| |
Acceleration of Age-Associated Methylation Patterns in HIV-1-Infected Adults.... adults appeared to be approximately 14 years older than their chronologic age
|
| |
| |
Download the PDF here
"We selected participants from the Multicenter AIDS Cohort Study (MACS), a study of the natural and treated history of HIV-1 infection in men who have sex with men [36]. There were two groups of participants in data set one: 24 of the samples were from individuals 20-24 years of age and 24 were from individuals 48-56 years of age......They compared epigenetic patterns that are strongly associated with aging to changes that occur during HIV-infection and found significant overlap in the two patterns, and used those overlapping patterns to estimate the biological age of HIV-infected, untreated adults. The researchers found that at the epigenetic level, the adults appeared to be approximately 14 years older than their chronologic age, said Jamieson, who also is director of the UCLA Flow Cytometry Core."
PLOS ONE Beth D. Jamieson et al UCLA
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0119201
Together, this data supports the theory that aging within the immune system plays an important role in aging of the overall organism, and our methylation data suggests that HIV-1 infection may be accelerating this process using similar mechanisms....the researchers suggest that HIV itself accelerates these aging related changes by more than 14 years......These results are an important first step for finding potential therapeutic approaches to mitigate the effects of both HIV and aging......our data show that the effects of HIV-1 infection at a subset of methylation sites appear to be additive with the effects of aging......"While we were surprised by the number of epigenetic changes that were significantly associated with both aging and HIV-infection, we were most surprised that the data suggests HIV-infection can accelerate aging-related epigenetic changes by 13.7 to 14.7 years," said Beth Jamieson, professor of medicine in the division of hematology/oncology at the David Geffen School of Medicine at UCLA and one of the study's senior authors. "This number is in line with both anecdotal and published data suggesting that treated HIV-infected adults can develop the diseases of aging mentioned above, approximately a decade earlier than their uninfected peers.".....Multiple studies have shown the effects of HIV-1 infection on the aging of the immune system [15,32-33]. Our own studies on telomere length and decreasing numbers of naïve T-cells showed that individuals with HIV-1 infection appeared immunologically similar to uninfected individuals twenty to thirty years older [32]. ......It has also been shown that HIV-1 infected individuals develop frailty ten years younger than uninfected individuals [17] and coronary artery calcium measurements show a coronary artery "age" that is accelerated by approximately 15 years with HIV-1 infection [53]......Thus, our data showing HIV-1 acceleration of aging by 13.7 years in data set one and 14.7 years in data set 2, as measured by methylation analysis, fits well with these other studies. As such, it would change if we used other modules or other epigenetic biomarkers of aging [14]. We have focused on module 3 as it showed the strongest positive correlation for both age and HIV-1 status.....Some of the changes in methylation in module 3 associated with HIV status can be attributed to a change in numbers of senescent CD57+CD28-CD8+ cells (S1 Table).
"We selected participants from the Multicenter AIDS Cohort Study (MACS), a study of the natural and treated history of HIV-1 infection in men who have sex with men [36]. There were two groups of participants in data set one: 24 of the samples were from individuals 20-24 years of age and 24 were from individuals 48-56 years of age. In each group, 12 of the samples were from HIV-1 seropositive (SP) men and 12 were from HIV-1 seronegative (SN) men. There were two groups of participants in data set two: 24 of the samples were from individuals 27-35 years of age and 24 were from individuals 36-56 years of age."
"Aging is associated with an increasing incidence of chronic, debilitating, diseases. While cardiovascular, skeletal, and neurodegenerative diseases are widely known and discussed in the general population, there is virtually no organ or tissue system that is not at risk......As we have documented, HIV-1-infection is also associated with aging-related changes in the CD4+ T-cell compartment [32]. There is a significant decrease in the number of CD31+ naïve CD4+ T-cells and shortening of telomeres in the overall naïve CD4+ T-cell subpopulation, rendering this cellular compartment more phenotypically similar to that of an uninfected adult 20-30 years older [32]. There is also evidence of telomere shortening within CD8+ T-cells and a significant increase in senescent CD8+ T-cells in HIV-1-infected individuals, similar to that observed in older seronegative individuals [33-34]. Together, these findings have led to the suggestion that HIV-1-infection and aging may interact in a mechanistic manner.What is missing are data regarding the mechanism(s) by which HIV-1-infection contributes to the earlier manifestation of age-inappropriate clinical outcomes and how those mechanisms overlap with, or are disparate from, the mechanisms responsible for aging-associated comorbidities in the absence of HIV-1-infection........to answer that question, researchers at the UCLA AIDS Institute and Center for AIDS Research and the Multicenter AIDS Cohort Study investigated whether the virus induces age-associated epigenetic changes - that is, changes to the DNA that in turn lead to changes in expression of gene levels without changing the inherited genetic code. These changes affect biological processes and can be brought on by environmental factors or by the aging process itself.......Using this model, we estimate that HIV-1 infection accelerates age-related methylation by approximately 13.7 years in data set 1 and 14.7 years in data set 2
We investigated whether HIV-1-infection would induce age-associated methylation changes. We evaluated DNA methylation levels at >450,000 CpG sites in peripheral blood mononuclear cells (PBMC) of young (20-35) and older (36-56) adults in two separate groups of participants. Each age group for each data set consisted of 12 HIV-1-infected and 12 age-matched HIV-1-uninfected samples for a total of 96 samples. The effects of age and HIV-1 infection on methylation at each CpG revealed a strong correlation of 0.49, p<1 x10-200 and 0.47, p<1x10-200.......We used two unique data sets to carry out a weighted correlation network analysis (WGCNA) of DNA methylation data and identified an age related co-methylation module, or aging module, that contains CpG sites that are hypermethylated with age. This is the first study that shows that an age-related co-module, a new biomarker of aging, detects accelerated aging epigenetic effects due to HIV infection"
Interestingly, we found that absolute numbers of effector/memory, activated, and senescent T-cells in both data sets positively correlated with module 3, the same module that correlated significantly with aging and HIV status. Among the many changes that occur in the immune system with aging, the accumulation of CD28-CD8+ T-cells, referred to as senescent CD8+ T-cells, is strongly correlated with age, reduced vaccine efficacy, increased autoimmunity and the development of aging-related comorbidities such as frailty, bone loss and cardiovascular disease [46-47,54]. Together with decreased T-cell responses to mitogen and reduced B-cell numbers, an inverted CD4+/CD8+ T-cell ratio and the accumulation of CD28-CD8+ T-cells was found to be associated with an increased risk of morbidities and mortalities in the OCTO/NONA studies of elderly individuals [55]. In addition to other functional and phenotypic perturbations, these CD28-CD8+ T-cells secrete IL-6 and TNF-alpha, thereby directly contributing to the systemic inflammatory environment within the elderly [47]. Inflammaging, the chronic inflammatory environment during aging, is itself associated with damage to organ systems and increased morbidity and mortality. In an intriguing proof-of-principal investigation into the role of senescent T-cells in aging related morbidity, Baker, et al. [56] demonstrated that removal of senescent cells expressing p16Ink-4a from a murine model of Progeria delayed the onset of aging-related diseases in several tissue and organ systems.
Tammy M. Rickabaugh1, Ruth M. Baxter2, Mary Sehl1,3, Janet S. Sinsheimer2,3,4, Patricia M. Hultin5, Lance E. Hultin1, Austin Quach2, Otoniel Martinez-Maza5,6, Steve Horvath2, Eric Vilain2, Beth D. Jamieson1*
1 Department of Medicine, Division of Hematology/Oncology, AIDS Institute, University of California Los Angeles, Los Angeles, California, United States of America, 2 Department of Human Genetics, University of California Los Angeles, Los Angeles, California, United States of America, 3 Biomathematics, David Geffen
School of Medicine, University of California Los Angeles, Los Angeles, California, United States of America, 4 Department of Biostatistics, School of Public Health, University of California Los Angeles, Los Angeles, California, United States of America, 5 Department of Epidemiology, School of Public Health, University of California Los Angeles, Los Angeles, California, United States of America, 6 Departments of Obstetrics and
Gynecology, and Microbiology, Immunology and Molecular Genetics, University of California Los Angeles, Los Angeles, California, United States of America
Abstract
Patients with treated HIV-1-infection experience earlier occurrence of aging-associated diseases, raising speculation that HIV-1-infection, or antiretroviral treatment, may accelerate aging. We recently described an age-related co-methylation module comprised of hundreds of CpGs; however, it is unknown whether aging and HIV-1-infection exert negative health effects through similar, or disparate, mechanisms. We investigated whether HIV-1-infection would induce age-associated methylation changes. We evaluated DNA methylation levels at >450,000 CpG sites in peripheral blood mononuclear cells (PBMC) of young (20-35) and older (36-56) adults in two separate groups of participants. Each age group for each data set consisted of 12 HIV-1-infected and 12 age-matched HIV-1-uninfected samples for a total of 96 samples. The effects of age and HIV-1 infection on methylation at each CpG revealed a strong correlation of 0.49, p<1 x10-200 and 0.47, p<1x10-200. Weighted gene correlation network analysis (WGCNA) identified 17 co-methylation modules; module 3 (ME3) was significantly correlated with age (cor=0.70) and HIV-1 status (cor=0.31). Older HIV-1+ individuals had a greater number of hypermethylated CpGs across ME3 (p=0.015). In a multivariate model, ME3 was significantly associated with age and HIV status (Data set 1: ßage= 0.007088, p=2.08 x 10-9; ßHIV= 0.099574, p=0.0011; Data set 2: ßage= 0.008762, p=1.27x 10-5; ßHIV= 0.128649, p= 0.0001). Using this model, we estimate that HIV-1 infection accelerates age-related methylation by approximately 13.7 years in data set 1 and 14.7 years in data set 2. The genes related to CpGs in ME3 are enriched for polycomb group target genes known to be involved in cell renewal and aging. The overlap between ME3 and an aging methylation module found in solid tissues is also highly significant (Fisher-exact p=5.6 x 10-6, odds ratio=1.91). These data demonstrate that HIV-1 infection is associated with methylation patterns that are similar to age-associated patterns and suggest that general aging and HIV-1 related aging work through some common cellular and molecular mechanisms. These results are an important first step for finding potential therapeutic targets and novel clinical approaches to mitigate the detrimental effects of both HIV-1-infection and aging.
Introduction
Aging is associated with an increasing incidence of chronic, debilitating, diseases. While cardiovascular, skeletal, and neurodegenerative diseases are widely known and discussed in the general population, there is virtually no organ or tissue system that is not at risk. The mechanisms underlying aging and its deleterious effects are poorly understood, but thought to be multifactorial and to involve epigenetic changes [1]. Epigenetics is the alteration of DNA through modifications that do not change the underlying nucleotide sequence [2], yet are important in controlling gene expression. There are many types of epigenetic regulation, including small RNAs, acetylation of histones and the methylation of cytosine residues [3]. Currently available array and sequencing technology has allowed genome-wide examination of methylation levels, and there now exists a body of literature showing both global and site-specific changes in methylation patterns in relation to age [4-14].
There are several lines of evidence to suggest that HIV-1-infection accelerates at least some aspects of the aging process [15-16]. Perhaps the most dramatic piece of evidence is that successfully treated HIV-1-infection is associated with a greater susceptibility to morbidities more commonly observed in older, uninfected, individuals such as frailty [17], non-Hodgkin's lymphoma [18], anal and cervical carcinomas [19-20], osteoporosis [21-22], liver [23-25] and renal impairment [26], cardiovascular disease [27-28], diabetes [29] and hypertension [29-30]. There is also evidence of faster disease progression in older HIV-1-infected adults [31].
As we have documented, HIV-1-infection is also associated with aging-related changes in the CD4+ T-cell compartment [32]. There is a significant decrease in the number of CD31+ naïve CD4+ T-cells and shortening of telomeres in the overall naïve CD4+ T-cell subpopulation, rendering this cellular compartment more phenotypically similar to that of an uninfected adult 20-30 years older [32]. There is also evidence of telomere shortening within CD8+ T-cells and a significant increase in senescent CD8+ T-cells in HIV-1-infected individuals, similar to that observed in older seronegative individuals [33-34]. Together, these findings have led to the suggestion that HIV-1-infection and aging may interact in a mechanistic manner.
To date, the vast majority of data examining the interrelationship between HIV-1-infection and aging have been obtained at the cellular and organismal level. Much of these data measure outcomes and are unable to address mechanisms or pathways. For example, the finding that HIV-1-infection renders individuals more likely to develop frailty 10 years earlier than their uninfected peers [17] is an outcome. What is missing are data regarding the mechanism(s) by which HIV-1-infection contributes to the earlier manifestation of age-inappropriate clinical outcomes and how those mechanisms overlap with, or are disparate from, the mechanisms responsible for aging-associated comorbidities in the absence of HIV-1-infection.
Using a systems biologic analysis approach we recently revealed a robustly defined age-related co-methylation module that is present in multiple human tissues, including saliva [35], blood, and brain [12]. These studies demonstrate that blood is a promising surrogate for other tissues when studying the effects of age on DNA methylation profiles [12,14] and that the aging module could be an important biomarker for detecting accelerated aging effects.
We used two unique data sets to carry out a weighted correlation network analysis (WGCNA) of DNA methylation data and identified an age related co-methylation module, or aging module, that contains CpG sites that are hypermethylated with age. This is the first study that shows that an age-related co-module, a new biomarker of aging, detects accelerated aging epigenetic effects due to HIV infection. We also show that our systems biologic analysis based on WGCNA leads to more pronounced biological insights than a standard differential methylation analysis that only considers marginal relationships between CpG sites and HIV infection. Comparison of this module to our previously found aging module [12] revealed that it can also be found in other solid tissues, notably human brain tissue, and may thus also measure organismal aging effects. These unique tools may aid in the elucidation of novel therapeutic targets for aging-related clinical diagnoses in HIV-infected and uninfected individuals.
Participants
We selected participants from the Multicenter AIDS Cohort Study (MACS), a study of the natural and treated history of HIV-1 infection in men who have sex with men [36]. There were two groups of participants in data set one: 24 of the samples were from individuals 20-24 years of age and 24 were from individuals 48-56 years of age. In each group, 12 of the samples were from HIV-1 seropositive (SP) men and 12 were from HIV-1 seronegative (SN) men. There were two groups of participants in data set two: 24 of the samples were from individuals 27-35 years of age and 24 were from individuals 36-56 years of age. In each group, 12 of the samples were from HIV-1 SP men and 12 were from HIV-1 SN men. Selection criteria included the following characteristics: age, visit date, anti-retroviral treatment (ART) naïve (self-reported during the semi-annual MACS study visits), history of smoking, and ethnicity. Each HIV-1 seropositive sample was individually matched to a seronegative control using the selection criteria.
We examined BMI data on our participants and we do not find any significant difference in BMI for our HIV-1 seronegative and HIV-infected participants. The seronegative participants had an average BMI of 23.9 and a median BMI of 23.8. The seropositive participants had an average BMI of 22.4 and a median BMI of 22.2. We also examined chronic co-infection with Hepatitis B and C, two common viruses that are known to cause chronic infection. In our first group of samples, two of the HIV-1-infected participants had chronic Hepatitis C and two of the HIV-1-infected participants had chronic Hepatitis B infection (none of the seronegatives in either group had a chronic infection).
Discussion
These are the first data to show that HIV-1 infection is associated with methylation patterns that are similar to those seen with aging in the general population. Examination of the effects of aging, and the effects of HIV, showed that aging has a stronger effect on changes in methylation. Marginal analysis of the data showed that HIV-1 infection does not have a significant global effect. However, in two separate data sets which significantly correlated with each other, a subset of CpG sites that are hypermethylated with age showed a further increase of methylation levels in individuals infected with HIV-1 (Figs. 1 and 2B). Additionally, a group of sites that show hypomethylation with age are further demethylated in the HIV-1 infected group (Figs. 1 and 2B). Thus, although we have not identified a strong methylation signature associated with HIV-1 infection, our data show that the effects of HIV-1 infection at a subset of methylation sites appear to be additive with the effects of aging.
Multiple studies have shown the effects of HIV-1 infection on the aging of the immune system [15,32-33]. Our own studies on telomere length and decreasing numbers of naïve T-cells showed that individuals with HIV-1 infection appeared immunologically similar to uninfected individuals twenty to thirty years older [32]. It has also been shown that HIV-1 infected individuals develop frailty ten years younger than uninfected individuals [17] and coronary artery calcium measurements show a coronary artery "age" that is accelerated by approximately 15 years with HIV-1 infection [53]. Thus, our data showing HIV-1 acceleration of aging by 13.7 years in data set one and 14.7 years in data set 2, as measured by methylation analysis, fits well with these other studies. As such, it would change if we used other modules or other epigenetic biomarkers of aging [14]. We have focused on module 3 as it showed the strongest positive correlation for both age and HIV-1 status.
Interestingly, we found that absolute numbers of effector/memory, activated, and senescent T-cells in both data sets positively correlated with module 3, the same module that correlated significantly with aging and HIV status. Among the many changes that occur in the immune system with aging, the accumulation of CD28-CD8+ T-cells, referred to as senescent CD8+ T-cells, is strongly correlated with age, reduced vaccine efficacy, increased autoimmunity and the development of aging-related comorbidities such as frailty, bone loss and cardiovascular disease [46-47,54]. Together with decreased T-cell responses to mitogen and reduced B-cell numbers, an inverted CD4+/CD8+ T-cell ratio and the accumulation of CD28-CD8+ T-cells was found to be associated with an increased risk of morbidities and mortalities in the OCTO/NONA studies of elderly individuals [55]. In addition to other functional and phenotypic perturbations, these CD28-CD8+ T-cells secrete IL-6 and TNF-alpha, thereby directly contributing to the systemic inflammatory environment within the elderly [47]. Inflammaging, the chronic inflammatory environment during aging, is itself associated with damage to organ systems and increased morbidity and mortality. In an intriguing proof-of-principal investigation into the role of senescent T-cells in aging related morbidity, Baker, et al. [56] demonstrated that removal of senescent cells expressing p16Ink-4a from a murine model of Progeria delayed the onset of aging-related diseases in several tissue and organ systems.
HIV-1-infection is thought to contribute to immunosenescence by the activation and expansion of CD8+ T-cells directed against both HIV-1 and CMV. The response elicited by HIV-1-infection is also highly inflammatory. Indeed, inflammation during HIV-1-infection is associated with an increased rate of progression to AIDS and mortality [47]. We recently reported that HIV-1-infection was associated with premature accumulation of an immunosenescent phenotype which in turn was associated with faster progression to AIDS [57] leading us to speculate that HIV-1-infection does recapitulate some of the aspects of aging and that these are likely through the inflammatory response and accumulation of senescent cells associated with HIV-1-infection. Some of the changes in methylation in module 3 associated with HIV status can be attributed to a change in numbers of senescent CD57+CD28-CD8+ cells (S1 Table). However, after adjusting for this subset, there was still an effect of HIV status on methylation changes in module 3, suggesting that there are other factors, such as premature aging of naïve T-cells [32], contributing to this effect. Together, this data supports the theory that aging within the immune system plays an important role in aging of the overall organism, and our methylation data suggests that HIV-1 infection may be accelerating this process using similar mechanisms.
We found several Polycomb Group Target Genes represented by CpG's in module 3. The Polycomb complexes contribute to organismal development by modifying the expression of their targets, the PCGT genes, through epigenetic modulation [58]. Expression of PCGT genes is repressed in stem cells through high levels of methylation as a mechanism of preventing differentiation [58]. Hypermethylation resulting in repressed gene expression is also well established in tumorigenesis (reviewed in [59]). The presence of multiple PCGT genes within the module 3 kME >0.85 group suggests that decreased expression of PCGT genes and the consequent return of cells to less differentiated forms contributes to the increased susceptibility to cancer observed during aging and HIV-1-infection. Most of the PCGT genes in the module 3 kME >0.85 group were represented by more than one CpG site, and, strikingly RAB32 was represented by seven CpG sites. The fact that seven CpGs associated with RAB32 show increased methylation strongly suggests that expression of this gene is suppressed in response to aging and HIV-1 infection. RAB32 is a small GTPase related to the oncogene RAS, and is involved in mitochondrial membrane dynamics and apoptosis [60]. Mitochondrial dysfunction is strongly related to aging and cancer [61], and decreased expression of RAB32 may result in disrupted mitochondrial dynamics and changes in apoptotic processes which would also contribute to the development of cancer. Indeed, hypermethylation of RAB32 has been identified in gastric and endometrial cancer [62].
We also evaluated the overlap between our module 3 and the aging module recently described by Horvath, et al. [12]. The degree of overlap with the aging module, found in human brain tissues and other solid tissues [12] is highly significant (Fisher-exact p-value = 5.6 x 10-6, odds ratio = 1.91). Interestingly, one of the overlapping genes represented by both modules (the aging module and our module 3) is RAB32, which we discussed above. These results strongly suggest that our module 3 is not specific to blood tissue. Future research should evaluate whether module 3 also reveals HIV related age acceleration effects in other solid tissues. Taken together, these data suggest that HIV-1-infection does accelerate some aspects of aging and that general aging and HIV-1 related aging work through at least some common mechanisms. These results are an important first step for finding potential therapeutic approaches to mitigate the effects of both HIV and aging.
|
|
| |
| |
|
|
|
