Coregulation of CD8+ T cell exhaustion by multiple inhibitory receptors during chronic viral infection
Published online: 30 November 2008 | doi:10.1038/ni.1679
Shawn D Blackburn1, Haina Shin1, W Nicholas Haining2,3, Tao Zou1, Creg J Workman6, Antonio Polley1, Michael R Betts5, Gordon J Freeman4, Dario A A Vignali6 & E John Wherry1
"reversal of T cell exhaustion could be improved by therapeutic targeting of multiple inhibitory receptor pathways simultaneously, and approaches to 'fine-tune' T cell responses to persisting pathogens might be developed by targeting of inhibitory receptors that regulate specific T cell properties."....."There has been some debate, however, about precisely what T cell functions are regulated by PD-1.....Thus, it may be necessary to consider the full repertoire of inhibitory receptors expressed when investigating the function of any individual inhibitory pathway."...."The diversity of inhibitory receptors expressed by exhausted CD8+ T cells is remarkable."
T cell exhaustion often occurs during chronic infection and prevents optimal viral control. The molecular pathways involved in T cell exhaustion remain poorly understood. Here we show that exhausted CD8+ T cells are subject to complex layers of negative regulation resulting from the coexpression of multiple inhibitory receptors. Exhausted CD8+ T cells expressed up to seven inhibitory receptors. Coexpression of multiple distinct inhibitory receptors was associated with greater T cell exhaustion and more severe infection. Regulation of T cell exhaustion by various inhibitory pathways was nonredundant, as blockade of the T cell inhibitory receptors PD-1 and LAG-3 simultaneously and synergistically improved T cell responses and diminished viral load in vivo. Thus, CD8+ T cell responses during chronic viral infections are regulated by complex patterns of coexpressed inhibitory receptors.
"The improved CD8+ T cell responses after blockade of inhibitory receptors could benefit from both the blockade of an inhibitory pathway and enhanced control of infection and lower viral load. Blockade of inhibitory receptors could therefore be a pivotal event that initially improves T cell function and viral control. The resulting lower viral load could then have a secondary positive effect on the severity of exhaustion. Given the potential for clinical applications, it will be necessary to understand the contribution of continued inhibitory receptor blockade versus lower viral load to improved T cell responses."
After acute viral infection, memory CD8+ T cells rapidly reactivate effector functions, have high proliferative potential and are maintained by antigen-independent homeostatic proliferation1, 2, 3. These properties allow memory T cells to confer protective immunity. In contrast, during chronic viral infection, antigen-specific CD8+ T cells initially acquire effector functions but gradually become less functional as the infection progresses. This loss of function, known as 'exhaustion', is hierarchical, with properties such as proliferative potential and production of interleukin 2 lost early, production of tumor necrosis factor (TNF) persisting for somewhat longer and production of interferon-gamma (IFN-gamma) lost only at the most extreme stages of exhaustion4. First described with lymphocytic choriomeningitis virus (LCMV)5, CD8+ T cell exhaustion during persistent infection is a common feature of many experimental models and chronic human infection, including infection with human immunodeficiency virus (HIV) and hepatitis B and C viruses4. T cell exhaustion is probably a chief reason for the ineffective viral control in these situations.
An important function for PD-1 and its ligand PD-L1 has been reported for CD8+ T cell exhaustion during chronic viral infection6. PD-1, an inhibitory receptor in the CD28 'superfamily'7, is highly expressed by exhausted CD8+ T cells from chronically infected mice but not by functional LCMV-specific CD8+ T cells from mice that have cleared the infection6. Blocking the PD-1-PD-L pathway in vivo increases virus-specific CD8+ T cell responses, enhances 'per-cell' function and decreases the viral load6. Involvement of the PD-1 pathway has also been shown during infection with HIV and hepatitis B and C viruses8, 9, 10, 11, 12, 13, 14, 15. Such studies have indicated that exhausted CD8+ T cells could be 'rejuvenated' to enhance antiviral immunity. However, functional restoration by PD-1-PD-L blockade is incomplete, and defects in CD8+ T cells remain after PD-1 pathway blockade6, which suggests the involvement of other negative regulatory pathways in CD8+ T cell exhaustion. Comparison of global gene expression profiles of exhausted CD8+ T cells with functional virus-specific effector and memory CD8+ T cells has identified upregulation of many inhibitory receptors in addition to PD-1 (ref. 16). HIV-specific CD4+ T cells have been shown to coexpress PD-1 and another inhibitory molecule, CTLA-4 (ref. 17). The effect of upregulation of these additional inhibitory receptors on CD8+ T cell dysfunction during chronic viral infection is not known. It is also unclear whether these inhibitory receptors are expressed by distinct subpopulations of exhausted CD8+ T cells or whether there are exhausted CD8+ T cells that coexpress multiple inhibitory receptors. Finally, there is little information about whether these potential inhibitory pathways serve different functions in regulating T cell exhaustion.
We addressed these issues with a mouse model of chronic LCMV infection. Exhausted CD8+ T cells coexpressed multiple inhibitory receptors, and the pattern of inhibitory receptor coexpression affected the functional quality of these virus-specific CD8+ T cells during chronic infection. The severity of chronic infection was associated with the number and 'intensity' of inhibitory receptors expressed. In addition, in vivo blockade of the PD-1 and LAG-3 inhibitory receptor pathways together led to much greater reversal of T cell exhaustion and viral control than did blockade of either pathway alone. These observations indicate that layers of negative regulation exist in exhausted CD8+ T cells as a result of the coexpression of multiple inhibitory receptors and have implications for therapeutic interventions during chronic infection.
Here we have investigated the expression of various inhibitory receptors by exhausted CD8+ T cells during chronic viral infection, with the following results. First, exhausted CD8+ T cells expressed up to seven distinct inhibitory receptors. Second, inhibitory receptors were coexpressed by exhausted CD8+ T cells, and complex patterns of coexpression were apparent. Third, the severity of viral infection affected the diversity and amounts of inhibitory receptors expressed. Fourth, the inhibitory receptors LAG-3 and PD-1 operated as at least partially nonoverlapping pathways that negatively regulated T cell responses during chronic viral infection. Finally, combined blockade of LAG-3 and PD-L1 acted in synergy to enhance T cell responses and viral control during chronic LCMV infection. These observations collectively indicate that multiple negative regulatory pathways contribute to CD8+ T cell exhaustion and suggest that the pattern of inhibitory receptor coexpression is a useful correlate of the amount of T cell exhaustion. Thus, reversal of T cell exhaustion could be improved by therapeutic targeting of multiple inhibitory receptor pathways simultaneously, and approaches to 'fine-tune' T cell responses to persisting pathogens might be developed by targeting of inhibitory receptors that regulate specific T cell properties.
Our results obtained with different LCMV infections suggest that the patterns of coexpression of inhibitory receptors are influenced by the severity and/or type of infection. Our studies indicate the pathogenesis of infection as an important parameter influencing inhibitory receptor expression and suggest the involvement of viral load. Additional studies are needed to investigate the influence of inflammation and CD4+ T cell help, as well as to investigate the patterns of the coexpression of inhibitory receptors during other types of persistent infections. Many studies have emphasized the relationship between the severity of infection and the extent of CD8+ T cell dysfunction6, 8, 23, 40, 41, 42. The improved CD8+ T cell responses after blockade of inhibitory receptors could benefit from both the blockade of an inhibitory pathway and enhanced control of infection and lower viral load. Blockade of inhibitory receptors could therefore be a pivotal event that initially improves T cell function and viral control. The resulting lower viral load could then have a secondary positive effect on the severity of exhaustion. Given the potential for clinical applications, it will be necessary to understand the contribution of continued inhibitory receptor blockade versus lower viral load to improved T cell responses.
The pathway of PD-1 and PD-L1 or PD-L2 has received considerable attention for its involvement in regulating T cell exhaustion. There has been some debate, however, about precisely what T cell functions are regulated by PD-1 (refs. 7,43). Our data provide a potential explanation for these apparently disparate results, if the coexpression of other inhibitory receptors differed in the T cell populations analyzed in different studies. For example, the greater effect of PD-1 pathway blockade alone on PD-1intCD8+ T cells36 could result from the presence of multiple additional inhibitory receptors, such as CD160, on the PD-1hi subset of exhausted CD8+ T cells. Thus, it may be necessary to consider the full repertoire of inhibitory receptors expressed when investigating the function of any individual inhibitory pathway.
The diversity of inhibitory receptors expressed by exhausted CD8+ T cells is remarkable. Among the set of inhibitory receptors examined here, exhausted CD8+ T cells expressed two inhibitory molecules in the CD28 'superfamily' (PD-1 and CTLA-4) but downregulated a third (BTLA). PD-1-PD-L1 is a chief pathway that regulates T cell exhaustion during many chronic viral infections7. During chronic LCMV infection, CTLA-4 does not seem to have a notable effect on exhausted CD8+ T cells6, but an effect of CTLA-4 on HIV-specific CD4+ T cells has been described17. LAG-3 binds to major histocompatibility complex class II molecules, consistent with homology between LAG-3 and CD4 (ref. 44). LAG-3 negatively regulates T cell activation and proliferation in several settings21, 45, but its function during chronic viral infection has remained unknown. Although a function for LAG-3 has been reported on regulatory T cells46, there is a CD8+ T cell-intrinsic function for LAG-3 in antitumor responses47. LAG-3 limits T cell proliferation and memory T cell homeostasis through cell cycle arrest25, 48, 49, 50. The function of LAG-3 as a negative regulator of cell cycle progression of CD8+ T cells25, 48, 49, 50 fits well with the higher Ki67 expression we noted after LAG-3 blockade during chronic LCMV infection. The in vivo studies presented here are consistent with a CD8+ T cell-intrinsic function for LAG-3 but do not exclude a function for LAG-3 on other cell types.
In addition to PD-1 and LAG-3, which have been shown by in vivo blocking studies to be involved in regulating dysfunction, exhausted CD8+ T cells coexpressed a variety of other inhibitory receptors that might be future therapeutic targets. Exhausted CD8+ T cells expressed PIR-B and GP49, both of which have been linked to negative regulation of mast cells, neutrophils, eosinophils, macrophages and natural killer cells26, 27. Notably, GP49A is not inhibitory, whereas GP49B is a negative regulator of Fc receptor signaling27, 51. Because antibodies do not distinguish between GP49A and GP49B, future studies are needed to further delineate this pathway. CD160 was also upregulated on exhausted CD8+ T cells; this molecule inhibits human CD4+ T cell responses20. Our preliminary in vitro studies are consistent with an inhibitory function for CD160 on exhausted CD8+ T cells, but the exact function of this molecule in natural killer cells and T cells and during viral infection remains poorly understood. Finally, 2B4 (also called CD244) is a SLAM family member that can be either activating or inhibitory and has been best characterized on T cells in the intestinal mucosa and natural killer cells19, 52, 53. The amount of 2B4 per cell and the expression of the 'downstream' adaptor molecule SAP are key determinants of whether 2B4 delivers an inhibitory or an activating signal53. Crosslinking of 2B4 on 2B4lo or 2B4int cells delivers a positive signal38. However, when 2B4 expression is high and SAP expression is low, 2B4 is inhibitory38. SAP amounts are probably limiting in exhausted CD8+ T cells but not effector CD8+ T cells on the basis of gene expression profiles16. The high expression of 2B4 by exhausted CD8+ T cells and the effects of in vitro CD48 blockade are consistent with an inhibitory function for 2B4 on 2B4hi exhausted CD8+ T cells. The diverse potential of these inhibitory receptors suggests that these pathways contribute in qualitatively different ways to T cell exhaustion. Indeed, the results from the dual blockade of PD-L1 and LAG-3 indicate that these two pathways impart distinct regulatory effects on exhausted CD8+ T cells.
A question that emerges from these and other studies is whether inhibitory receptors cause T cells to become exhausted in the first place. Functional effector CD8+ T cells upregulate PD-1 and LAG-3 during the first week of an acute viral infection6, 16, however, and some of these cells form functional memory T cells once the infection is cleared and expression of inhibitory receptors decreases. Although prolonged expression of inhibitory receptors could be involved, the pathways involved in the induction of T cell exhaustion and the lineage fate 'decisions' of T cell exhaustion and T cell memory remain poorly understood. Published studies have identified about 500 genes that are differently expressed in exhausted CD8+ T cells and naive, effector and memory T cells16, and it is likely that many pathways in addition to those activated by inhibitory receptors are involved in the early events controlling T cell exhaustion.
In summary, here we have identified coexpression of inhibitory receptors as a key feature of exhausted CD8+ T cells responding to chronic viral infection. Blockade of the PD-1-PD-L pathway identified a previously unknown function for coregulation of exhausted CD8+ T cells by a second inhibitory receptor, LAG-3, and the complex pattern of coexpression of these and other inhibitory receptors suggests a 'tunable' array of inhibitory pathways acting on exhausted CD8+ T cells. Combination therapies for chronic viral infection have considerable potential, and studies combining blockade of PD-1-PD-L or the interleukin 10 receptor with therapeutic vaccination have shown promising results54, 55. Our studies have shown synergistic effects of dual blockade of two inhibitory receptors during chronic viral infection. Future studies should help define the intracellular targets of these inhibitory pathways and test the potential of therapeutic interventions that target combinations of negative regulatory pathways during chronic viral infection.
Inhibitory receptors expressed by exhausted T cells
To identify the genes with expression patterns closest to that of Pdcd1 (which encodes PD-1) in naive, effector, memory and exhausted CD8+ T cells, we did a nearest-neighbor analysis18 using published gene expression data16. Among the top 100 neighbors of Pdcd1, we found several genes encoding surface receptors with inhibitory functions, including Cd244 (which encodes 2B4)19, Cd160 (ref. 20), Ctla4 and Lag3 (ref. 21; Supplementary Table 1 online); the expression of several of these potential inhibitory genes was highly correlated with Pdcd1 expression (Fig. 1a). In contrast, the expression of genes encoding other potential inhibitory receptors, such as Klra3, Cd80 and Klrg1, was either neutrally or negatively correlated with Pdcd1 expression. Thus, at the population level, exhausted cells showed coordinated upregulation of specific genes encoding inhibitory receptors.
We then examined the expression of inhibitory and natural killer cell receptors by flow cytometry of naive and LCMV-specific memory and exhausted CD8+ T cells. We used various strains of LCMV to generate either acute infection and functional memory CD8+ T cells or chronic infection and exhausted CD8+ T cells. Initially, we used the Armstrong (Arm) strain of LCMV, which results in acute infection and functional memory T cells, and the clone 13 strain, which results in chronic infection and T cell exhaustion22, 23, 24 (discussed below). Exhausted CD8+ T cells specific for the H-2Db-restricted gp33-41 epitope of the LCMV glycoprotein (amino acids 33-41; called 'gp33' here) from clone 13-infected mice had higher expression of PD-1, LAG3, 2B4, CD160, CTLA-4, PIR-B and GP49 and lower expression of KLRG1, NKG2A, BTLA and the potentially activating receptor NKG2D than did the functional H-2Db gp33 tetramer-positive memory CD8+ T cells from Arm-infected mice (Fig. 1b). Each of these upregulated receptors has been reported to negatively regulate the function of T cells or other hematopoietic cells19, 20, 25, 26, 27. Thus, multiple inhibitory receptors were expressed by exhausted CD8+ T cells but not by memory or naive CD8+ T cells. In addition, the upregulation of inhibitory receptors by exhausted CD8+ T cells was selective, as the inhibitory pathways KLRG1, NKG2A and BTLA were not enhanced.
Varying severity of LCMV infection in mice
We next used four types of LCMV infection to examine how the severity of infection affects virus-specific CD8+ T cell responses and expression of inhibitory receptors. In addition to the infections with Arm and clone 13 mentioned above, we used infection with LCMV T1B, as well as infection with clone 13 plus transient depletion of CD4+ T cells with antibody to CD4 (anti-CD4). We first defined the kinetics of viral replication with these four infection types. Infection of adult C57BL/6 mice with Arm resulted in high viral replication on day 3 after infection and complete clearance by days 8-10 after infection28 (Fig. 2a,b). For infection with T1B, systemic viral replication lasted about 3-4 weeks, whereas for clone 13, viremia lasted about 2-3 months (Fig. 2a). Anti-CD4 plus infection with clone 13 resulted in lifelong viremia5, 23, 29 (Fig. 2a). Thus, the use of Arm, T1B, clone 13 alone and clone 13 plus anti-CD4 would allow us to compare CD8+ T cell function and the expression of inhibitory receptors during infections of different severity.
All four types of LCMV infection generated robust LCMV-specific CD8+ T cell responses that peaked at about 1-2 weeks after infection, contracted and were then maintained (Fig. 2c,d). After infection with Arm, virus-specific CD8+ T cells followed a typical pattern of functional memory CD8+ T cell differentiation30, 31, 32, including upregulation of CD127 and CD62L over time (Fig. 2e). During the three chronic viral infections, with the exception of some expression of CD127 on day 100 after infection with T1B, the expression of these markers of normal memory CD8+ T cell differentiation remained low for at least about 100 d (refs 24,33,34; Fig. 2e).
Infection severity and expression of inhibitory receptors
To assess the regulation of the expression of inhibitory receptors on exhausted CD8+ T cells, we examined the expression of PD-1, LAG-3, 2B4 and CD160 on virus-specific CD8+ T cells with the four infections described above. On day 30 after infection with Arm, H-2Db gp33+CD8+ T cells had mostly low expression of PD-1, LAG-3 and CD160 and low or intermediate expression of 2B4 (Fig. 3). Most virus-specific CD8+ T cells produced both IFN-gamma and TNF after peptide stimulation, an indicator of functional competence23, 35. During infection with T1B, most H-2Db gp33+CD8+ T cells expressed PD-1 but expressed somewhat less 2B4 than after infection with Arm. Expression of LAG-3 and CD160 on these cells was low. Approximately half of the gp33-specific CD8+ T cells from T1B-infected mice produced both IFN-gamma and TNF. During infection with clone 13, H-2Db gp33+CD8+ T cells had high expression of PD-1 and 2B4 and substantial expression of LAG-3, and some cells expressed CD160. Only about 10% of these gp33-specific CD8+ T cells produced both IFN-gamma and TNF. In the most severe condition (anti-CD4 plus clone 13), all four inhibitory receptors were expressed, and these virus-specific CD8+ T cells were very exhausted. As the severity of infection increased, we noted a higher frequency of inhibitory receptor-positive, virus-specific CD8+ T cells, and the H-2Db gp33+CD8+ T cells had substantially higher expression of each inhibitory receptor per cell and produced less IFN-gamma per cell (Fig. 3c,d). When we examined the PD-1int and PD-1hi subsets of exhausted CD8+ T cells36, it was apparent that PD-1hi, H-2Db gp33+ exhausted CD8+ T cells expressed more 2B4, LAG-3 and CD160 than did their PD-1int counterparts (Supplementary Fig. 1 online). Thus, virus-specific CD8+ T cells with high expression of PD-1 also coexpressed other inhibitory receptors.
Patterns of coexpression of inhibitory receptors
We next used multiparameter flow cytometry to analyze coexpression of PD-1, LAG-3, CD160 and 2B4. On day 8 after infection, a small proportion of H-2Db gp33+, CD8+ T cells from Arm-infected mice expressed more than one inhibitory receptor (Fig. 4a). Conversely, most H-2Db gp33+, CD8+ T cells from mice infected with T1B or clone 13 alone or clone 13 plus anti-CD4 coexpressed three or four inhibitory receptors. On day 30 after infection, the percentage of H-2Db gp33+, CD8+ T cells from Arm-immune mice that expressed two or more inhibitory receptors had decreased. The initially high inhibitory receptor coexpression during T1B infection also decreased to some extent by day 30. In contrast, the more severe infections with clone 13, alone or with anti-CD4, resulted in high coexpression of three or four inhibitory receptors on day 8 after infection, which persisted on day 30. Coexpression of inhibitory receptors was similar for gp33-specific CD8+ T cells from P14 mice (Supplementary Fig. 2 online), which suggested that T cell antigen receptor (TCR) diversity is not the main determinant of the pattern of coexpression of inhibitory receptors. Thus, during chronic LCMV infection, there was substantial coexpression of multiple inhibitory receptors on the same virus-specific CD8+ T cells. Moreover, the extent of inhibitory receptor coexpression was associated with the severity of infection.
Multiparameter staining allowed us to determine which subpopulations of exhausted CD8+ T cells expressing various combinations of inhibitory receptors were present during chronic infection. Of the 16 possible combinations, we detected 11, but over 90% of the exhausted H-2Db gp33+, CD8+ T cells from clone 13-infected mice were in five subpopulations (Fig. 4b). Notably, during infection with clone 13 (and infection with clone 13 plus anti-CD4), PD-1 was expressed by most virus-specific CD8+ T cells, followed by LAG-3 and 2B4 (Fig. 4b and Supplementary Fig. 3 online). CD160 was expressed by the lowest percentage of exhausted CD8+ T cells and was associated with coexpression of the other three molecules (Fig. 4b and Supplementary Fig. 3).
To determine whether the pattern of inhibitory receptor coexpression was associated with T cell function, we did intracellular cytokine staining (ICS) for IFN-gamma and TNF after peptide stimulation and stained for the four inhibitory receptors. During infection with clone 13, the degree of CD8+ T cell exhaustion was directly associated the number of inhibitory receptors coexpressed (Fig. 4c). Thus, during chronic viral infection, coexpression of more inhibitory receptors was associated lower T cell function.
In vitro blockade of inhibitory pathways
To assess the effect of inhibitory receptors on exhausted CD8+ T cells, we investigated short-term (less than 18 h) changes in T cell properties in the presence of blocking antibodies with three separate in vitro assays: an in vitro killing assay37, an in vitro apoptosis assay36 and an ICS assay for cytokine production (Supplementary Fig. 4 online). In vitro cytotoxicity was enhanced by anti-CD160 but not by anti-PD-L1 or anti-LAG-3 or by blockade of the 2B4 ligand CD48. To examine T cell apoptosis and survival, we separated exhausted CD8+ T cells into PD-1int and PD-1hi subsets as described36. Expression of active caspase 3 was lower in the PD-1int subset in the presence of anti-PD-L1 but not in the presence of anti-LAG-3 or anti-CD160. In contrast, anti-CD160, but not anti-PD-L1 or anti-LAG-3, improved the survival of the PD-1hi subset. Moderate 2B4 expression can deliver a costimulatory signal, whereas high 2B4 expression can be inhibitory38. Blocking the 2B4 pathway with anti-CD48 decreased cytokine production by 2B4int LCMV-specific CD8+ T cells from chronically infected mice but significantly increased IFN-gamma production by exhausted 2B4hi CD8+ T cells (P = 0.026). CD48 can also bind CD2 and deliver a costimulatory signal. Blockade of CD2 decreased IFN-gamma production from exhausted 2B4int and 2B4hi CD8+ T cells. These observations are consistent with a costimulatory function for CD48 binding to either CD2 or 2B4 on 2B4int virus-specific CD8+ T cells and an inhibitory function of 2B4 through interactions with CD48 on exhausted 2B4hi CD8+ T cells. Although additional in vivo studies are warranted, these in vitro experiments suggest that the CD160 and 2B4 pathways can contribute to CD8+ T cell exhaustion.
Simultaneous blockade of PD-L1 and LAG-3 in vivo
It is unclear whether different inhibitory receptors have overlapping functions or whether these distinct pathways contribute independently to T cell dysfunction during chronic viral infection. To distinguish between those possibilities, we blocked two pathways simultaneously with blocking antibodies to PD-L1 and LAG-3, either alone or in combination during chronic LCMV infection in vivo (Fig. 5a). Two weeks of treatment with anti-LAG-3 alone did not improve the frequency or absolute number of tetramer-positive CD8+ T cells specific for the H-2Db-restricted gp276-286 epitope of the LCMV glycoprotein (amino acids 276-286; H-2Db gp276), whereas blocking the PD-1 pathway augmented the H-2Db gp276 response (Fig. 5b,c). Despite the lack of effect of anti-LAG-3 alone on T cell numbers, combining LAG-3 blockade with PD-1-PD-L1 blockade resulted in an 81% higher frequency of H-2Db gp276+CD8+ splenocytes relative to that of the group treated with anti-PD-L1 and a five-fold improvement relative to the untreated group (Fig. 5b), with corresponding increases in absolute numbers (Fig. 5c). Dual blockade also resulted in significantly greater CD8+ T cell responses to the np396-404 epitope of the LCMV nucleoprotein (amino acids 396-404; np396), the gp33 and gp276 peptides and a pooled peptide mixture containing 20 known LCMV epitopes than did either single treatment or control treatment, as measured by ICS instead of tetramer staining (Fig. 5d).
We also measured the frequency of LCMV-specific CD8+ T cells in the blood before and after treatment (Fig. 5e,f). The frequency of circulating CD8+ T cells positive for H-2Db gp276 tetramers remained essentially constant in control mice and mice treated with anti-LAG-3 only. PD-1-PD-L1 blockade alone resulted in responses that were about 1.8-fold higher (P = 0.0456), whereas dual blockade of both LAG-3 and PD-1-PD-L1 led to a circulating pool of H-2Db gp276+CD8+ T cells that was about 3.7-fold larger (P = 0.0307). Thus, blocking the PD-1-PD-L pathway in vivo identified substantial involvement of the LAG-3 pathway in coregulating CD8+ T cell exhaustion during chronic viral infection.
Greater function with LAG-3 and PD-L1 coblockade
To determine whether inhibitory receptor blockade improved the quality of the antiviral CD8+ T cells, we examined function on a 'per-cell' basis. Treatment with anti-PD-L1 alone resulted in a significantly higher percentage of H-2Db gp276 tetramer-positive CD8+ T cells that produced IFN-gamma (Fig. 6a). Although treatment with anti-LAG-3 alone had little effect on the severity of exhaustion, combined treatment with anti-LAG-3 and anti-PD-L1 led to substantially greater 'per-cell' function than did anti-PD-L1 alone, anti-LAG-3 alone or control treatment (Fig. 6a,b). Combined treatment with anti-LAG-3 and anti-PD-L1 also resulted in a significantly greater percentage of dual-function (CD107+IFN-gamma+) gp276-specific CD8+ T cells than did treatment with anti-PD-L1 or anti-LAG-3 alone (Fig. 6b,c). Although the percentage of cells producing both IFN-gamma and TNF was only slightly different in the groups (Supplementary Fig. 5 online), the absolute number of dual-cytokine producers and CD107+IFN-gamma+ virus-specific CD8+ T cells was much greater after in vivo blockade (data not shown). Thus, combining LAG-3 blockade with PD-1-PD-L1 blockade resulted in substantially better reversal of exhaustion than did PD-1-PD-L1 blockade alone.
Despite the lack of effect of anti-LAG-3 alone on antiviral T cell numbers or effector function, this treatment seemed to have an effect on exhausted CD8+ T cells. Blocking LAG-3 alone, as well as PD-L1 alone or both LAG-3 and PD-L1, resulted in higher expression of PD-1 on virus-specific exhausted CD8+ T cells in vivo (Supplementary Fig. 6 online). As TCR signals can increase PD-1 expression39, these results suggest that blockade of either LAG-3 or PD-L1 or both can modulate TCR signaling by persistent antigen. Although it is possible that the PD-L1 and LAG-3 blockade operates through other cell types, our results are consistent with the idea that blockade of these inhibitory receptors acts directly on exhausted, virus-specific CD8+ T cells.
We next examined whether the blockade treatments were associated with more T cell division. Although about 30% of the H-2Db gp276+CD8+ T cells from control mice expressed the proliferation marker Ki67, about 50% of the virus-specific CD8+ T cells in all three treatment groups expressed Ki67 (Fig. 6d,e). In contrast to the minimal effect of anti-LAG-3 alone on effector function, blockade of LAG-3 and PD-1-PD-L1 alone or together resulted in similarly improved entry of exhausted CD8+ T cells into the cell cycle. However, many more virus-specific CD8+ T cells accumulated after treatment with anti-PD-L1, and especially after treatment with both anti-LAG-3 and anti-PD-L1, than after treatment with anti-LAG-3 alone.
Improved viral control with blockade of both PD-L1 and LAG-3
To determine whether the blockade of inhibitory receptors affected viral control during treatment with anti-CD4 and infection with clone 13, we measured viral load in the blood before and after treatment. The viral load of control mice treated with immunoglobulin G2b (IgG2b) was only slightly lower (by about 22%) during this time. In contrast, viral load was 70% lower in mice treated with anti-LAG-3 alone, 48% lower with anti-PD-L1 and 81% lower with dual treatment (Fig. 7a). Treatment with anti-LAG-3 or anti-PD-L1 alone also resulted in significantly lower viral loads in the spleen, liver and kidney (Fig. 7b). However, combined blockade of LAG-3 and PD-L1 resulted in viral loads that were significantly lower in several organs than did blockade of LAG-3 or PD-L1 alone.
Mice, viruses and infections.
Four-week-old C57BL/6 mice were from Jackson Laboratories. Ly5.1+ P14 mice bearing H-2Db gp33-specific TCR were backcrossed ten generations or more to C57BL/6 mice and were maintained in the Wistar Institute animal colony. LCMV strains were propagated, titered and used as described23. C57BL/6 mice were infected intraperitoneally with 2 times 105 plaque-forming units (PFU) of LCMV Armstrong, intravenously with 2 times 106 PFU LCMV T1B or intravenously with 2 times 106 PFU LCMV clone 13. In experiments in which mice depleted of CD4+ cells were used, C57BL/6 mice were injected intraperitoneally with 200 mug anti-CD4 (GK1.5; BioXcell) on days -1 and +1 relative to infection with LCMV clone 13 on day 0. Chimeric P14 mice were generated by adoptive transfer of 5 times 102 naive TCR-transgenic T cells into naive C57BL/6 mice, followed by infection with LCMV Armstrong, T1B or clone 13. All mice were used in accordance with guidelines of the Institutional Animal Care and Use Committee of the Wistar Institute.
Flow cytometry and intracellular cytokine staining.
Lymphocytes were isolated from the spleen and peripheral blood as described23. Tetramers of major histocompatibility class I peptide were made and used as described23. Antibodies were from eBioscience, BioLegend, Invitrogen, Abcam, Accurate Chemical, R&D Systems and BD Biosciences (Supplementary Table 2 online). Lymphocytes were stained and analyzed as described23. ICS and the CD107a assay were done as described6, 23. Cells were stimulated with individual LCMV peptides or a pool of 20 LCMV epitopes56 (Supplementary Table 3 online). Cells were then analyzed on a LSR II flow cytometer (BD Immunocytometry Systems). Approximately 1 times 106 events were collected per sample. Data were analyzed with FlowJo v.8.2 (TreeStar). Doublets were removed by gating on a plot of forward-scatter area versus forward-scatter height. Dead cells were removed by gating on Live/Dead Aqua (Invitrogen) versus forward-scatter area. Cells were subjected to a lymphocyte gate on a plot of forward-scatter area versus side scatter, and then were sequentially gated on CD4-, CD8+ and H-2Db gp33+ events. Positive and negative gates for each inhibitory receptor on H-2Db gp33+, CD8+ T cells were drawn on the basis of 'fluorescence-minus-one' controls. The Boolean gating function of FlowJo was used to assess each possible inhibitory receptor expression pattern. Coexpression patterns were analyzed with the SPICE program from the Exon distribution server.
In vivo antibody blockade.
Rat anti-PD-L1 (200 mug; 10F.9G2; prepared 'in house'), rat IgG2b isotype-matched control antibody (200 mug; BD Biosciences) or rat blocking anti-mouse LAG-3 (200 mug; C9B7W; prepared 'in house') was administered intraperitoneally every third day for 2 weeks. Anti-PD-L1 and anti-LAG-3 do not produce depletion in vivo6, 46. For dual blockade of PD-L1 and LAG-3, 200 mug antibody to each was used. Control groups were left untreated or were injected with 200 mug of isotype-matched control antibody; no differences were noted between these control mice (data not shown).
Nearest-neighbor analysis was done essentially as described18 with the GeneNeighbor marker analysis algorithm to calculate the nearest neighbors to PD-1 in gene expression profiles. Gene expression profiles from the naive, effector, memory and exhausted CD8+ T cells used here have been published16.
P values were calculated by pairwise t-tests and analysis of variance (where more than two groups were compared). Pearson correlation coefficients were calculated with the GeneNeighbor marker analysis algorithm.