Clonal hematopoiesis of indeterminate potential (CHIP) is defined as the presence of myeloid cancer-associated, somatic mutations with a variant allele frequency (VAF) of ≥2% in the hematopoietic cells of individuals without hematologic malignancy.1 Its prevalence steeply increases with age, affecting about 15-20% of individuals of ≥60 years, indicating an association with aging processes. DNMT3A, TET2, and ASXL1 are the most frequently mutated genes and can all be classified as epigenetic modifiers.2-4 CHIP is associated with various age-related, adverse health conditions, e.g., cardiovascular diseases and stroke.2 Additionally, CHIP bares the inherent risk of progression to hematologic malignancy.1-3 Together, these adverse effects translate into an increased overall mortality.2,3
There are nine well-accepted hallmarks of aging, which comprise genomic instability and epigenetic alterations.5 As CHIP can be interpreted as genomic instability, matching CHIP state with other established markers of aging is of high interest. DNA methylation patterns change over an individual’s life, allowing the establishment of “epigenetic clocks” that estimate epigenetic age through regression of methylation ratios at predefined CpG sites on chronological age.6-8 Epigenetic age acceleration has been shown to correlate with various age-associated conditions and overall mortality (reviewed by Horvath and Raj6). Due to its interrelatedness with older age and aging-associated conditions, we asked whether the presence of CHIP itself is a surrogate marker of an enhanced biological aging process. We therefore determined epigenetic age measures using an established adaption of the epigenetic clock9 originally published by Vidal-Bralo and colleagues8 in a cohort of 381 individuals with previously determined CHIP status.10
Blood samples were collected from 417 inpatients from the Departments of Cardiology, Nephrology, Musculoskeletal Surgery, and Oncology at Charité – Universitätsmedizin Berlin as reported previously.10 Tumor patients were included before starting (radio-)chemotherapy. Individuals with hematologic malignancy were excluded. The study was conducted in accordance with the Declaration of Helsinki and patients gave informed consent.
Whole blood DNA of the study participants was screened for CHIP (VAF ≥2%) as performed and published previously.10 Methylation status of CpG sites of interest was measured according to a protocol for the methylation-sensitive single-nucleotide-primer-extension method (ms-SNuPE) developed by Vidal-Bralo and colleagues8 and adapted by Vetter and colleagues.9 DNA methylation (DNAm) age was calculated with the 7-CpG clock using a regression model from multiple linear regression of chronological age on methylation fractions of the Berlin Aging Study II cohort. Age acceleration (AA) was determined as the residuals from regressing DNAm age on chronological age. Intrinsic epigenetic age acceleration (IEAA) was determined as the residuals from regressing DNAm age on chronological age and blood cell counts.9 Alpha-level was set at 0.05, and IBM SPSS Statistics software (IBM, USA) was used.
Methylation ratios were successfully determined for 381 of 417 individuals (=91%) (Online Supplementary Figure S1A). Mean chronological age was 74.7 years (standard deviation [SD]= ±7.9 years, range, 55-98 years) and 45% of individuals were female. Our cohort was characterized by a high prevalence of clinically relevant morbidities as detailed in Table 1. Of the 381 individuals, 106 (=28%) had evidence of CHIP. Median VAF was 6% (Figure 1A), and an overview of all mutations is presented in Figure 1B and the Online Supplementary Table S2. The mean DNAm age was 76.2 years (±9.7 years), mean DNAm age acceleration was ±8.4 years, and mean IEAA was ±8.1 years.
AA (n=381) and IEAA (n=297) were analyzed with regard to baseline and clinical characteristics. In line with existing data6, AA and IEAA were significantly higher in males than in females (AA: P=0.006; IEAA: P=0.001) (Figure 1C). No statistically relevant differences in AA could be noticed with regard to medical conditions, except for hyperuricaemia (AA: P=0.001; IEAA: P<0.001) and kidney failure requiring hemodialysis (AA: P<0.001; IEAA: P=0.018) (Online Supplementary Table S1).
In our study, the general presence of CHIP was not associated with age acceleration (AA: P= 0.183; IEAA: P=0.513), nor were individual CHIP mutations (e.g., DNMT3A, TET2, ASXL1), functional CHIP groups or the presence of more than one mutation when compared to individuals without CHIP (Figure 2A to C; Online Supplementary Figure S1A to C). In order to obtain a more differentiated view, we investigated the relationship between CHIP clone size and epigenetic age acceleration. First, patients with CHIP were subdivided into three groups according to clone size.
Clones with a VAF <10% were considered “small to intermediate”, while clones with a VAF of ≥10% were considered “large”. A third group was defined by the absence of CHIP. Using ANOVA, significant differences between the individual groups were detected (P=0.023). This finding remained significant after correction for sex and age using ANCOVA (P=0.022). Differences in AA and IEAA were most pronounced when comparing individuals without CHIP to those with large clones (contrast analysis: P=0.01) (Figure 2D; Online Supplementary Figure S2D). The difference in mean AA amounted to 4.5 years. A comparison of AA between individuals without CHIP and individuals with clones of small to intermediate size showed no differences (P=0.718). Next, we sought to identify to which extent smaller clones still affect AA/IEAA. The lowest VAF with significant differences in AA/IEAA compared to AA/IEAA in individuals without CHIP was ≥5% (Kruskal-Wallis-Test: P=0.008; Mann-Whitney-Test: P=0.008). This effect was likewise significant for IEAA (Kruskal-Wallis-Test: P=0.016; Mann-Whitney-Test: P=0.044).
In order to further explore the relation between CHIP clone size and DNAm AA while accounting for the continuous character of VAF, we performed Spearman’s correlations for DNAm age acceleration measures and VAF of individuals with CHIP. For individuals with more than one CHIP mutation, VAF of the largest clone was used. Our analysis revealed a statistically significant positive correlation between VAF and AA/IEAA (AA: RS=0.328, P<0.001, IEAA: RS=0.420, P<0.001) (Figure 2E and F). This association remained significant after correction for sex and age (AA: RS=0.323, P<0.001, IEAA: RS=0.411, P<0.001), further supporting a connection between CHIP clone size and accelerated aging processes.
Two previous studies consistently described a significant association of CHIP and epigenetic age acceleration based on methylation array data.11,12 However, the whole genome sequencing (WGS) methodology applied for detection of CHIP in these investigations is characterized by low coverage, prohibiting the reliable detection of small to intermediate sized CHIP clones. This is reflected by comparably low CHIP frequencies of about 6% in both studies and no CHIP clones with a VAF <6% were included in Nachun and colleagues.11 However, clones with a VAF of ≥2 and <6% account for a substantial part of CHIP.13 In our cohort, 54% of CHIP carriers had CHIP mutations with VAF solely between 2-6%, which is within the expected frequency range.
Our key finding is the correlation of CHIP clone size with epigenetic age acceleration, showing that an increase in clone size is associated with increased epigenetic age acceleration. In contrast to the two previous studies, CHIP clones were identified by targeted sequencing with robust coverage (average depth: 2,400x), allowing reliable detection of small- and intermediate-sized CHIP clones. We could therefore for the first time demonstrate that a significant epigenetic age acceleration can only be detected with CHIP clones of a VAF of 5% or greater, while smaller clones – contributing to a relevant proportion of CHIP in general – do not seem to be associated with acceleration of the epigenetic age. This association matches the results of multiple studies that report an increasing likelihood of CHIP-associated adverse conditions with ascending VAF.2,13 To the best of our knowledge, the CpG of the clock implemented here are no major targets of the frequently mutated DNMT3A14 and TET2.15 Accelerated epigenetic age in CHIP carriers can therefore be interpreted as a reflection of actual accelerated biological aging rather than being an epiphenomenon of altered methylation processes caused by mutations. However, large clones might result from a predominant pool of HSC, possibly implicating higher cycle numbers and mitotic aging. Even though mitotic and epigenetic aging probably do not address the same aging processes,9,16 we acknowledge that further research is needed to explore these relations.
Compared to the two previous studies that used methylation array based epigenetic clocks (e.g., by Horvath), our epigenetic age estimations relied on relatively few CpG sites. Nevertheless, our clock performs comparably well and high intercorrelation with Horvath clock was shown.17 Unlike the previous studies, we could not detect an association of individual mutations with a particularly accelerated epigenetic age. However, due to rather small sample sizes in our subgroups, we cannot fully exclude type II errors in this place.
While the study population is heterogeneous with respect to the underlying medical conditions, the overall disease burden of the study cohort is above the population average. Therefore, the study design is not suited to evaluate the impact of individual medical conditions on epigenetic aging. While a generally high level of comorbidities likely restores comparability with regard to CHIP as an independent factor, our study is not powered to investigate associations between medical conditions and combined CHIP/AA status. In conclusion, our study revealed a correlation of CHIP clone size and accelerated epigenetic aging. This finding matches and extends our knowledge on the role of CHIP clone size and sets the foundation for future investigations exploring the interrelatedness of CHIP and aging.
- Received October 14, 2021
- Accepted February 21, 2022
JF, VV, CMA, TL, ID, and MF declare no conflicts of interest. LB received honoraria from Seattle Genetics, Sanofi, Astellas, Amgen, consultancy fee from Gilead, Hexal, and Menarini, consultancy fee and Honoraria Abbvie, BMS/Celgene, Daiichi Sankyo, Janssen, Jazz Pharmaceuticals, Novartis and Pfizer, and research funding from Bayer and Jazz Pharmaceuticals. FD reports personal fees from AbbVie, Astra Zeneca, Gilead, Novartis, and Roche outside the submitted work.
JF and VV performed experiments, analyzed and interpreted data; MF, ID, and FD designed and supervised research and experiments, CMA collected samples and clinical data, performed experiments, analyzed and interpreted data; JF and MF wrote the manuscript; VV, CMA, TL, LB, FD and ID revised the manuscript. All authors approved the final version.
Data sharing statement
The data that support the findings of this study are available from the corresponding author, upon reasonable request.
- Steensma DP, Bejar R, Jaiswal S. Clonal hematopoiesis of indeterminate potential and its distinction from myelodysplastic syndromes. Blood. 2015; 126(1):9-16. Google Scholar
- Jaiswal S, Fontanillas P, Flannick J. Age-related clonal hematopoiesis associated with adverse outcomes. N Engl J Med. 2014; 371(26):2488-2498. Google Scholar
- Genovese G, Kähler AK, Handsaker RE. Clonal hematopoiesis and blood-cancer risk inferred from blood DNA sequence. N Engl J Med. 2014; 371(26):2477-2487. Google Scholar
- Frick M, Chan W, Arends CM. Role of donor clonal hematopoiesis in allogeneic hematopoietic stem-cell transplantation. J Clin Oncol. 2019; 37(5):375-385. Google Scholar
- Lopez-Otin C, Blasco MA, Partridge L, Serrano M, Kroemer G.. The hallmarks of aging. Cell. 2013; 153(6):1194-1217. Google Scholar
- Horvath S, Raj K.. DNA methylation-based biomarkers and the epigenetic clock theory of ageing. Nat Rev Genet. 2018; 19(6):371-384. Google Scholar
- Horvath S. DNA methylation age of human tissues and cell types. Genome Biol. 2013; 14(10):R115. Google Scholar
- Vidal-Bralo L, Lopez-Golan Y, Gonzalez A.. Simplified assay for epigenetic age estimation in whole blood of adults. Front Genet. 2016; 7:126. Google Scholar
- Vetter VM, Meyer A, Karbasiyan M, Steinhagen-Thiessen E, Hopfenmuller W, Demuth I.. Epigenetic clock and relative telomere length represent largely different aspects of aging in the Berlin Aging Study II (BASE-II). J Gerontol A Biol Sci Med Sci. 2019; 74(1):27-32. Google Scholar
- Arends CM, Galan-Sousa J, Hoyer K. Hematopoietic lineage distribution and evolutionary dynamics of clonal hematopoiesis. Leukemia. 2018; 32(9):1908-1919. Google Scholar
- Nachun D, Lu AT, Bick AG. Clonal hematopoiesis associated with epigenetic aging and clinical outcomes. Aging Cell. 2021; 20(6):e13366. Google Scholar
- Robertson NA, Hillary RF, McCartney DL. Age-related clonal haemopoiesis is associated with increased epigenetic age. Curr Biol. 2019; 29(16):R786-R787. Google Scholar
- Abelson S, Collord G, Ng SWK. Prediction of acute myeloid leukaemia risk in healthy individuals. Nature. 2018; 559(7714):400-404. Google Scholar
- Qu Y, Lennartsson A, Gaidzik VI. Differential methylation in CN-AML preferentially targets non-CGI regions and is dictated by DNMT3A mutational status and associated with predominant hypomethylation of HOX genes. Epigenetics. 2014; 9(8):1108-1119. Google Scholar
- Fazila A, Vasu P, Jesper C. Genome-wide profiling identifies a DNA methylation signature that associates with TET2 mutations in diffuse large B-cell lymphoma. Haematologica. 2013; 98(12):1912-1920. Google Scholar
- Marioni RE, Harris SE, Shah S. The epigenetic clock and telomere length are independently associated with chronological age and mortality. Int J Epidemiol. 2018; 45(2):424-432. Google Scholar
- Vetter VM, Kalies CH, Sommerer Y. Relationship between five epigenetic clocks, telomere length and functional capacity assessed in older adults: cross-sectional and longitudinal analyses. J Gerontol A Biol Sci Med Sci. 2022. Google Scholar
Figures & Tables
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.