptin (P = 0.011) and resistin levels (P = 0.048) after 6 weeks of HFD, consistent with increased adiposity (Figure 1F and Supplemental Figure 1G). Even without HFD challenge, plasma leptin was mildly elevated in both Dnmt3a+/– and Dnmt3a+/RH-CH animals 6 months after transplantation (corresponding to 4 months after diet randomization) (Supplemental Figure 1, H and I), suggesting a direct obesogenic effect of Dnmt3a CH.
Figure 1Experimental Dnmt3a CH promotes obesity, impaired glucose metabolism, inflammation, and severe steatohepatitis. (A) CH and incident obesity risk; Cox proportional hazard regression. (B) Experimental workflow. (C) Body weight gain; mixed-effects analysis vs. Dnmt3a+/+ control; n = 7–8. (D) Food intake, averaged per cage. (E) Adipocyte size (SuperPlot, n = 3/group, 2 fields/animal); unpaired 2-tailed t test. (F–H) Plasma leptin (F), fasting glucose (G), and insulin (H) levels on HFD, plotted as quartiles. (I) Blood glucose levels after glucose load in fasted animals; mean ± SEM; 2-tailed Student’s t test vs. Dnmt3a+/+ controls. (J) Inflammatory monocytes in donor (CD45.2) or wild-type competitor/host (CD45.1) myeloid cells in spleen; paired 2-tailed Student’s t test. (K) Plasma MIP-1α and CXCL1; unpaired 2-tailed t test. (L) Liver histopathology (H&E; black arrow, acidophilic body; white arrows, ballooned hepatocytes; yellow arrows, inflammatory infiltrates; scale bar: 50 μm). (M) Steatohepatitis severity score (n = 3/group, 5 fields/animal). (N) Liver fibrosis (Masson’s trichrome staining) and percentage of positive area; unpaired 2-tailed t test with Welch’s correction (SuperPlot, n = 3/group, 4 fields/animal).
In line with a well-established link between overweight, insulin resistance, and diabetes, when maintained on HFD these mice exhibited a greater propensity toward impaired glucose metabolism, with elevated fasting blood glucose and insulin levels and impaired glucose tolerance (Figure 1, G–I, and Supplemental Figure 1J). On normal chow, almost half of Dnmt3a+/–-CH mice had their fasting glucose reach prediabetic levels 7 months after transplantation despite unperturbed plasma insulin and glucose tolerance (Supplemental Figure 1, K–M). These findings suggest that Dnmt3a CH, particularly that driven by Dnmt3a LOF, directly contributes to metabolic dysfunction similar to obesity and T2D.
Given the emerging link between CH, inflammation, and changes in myelopoiesis (6), next we examined lineage composition in the hematopoietic system and cytokine profiles. Dnmt3a+/–-CH mice showed increased abundance of donor-derived (CD45.2) inflammatory monocytes in spleens and peripheral blood compared with both wild-type competitor cells (CD45.1) within the same animal and with Dnmt3aWT-CH (CD45.2) controls, a finding that was most pronounced under HFD; the Dnmt3a+/RH-CH group exhibited an intermediate phenotype (Figure 1J and Supplemental Figure 2, A and B). Consistently, after 10 weeks of HFD, Dnmt3a+/–-CH mice had elevated inflammation-related cytokines (MIP-1α, CXCL1, IL-1α, and IL-15) compared with Dnmt3aWT-engrafted controls, although levels of classical proinflammatory IL-1β and IL-6 were less perturbed (Figure 1K and Supplemental Figure 2C).
The meta-inflammatory state was further evident in the livers of Dnmt3a+/–- and Dnmt3a+/RH-CH animals. On normal chow, both Dnmt3a-CH groups exhibited greater lobular hepatic inflammation, with prominent immune infiltrate, compared with Dnmt3aWT animals (Figure 1L and Supplemental Figure 2, D and E). Furthermore, Dnmt3a+/–-CH animals developed signs of metabolic dysfunction–associated steatotic liver disease (MASLD) characterized by larger and more numerous macrovesicular fat droplets (Figure 1L and Supplemental Figure 2F). Under HFD, the Dnmt3a+/–-CH group progressed to severe MASLD, with marked hepatitis reflected by high nonalcoholic fatty liver disease activity score (3) and extensive fibrosis (Figure 1, L–N).
Collectively, our epidemiological and experimental findings indicate that Dnmt3a-CH promotes the development of obesity and metabolic disease. In mice, these phenotypes are exacerbated by HFD, most strongly in the Dnmt3a+/– context. While previous studies focused on inflammation and obesity driving expansion of the CH clone, our findings indicate this relationship is bidirectional, wherein CH directly contributes to obesity and metabolic syndrome, suggesting a vicious molecular etiology cycle. Our study highlights the differential impact of various CH mutations, urging detailed mutation- and gene-specific investigation of CH in human chronic disease pathogenesis, risk stratification, and mitigation through pharmacologic, lifestyle, and dietary interventions.
For detailed methods, information regarding sex as a biological variable, statistics, study approval, author contributions, and acknowledgments, see the Supplemental Methods.
Copyright © 2025 American Society for Clinical Investigation
ISSN: 0021-9738 (print), 1558-8238 (online)