|Home | About | Journals | Submit | Contact Us | Français|
During persistent viral infection, adaptive immune responses are suppressed by immunoregulatory factors, contributing to viral persistence. Although this suppression is mediated by inhibitory factors, the mechanisms by which virus-specific T cells encounter and integrate immunoregulatory signals during persistent infection are unclear. We show that a distinct population of IL-10-expressing immunoregulatory antigen presenting cells (APC) is amplified during chronic versus acute lymphocytic choriomeningitis virus (LCMV) infection and suppresses T cell responses. Although acute LCMV infection induces the expansion of immunoregulatory APC, they subsequently decline. However, during persistent LCMV infection, immunoregulatory APC are amplified and parallel the viral replication kinetics. Further characterization demonstrates that immunoregulatory APC are molecularly and metabolically distinct, and exhibit increased expression of T cell-interacting molecules and negative regulatory factors that suppress T cell responses. Thus, immunoregulatory APC are amplified during viral persistence and deliver inhibitory signals that suppress antiviral T cell immunity and likely contribute to persistent infection.
The immune system resolves the majority of viral infections through a concerted effort of both innate and adaptive mechanisms. These multifaceted responses result in effective elimination of viral pathogens and the establishment of long lasting protective immunity. However, some viruses, including human immunodeficiency virus (HIV) and hepatitis C virus (HCV) in humans and lymphocytic choriomeningitis virus (LCMV) in mice, are capable of enduring the initial immune onslaught and establish persistent infections (Klenerman and Hill, 2005; Wilson and Brooks, 2010). The prolonged and elevated viral titers associated with these types of infection progressively alter T cell responses in a phenomenon known as exhaustion (Fahey and Brooks, 2010). While immune exhaustion is counterproductive to viral clearance it is likely necessary to prevent harmful bystander immunopathology that is associated with prolonged T cell responses in the face of unresolved high level virus replication (Barber et al., 2006; Yi et al., 2009). Exhausted T cell responses have a unique developmental program characterized by decreased proliferation and the diminished ability to produce antiviral and immunostimulatory cytokines that are associated with acute viral infections (Fahey et al., 2011; Wherry et al., 2003; Wherry et al., 2007). Importantly, some degree of lingering T cell activity is actively maintained in persistent infection and while these residual responses differ in cytokine production and magnitude from what is considered to be fully productive antiviral T cell responses, they are critical for the long term control of viral replication (Agnellini et al., 2007; Elsaesser et al., 2009; Fahey et al., 2011; Frohlich et al., 2009; Yi et al., 2009). Thus, elucidating the control mechanisms that modulate T cell responses will be important toward understanding how these pathogens subvert the immune response to persist.
At the onset of an infection, T cells are primed by specialized antigen presenting cells (APC) called dendritic cells (DC) (Banchereau and Steinman, 1998). During the initial priming, multiple interactions, including signals from surface bound and soluble co-stimulatory and/or inhibitory molecules, function in concert to stimulate and fine tune T cell responses. However, since these initial interactions cannot forecast the long-term immune requirements needed to fight a particular infection, cellular responses are pliant to local signals, and T cell functions are continually modulated in response to the needs of the current antigenic environment (Brooks et al., 2006b). As a result, multiple APC (including macrophages and B cells) and infected cell populations (Mueller et al., 2007) that are not potent inducers of T cell activation likely have important roles in the modulation of the immune response as infection progresses. This is particularly relevant during persistent infections both early as the initially productive T cell responses are suppressed, as well as during viral persistence to modulate T cell activity and protect from immunopathology while continuing to battle the infection (Barber et al., 2006; Brooks et al., 2006c).
The host derived immunoregulatory cytokine IL-10 is crucial in driving T cell exhaustion and viral persistence following LCMV infection (Brooks et al., 2006c; Ejrnaes et al., 2006). Early disruption of IL-10-mediated suppression prevents the loss of T cell activity in response to an otherwise persistent LCMV infection leading to rapid virus clearance (Brooks et al., 2006c; Ejrnaes et al., 2006). Similarly, elevated IL10 levels correlate with HIV replication in humans and recently a link between IL-10 expression and transition into persistent HCV infection was identified (Brockman et al., 2009; Flynn et al., 2011). Ex vivo IL-10 blockade enhanced anti-HIV and anti-HCV T cell activity (Clerici et al., 1994; Landay et al., 1996; Rigopoulou et al., 2005), further indicating the important and conserved suppressive role of IL-10 during many persistent virus infections.
In addition to IL-10, multiple immunoregulatory pathways actively, simultaneously and increasingly suppress T cell activity during viral persistence, including PDL1/PD1, Lag3, Tim3, CTLA4, indoleamine 2,3 dioxygenase (IDO) and TGFβ (Wilson and Brooks, 2011). Consistent with graded levels of T cell exhaustion, wherein increasing amounts of inhibitory signals in combination achieve a threshold necessary for functional T cell suppression, the concurrent blockade of multiple suppressive signals additively enhances antiviral T cell activity (Blackburn et al., 2009; Brooks et al., 2008; Jin et al., 2010). Yet, it is unclear how the concurrent inhibitory signals needed to achieve the suppressive threshold are delivered to a specific T cell in the context of an overwhelming virus infection and inflammatory environment. Herein we identify specific populations of immunoregulatory APC (which for simplicity in nomenclature will be called iAPC) that simultaneously produce multiple T cell interacting and immune-inhibitory molecules to suppress antiviral T cell activity. Consistent with the expression of suppressive factors, iAPC are invoked by virus replication but rapidly wane in the stimulatory environment of an acute infection. Conversely, iAPC are potentiated and highly amplified during persistent infection to facilitate dampening of T cell responses and maintenance of the suppressive environment. Thus, the amplification and clustering of multiple immunoregulatory factors on a single iAPC is a mechanism to simultaneously deliver potent inhibitory signals and in a single interaction tip the balance toward exhausted T cell responses and viral persistence.
To elucidate the dynamics of IL-10 expression in viral infection we utilized the LCMV model system. Infection with the Armstrong (Arm) variant of LCMV induces robust CD4 and CD8 T cell responses that clear the virus within the first two weeks after infection. Alternatively, the Clone 13 (Cl 13) variant of LCMV replicates to substantially higher titers and, although initially inducing productive T cell responses, rapidly elicits the expression of multiple host immunoregulatory factors that suppress T cell activity to generate a persistent infection (Ahmed et al., 1984; Wilson and Brooks, 2011). A unique aspect of the LCMV system is the ability to directly compare an acute and persistent infection with the same virus allowing for the differentiation of factors that are increased and important in persistent infection from those that are the result of a common response to viral infection.
To specifically characterize mechanisms of IL-10 mediated immunosuppression we utilized the Vert-X IL-10 reporter mouse which expresses green fluorescent protein (GFP) linked by an internal ribosome sequence (IRES) to the IL-10 locus (Madan et al., 2009). Vert-X mice allow for direct identification of IL-10 expressing cells without in vitro manipulation or cell fixation. Regulation of IL-10 expression is not perturbed in these mice and the course of LCMV infection in this strain is analogous to WT mice (data not shown). Importantly, IL-10 RNA and protein expression was dramatically elevated in the GFP+ cells (Fig. S1). Additionally, analyses described herein were confirmed using a second IL-10 reporter (10BiT) mouse (Maynard et al., 2007). The 10BiT mouse contains a bacterial artificial chromosome (BAC) with a Thy1.1 reporter cassette inserted in frame in the first exon of the IL-10 locus along with a stop codon. Thus, enabling identification of IL-10 expressing cells without any manipulation of the two functioning (i.e., no IRES) endogenous IL-10 genes. All data were reproducible and analogous in the 10BiT mice (data not shown).
IL-10 protein was detectible in the plasma from both Arm and Cl13 infected mice at early time points; however, upon the resolution of acute infection IL-10 levels wane, whereas they remain elevated throughout the course of a persistent infection (Fig 1A). While many cell types can produce and respond to IL-10 in varying situations, the relevant sources in persistent viral infection are still controversial (Wilson and Brooks, 2011). To begin to define the relevant sources of IL-10 in persistent infection we generated reciprocal mixed bone marrow chimeras of B6 and IL-10 deficient mice. Following persistent Cl13 infection, plasma IL-10 levels were elevated only when the hematopoietic compartment was IL-10 sufficient (Fig 1B). On the other hand, IL-10 levels were low to undetectable when the hematopoietic compartment was derived from IL-10 deficient mice, despite the rest of the mouse being IL-10 sufficient, indicating that the induction of IL-10 during viral persistence is predominantly from hematopoietic derived cells. Production of IL-10 from hematopoietically derived cells was elevated in multiple organs throughout persistent compared to acute infection (Fig. 1C and D). GFP+ gates were based on control non-transgenic B6 mice (Fig. S1). Interestingly, the absolute number of GFP expressing cells was elevated in persistent infection, despite the overall decrease in splenic cellularity relative to acute infection (Arm: 3.9 × 107 +/− 1.2 × 107 vs. Cl13: 2.5 × 107 +/−9 × 106; p=0.02). In other tissues, the percentage of IL-10 expressing cells was elevated during persistent compared to acute infection, but the lower overall levels of cellularity during persistent infection led to equal overall numbers of iAPC (Fig. 1D). Levels of IL-10 expressing cells rose and peaked in the spleen around day 9 after LCMV-Cl 13 infection (Fig. 1C and D) in conjunction with the transition from functional to exhausted T cell responses (Brooks et al., 2006b). IL-10 expressing cells in the lymph nodes and in multiple non-lymphoid organs were initially lower, but then maintained or increased through persistent infection, whereas they remained low in acute infection (Fig. 1C and D).
To define the mechanisms of how IL-10 suppresses the immune response to facilitate viral persistence, specific cell subsets that exhibit enhanced IL-10 expression in persistent as compared to acute infection were quantified. At the peak of IL-10 expression in the spleen, multiple immune subsets produced IL-10 including dendritic cells (DC); macrophages; B cells; CD4+ and CD8+ T cells (Fig. 2A). GFP positive gates were again set using non-transgenic B6 control mice (Fig. S2). Interestingly both numerically and as a percentage, macrophages comprised the highest amount of GFP+ (IL-10 producing) cells. Of note, the level of IL-10 producing CD4 and CD8 T cells were similar at the initial stages of acute and persistent infections and the vast majority of virus-specific T cells (i.e., tetramer positive) did not produce IL-10. GFP expression was not observed in NK cells or neutrophils (data not shown). While most of the cell subsets contained IL-10 producing populations at day 9 post infection, the percent and number of IL-10 producing APC (DC, macrophages, and B cells) were substantially higher in persistent infection and continued to remain elevated throughout viral persistence (Fig 2B). Thus, although multiple cells subsets produce IL-10 in response to viral infection, there was a distinct difference in the regulation of APC populations during acute and persistent infection.
DC are required to initially activate naïve antiviral CD4+ and CD8+ T cell responses in vivo (Jung et al., 2002; Probst and van den Broek, 2005). However, depletion of DC after priming (but before loss of T cell function) following LCMV-Cl 13 infection did not prevent T cell exhaustion (Fig. S2). We previously demonstrated that T cells retain function when removed from LCMV-Cl 13 infection at this same time (Brooks et al., 2006b), indicating that as infection progresses alternative APC in addition to DC, such as macrophage and B cells, play important regulatory roles to modulate previously activated T cell responses and adjust the immune response to the needs of the evolving antigenic environment. Therefore, we next sought to investigate the regulatory properties of IL-10 producing APC during persistent infection. The majority of IL-10 producing DC were CD11cbright, CD11bhi CD8 αlo, B220neg so we subsequently compared IL-10+ vs. IL-10− DC within this subset, however a smaller yet evident population of CD11cbright, CD11blo CD8 αhi, B220neg DC also produced IL-10 (Fig. S2). Of note, similar results as described below were observed when IL-10+ vs. IL-10− CD8+ DC or total DC were compared (data not shown). IL-10 expressing B cells are not limited to a single B cell subset and are observed in fractions of plasma cells, memory cells, marginal zone and germinal center compartments (data not shown). Compared to their non-IL-10 producing counterparts, IL-10-producing DC, macrophages, and B cells expressed equal if not enhanced levels of MHC class I and II proteins as well as the co-stimulatory molecules CD80 (B7-1) and CD86 (B7-2) (Fig 3A–C). The small numbers of IL-10 expressing APC present in acute infection (Figure 2) were of similar phenotype and DC subset distribution (Fig. S2 and S3). Taken together, these results indicate that IL-10 producing APC are amplified in persistent infection and represent mature populations capable of interacting with and potentially modulating T cell responses.
Numerous regulatory mechanisms in addition to IL-10 have evolved to restrain immune activity during persistent infection (Wilson and Brooks, 2011). Exemplary of these are PD-1/PDL1 interactions and the critical role these host-derived proteins play in sustaining T cell suppression in persistent viral infections (Barber et al., 2006). Interestingly, both PDL1 and PDL2 were significantly up-regulated on all IL-10 expressing APC populations compared to their non-IL-10 producing counterparts (Fig. 4A and S4) and subsequently maintained throughout the course of persistent infection (Fig 4A), suggesting that multiple inhibitory factors may be centralized onto single iAPC subsets.
To further explore the functions of iAPC populations, we assessed expression of other immunosuppressive molecules implicated in suppressing immune responses during persistent infection. Indoleamine 2,3-dioxygenase (IDO) dampens T cell responses via modulation of tryptophan catabolism (Mellor and Munn, 2004). IDO expression was substantially elevated in IL-10 producing DC, macrophages, and B cells relative to their non IL-10 producing counterparts (Fig 4B). TGFβ is another immuno-regulatory cytokine recently indicated to limit antiviral T cell responses and facilitate LCMV persistence (Tinoco et al., 2009). However, unlike other immunosuppressive factors, iAPC were not enriched for TGFβ transcripts (Fig 4B), indicating that iAPC do not universally up-regulate all immunosuppressive genes important during viral persistence. APC production of the potent pro-inflammatory cytokine IL-12 is a key component in the induction of IFNγ producing T cell responses (O’Garra, 1998). RT-PCR analysis in iAPC subsets indeed revealed a dramatic reduction in IL-12p35 levels when iAPC were compared to their non-IL-10 expressing counterparts (Fig 4B). Thus, in addition to increased expression of negative regulatory factors, iAPC also produced decreased levels of positive stimulatory factors.
The increased expression of T cell interacting proteins and immunosuppressive factors in conjunction with lower levels of stimulatory cytokines suggested that iAPC in particular would diminish instead of stimulate T cell responses. In accordance with their inability to prime naïve T cells, neither B cells nor macrophages derived directly from persistent infection (regardless of IL-10 expression) were able to stimulate naïve T cells ex vivo (data not shown). However, consistent with their immunosuppressive phenotype, iDC suppressed virus-specific T cells responses ex vivo compared to their non-IL-10 producing counterparts (Fig 4C and S4). T cell stimulatory capacity was enhanced in iDC, but not non-iDC, when IL-10, PDL1 or IDO signaling was blocked (Fig. 4C and S4). Thus, iAPC express concentrated levels of inhibitory molecules with the ability to independently suppress virus-specific T cell responses.
Multiple immunosuppressive molecules are similarly up-regulated early in acute and persistent virus infection; however, expression of these factors diminishes in an acute infection allowing T cells to maintain functionality (Barber et al., 2006; Brooks et al., 2006c). Conversely, at the same time when negative regulatory factors wane during acute infection, their expression is sustained and amplified in persistent infection promoting T cell exhaustion and viral persistence [Fig 1A and (Brooks et al., 2006c; Ejrnaes et al., 2006)]. Yet, the reasons for the simultaneous sustained and increased expression of these multiple factors are unclear. Interestingly, the expression kinetics of iAPC in acute and persistent infection parallels this course, being initially equal when virus titers are similar (day 5) and then waning to very low levels with decreased virus replication in acute infection and expanding in conjunction with the increasing virus replication in what will become a persistent infection (Fig. 5A). Inversely, levels of conventional (non-IL-10 producing APC) dramatically increased during acute infection, whereas they remained low in response to prolonged virus replication in persistent infection (Fig 5A.). As the persistent infection ensues, contraction of iAPC mirrors the decrease in virus titers (Fig. 5A). Thus, iAPC are invoked in response to elevating virus expression, leading to sustained and heightened expression of multiple immunoregulatory factors associated with immune suppression during viral persistence.
Because of the tight correlation between viral load and iAPC numbers, we sought to determine whether direct infection of APC was responsible in inducing the iAPC phenotype. While there was a slight increase in the percentage of iAPC that were infected relative to their non-producing counterparts, the vast majority of iAPC were not productively infected (Fig 5B & data not shown), indicating that direct infection is not driving the iAPC phenotype.
Recent studies have implicated an APC’s underlying metabolic program in regulating their fate and function. Oxidative metabolism appears to be required for the acquisition of an anti-inflammatory or immunoregulatory phenotype (Vats et al., 2006). Conversely, TLR activation of canonical pro-inflammatory programs in DCs requires a metabolic switch to aerobic glycolysis (Krawczyk et al., 2010). Interestingly, IL-10 signals inhibited the acquisition of TLR-induced glycolytic programs in APC (Krawczyk et al., 2010). To determine if this metabolic paradigm was operative in the context of a persistent viral infection, we quantified the mitochondrial membrane potential of IL-10 producing and IL-10 negative macrophages, DCs and B cells on day 9 after LCMV-Cl 13 infection. Strikingly, GFP+ macrophage, DCs and B cells exhibited increased Mitotracker Red staining relative to their GFP− counterparts ex vivo, indicating higher mitochondrial membrane potential (Fig. 6A).
The increased mitochondrial potential in IL-10 producing APCs could be indicative of a heightened oxidative metabolic state or the result of increased mitochondrial mass. To distinguish between these two possibilities, we quantified expression of the mitochondrially-encoded gene cytochrome c oxidase II (Mt-co2), relative to the nuclear housekeeping gene nuclear receptor interacting protein 1 (Rip140). The ratio of Mt-co2 to Rip140 was not statistically different between IL-10 producing and IL-10 negative DC and macrophage (Fig. 6B), suggesting an equivalent mitochondrial mass in both populations of APC. The metabolic state of IL-10 producing B cells is less clear, as Mt-co2 levels are significantly elevated in these populations- indicating an increase in mitochondrial number as well as potential. We further examined the expression levels of key glycolytic genes. IL-10 producing macrophages expressed markedly lower levels of the glycolysis genes hexokinase 2 (HK2), lactate dehydrogenase A (Ldha) and pyruvate kinase (Pkm2) (Fig. 6C), indicative of an oxidative/suppressive metabolic program. Interestingly, GFP+ and GFP− DC populations expressed similar levels of the glycolysis genes, suggesting a complex metabolic regulation. Hence, iAPC are metabolically distinct subsets of cells that simultaneously express multiple immunoregulatory factors with heightened suppressive activity.
The metabolic, phenotypic and functional attributes of the immunoregulatory macrophages that arise during infection are consistent with those of M2/alternatively activated macrophages (Lawrence and Natoli, 2011). The IL-10 expressing macrophages induced during persistent viral infection express low levels of GR1 [correlating with M2 development, (Lin et al., 2009) and are highly enriched for the M2 associated transcripts Arg1 and CD206 (Fig. 6D). iAPC populations also exhibit a distinct molecular profile. Interferon regulatory factors (IRFs) comprise a family of signaling molecules that play important roles in the regulation of M2 macrophage polarization and DC ontogeny (Savitsky et al., 2010) (Satoh et al., 2010). Expression of IRF4 is significantly elevated in IL-10 producing macrophages and DC, whereas expression of the related interferon regulatory factor IRF3 is similar compared to IL-10 non-producers (Fig 6E). B-lymphocyte induced maturation protein 1 (Blimp-1) is a transcriptional repressor in T and B cell development (Xin et al., 2011), promotes IL-10 production and is linked to maintenance of tolerogenic DC populations (Chan et al., 2009; Kim et al., 2011; Martins et al., 2006). Interestingly, Blimp1 expression was highly elevated in IL-10 expressing, compared to non-expressing, DC and macrophages (Fig. 6E). In accordance with the antagonism between Bcl6 and Blimp1, Bcl6 expression was decreased in iAPC (Fig. 6E). Thus, persistent infection promotes the amplification of iAPC populations with distinct molecular and metabolic profiles that simultaneously express multiple immunosuppressive and T cell modulating factors (IL-10, PDL1, PDL2, B7-1, B7-2 and IDO) capable of negatively regulating antiviral T cell activity.
Activation of effective CD4 and CD8+ T cell responses in the face of persistent viral infection is a major therapeutic goal. However, the multiple immunoregulatory pathways that are elevated and sustained during persistent infection, resulting in T cell suppression, must be overcome while not engendering wide-spread immunopathology. In this study we demonstrate that many immune cell types simultaneously produce IL-10, potentially all playing important and distinct roles in suppressing the host immune response. Interestingly, within each APC population we identified a distinct subset that co-expressed many of the dominant factors implicated in suppressing T cell responses in viral persistence. These immunoregulatory APC subsets have similarities to other previously described populations (i.e., M2 macrophages) that can arise during other states of chronic inflammation. Importantly, these specialized iAPC are dramatically amplified during persistent infection and peak during the transition into T cell dysfunction.
Our results demonstrate that invocation of iAPC is a fundamental and previously unrecognized response to viral replication. Paradoxically, the same inhibitory factors that are responsible for inducing viral persistence are also present early in acute infection, but then rapidly disappear to facilitate productive T cells responses (Fig. 1A, ,5A,5A, and (Barber et al., 2006; Brooks et al., 2006c). On the other hand, these same factors are sustained and amplified in what will become a persistent infection, promoting immune suppression and viral persistence. Thus, an inflection point exists when the immune response determines whether it is winning or losing the battle with viral replication and then adjusts itself accordingly. Interestingly, dynamic regulation of iAPC at the inflection point between virus clearance and persistence clarifies the enigmatic expression kinetics of multiple inhibitory factors. Based on the important role of these multiple inhibitory factors in inducing and maintaining T cell/immune exhaustion, the co-expression of these molecules by single cells that can themselves can be rapidly regulated is a mechanism whereby the immune response can be quickly adjusted to appropriately modulate aggressive immunity when the battle is being won, with suppressive responses that prevent excessive immunopathology when the battle is lost. However, the immune response is not binary (i.e., vigorous effector responses or complete exhaustion) and by varying the levels of iAPC, it may be possible to achieve the spectrum of T cell (and overall immune) quality/quantity required to fight a given pathogen. Importantly, targeting of these cells may represent an effective therapeutic approach to delete immunosuppressive cells while maintaining stimulatory APC to sustain ongoing immune responses.
iAPC exhibit a separate molecular, metabolic, phenotypic and functional profile compared to their conventional APC counterparts residing in the same environment, indicating that they are a distinct group of cells that naturally arise during viral infection and are specifically and dramatically amplified in response to persistent infection. Further, despite belonging to different APC cell types (i.e., DC, macrophages, B cells), the similar molecular profile of iAPC, suggests that they may share common triggers and differentiation programs in response to ongoing virus replication. The direct correlation of iAPC numbers with virus replication kinetics suggests that iAPC may not be sustained individually long-term as much as continually stimulated. Based on the known plasticity of macrophages (and potentially DC, either through functional modification or rapid turnover), it is possible that APC are continually recruited into the immunoregulatory subset to appropriately modulate the current needs of the immune environment. However, the fact that blockade of IL-10, PDL1 and other negative regulatory factors successfully enhances effector T cell responses during persistent infection indicates that these iAPC populations continue to have an important immune-modulatory role throughout the course of such infections.
The concurrent blockade of multiple suppressive signals additively enhances antiviral T cell activity (Blackburn et al., 2009; Brooks et al., 2008; Jin et al., 2010), suggesting the existence of a threshold level of duration and/or strength of signals that must be reached by a particular cell to initiate the exhaustion program. It is important to note that inhibitory factors may possess different levels of activity during persistent infection. In vivo inhibition of IL-10 or PDL1 can prevent or decrease persistent virus replication, while blockade of other immunoregulatory molecules in vivo such as CTLA4 have less obvious impact (Barber et al., 2006; Kaufmann et al., 2007), potentially serving to fine-tune the ongoing response. Thus, immunosuppression is a multifactorial process in which many factors contribute to varying degrees (likely, through targeting different cells and distinct cellular functions) to cumulatively restrain immune activity (Blackburn et al., 2009). By this mechanism, the clustering of multiple immunoregulatory factors on iAPC provides a method for rapid delivery of numerous potent inhibitory signals directly to individual T cells.
Antibody blockade of IL-10, PDL1 or other factors during persistent infections enhances T cell activity (Barber et al., 2006; Brooks et al., 2006c; Ejrnaes et al., 2006; Fahey and Brooks, 2010). This underscores the fact that maintenance of exhaustion as an active process and that dynamic alterations in post-priming immune environment are critical modulators of T cell activity. In addition to IL-10 expressing DC, we also identified the emergence of B cells and a large percentage of macrophages that express IL-10 in concert with many other inhibitory factors that suppress antiviral immunity during persistent infection. When T cells are removed after priming, but prior to loss of function, from what will become a persistent infection, they retain effector activity and differentiate into memory cells (Brooks et al., 2006b). Paralleling this, our finding that DC deletion after priming, but before loss of T cell activity does not prevent T cell exhaustion, indicates that multiple iAPC populations modulate ongoing T cell responses to potentiate T cell exhaustion. Consequently, while not capable of priming naïve T cells, IL-10 expressing macrophages and B cells likely play critical roles as infection progresses. In this manner, IL-10 expressing macrophages and B cells express high levels of MHC molecules, fostering their ability to interact with T cells. How each of these iAPC target distinct aspects of the immune response to promote persistent infection is undoubtedly complex, but will ultimately be of tremendous importance for the development of targeted therapies to overcome the mechanisms viruses utilize to subvert immune clearance.
Studies examining the metabolic programs adopted by APCs in the context of infection and immunity are limited. Elegant studies by Chawla and colleagues demonstrated that expression of an oxidative metabolic program as an absolute requirement for the polarization of M2 or alternatively activated macrophages in chronic inflammatory models (Vats et al., 2006). In contrast, activation of DCs by pro-inflammatory signals induces a metabolic switch from oxidative to glycolytic metabolism which, if pharmacologically inhibited, alters T cell stimulatory capacity (Krawczyk et al., 2010). In our study, we observed that IL-10 producing macrophages have significantly higher membrane potential (with no appreciable change in mitochondrial mass and decreased glycolytic gene expression) relative to their IL-10 negative counterparts, indicative of an oxidative metabolic program. Further, IL-10 expressing macrophages also expressed high levels of multiple phenotypic and molecular factors associated with M2/alternatively activated macrophage differentiation. Unlike acute LCMV infection, during persistent LCMV infection, these macrophages represent a high proportion of the total macrophage population, suggesting a high likelihood that any interaction between a T cell and macrophage will include these regulatory macrophages and result in the delivery of multiple potent inhibitory signals that decrease antiviral responses in order to limit immunopathology. Interestingly, these data support the concept that emergence of oxidative, regulatory macrophages is a general feature of chronic inflammatory events.
IL-10 expressing DCs similarly exhibit increased mitochondrial membrane potential with no notable change in mitochondrial mass. In vitro studies on demonstrated an antagonistic role for IL-10 in the acquisition of the glycolytic program in response to TLR signaling. However, we did not find significant differences in the glycolytic program of IL-10 positive and negative DCs and B cells in vivo, suggesting that IL-10 signals may not directly suppress glycolytic gene expression in these cells. Alternatively, IL-10 signaling may similarly suppress the glycolytic program in both IL-10 expressing and non-expressing DC and B cells. Importantly, our study identifies a metabolic and gene expression pathway associated with naturally arising iAPC during a persistent infection.
The tight correlation of iAPC with viral titers and their similar molecular/transcriptional profiles suggests the existence of specific host sensors of viral replication that, in turn, regulate execution of the iAPC differentiation program. Direct targeting of iAPC, or such sensors, may provide a potent therapeutic strategy for rapidly and specifically dampening suppressive signals and bolstering T cell responses during viral persistence- without perturbing more global regulatory processes. A further understanding of how iAPC are generated, and the mechanisms by which they modulate T cell activity, may facilitate development of therapeutic strategies for the restoration of immune responses during persistent infection.
C57BL/6 (wild type), IL-10-deficient and CD11c-DTR mice were purchased from The Jackson Laboratory (Bar Harbor, ME). Vert-X, LCMV-GP61-80-specific CD4+ TCR transgenic (SMARTA) and the IL-10/Thy1.1 (10BiT) reporter mice [generously provided by Casey Weaver at the University of Alabama, Birmingham (via Gislaine Martins at Cedars-Sinai Medical Center, Los Angeles, California)] have been described previously (Madan et al., 2009; Maynard et al., 2007; Oxenius et al., 1998). All mice were housed under specific pathogen-free conditions and mouse handling conformed to the requirements of the University of California, Los Angeles Animal Research Committee guidelines. In all experiments mice were infected intravenously (i.v) via the retro-orbital sinus with 2×106 plaque forming units (PFU) of LCMV-Arm or LCMV-Cl 13. Virus stocks were prepared and viral titers were quantified as described previously (Brooks et al., 2005). The frequency of productively infected IL-10 expressing splenic DCs (GFP+, CD45+, CD3–, NK1.1–, CD11c+), B cells (GFP+, B220+) and macrophages (GFP+, CD45+, CD3–, NK1.1–, CD11c−, F4/80+) was determined by FACSorting the individual population and performing limiting dilution plaque assays (infectious center assays) on Vero cells (Brooks et al., 2006a). 100ng of diphtheria toxin (List Biological Laboratories, Cambell CA) was administered intraperitoneally on day 5 and 7 after LCMV-Cl 13 infection.
C57BL6 or IL-10-deficient recipient mice were lethally irradiated with 950 rads and on the same day received 20 million total bone marrow cells intravenously isolated from the femurs and tibia of either IL-10 deficient or WT donor mice. Recipient mice were treated with antibiotics (Sulfamethoxazole and Trimethoprim in the drinking water) for 3 wk to prevent infection and allow for hematopoietic reconstitution. Eight weeks following bone marrow transfer mice were bled to confirm consistent reconstitution and then infected with LCMV Cl13.
Analysis of IL-10 expressing (GFP+) and non-expressing (GFP−) cells was performed by staining directly ex vivo for surface expression of CD45-Pacific Orange or Pacific Blue, CD11c-Pacific Blue or PE, Thy1.1-FITC or PE, Thy1.2-PerCP, NK1.1-PerCPCy5, B220-APCCy7, CD11bPeCy7, F4/80-PE or APC, PDL1-PE or biotin followed by streptavidin-APC, PDL2-PE, MHC Class II-PE, MHC Class I-PE, CD80-PE, CD86-PE, CD4-Pacific Blue, CD8-Pacific Blue. Mitochondrial membrane potential was accessed by staining with Mitotracker FM Red (Invitrogen) according the manufacturer’s instructions. MHC tetramers were obtained from the NIH. Flow cytometric analysis was performed using the Digital LSR II (Becton Dickinson) in the UCLA Jonsson Comprehensive Cancer Center (JCCC)/Center for AIDS Research Flow Cytometry Core Facility (National Institutes of Health grant CA-16042).
IL-10 producing and non-producing DC and macrophages were sorted from spleen following B cell depletion (CD19 MACS beads, Miltenyi) as follows: IL-10 expressing DC (GFP+, CD45+, Thy1.2-;, NK1.1-, CD11c+ bright, CD11b+), non-IL-10-producing DC (GFP−, CD45+, Thy1.2-, NK1.1-, CD11c+ bright, CD11b+); IL-10 expressing macrophage (GFP+, CD45+, Thy1.2-, NK1.1-, F4/80+), and non-IL-10 producing macrophages (GFP−, CD45+, Thy1.2-, NK1.1-, F4/80+). The depleted B cell (CD19+) fraction was subsequently sorted for IL-10 producing (GFP+, CD45+, Thy1.2-, NK1.1-, B220+), and non-producing (GFP−, CD45+, Thy1.2-, NK1.1-, B220+) populations. Cells were sorted using a FACS Aria or Vantage fluorescence-activated cell sorter (Becton Dickinson). Post sort purity was >98%.
RNA purified from whole splenocytes or sorted APC populations was isolated with the RNeasy extraction kit (Qiagen). RNA was normalized for input and amplified directly using the One-Step RT-PCR kit (Qiagen). IDO, TGFb, IL-10, IL-12, Blimp1, IRF4, IRF3, BCL6 and HPRT were amplified using Applied Biosystems Assays-on-Demand TaqMan pre-made expression assays. IDO, TGFb and IL-12 expression were normalized to HPRT. Metabolic gene expression was determined from cDNA with the exception that SYBR green (Roche) real-time quantitative PCR assays were performed using the Roche Light cycler 480 II (Roche). Genes are normalized to ribosomal housekeeping gene Rip140. The primer sequences for the metabolic genes are available upon request.
Antigen specific naïve CD4+ T cells or CD8+ T cells were isolated from the spleens of SMARTA or P14 mice, respectively, and purified by negative selection (StemCell Technologies). T cells were then labeled with 2.5 uM CFSE (Invitrogen) and cultured for three days with sorted populations of IL-10 producing (GFP+) or non IL-10 producing (GFP−) APC at a ratio of 2:1 (40,000 T cells/20,000 APC). In some situations 10μg/ml anti-IL10R antibody (clone 1B1.3A; BioXcell) or anti-PDL1 antibody (clone 1F.9G2, BioXcell), or 200uM 1-methyl tryptophan (1-MT; Sigma-Aldrich) was added to the cultures. The frequency of SMARTA or P14 cells that proliferated during culture was determined as described (Gett and Hodgkin, 2000). Briefly, the number of cells in each division peak was divided by 2i (where i equals the number of divisions). The total number of SMARTA cells in each peak was then summed and this number was divided by the total number of SMARTA cells in the culture. This number was then expressed as the percent of cells that had proliferated. To obtain the stimulation index the total number of T cells that proliferated in the cultures of GFP+ culture was divided by the number of T cells that proliferated in the GFP− culture, multiplied by 100.
Plasma IL-10 levels were determined using the Quantikine IL-10 Elisa kit (R & D Systems, Minneapolis, MN). Optical density values were read using a Synergy 2 plate reader (BioTek, Winooski, VT) at 450nm.
Student’s t-tests (two-tailed, unpaired) were performed using the GraphPad Prism 5 software (GraphPad Software Inc.).
We thank L. Huang and A. L. Mellor at the Georgia Health Sciences University for helpful discussions and technical assistance regarding IDO activity. Our work was supported by the National Institutes of Health (Grants AI085043 and AI082975 to D.G.B.; AI060567 to E.B.W.; AI007323 to L.M.R) and the UCLA Center for AIDS Research (P30 AI028697).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.