Pathogen-specific T cell susceptibility to HIV influences the natural history of opportunistic infections

Abstract

During HIV infection, the timing of opportunistic infections is not always associated with severity of CD4 T cell depletion and different opportunistic pathogens reactivate at different CD4 T cell thresholds. Here we review how differences in the phenotype and function of pathogen-specific CD4 T cells influence susceptibility to HIV infection. By focusing on three common opportunistic infections (Mycobacterium tuberculosis, human papillomavirus, and cytomegalovirus) we examine how differential depletion of pathogen-specific CD4 T cells impacts the natural history of these pathogens in HIV infection. A broader understanding of this relationship can better inform treatment strategies against co-pathogens.

Introduction

Chronic, untreated HIV infection is characterized by progressive depletion of CD4 T cells increasing the risk of co-infection with pathogens typically controlled by innate and adaptive immune responses. While the mechanisms leading to CD4 T cell depletion are complex, susceptibility to opportunistic infections has generally been associated with progressive declines in CD4 T cell counts. However, the timing of opportunistic infections is not always associated with severity of CD4 depletion. Tuberculosis (TB) infection can reactivate when CD4 T cell counts are high and infection with Candida species can occur at any CD4 T cell count (Figure 1).

Figure 1. Incidence of select opportunistic infections stratified by CD4 count.

Data is adapted from several cohort studies of HIV positive individuals.(Anglaret et al., 2012; Minga et al., 2008; Mocroft et al., 2013; Mocroft et al., 1998; Munoz et al., 1993; Yazdanpanah et al., 2001) The bar color indicates the approximate incidence rate (IR) per 100 person-years for each respective opportunistic pathogen. Dark orange represents IR >10, orange = IR between 5 and 10, light orange = IR between 1 and 5, and white = IR <1. As CD4 counts increase, differences in the incidence rates of opportunistic infections emerge. The incidence of CMV EOD is rare when CD4 counts are >200, while the incidence of esophageal candidiasis remains moderately elevated at CD4 counts between 200 and 350, and cases of pulmonary tuberculosis continue to occur at CD4 counts >500.

Initiation of anti-retroviral therapy (ART) halts HIV replication and raises CD4 T cell counts, but does not always restore pathogen-specific immunity to normal levels. For example, HIV positive individuals on ART with CD4 > 700/mm³ have 4-fold higher rates of TB disease than HIV uninfected individuals in the same community.(Gupta et al., 2012) The impact of ART on human papillomavirus (HPV) infection has been controversial, with some studies demonstrating reduced HPV prevalence and regression of HPV-associated squamous intraepithelial lesions (SIL), while others fail to show any impact on HPV-associated disease.(Adler, 2010)

It is thought that functional defects and depletion of pathogen-specific CD4 T cells by HIV occur at varying rates accounting for differences in pathogenesis of specific opportunistic infections.(Geldmacher and Koup, 2012) However, it remains unclear whether differences in pathogenesis are due to differences in pathogen-specific CD4 T cell susceptibility to HIV infection or other factors. Here, we review the immunopathogenesis of three infections causing substantial morbidity and mortality in HIV-infected individuals: TB, HPV, and cytomegalovirus (CMV). Understanding the complex interplay between HIV and these pathogens provides insight into differences in disease prevalence and impact of ART on the natural history of infection.

HIV-TB

Epidemiology and burden of disease

HIV and TB co-infection remains a serious global health problem. According to the World Health Organization, there were 8.7 million new cases of TB and 1.4 million deaths due to TB disease in 2011.(World Health Organization. and Global Tuberculosis Programme.) TB is a leading cause of death among HIV infected individuals, especially in Africa where over 50% of deaths in persons with HIV are due to TB disease.(Bates et al., 2013) HIV is a leading risk factor for TB disease with rates of active TB doubling within one year of HIV seroconversion and increasing more than 4-fold in chronic HIV infection.(Lodi et al., 2013; Sonnenberg et al., 2005) Although ART reduces the incidence of TB disease, rates of TB in individuals with reconstituted immune systems remain higher than the general population.(Gupta et al., 2012) This suggests that HIV infection induces functional defects in the immune response to TB that persist despite immune reconstitution.

Cell-mediated immune response to TB

Interactions between the host's innate and adaptive immune system and the organism dictate the outcome of infection with Mycobacterium tuberculosis (Mtb). Although innate immune cells are an important component of the immune response to TB infection (van Crevel et al., 2003) it is clear that T cells are essential for containing Mtb. Mice deficient in CD4 T cells have reduced survival and greater bacterial burden following aerosol exposure to Mtb than their wild type counterparts.(Caruso et al., 1999) Antibody mediated depletion of CD4 T cells in mice results in rapid reactivation of persistent TB infection and reduced survival.(Mogues et al., 2001) Similarly, non-human primates (NHP) depleted of CD4 T cells have an increased incidence of active TB disease following Mtb exposure and a higher rate of reactivation TB compared to non-CD4 depleted monkeys.(Lin et al., 2012) SIV infection of NHP with latent TB infection results in reactivation of TB in all infected monkeys, albeit at different rates.(Diedrich et al., 2010) Monkeys reactivating earlier exhibited a greater initial decline in CD4/8 T cells following SIV infection and had fewer CD4 T cells within their airways. Taken together, these animal studies support the critical role CD4 T cells play in controlling TB infection.

Why are CD4 T cells important? For one, they are a major source of IFNγ, which is necessary for the production of reactive nitrogen intermediates and killing of intracellular mycobacteria by macrophages. In fact, IFNγ specifically from CD4 T cells is required for a robust CD8 T cell response and for inhibiting intracellular replication of tubercle bacilli within macrophages.(Green et al., 2013) The critical role of IFNγ in controlling Mtb is best demonstrated in mice devoid of IFNγ or with impaired IFNγ signaling, which rapidly succumb to TB disease.(Flynn et al., 1993; Kamijo et al., 1993) In the same vein, humans with mutations in genes encoding IFNγ receptor 1 or with polymorphisms in the IFNγ gene are at increased risk of mycobacterial disease.(Lopez-Maderuelo et al., 2003; Newport et al., 1996) CD4 T cells are also crucial in maintaining the integrity and architecture of granulomas in TB infection.(Heuts et al., 2013; Saunders et al., 2002) Expression of CXCR5, a chemokine receptor important in B and T cell homing to lymphoid tissue, appears to be an important mediator of CD4 T cell localization within granulomas, facilitating lymphoid follicle formation and better protective outcomes.(Slight et al., 2013)

Studies of chronic viral infections in animals and humans suggest that with viral control the ability of CD4/8 T cells to produce multiple cytokines (polyfunctional T cells) increases.(Harari et al., 2005; Pantaleo and Harari, 2006) Whether this paradigm holds true for TB infection remains debatable. TB vaccine studies in animals have equated increased frequencies of polyfunctional T cells to protection from aerosol Mtb challenge.(Aagaard et al., 2009; Forbes et al., 2008) Polyfunctional T cells are induced in humans following the administration of novel TB vaccine formulations (Abel et al., 2010; Scriba et al., 2012) but whether this results in enhanced protection remains to be seen. Individuals with smear positive TB disease or untreated TB disease have reduced frequencies of polyfunctional CD4 T cells compared to individuals with smear negative TB disease or with latent or treated TB infection.(Day et al., 2011; Harari et al., 2011) These data suggest that as mycobacterial burden is reduced or controlled, T cell responses are characterized by the ability to produce multiple cytokines. However, not all authors have reached similar conclusions.(Caccamo et al., 2010; Sutherland et al., 2009) It is possible that differences in methodology or patient populations may account for the discordant results.

Impact of HIV infection on M. tuberculosis cell-mediated immunity

HIV co-infection adversely affects T cell responses to Mtb. Peripheral T cells from HIV-TB co-infected individuals have impaired proliferation and reduced IFNγ production in response to Mtb antigen stimulation compared to TB mono-infected patients.(Geldmacher et al., 2008; Zhang et al., 1994) Similar defects have been observed in Mtb-specific CD4 T cells isolated from lung lavage samples of HIV infected individuals.(Bonecini-Almeida Mda et al., 1998; Kalsdorf et al., 2009) In addition, HIV co-infection alters the quality of the T cell response, resulting in reduced frequencies of polyfunctional T cells. (Day et al., 2008; Jambo et al., 2011; Kalsdorf et al., 2009)

How soon after HIV infection are impairments in T cell responses seen? Dramatic decreases in Mtb-specific memory CD4 T cells (CD27+CD45RO+) were observed within one year of HIV seroconversion in 4 out of 5 individuals with latent TB infection.(Geldmacher et al., 2008) The early depletion of Mtb-specific CD4 T cells can be explained by distinct functional and phenotypic characteristics of these cells. First, Mtb-specific CD4 T cells have increased expression of the chemokine receptor CCR5, which also serves as a HIV co-receptor, leading to increased HIV entry into CD4 T cells.(Geldmacher et al., 2010) More importantly, Mtb-specific CD4 T cells preferentially produce IL-2 upon stimulation. IL-2 production not only supports proliferation of antigen-specific T cells but promotes HIV replication.(Geldmacher et al., 2010) In contrast, CMV-specific CD4 T cells tend to secrete beta-chemokines, such as MIP-1β, which can reduce susceptibility to HIV infection by hindering HIV binding to CCR5. Lastly, most Mtb-specific CD4 T cells express CD27, a cell surface maker associated with an early differentiated phenotype and replicative potential. CMV-specific CD4 T cells, on the other hand, rarely express CD27 and more often express CD57. (Geldmacher and Koup, 2012) Antigen-specific CD4 T cells expressing CD57 are more terminally differentiated, having reduced proliferative capacity and thus harboring lower cellular HIV loads. (Brenchley et al., 2004)

While we believe the early depletion of Mtb-specific CD4 T cells following HIV infection is an important contributor to the increased risk of TB disease, other mechanisms increasing TB reactivation risk in HIV-TB co-infected individuals should not be overlooked. HIV infection of macrophages reduces the apoptotic response to Mtb and down-regulates macrophage autophagy. (Patel et al., 2009; Zhou and Spector, 2008) This not only results in bacterial persistence, but could limit antigen presentation to T cells. Observational studies have noted a high frequency of vitamin D deficiency among HIV infected individuals and vitamin D deficiency has been associated with active TB disease. (Viard et al., 2011; Wilkinson et al., 2000) The bioactive form of vitamin D, 1,25 dihydroxyvitamin D_3, restricts Mtb growth by inducing the expression of cathelicidin antimicrobial peptide which is important in the induction of macrophage autophagy. (Campbell and Spector, 2012; Liu et al., 2006) Finding ways to manipulate this pathway to modulate the host immune response to TB is of great interest. One additional factor contributing to increased TB risk is the genetic diversity of the pathogen itself. Strains of Mtb vary in their inhibition of innate immune responses and in virulence. A recent study suggests that strains of the modern lineage induce a lower inflammatory response, selecting for more rapid progression to active TB than strains of the ancient lineage. (Portevin et al., 2011) Further research on the immunobiology and clinical outcomes of various Mtb strains is needed to more fully understand the consequences of strain diversity in Mtb.

Impact of ART on TB disease

Suppression of HIV replication by ART leads to increases in CD4 T cell count and demonstrable mortality and morbidity benefits in patients co-infected with TB. This is especially true in individuals with CD4 counts less than 50 cells/mm³ if ART is started after 2 weeks of TB treatment. (Havlir et al., 2011) ART reduces the risk of TB disease by 67%. (Lawn et al., 2011) This may be due to increases in naïve and central memory CD4 T cell pools along with increases in IFNγ secretion or polyfunctional T cell responses following ART. (Sutherland et al., 2010; Wilkinson et al., 2009) However, despite HIV viral load suppression and reconstitution of CD4 T cell counts, ART-treated patients continue to have risks of TB disease that exceed that of the HIV-uninfected population. (Girardi et al., 2005) One possible explanation for this is that ART does not restore Mtb-specific immune responses to levels observed in HIV-uninfected individuals. (Schluger et al., 2002; Sutherland et al., 2006)

HIV-HPV

Epidemiology and burden of disease

HPV is the most common sexually transmitted infection (STI) globally. Persistent infection with oncogenic HPV types is linked to development of dysplastic changes in the cervix, vulva, and anorectal regions. If persistent, these dysplastic changes can progress to invasive squamous cell carcinoma.

Infection with HIV is associated with an increased prevalence of HPV infection, infection with multiple HPV types, and increased incidence of cervical cancer and other anogenital malignancies. (Luchters et al., 2010; Singh et al., 2009) Lower CD4 T cell counts and higher HIV viral loads are independent risk factors for HPV infection and infection with multiple HPV types. (Palefsky et al., 1999)

Natural history studies demonstrate that 70–80% of sexually active young women who develop incident cervical HPV infection clear their infection within 12 to 30 months. (Evander et al., 1995; Ho et al., 1998) Young women with low grade cervical lesions (CIN I) show spontaneous regression rates ranging from 45% to 90%, while women with pre-cancerous lesions (CIN III) show only an 11% regression rate. (Insinga et al., 2009) Although factors influencing the natural history of these different lesions are not completely known, cell-mediated immune responses are thought to play role.

Cell-mediated immune response to HPV infection

Immunohistopathological studies reveal that at the time of regression, HPV-infected genital lesions are infiltrated with CD8 cytotoxic T cells (CTL), CD4 T cells, and macrophages, suggesting that a cell-mediated immune response is integral to the control of HPV infection. (Coleman et al., 1994) Further evidence implicating T cells in the control of HPV infection comes from studies evaluating functional responses of T cells to HPV early antigens. Cellular immune responses directed against HPV16 E6 and E7 are associated with regression of HPV16-associated cervical lesions. (Kadish et al., 2002; Sarkar et al., 2005) HPV16-specific T cell responses are more frequently detected in women with resolved cervical HPV16 infection than with persistent infection. (Farhat et al., 2009; Nakagawa et al., 2000) Finally, HPV16-specific responses are largely absent or severely impaired in women with cervical cancer. (de Jong et al., 2004) These data imply that control of HPV infection requires a robust systemic T cell response. This observation is further supported by a recent study in which vaccine-induced regression of HPV16 high-grade vaginal intraepithelial neoplasia (VIN) was more common in women exhibiting strong and broad IFNγ-associated CD4 T cell responses to HPV 16 E6 and E7 proteins. (Welters et al., 2010)

Impact of HIV infection on HPV cell-mediated immunity

Evidence of direct biological effects of HIV on HPV infection comes from two studies of HPV reactivation. Strickler et al. evaluated the rate of new HPV DNA detection in sexually inactive (18 months of sexual abstinence) HIV positive and HIV negative women. HPV DNA was detected in 5% of sexually inactive HIV negative women compared to 22% of HIV positive women with advanced disease (CD4 T cell count <200 cells/mm³). (Strickler et al., 2005) Higher HPV reactivation rates in HIV infected women were also reported by Thieler and colleagues, who noted an association between HPV reactivation and declining CD4 T cell counts. (Theiler et al., 2010) These studies not only support the concept of HPV reactivation, but show that progressive CD4 depletion increases the risk of HPV reactivation.

If the risk of HPV infection is related to CD4 depletion, the risk should increase soon after HIV seroconversion. Why would this be the case? It is well known that during acute HIV and SIV infection there is rapid depletion of memory CD4 T cells within the mucosa, including the endocervical compartment. (Mattapallil et al., 2005; Veazey et al., 1998; Zhang et al., 1999) If CD4 T cells are critical to the control of HPV infection, the profound depletion of CD4 T cells within the genital tract soon after HIV infection should increase the risk of HPV infection.

Recently, two groups evaluated the impact of incident HIV on the detection of HPV infection. Relative to HIV uninfected women, HIV infected women had higher rates of HPV detection at their first visit following HIV seroconversion. (Nowak et al., 2011; Wang et al., 2011) HIV seroconverters had a 5-fold increased risk of HPV infection with multiple HPV types and a 2.5-fold increased risk of infection with any single HPV type. (Nowak et al., 2011) Moreover, HIV seroconversion was associated with a 2.5-fold increased risk of low-grade cervical cytological abnormalities. (Wang et al., 2011) These studies provide evidence of increased rates of HPV infection following HIV seroconversion which may be driven in part by the rapid HIV-associated CD4 T cell depletion within the cervical mucosa.

Why would CD4 T cells in the genital tract be subject to rapid depletion? A recent study of cervical CD4 T cells from female sex workers in Kenya characterized a subset of activated cervical CD4 T cells that produce IL-22 and IL-17a, identifying a Th17 cell population within the cervix. (McKinnon et al., 2011) Th17 cells are consistently depleted from the gut mucosa in HIV positive individuals. (Brenchley et al., 2008; Raffatellu et al., 2008) The authors showed that cervical Th17 cells express HIV susceptibility markers CCR5 and integrin α₄β₇. Integrin α₄β₇ is a lymphocyte mucosal homing marker which also serves as a cellular receptor for HIV-1 gp120. (Cicala et al., 2011) While integrin α₄β₇ is not required for HIV infection, it can enhance CD4 T cell susceptibility to HIV infection. McKinnon and colleagues showed that co-expression of these receptors by cervical CD4 T cells increased susceptibility to HIV infection and targeted depletion. (McKinnon et al., 2011) Although the rapid depletion of cervical CD4 T cells is driven by productive HIV infection of these cells, increased immune activation within genital mucosa may also contribute to CD4 T cell loss. (Jaspan et al., 2011) While CD4 T cell depletion within the genital mucosa may contribute to the increased risk of HPV infection, it remains unclear whether HPV-specific T cells are being preferentially targeted and depleted. Further research assessing the effects of HIV on the kinetics of HPV-specific T cells within the genital mucosa is needed.

Impact of ART on HPV Infection

The impact of ART on the natural history of HPV infection and its associated complications remains uncertain. The partial immune reconstitution observed with ART has decreased rates of some AIDS-defining malignancies, namely Kaposi's sarcoma and non-Hodgkin lymphoma; however, the same benefits have not been observed with respect to HPV-associated cervical cancer (Table 1). One theoretical explanation is that by prolonging survival of HIV infected women ART increases the cumulative exposure to oncogenic HPV types allowing for progression of HPV-related disease.

Table1.

Incidence of AIDS and non-AIDS defining malignancies in HIV positive individuals and the influence of ART.

Cancer Type	Mean Standardized Incidence Ratio (Range)			Incidence Rate per 100 000 Person-years
	Pre ART Era	Early ART Era	Late-ART Era	0–6 mos post-ART initiation	6mo-1 Ovr post-ART intiation
AIDS Defining Cancers
Kaposi Sarcoma	1555.5 (246 – 2628.5)	448.3 (47.8 – 848.8)	317.1 (22.9 – 572)	1342	164
Non-hodgkin lymphoma	537.3 (103 – 1011.8)	260.4 (26.7 – 494.4)	107.3 (16.2 – 212.2)	357	134
Non-hodgkin lymphoma (CNS)				160	24
Cervical	69.8 (8.4 – 149.9)	99.2 (3.7 – 194.6)	88 (41.5 – 134.5)	n/a	n/a
Non-AIDS Defining Cancers
Anal	44.5 (19 – 97.9)	90.1 (48.3 – 112.0)	78.4 (44 – 141.4)	72	69
Hodgkin lymphoma	16.5 (4.5 – 34.3)	28.9 (11.1 – 54.7)	36.4 (20.7 – 64.4)	144	47
Liver	9.1 (0 – 19.9)	17.5 (5.9 – 35.9)	13.7 (6.1 – 35.4)	18	39
Lung	24.3 (0 – 91.9)	33.2 (2.8 – 93.8)	23.5 (2.4 – 84.9)	54	56
Prostrate	5.3 (0 – 14.7)	13.3 (0 – 38.0)	9.9 (0 – 37.5)	22	58
Breast	19.1 (0.6 – 56.0)	35.5 (1.2 – 69.9)	32.5 (0.6 – 96.0)	177	122
Head and neck	9.2 (1.4 – 29.0)	12.1 (2.2 – 31)	10.6 (1.5 – 36.9)	n/a	n/a
Melanoma	4.4 (0 – 15.6)	8.5 (0 – 24.8)	10.7 (0.6 – 37.5)	n/a	n/a
HPV-related cancers				108	103

The mean standardized incidence ratios (SIR) were calculated by averaging SIR reported in several studies of cancer risk in persons with HIV.(Dal Maso et al., 2009; Franceschi et al., 2010; Patel et al., 2008; Powles et al., 2009) The pre ART era covers the years prior to 1996, the early ART era covers 1996–2001, and the late ART era covers 2002–2007, though there is some overlap between the studies. The incidence ratios of Kaposi sarcoma (KS) and non-hodgkin lymphoma (NHL), both AIDS defining cancers, decrease considerably from the pre ART era to the late ART era. This is not seen with cervical cancer. In a separate study, administration of ART for >6 months reduced the incidence of both KS and NHL while no difference in incidence of HPV-related malignancies was noted.(Yanik et al., 2013)

A number of clinical studies have evaluated the impact of ART on HPV infection and HPV-related cervical disease and are reviewed in detail elsewhere. (Adler, 2010) Importantly, two large, longitudinal cohort studies of HIV positive women offered conflicting data on the effects of ART on HPV disease prevalence and HPV-related cervical disease. (Adler, 2010) However, in a very recent publication, Blitz et al. found that ART increases the likelihood of cervical SIL regression (HR 3.3), while also increasing clearance rates of oncogenic HPV types, although this was type specific. (Blitz et al., 2013) The conflicting results may be due to differences in primary end-points, cohort design, or definitions of cytological progression/regression. One additional factor to consider is the assessment of ART efficacy. Minkoff and colleagues showed that women who were adherent to ART had a more significant decline in detection of oncogenic HPV and more rapid clearance of HPV SIL, compared to non-adherent women.(Minkoff et al., 2010) While there seems to be mounting evidence that ART may have a beneficial impact on the natural history of HPV the extent and magnitude of this effect is not clear and the influence of ART on the kinetics of HPV-specific immunity remains largely unknown.

Impact of HPV on HIV

Sexually transmitted infections such as herpes simplex virus-2 and syphilis have a profound impact on HIV acquisition. A recent review assessed the association of HPV infection with HIV acquisition. Women with prevalent HPV infection, irrespective of HPV type, had a 2-fold higher risk of HIV infection. An increased risk of HIV acquisition was also observed in men with HPV infection, though this data was generated from only two observational studies.(Houlihan et al., 2012).

Biologically, it is plausible that HPV infection could affect HIV acquisition and dissemination. As just described, T cells are an integral component of the host immune response to HPV. Given that CD4 T cells are primary targets for HIV, a robust T cell response to HPV infection could increase risk of HIV acquisition and dissemination. Langerhans cells are found in the mucosal layer of the genital tract and are generally refractory to HIV transmission.(de Witte et al., 2007) Depletion of Langerhans cells within the cervical epithelium in HPV infection could contribute to increased HIV acquisition.(Connor et al., 1999; Tay et al., 1987) Additionally, E-cadherin, an epithelial adhesion molecule, is down-regulated during HPV infection (Matthews et al., 2003; Vessey et al., 1995) which could potentially increase genital mucosa permeability to HIV.

Given that HPV is the most common STI worldwide, a more thorough understanding of the contribution of HPV infection to the HIV epidemic is required. Further research into the biological mechanisms underlying this increased risk should be pursued. With the implementation of routine HPV vaccination, studies evaluating the impact of HPV vaccination on the HIV epidemic should be considered.

HIV-CMV

Epidemiology and burden of disease

CMV is a member of the human herpesvirus family. More than half of the adult human population is infected with CMV with higher rates of infection in HIV positive individuals.(Bate et al., 2010; Lang et al., 1989) Following primary infection, the virus is not eliminated but remains in a latent state within many tissues of the host. In healthy individuals, CMV can reactivate intermittently with little consequence; however, in immunocompromised hosts reactivation can lead to end-organ disease (EOD).

In HIV infection, CMV EOD typically occurs when CD4 cell counts decrease to <50 cells/mm³. CMV retinitis, the most common form of EOD in HIV infection, occurs in 40% of individuals with CD4 cell counts <50 cells/mm³.(Pertel et al., 1992) The incidence of CMV EOD has decreased substantially with the introduction of ART.(Ledergerber et al., 1999; Palella et al., 1998) The increased incidence of CMV disease in patients with AIDS and restoration of protective immunity to CMV with ART suggests that cellular immunity is critical to the control of CMV infection.

Cell-mediated immune response to CMV

The immune system utilizes tremendous resources to control CMV. It is estimated that approximately 10% of the total memory CD4 and CD8 T cell pool in CMV seropositive healthy donors is specific for CMV antigens, with higher frequencies in certain populations.(Khan et al., 2004; Sester et al., 2002; Sylwester et al., 2005) The role of CD8 T cells in controlling CMV viral replication has been well described in animal models. (Barry et al., 2007; Polic et al., 1998; Reddehase et al., 1985) In humans, seminal works by Riddell and Walter showed that infusion of donor-derived CMV-specific CD8 T cells not only restored CMV-specific cellular immunity but also provided protection against CMV clinical disease.(Riddell et al., 1992; Walter et al., 1995)

While CD8 T cells have a clear role in controlling CMV infection, CD4 T cells are also critical to immune control. During primary CMV infection, there is a major expansion of CMV-specific CD4 T cells and these responses appear before CD8 T cell responses.(Harari et al., 2004) Delays in CD4 T cell response during primary infection have been associated with symptomatic infection and prolonged viral shedding.(Gamadia et al., 2003; Tu et al., 2004) It has also been shown that recovery from primary infection requires CMV-specific CD4 T cell responses.(Gamadia et al., 2003) Why might this be the case? Komanduri et al. demonstrated that while CD4 and CD8 T cell responses to CMV are sustained over time, individuals maintaining robust CD4 T cell responses are more likely to exhibit higher frequencies of CMV-specific CD8 T cells.(Komanduri et al., 2001) Thus, CMV-specific CD4 T cells are important for maintaining and expanding the CMV-specific CD8 T cell pool.

Apart from providing help to CD8 T cells, CD4 T cells may control infection through direct killing of virus-infected cells. Casazza and colleagues demonstrated that certain populations of CMV-specific CD4 T cells contain granzyme A, granzyme B, and perforin, and that these cells are capable of degranulating and killing cells expressing their cognate antigen.(Casazza et al., 2006)

Impact of HIV on CMV immunobiology

Just as loss of Mtb-specific CD4 T cells contributes to the increased risk of TB disease in HIV-TB co-infection, loss of CMV-specific T cells contributes to the pathogenesis of CMV EOD in HIV infected individuals. Patients with active CMV disease have diminished CMV-specific T cell responses compared to individuals with immune recovery from CMV EOD.(Jacobson et al., 2001; Komanduri et al., 1998) In a study evaluating the dynamics of CMV-specific T cell responses, significant reductions in CMV-specific IFNγ-producing CD4 T cells were seen in the year prior to onset of CMV EOD.(Bronke et al., 2005) Coincident with the reduction in CMV-specific CD4 T cells was a decline in CMV-specific CD8 T cells producing IFNγ. This is not surprising as CD4 T cell help is required for maintaining and expanding the CD8 T cell pool.

Unlike Mtb-specific CD4 T cell responses, the frequency of CMV-specific CD4 T cell responses is generally maintained in chronic HIV infection and is similar to that observed in CMV seropositive, HIV negative individuals.(Komanduri et al., 1998) In a recent study evaluating CMV-specific T cell responses across various HIV disease states, CMV-specific CD4 and CD8 T cell responses were higher in chronic, untreated HIV infection than in acute HIV infection.(Naeger et al., 2010)

Why is it that following HIV seroconversion Mtb-specific CD4 T cells are lost whereas CMV-specific CD4 T cells are generally maintained? CMV-specific CD4 T cells produce MIP-1β and MIP-1α upon stimulation and bind to CCR5, which also serves as an HIV co-receptor. In addition, CMV-specific CD4 T cells producing MIP-1 β have lower surface expression of CCR5 and are less frequently infected with HIV than CMV-specific CD4 T cells that do not produce MIP-1β.(Casazza et al., 2009) We proposed that the autocrine production of beta-chemokines by CMV-specific CD4 T cells blocks R5-tropic HIV entry, either by direct competition or down-regulation of CCR5 surface expression. Recently, Hu and colleagues have extended this observation. In a novel study, they showed that neutralization of beta-chemokines led to enhanced HIV infection in tetanus toxoid- and C. albicans-specific CD4 T cells, while HIV infection of CMV-specific CD4 T cells remained low suggesting post entry restriction of HIV by CMV-specific CD4 T cells.(Hu et al., 2013) Microarray analysis revealed that a broad array of innate antiviral genes was activated in CMV-specific CD4 T cells, including several type-I IFN response genes and HIV/SIV restriction factors (TRIM22 and TRIM5). This implies that other mechanisms may be involved in determining differential susceptibility of pathogen-specific CD4 T cells to HIV infection.

Conclusion

The timing of opportunistic infections in HIV infection is not always associated with severity of CD4 T cell depletion and pathogen-specific differences in the incidence of infection within CD4 strata exist. Recent data suggest that functional or phenotypic differences in pathogen-specific CD4 T cells may influence the timing of disease reactivation (figure 2). For example, Mtb-specific CD4 T cells have high expression of CCR5 and preferentially produce IL-2 leading to increased susceptibility to HIV infection. Cervical CD4 T cells also have increased CCR5 expression but also express integrin α₄β₇ increasing their susceptibility to HIV infection. In contrast, CMV-specific CD4 T cells preferentially produce beta-chemokines, and express type I interferon response genes and genes encoding HIV restriction factors all of which reduce their susceptibility to productive HIV infection and delay disease reactivation until there is a profound loss of CD4 T cells. While ART has reduced the burden of many opportunistic infections, pathogen-specific immunity is not always fully restored as the risk of infection from certain opportunistic pathogens in immune reconstituted HIV positive individuals continues to exceed that of the general population.

Figure 2. Proposed model for HIV-associated depletion of pathogen-specific CD4 T cells.

(Top panel) Mtb-specific CD4 T cell is activated upon binding to its cognate antigen on an HIV infected antigen-presenting cell (APC). T cell activation induces secretion of IL-2 and up-regulation of CD25 (IL-2 receptor) and CCR5 (HIV co-receptor). Binding of IL-2 to CD25 induces cellular proliferation (expression of proliferation marker Ki67) and promotes viral replication leading to productive HIV infection and cell death. (Middle Panel). Upon activation, CMV-specific CD4 T cells produce MIP-1β, which binds to its ligand, CCR5, an HIV co-receptor, blocking HIV entry into the cell. Binding of MIP-1β to CCR5 may also down-regulate CCR5 further impeding HIV entry. CMV-specific CD4 T cells express CD57, a marker to replicative senescence, and do not enter the cell cycle, limiting HIV viral replication. Expression of TRIM and Type I IFN may further restrict HIV replication. These cells are less susceptible to productive HIV infection and targeted depletion. (Bottom Panel). Activation of cervical CD4 cells leads to increased expression of both CCR5 and the mucosal homing receptor, integrin α₄β₇, which binds to HIV gp120. Co-expression of these receptors increases susceptibility to HIV infection. These cells are also highly activated (express Ki67), promoting HIV replication, leading to productive infection and cell death.

The immunobiology of HIV and co-pathogen infection remains quite complex and the mechanisms leading to differential susceptibility of pathogen-specific CD4 T cells to HIV infection are only now beginning to be understood. Powerful technologies such as mulitparameter flow cytometry and microarrays have advanced our understanding but novel platforms allowing analysis at the single cell level may provide us with a better appreciation of how these pathogen-specific differences are generated. As ART is initiated even earlier in HIV infection, it will be important to assess whether earlier implementation improves pathogen-specific immunity. Although mechanistic studies of HIV and co-pathogen infection are complex, a more thorough understanding of this relationship will better inform treatment strategies against co-pathogens.