|Home | About | Journals | Submit | Contact Us | Français|
TLR7 is an innate signaling receptor that recognizes single-stranded viral RNA and is activated by viruses that cause persistent infections. We show that TLR7 signaling dictates either clearance or establishment of life-long chronic infection by lymphocytic choriomeningitis virus (LCMV) Cl 13 but does not affect clearance of the acute LCMV Armstrong 53b strain. TLR7−/− mice infected with LCMV Cl 13 remained viremic throughout life from defects in the adaptive anti-viral immune response, notably diminished T cell function, exacerbated T cell exhaustion, decreased plasma cell maturation, and negligible anti-viral antibody production. Adoptive transfer of TLR7+/+ LCMV immune memory cells that enhance clearance of persistent LCMV Cl 13 infection in TLR7+/+ mice fail to purge LCMV Cl 13 infection in TLR7−/− mice, demonstrating that a TLR7-deficient environment renders anti-viral responses ineffective. Therefore, methods that promote TLR7 signaling are promising treatment strategies for chronic viral infections.
Chronic viral infections such as human immunodeficiency virus, hepatitis B virus and hepatitis C virus (HCV) result in dysfunctional immune responses, including altered innate immune responses, T cell exhaustion and defective B cell responses (Frebel et al., 2010; Liu et al., 2009; Oliviero et al., 2011; Urbani et al., 2006). Understanding the mechanisms that cause viral persistence should lead to optimally planned therapies to overcome such infections. For example, the subversion of host T and B cell immune responses through up-regulation of host negative immune regulators (NIRs) effectively exacerbated persistence and blockade of NIR signaling such as IL-10, PD-1, TGF-β, and LAG-3 resurrected T cell function that enhanced viral clearance and terminated the persistent infection (Barber et al., 2006; Blackburn et al., 2009; Brooks et al., 2006; Tinoco et al., 2009).
LCMV is a negative-strand virus containing two RNA segments (Buchmeier et al., 2007). The LCMV strain Armstrong (Arm) 53b, when inoculated into mice intravenously (i.v.), generates an acute infection. As a result, a robust anti-LCMV-specific CD8+ T cell response was developed within 7 days and terminated the infection (Brooks et al., 2006). Instillation of LCMV Arm 53b into newborn mice resulted in a lifelong persistent infection (Oldstone, 2002). Isolation of LCMV virus clones from splenic lymphoid cells of such persistently infected mice resulted in the generation and characterization of LCMV Cl 13, an LCMV variant that differs from LCMV Arm 53b by three amino acid residues (Ahmed et al., 1984; Bergthaler et al., 2010; Salvato et al., 1991; Sullivan et al., 2011). Inoculation of LCMV Cl 13 (2×106 PFU i.v.) caused a systemic persistent virus infection that lasted for > 90 days (Ahmed et al., 1984).
Host immune factors exist to inhibit the dissemination of microbes, terminate infection, and reduce harm to infected tissues. However, host constituents like NIRs, which normally function to retard and suppress an exaggerated, tissue injurious, antiviral immune response are also utilized by viruses for their own strategies to establish and maintain persistence. In addition, disruption of vital innate signaling molecules such as interferon (IFN)-α/β receptor and the myeloid differentiation primary response gene 88 (MyD88) caused non-persistent strains of LCMV to avoid elimination and persevere (Jung et al., 2008; van den Broek et al., 1995). MyD88 is an adaptor molecule for toll-like receptors (TLR), a class of signaling molecules of the innate immune system that recognize pathogen-associated molecular patterns derived from microbes (Moresco et al., 2011). These receptors form the first line of defense against pathogens. MyD88-dependent TLRs are TLR1, 2, 5, 6, 7, 8 and 9. Signaling through such receptors stimulates the production of inflammatory mediators, type I IFNs, and cytokines with potent anti-microbe activity (Moresco et al., 2011). LCMV stimulates type I IFN production by MyD88-dependent signaling pathways and reportedly involves TLR2, 7 and 9 (Borrow et al., 2010). The contribution of individual TLRs that require MyD88 for signaling to control and eliminate LCMV infection is unclear.
TLR7 was first identified as a receptor for single-stranded RNA from vesicular stomatitis and influenza viruses as well as the chemical ligands imiquimod and R-848 (Hemmi et al., 2002; Lund et al., 2004). TLR7 is primarily expressed by plasmacytoid DCs, but also emanates from other DC subsets and myelomonocytic cells, T cells and B cells (Hammond et al., 2010; Hemmi et al., 2002; Hornung et al., 2002; Kadowaki et al., 2001). Signaling through TLR7 results in translocation of IFN regulatory factor 7 (IRF7) and nuclear factor-kappaB (NF-κB) to the nucleus where IRF7 stimulates expression of type I IFN as well as IFN inducible genes, and NF-κB elicits production of inflammatory cytokines. TLR7 stimulates inflammatory responses in DCs and macrophages, enhances cytolytic activity in CD8+ T cells and augments B cell class switching (Ambach et al., 2004; Edwards et al., 2003; Heer et al., 2007; Hemmi et al., 2002). Mice deficient in TLR7 are more susceptible to murine cytomegalovirus, West Nile virus (WNV), influenza virus and Friend virus infections (Browne, 2011; Koyama et al., 2007; Town et al., 2009; Zucchini et al., 2008).
Given the importance of TLR7 in recognition of viruses and stimulation of multiple arms of the immune system, we investigated the role of TLR7 during LCMV Cl 13 infection of mice, the prototypic model for analyzing persistent virus infection. Here we demonstrate that TLR7 sufficiency is essential for spontaneous termination of LCMV Cl 13 infection; in contrast TLR7 deficiency resulted in life-long persistence of this virus. The mechanism for TLR7 deficiency leading to lifelong persistent viral infection is dysfunction in T and B cell anti-viral immune responses as well as exacerbated T cell exhaustion. Introduction of TLR7+/+ LCMV immune memory, despite enhancing viral clearance in TLR7+/+ mice, failed to clear LCMV Cl 13 infection in TLR7−/− mice.
TLR7+/+ and TLR7−/− mice infected with LCMV Arm developed an acute infection that was eliminated from the serum (data not shown), liver (Fig. 1A), brain (Fig. S1A) and kidney (Fig. S1B) by day 9 post-infection (p.i.). To the contrary, TLR7+/+ and TLR7−/− inoculated with LCMV Cl 13 acquired a persistent viral infection that was detectable in the liver (Fig. 1A) and serum (Fig. 1B) past days 70 and 35, respectively. Although most organs of TLR7+/+ mice were negative for infectious LCMV Cl 13 progeny according to plaque assays on day 100 p.i., in contrast, LCMV Cl 13 persisted in the brain and kidneys for up to 200 days and 2–3 years, respectively (Oldstone, 2002). TLR7+/+ mice infected with LCMV Cl 13 harbored replicating virus in the brain (Fig. S1A) and kidney (Fig. S1B) at day 131 p.i.; however the viral levels in TLR7+/+ declined between days 35 and 131 p.i. demonstrating that the hosts’ immune response was clearing virus from these organs. TLR7−/− mice infected with LCMV Cl 13 retained high levels of virus in their sera at all time-points assayed (Fig. 1B) and these levels remained elevated throughout their lives (>day 320 p.i.). The liver (Fig. 1A), brain (Fig. S1A) and kidney (Fig. S1B) of LCMV Cl 13-infected TLR7−/− mice maintained >106 PFU/g tissue throughout the course of infection. This viral burden was significantly greater than that of virus levels in all tissues from TLR7+/+ mice at days 70 and 131 p.i. Lifelong virus persistence observed in TLR7−/− mice was not caused by disrupted early control of LCMV Cl 13 replication given that virus titer in the spleen, liver, lung and kidney on day 3 p.i. were not significantly different when evaluated against TLR7+/+ mice (Fig. S1C). Even though tissue virus burden was unaltered, reductions in IFN-α (Fig. S1D), IFN-β (Fig. S1E), CCL2 and CXCL10 (Fig. S1F) were detected in LCMV Cl 13-infected TLR7−/− mice. However, the levels of these cytokines became equivalent to TLR7+/+ mice by day 2 p.i. (Fig. S1G). Although, the levels of some cytokines were reduced early in LCMV Cl 13-infected TLR7−/− mice, they did not adversely affect early control of LCMV replication. These results demonstrate a biased requirement for TLR7 in controlling chronic LCMV Cl 13 infection.
Since T cells are essential for LCMV clearance, and functional exhaustion of T cells correlates with viral persistence (Berger et al., 2000; Zajac et al., 1998), we next investigated these factors in TLR7+/+ and TLR7−/− mice. The spleen of TLR7−/− mice had significantly more LCMV-specific CD8+ T cells (GP33–41, H-2Db, Fig. 2A) and CD4+ T cells (GP66–77, I-Ab, Fig. 2B) than those of TLR7+/+ mice on days 7 and 15 p.i. with LCMV Cl 13. By day 35 p.i. and thereafter, both strains had equivalent numbers of LCMV-specific CD8+ (Fig. 2A) and CD4+ (Fig. 2B) T cells in the spleen. Elevated numbers of LCMV-specific CD8+ T cells within the spleens of TLR7−/− mice was not caused by a failure to exit lymphoid tissue and/or enter infected peripheral tissue since the numbers of such cells within the spleen (Fig. S2A), liver (Fig. S2B) and brain (Fig. S2C) of TLR7−/− mice were at levels either equal to or greater than the content found in TLR7+/+ mice. Thus, TLR7−/− mice displayed normal T cell homing to infected tissues.
Since the TLR7−/− host response against LCMV Cl 13 resulted in enhanced T cell expansion, but failure in viral clearance, life-long persistence may be caused by disrupted antiviral T cell function. To evaluate this point, we characterized T cell antiviral poly-functionality by analyzing cytokine and granzyme B production by LCMV-specific T cells. To accomplish this, splenocytes from TLR7+/+ and TLR7−/− mice were cultured ex vivo with peptides for the CD8+ and CD4+ T cell immunodominant LCMV epitopes GP33–41 and GP61–80, respectively. GP33–41 peptide stimulation of CD8+ T cells from TLR7−/− mice infected with LCMV Arm and Cl 13 had decreased frequencies of TNF-alpha and IL-2 expressing cells in IFN-γ+ gated cells compared to identically stimulated CD8+ T cells from TLR7+/+ mice (Fig. 2C, left panel). A significantly greater percentage of IFN-γ-only producing and a significant reduction in frequencies of multifunctional cytokine producing CD8+ T cells, i.e., fewer IFN-γ, TNF-α dual producers (LCMV Arm only) and IFN-γ, TNF-α, IL-2 triple producers, were detected in TLR7−/− mice when compared to TLR7+/+ mice on day 9 p.i. (Fig. 2C, right panel). Ex vivo analysis of TLR7−/− mice found significantly diminished frequencies of granzyme B+ GP33–41 tetramer-positive CD8+ T cells on days 15 and 35 p.i. when judged against TLR7+/+ mice (Fig. 2D). Granzyme B LCMV-specific CD8+ T cell frequencies in TLR7+/+ mice became equivalent to and significantly less than in TLR7−/− mice on days 70 and 131 p.i., time points at which viremia was spontaneously terminated in TLR7+/+ mice. Although elevated numbers of virus-specific T cells were observed in TLR7−/− mice, these cells produced lesser amounts of cytokines and granzyme B, thereby exhibiting diminished antiviral potential.
Next we determined if TLR7 deletion within T cells or extrinsic factors caused by a TLR7-deficient environment enhanced T cell proliferation and reduced antiviral cytokine production. For that purpose, TLR7+/+ T cell receptor (TCR) transgenic (Tg) CD8+ (P14, LCMV GP33–41-specific, H-2Db) and CD4+ (SMARTA, LCMV GP66–77-specific, I-Ab) T cells were adoptively transferred into TLR7+/+ and TLR7−/− mice 1 day prior to infection with LCMV Arm or Cl 13. These TLR7+/+ TCR Tg CD8+ and CD4+ T cells expressed the congenic markers Thy1.1 and Ly5.1, respectively, which assured the cells’ identification when adoptively transferred into Thy1.2+ Ly5.2+ TLR7+/+ and TLR7−/− mice. A significantly greater number of TCR Tg Thy1.1+ CD8+ T cells were retrieved from the spleens of TLR7−/− mice infected with either LCMV Arm or Cl 13 when compared to TLR7+/+ mice on day 9 p.i. (Fig. 2E). Additionally, GP61–80 (Fig. 2F) and GP33–41 (Fig. 2G) peptide stimulation of total splenocytes significantly increased the frequencies of IFN-γ-only producing, as well as significantly reduced frequencies of multiple antiviral cytokine producing endogenous and exogenously administered TLR7+/+ TCR Tg T cells from TLR7−/− mice when compared to TLR7+/+ mice 8 days after LCMV Cl 13 infection. Therefore, the enhanced expansion and diminished poly-functionality of LCMV-specific T cells observed within TLR7−/− mice were caused by environmental factors, not an intrinsic property of T cells.
The cause of defective antiviral T cell responses may be inherent to TLR7 deficiency in DCs, cells that preferentially express and respond to TLR7 signaling. To evaluate this possibility, DC subsets were analyzed for expression of functional markers, virus antigen and the ability to induce proliferation at various time points following infection with LCMV Cl 13. Increased and decreased expression of markers analyzed were noted in plasmacytoid DCs (pDCs; Fig. S2D), myeloid-derived DCs (mDCs; Fig. S2E) and CD8α+ DCs (Fig. S2F). In addition, significantly greater frequencies of infected CD8α+ DCs (Fig. S2G) and mDCs (Fig. S2H) were detected in TLR7−/− mice at several different time points. DCs from the spleens of TLR7+/+ and TLR7−/− mice on day 35 p.i. induced proliferation of TLR7+/+ TCR Tg CD8+ (P14) and CD4+ (SMARTA) T cells equivalently (Fig. S2I). In addition, DCs from TLR7−/− mice on day 317 p.i. induced proliferation of TLR7+/+ TCR Tg CD8+ T cells (P14) with or without the addition of LCMV GP33–41 peptide (Fig. S2J). In vivo instillation of TLR7+/+ TCR Tg CD8+ T cells (P14) into TLR7−/− mice on day 85 p.i. resulted in CD8+ T cell proliferation (Fig. S2K). This proliferation was compared to TLR7+/+ mice that had cleared LCMV Cl 13 systemic infection, but received 1×105 PFU LCMV Arm 1 day following TCR Tg CD8+ T cell (P14) transfer to stimulate cell proliferation; however, the proliferative response was poor and well below the expansion observed in TLR7−/− mice (Fig. S2K). Despite elevated infection and altered expression of activation markers on TLR7−/− DCs, DCs retained the ability to stimulate T cell proliferation in vitro and in vivo.
Reduced anti-viral functionality of CD8+ T cells from TLR7−/− mice may have resulted from the cells’ exacerbated and/or extended functional exhaustion. To address this issue, markers of functional exhaustion were evaluated in TLR7+/+ and TLR7−/− mice infected with LCMV Cl 13. Accordingly, we determined that programmed death-1 (PD-1) on LCMV-specific CD8+ T cells in TLR7+/+ mice peaked on day 15 p.i., then dissipated over the course of infection (Fig. 3A). Compared to TLR7+/+ mice, TLR7−/− mice expressed significantly greater levels of PD-1 on LCMV-specific CD8+ T cells on days 35, 70 and 131 p.i. (Fig. 3A). Additional analysis of several NIR molecules on LCMV-specific CD8+ T cells from TLR7−/− mice 225 days after LCMV Cl 13 infection revealed high levels of PD-1, LAG-3, 2B4 and CD160 (Fig. 3B). This expression was greater than that observed in a TLR7+/+ mouse at the same time-point when their systemic LCMV Cl 13 was purged. Clearly, TLR7−/− mice infected with LCMV Cl 13 lacked antiviral functionality and suffered from immune system exhaustion, both of which indicate a severely defective T cell compartment.
The TLR7 deficiency markedly impeded humoral immunity when evaluated during influenza and Epstein-Barr virus infections (Heer et al., 2007; Iskra et al., 2010). We therefore analyzed possible effects on the B cell response during chronic LCMV Cl 13 infection. The generation of CD138+ plasma (Fig. 4A) and IgM secreting B cells (Fig. 4B) in spleens of TLR7−/− mice was significantly reduced compared to that of TLR7+/+ mice on day 20 p.i., the time point that germinal center B cells were detected in TLR7+/+ mice. In addition, TLR7−/− mice had a significantly decreased frequency of germinal center B cells on days 20 and 35 p.i. (Fig. 4C). The absence of LCMV Cl 13 in TLR7+/+ mice yielded waning frequencies of germinal center B cells on days 70 and 165 p.i., whereas, germinal center formation increased in TLR7−/− mice on day 165 p.i. In addition, ELISpot analysis revealed significantly lower numbers of IgG secreting LCMV-specific B cells in TLR7−/− mice compared to the content in TLR7+/+ mice on day 33 p.i. (Fig. 4D). Diminished germinal center B cell formation, a reduction of plasma cells and fewer B cells secreting LCMV-specific IgG correlated directly with a significant decrease in LCMV-specific antibodies on days 35 (Fig. 4E) and 74 (Fig. 4F) p.i., demonstrating severely impeded B cell antibody production in LCMV Cl 13 infected TLR7−/− mice. Peptide-pulsed B cells from TLR7−/− mice on day 35 p.i. induced TCR tg CD4+ (Fig. S3A) and CD8+ (Fig. S3B) T cell proliferation equivalently to that observed in TLR7+/+ mice. Therefore, TLR7−/− B cells maintained the ability to stimulate T cell proliferation, but are defective in anti-LCMV antibody production.
To determine whether this defect in the humoral immune response in TLR7−/− mice was caused by TLR7 deletion within B cells, bone marrow chimeras were generated. Lethally irradiated mice were reconstituted with a 1:1 mixture of cells from Ly5.1+ TLR7+/+ mice and Ly5.2+ TLR7−/− mice and infected with LCMV Cl 13 2 months after reconstitution. The frequency of TLR7−/− splenic germinal center B cells was reduced significantly below that of Ly5.1+ TLR7+/+ cells on day 30 after infection, effectively demonstrating that this defect in germinal center B cell formation stemmed from their intrinsic TLR7 deficiency (Fig. 4G). Additionally, far fewer Ly5.2+ TLR7−/− B cells produced LCMV-specific IgG antibody compared to Ly5.1+ TLR7+/+ B cells harvested from the same mice, and resembled that observed in TLR7−/− mice that were neither irradiated nor reconstituted (Fig. 4H). Therefore, B cells required TLR7 signaling for germinal center reactions, conversion to plasma cells and production of LCMV-specific IgG antibody during persistent LCMV Cl 13 infection.
TLR7−/− mice have negligible LCMV-specific antibody response, diminished anti-viral T cell function, and extended as well as heightened T cell exhaustion following LCMV Cl 13 infection. We reconstituted the TLR7−/− mice with TLR7+/+ LCMV immune memory cells as a means of augmenting their ability to purge persistent LCMV Cl 13 infection. Because TLR7−/− mice displayed an intrinsic B cell defect, immune memory splenic B cells from TLR7+/+ mice were administered to TLR7+/+ or TLR7−/− mice on day 15 p.i. with LCMV Cl 13. Subsequently, TLR7+/+ mice that received TLR7+/+ immune memory B cells terminated infection earlier than TLR7+/+ mice that did not receive B cells demonstrating that these cells enhanced LCMV Cl 13 clearance (Fig 5A). In contrast, TLR7−/− mice that received TLR7+/+ immune memory B cells sustained viremia for the duration of the experiment that was significantly greater (p < 0.0009, day 42 post-transfer) than TLR7+/+ mice that did not receive immune memory B cells. Since other cell types may be required for the elimination of LCMV Cl 13 infection, the experiment was repeated with adoptive transfer of TLR7+/+ or TLR7−/− total immune memory splenocytes into TLR7+/+ or TLR7−/− mice 15 days after infection with LCMV Cl 13. Viremia was terminated in TLR7+/+ mice that received TLR7+/+ immune memory cells by day 30 post-transfer (solid black circles, Fig. 5B). In contrast, TLR7+/+ mice that received TLR7−/− immune memory (shaded circle in Fig. 5B) cleared LCMV Cl 13 by 68 days after transfer, therefore, these mice did not reduce viral burden as efficiently as TLR7+/+ recipients of TLR7+/+ LCMV immune memory cells (Fig. 5B). Consequently, TLR7−/− immune memory cells cleared virus in vivo, but are less efficient in reducing viral load than immune memory cells generated in TLR7+/+ mice (Fig. 5B), correlating with the deficient functional T cell capacity of TLR7−/− mice infected with LCMV Arm (Fig. 2C). In contrast, unmanipulated TLR7−/− mice (solid black diamonds) and TLR7−/− mice that received total TLR7+/+ or TLR7−/− immune memory splenocytes (designated by and , respectively) maintained viremia and showed no evidence of clearing LCMV Cl 13 infection (Fig. 5B). By day 73 post-transfer, no virus was detectable in the serum, spleen or liver of TLR7+/+ mice that received either TLR7+/+ or TLR7−/− immune memory cells. In the brain and kidney of these mice, virus was still detectable, but at lower levels than TLR7−/− mice persistently infected with LCMV Cl 13 from the same time point (Fig. 5C). In contrast persistently infected TLR7−/− mice that received TLR7+/+ or TLR7−/− immune memory cells sustained high levels of virus in their sera and all tissues indicating that the adoptively transferred immune memory cells failed to purge the infection (Fig. 5C). Adoptively transferred Ly5.2+ TLR7−/− cells were clearly evident in the sera of Ly5.1+ TLR7+/+ recipients on day 45 post-transfer; indicating that these cells had not been eliminated and failure of clearance was not due to histoincompatability (Fig. S4). Although the adoptively transferred LCMV immune memory cells enhanced viral clearance in TLR7+/+ mice, these same cells failed to eliminate viremia in TLR7−/− mice. Consequently, our overall results highlight the importance of TLR7 signaling in eliminating chronic viral infection. Moreover, the numerous immune defects in TLR7−/− mice, both cell intrinsic and environmental, prevented viral clearance even after adoptive immunotherapy.
The signaling receptor TLR7 is critical for the efficient control of acute and persistent virus infections. Here we describe defects of the antiviral immune response associated with the absence of TLR7 during a persistent virus infection. We documented that deficient TLR7 signaling contributes to significant defects in both T and B cells, but not early control of virus replication or DCs, demonstrating a dysfunctional adaptive immune response. Our investigation now expands the known role of TLR7 as an essential modulator of T and B cell responses that are required to clear persistent virus infection.
TLR7−/− mice exhibited multiple aberrations within the T cell compartment. Soon after becoming infected with LCMV Cl 13, these mice made significantly greater LCMV-specific T cells, yet the infection does not resolve. We initially conjectured that this phenomenon was the failure of T cells to migrate appropriately to infected tissues as described in TLR7−/− mice infected with WNV (Town et al., 2009). LCMV-specific CD8+ T cells migrated into the brain and liver similarly in both TLR7−/− and TLR7+/+ mice (Fig. S2A–C). Our data support the likelihood that increased T cell proliferation is dependent on the TLR7−/− inflammatory environment, since adoptively transferring TLR7+/+ TCR Tg CD8+ T cells into TLR7−/− mice resulted in significantly enhanced expansion of the T cell population over that in TLR7+/+ recipients. Although we noted modulation of DC markers required for T cell expansion (Fig. S2D–F), those would not account for the great disparity in T cell numbers between TLR7+/+ and TLR7−/− mice early following LCMV infection. In addition, when LCMV-specific T cells were co-cultured with DCs sorted from TLR7+/+ and TLR7−/− mice (Fig. S2I and J), the proliferation profiles of both groups were similar. Moreover, the increased frequencies of infected DCs in TLR7−/− mice (Fig. S2G and H) could heighten the concentration of MHC-peptide complexes resulting in augmented T cell proliferation. Further, a secreted factor rather than interactions with antigen presenting cells could promote this T cell expansion.
Although LCMV-specific T cell numbers increased, defects in TLR7−/− T cell antiviral cytokine responses were detected following infection with both LCMV Arm and Cl 13 (Fig. 2C). Analysis of early cytokine content in the serum revealed significantly reduced type I IFN within TLR7−/− mice on day 1 following infection (Fig. S1D and E). Our results agree with two studies in this issue where an early, transient decrease in serum content of type I IFN within LCMV-Cl 13-infected TLR7−/− mice was observed when compared to TLR7+/+ mice (Macal et al., 2012; Wang et al., 2012). Reduced type I IFN correlated with diminished activation of T cells 1 day following infection (Macal et al., 2012). Several groups have shown that abrogation of type I IFN signaling within CD8+ T cells inhibited their clonal expansion and differentiation during acute LCMV infection (Kolumam et al., 2005; Wiesel et al., 2012). Reduced, but not abrogated, type I IFN may contribute to diminished CD8+ T cell anti-viral potential observed within TLR7−/− mice. Others have shown that ex vivo TLR7 stimulation of TLR7+ CD8+ T cells from HIV-infected patients induced IFN-γ production, but required accessory cells (Song et al., 2009), suggesting that TLR7 stimulation of accessory cells likely enhanced CD8+ T cell cytokine production. Nonetheless, TLR7 signaling did facilitate IFN-γ production by CD8+ T cells. We also documented that TLR7+/+ TCR Tg T cells adoptively transferred into TLR7−/− mice 1 day prior to infection had a diminished capacity to produce multiple antiviral cytokines when compared to similarly treated TLR7+/+ mice (Fig. 2F and G). In addition, the defect in antiviral cytokine production observed in LCMV-specific TCR Tg T cells within TLR7−/− mice was equivalent to that of endogenous TLR7−/− T cells. Therefore, enhanced expansion and the inhibited generation of poly-functional LCMV-specific T cells in TLR7−/− mice was due to environmental factors and was not an intrinsic property of the T cells.
CD8+ T cells are the only cells of the adaptive immune response necessary to purge an acute LCMV Arm infection (Matloubian et al., 1994; Tishon et al., 1995). CD8+ T cells from LCMV Arm-infected TLR7−/− mice were poor producers of anti-viral cytokines, however, the increased numbers of LCMV-specific CD8+ T cells in these mice likely overcame the cytokine defect and contributed virus eradication by day 9 p.i. That defect in T cell antiviral cytokine production was noted only before the establishment of T cell exhaustion caused by LCMV Cl 13 infection. Greater than 9 days after infection, diminished anti-viral activity caused by T cell exhaustion resulted in equivalent cytokine production by TLR7+/+ and TLR7−/− mice (data not shown). However, CD8+ T cells from LCMV Cl 13-infected TLR7−/− mice after day 9 post-infection had reduced granzyme B content, a molecule involved cytotoxic lymphocyte activity, demonstrating reduced antiviral potential (Fig. 2D). These findings correlate with a previous study in which production of perforin, a molecule essential for CD8+ T cell lytic activity, was enhanced in human peripheral blood mononuclear cells (PBMCs) exposed to a TLR7 agonist (Ambach et al., 2004); however it is uncertain whether TLR7 stimulation of APCs present in the isolated PBMCs promoted CD8+ T cell expression of perforin. The cause of the diminished antiviral potential observed in T cells from TLR7−/− mice may be caused by the antibody deficient environment, although, as yet, there is no definitive evidence to support this conclusion.
TLR7−/− mice exhibited continuously heightened PD-1 expression, a NIR molecule necessary for LCMV Cl 13 persistence (Barber et al., 2006), on LCMV-specific CD8+ T cells; this expression exceeded that of TLR7+/+ mice starting at day 35 p.i. (Fig 3A). In addition, on day 225 after infection, TLR7−/− mice expressed high levels of multiple NIRs including PD-1, LAG-3, 2B4 and CD160 (Fig. 3B). Although, this expression of NIRs cannot be compared to that of TLR7+/+ mice at that same time point, since the latter mice had previously cleared LCMV Cl 13, the amount of NIR expression was comparable to CD4−/− mice with lifelong LCMV Cl 13 infection (Blackburn et al., 2009). Reduced T cell antiviral potential and increased exhaustion of immune responsiveness ultimately contributed to LCMV Cl 13 persistence in TLR7−/− mice.
Elsewhere, mice deficient in B cells failed to clear LCMV Cl 13 infection; however those mice had dysfunctional CD4+ T cells, which are essential for CD8+ T cell-mediated clearance of LCMV Cl 13 (Bergthaler et al., 2009; Homann et al., 1998; Matloubian et al., 1994). Adoptive transfer of immunoglobulins, even when concentrated 10-fold over levels found in sera of mice that spontaneously cleared LCMV Cl 13 infection, failed to purge virus from LCMV Cl 13 persistently infected mice (unpublished observation). Similarly, transfer of B cells from immune mice or immune B cells and CD4+ T cells into LCMV Cl 13 persistently infected mice failed to abort that infection (unpublished observations), although, similar transfer of immune B and CD4+ T cells into mice persistently infected with measles virus ably cleared the infection (Tishon et al., 2006). Other studies utilizing IL-21−/− and IL-6−/− mice have demonstrated that antiviral antibody responses are associated with LCMV Cl 13 clearance, although neutralizing antibodies from these mice have either not been identified or found at serum neutralizing titers of less than a 1 to 5 dilution. (Elsaesser et al., 2009; Harker et al., 2011). Nonetheless, transfer of TLR7+/+ immune memory B cells into TLR7+/+ mice enhanced clearance of LCMV Cl 13 (Fig. 5A) suggesting that these cells contribute to virus elimination, although the exact mechanism of how this occurs is unclear. Although, as documented in this report, TLR7 sufficiency was required for the generation of anti-LCMV IgG secreting B cells (Fig. 4), but did not alter the ability of B cells to stimulate T cell proliferation (Fig. S3). Thus plausible, that TLR7 signaling within B cells is required for regulation of BCL-6, Blimp-1 and AID, genes required for the maturation of the antibody response (McHeyzer-Williams et al., 2012).
Mice infected at birth with LCMV develop a life-long persistent virus infection (Oldstone, 2002). Introduction of LCMV immune memory from TLR7+/+ mice into mice infected at birth with LCMV induced viral clearance that required TLR7+/+ CD4+ T, TLR7+/+ CD8+ T cells and TLR7+/+ B cells (Berger et al., 2000; Planz et al., 1997). However, transferring TLR7+/+ immune memory cells into TLR7−/− mice failed to remove LCMV Cl 13 virus (Fig. 5B), although TLR7−/− DCs effectively primed and expanded such cell populations (Fig. S2I and J). Presumably, then, an undefined cell type or non-hematopoietic cells may also play a role in viral clearance. In addition, it is conceivable that TLR7 signaling within LCMV Cl 13-infected cells is required for the elimination of those cells by immune memory. Nonetheless, immune memory cells enhanced LCMV Cl 13 clearance from TLR7+/+ mice, but lacked the requirements to purge LCMV Cl 13 from TLR7−/− mice.
TLR7 is essential for the development of an adaptive immune response sufficient to eliminate persistent LCMV Cl 13 infection. Interestingly, polymorphisms within the TLR7 gene that diminished IFN-α production following stimulation correlated directly with the accelerated progression of HIV (Oh et al., 2009). In addition, hepatitis C virus ssRNA stimulated TLR7 signaling, and administration of a synthetic TLR7 agonist, isatoribine, reduced the viral burden in sera of infected patients (Horsmans et al., 2005; Zhang et al., 2009) again relating TLR7 signaling to the clearance of persistent virus infection in mouse models and in humans. Understanding how TLR7 signaling contributes to the generation of sufficient antiviral T and B cell responses to clear acute virus infection and which molecules downstream of TLR7 terminate persistent infection remains essential. Dissecting the TLR7 pathway by using chemically tractable molecules that can restore TLR7 signaling provides potential for the effective treatment of persistent viral infections of humans.
C57Bl/6 (TLR7+/+), C57Bl/6 Ly5.1+ (B6.SJL-Ptprc(a)Pepc(b)/BoyJ), TLR7−/−, C57Bl/6 Thy1.1+ DbGP33–41 TCR tg (P14) (Pircher et al., 1989) and C57Bl/6 Thy1.1+ I-AbGP66–77 TCR tg (SMARTA) (Oxenius et al., 1998) male and female mice (6–12 weeks of age) were used. TLR7−/− mice was generated on the 129 background and bred to C57Bl/6 mice for 10 generations. Histocompatability between TLR7−/− and C57Bl/6 mice was validated by lack of a proliferative response in a mixed lymphocyte reaction assay (data not shown) and by the survival of adoptively transferred TLR7−/− cells into TLR7+/+ C57Bl/6 mice 60 days following transfer (Fig. S4). Mice were maintained in pathogen-free conditions, and handling conforms to the requirements of the National Institutes of Health and The Scripps Research Institute animal research committee. LCMV Cl 13 and Arm strains were grown, stored and quantified according to published methods (Borrow et al., 1995). Serum from blood drawn from the retro-orbital sinus, liver, lung, spleen, brains and kidneys were used to perform plaque assays as previously described (Ahmed et al., 1984).
Cells were obtained by mechanical disruption of spleens, liver and brain through a 100-μm filter. Red blood cells were lysed by exposing disrupted tissue to a selective red blood cell lysis buffer (Ammonium Chloride: Tris HCl 0.02M, NH4Cl 0.14M). Cells per tissue were quantified using a hemocytometer and the trypan blue exclusion method. For purification of splenic B and TCR Tg T cells, cell enrichment kits were used (STEMCELL Technologies Inc, Vancouver, Canada). Cells were isolated to a purity of > 95%. For TCR tg T cells, 1 × 104 purified cells were adoptively transferred i.v. into recipients one day prior to infection. 2×107 immune memory B cells or 3×107 total immune memory splenocytes from TLR7+/+ or TLR7−/− mice that were infected 45–50 days earlier with 2×105 PFU of LCMV Arm and restimulated with the same dose of LCMV Arm 3 days prior to isolation were transferred i.p. into recipients15–20 days p.i. with LCMV Cl 13.
Isolated cells were stained with antibodies raised against murine antigens. Antibodies were added to cells at dilutions of 1:100-1-400. For tetramer staining, cells were incubated for 1 hr on ice with MHC class I tetramers for DbGP33–41-APC and MHC II tetramer with I-AbGP66–77-APC for 1 hr at rt. Biotinylated monomers were obtained from the NIH Tetramer Core Facility and tetramerized using streptavidin-APC (Invitrogen, Carlsbad, CA, USA) according to the NIH Tetramer Core Facility protocol.
Splenocytes were stimulated for 5 h with 2 μg/ml of the MHC class I restricted LCMV-GP33–41 or the MHC class II restricted LCMV-GP61–80 peptide (>99% pure; American Peptide and The Scripps Research Institute, Center for Protein Science, La Jolla, CA, USA) in the presence of 4 μg/ml brefeldin A (Sigma, St Louis, MO, USA). Cells were fixed, permeabilized with 2% saponin, and stained intracellulary with antibodies to IFN-γ (XMG1.2), TNF-α (MP6-XT22) and IL-2 (JES6-5H4). Unstimulated tetramer stained (GP33–41) splenocytes were stained intracellularly with a 1:50 dilution of antibody against human granzyme B that cross-reacts with mouse granzyme B (catalog #: MHGB04, Invitrogen, Carlsbad, CA, USA). B cells were stained intracellularly with an anti-mouse IgM antibody conjugated to Pacific Blue kindly provided by the David Nemazee laboratory. Absolute numbers of cells were determined by multiplying the frequency of specific cell populations by the total number of viable cells.
Bone marrow was extracted from femurs of Ly5.1+ TLR7+/+ and Ly5.2+ TLR7−/− mice, disrupted through a 100 micron mesh screen and red blood cells were lysed as described previously. 1×107 TLR7+/+ and TLR7−/− isolated bone marrow cells were adoptively transferred into lethally irradiated (1100 rads) Ly5.2+ TLR7+/+ mice. Bone marrow chimeric mice were maintained on water supplemented with antibiotics to prevent opportunistic infections for 2 months. Mice were then infected with 2×106 PFU LCMV Cl 13, euthanized on day 30 p.i., after which spleens were harvested to analyze TLR7+/+ (Ly5.1+) and TLR7−/− (Ly5.2+) B cell responses.
For determination of the number of LCMV-specific antibody secreting B cells, 5×106 PFU of LCMV Arm or 100 ul of PBS containing 10–15 ug of BHK cell lysates infected with LCMV was plated per well in a 96-well nitrocellulose bottom plate overnight (o/n) at 4 °C. Plates were blocked with 5% FBS for 1 h at room temperature (rt);1×106 sorted splenic B cells were plated per well and 3-fold serial dilutions were performed 8 times. The plate was incubated at 37 °C at 5% CO2 for 5 h. 100 ul of 2 ug/ml of biotinylated anti-mouse IgG was added to the wells and incubated o/n at 4°C. 100 ul of 5 ug/ml of horse radish peroxidase (HRP)-conjugated avidin was added to the wells for 1 h at rt followed by incubation with AEC chromogen substrate until red dots appeared. The reaction was stopped by rinsing the plate in tap water and dots were counted manually using a dissection microscope. For determination of anti-LCMV antibody titer, ELISA clear bottom plates were coated with antigen and blocked as described above. 10-fold serial dilutions of serum isolated from mice were incubated in the wells o/n at 4 °C. 100 ul of 2 ug/ml of biotinylated anti-mouse IgG was added to the wells and incubated o/n at 4°C. 100 ul of 5 ug/ml of HRP-conjugated avidin was added to the wells for 1 h at rt followed by 100 ul of substrate reagent (R & D Systems, Minneapolis, MN, USA) for 20 min. The reaction was stopped by addition of 2N H2SO4, and the plate was read at 450 nm.
ANOVA or an unpaired two-tailed Student’s t-test were calculated using Excel or GraphPad Prism 5, when appropriate, to determine significance, which was set at 5%.
This is Publication Number 21492 from the Department of Immunology and Microbial Science, The Scripps Research Institute (TSRI). This research was supported by NIH grant AI009484 (MBAO), and NIH training grants NS041219 (KW), AI007244 (KW) and AI007364 (JT). KT was supported by Scripps Health Graduate Medical Education Foundation.
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.