Julie AE Nelson, PhD, UNC Center for AIDS Research, CB 7295, Chapel Hill NC 27599, USA. Tel: +1 919 843 9252; fax: +1 919 966 9873; e-mail: firstname.lastname@example.org
The p16INK4a tumor suppressor gene is a mediator of cellular senescence and has been suggested to be a biomarker of ‘molecular’ age in several tissues including T cells. To determine the association of both active and suppressed HIV infection with T-cell aging, T-cell p16INK4a expression was compared between 60 HIV+ suppressed subjects, 23 HIV+ untreated subjects, and 18 contemporaneously collected HIV-negative controls, as well as 148 HIV-negative historical samples. Expression did not correlate with chronologic age in untreated HIV+ patients, consistent with an effect of active HIV replication on p16INK4a expression. In patients on cART with suppressed viral loads, however, p16INK4a levels were similar to uninfected controls and correlated with chronologic age, with a trend toward an inverse correlation with CD4 count. These data show that p16INK4a is a reliable biomarker of T-cell aging in HIV+ patients with suppressed viral loads and suggest that poor CD4 cell recovery on cART may be associated with increased T-cell expression of p16INK4a, a marker of cellular senescence.
Assessment of the impact of age, chronic HIV-1 infection and persistent immune activation on immune function in HIV+ individuals remain critically important despite effective combination antiretroviral therapy (cART). Immune function and T-cell replicative capacity decline with age (Haynes & Maue, 2009). If active HIV replication (prior to cART) and persistent inflammation and activation while on cART accelerate this decline, then older HIV+ individuals may be at greater risk of the consequences of accelerated immune aging (aka ‘immunosenescence’), and younger successfully treated HIV+ individuals may have immune systems more comparable to older uninfected individuals.
Tumor suppressor mechanisms can contribute to aging of the immune system by decreasing replicative capacity in immune cells (Liu & Sharpless, 2009). The p16INK4a tumor suppressor gene induces a permanent growth arrest termed ‘cellular senescence’, and expression of p16INK4a increases exponentially with age in most tissues in every mammalian species thus far tested (reviewed by Sharpless & DePinho, 2007). Expression of p16INK4a is not an epiphenomenon of aging but appears to play a causal role in the age-associated replicative decline of several tissues, including T cells (Liu et al., 2011). Ablation of p16INK4a-expressing cells in genetically engineered mice delays some age-related phenotypes such as cataracts and sarcopenia (Baker et al., 2011).
Expression of p16INK4a in human peripheral T cells increases with chronologic age as well as age-modifying factors and interventions (e.g. smoking and sedentary lifestyle) (Liu et al., 2009; Muss et al., 2011), but the effects of HIV infection on this biomarker are unknown. Given that chronic HIV infection may induce some features of immunosenescence (Deeks, 2011), we sought to determine the association of active and suppressed HIV replication with T-cell p16INK4a expression in infected patients.
Results and discussion
To determine the association of untreated HIV infection with p16INK4a levels in T cells, we compared untreated HIV+ patients to HIV-negative controls (see Appendix S1). As shown in Fig. 1A, there was a significantly higher level of age-adjusted p16INK4a expression in HIV+ patients compared to HIV-negative controls (P = 0.003). Because of this marked increase in expression, there was no correlation between chronological age and p16INK4a levels in untreated HIV+ patients (r = 0.19 [95% CI −0.24, 0.56], P = 0.35). These data suggest that p16INK4a expression in untreated HIV+ patients is driven by HIV replication rather than chronological age. We next compared T-cell expression of p16INK4a in HIV+ patients on cART with suppressed viral loads to expression in HIV-negative controls. There was no significant difference in age-adjusted p16INK4a expression between HIV+ suppressed and HIV-negative individuals (P = 0.49) (Fig. 1B). Chronological age and p16INK4a levels were correlated in suppressed HIV+ patients (r = 0.29 [95% CI 0.04, 0.51], P = 0.02), indicating that control of viral replication re-establishes this relationship. Comparison of the suppressed and untreated HIV+ patients demonstrated a significant difference in age-adjusted p16INK4a levels (P = 0.04); however, this difference was only marginally significant after adjusting for smoking and CD4 count (P = 0.10), and not significant after adjusting for smoking and CD4/CD8 ratio (P = 0.39), which may be a marker of ongoing immune activation.
A multivariate linear regression model was fit with age, smoking and CD4 as predictor variables and p16INK4a as the response variable in HIV+ patients with suppressed viral loads. There was a significant linear association between age and p16INK4a (partial correlation r = 0.32 [95% CI 0.07, 0.53], P = 0.01), and smoking was associated with a 0.56 (95% CI 0.00, 1.12, P = 0.05) increase in p16INK4a expression. Gender and race were not significantly associated with p16INK4a (P > 0.2 for either variable) after adjusting for age and smoking, consistent with previous work (Liu et al., 2009). After adjusting for age and smoking, there was a trend toward a negative association between CD4 and p16INK4a expression in treated patients (partial correlation r = −0.21 [95% CI −0.44, 0.05], P = 0.10), although this association was weaker than in untreated HIV+ patients (Fig. 2).
These data suggest that untreated HIV replication leads to p16INK4a expression that is substantially above age-associated levels in uninfected individuals. Other stimuli have also been reported to induce p16INK4a expression in an age-independent fashion (e.g. tobacco and cytotoxic chemotherapy). Given recent work suggesting that p16INK4a expression compromises T-cell replicative capacity (Lemster et al., 2008; Liu et al., 2011), these results are consistent with the possibility that HIV replication is associated with premature T-cell senescence and may explain the observation that active HIV replication is associated with immune dysfunction over and above the effect on CD4 depletion (Valdez et al., 2000). The possible contribution to p16INK4a expression from relative CD8+ population enrichment and activation in untreated HIV+ individuals cannot be excluded. Our data also suggest that effective antiretroviral treatment that suppresses HIV replication results in p16INK4a levels similar to age-adjusted p16INK4a expression levels in uninfected individuals. Our data are consistent with the murine finding that p16INK4a expression leads to reduced T-cell replicative capacity with aging (Liu et al., 2011), because p16INK4a expression was marginally associated with lower CD4 counts even when controlling for age in treated patients. Investigating the relationship of this marker of immune aging with markers of immune activation and inflammation that have been shown to be elevated in many HIV+ patients suppressed on antiretroviral therapy may be helpful in understanding HIV-associated immune senescence (Deeks, 2011). Ultimately, if validated and correlated with other markers of immune health and function, p16INK4a levels could be used to identify HIV+ patients with premature immune aging, identify risk factors for premature loss of immune function, and to test therapeutic strategies to prevent or reverse HIV induced premature immunosenescence.
We thank Scott Gremillion, Christina Pieringer, and Tyler Hawkins for technical assistance and Kathleen Tompkins for recruiting HIV+ patients. This work was funded by the UNC CFAR (P30 AI50410) and by R01 AG024379 (to NES).