The main function of the innate immune system is not only to limit the early replication and spread of the invading pathogen, but also to initiate an adaptive response to clear the infection and establish long-lasting immunological memory. TLR7 is one of the major RNA sensing PRRs involved in the detection of IAV infection. Currently, it is understood that TLR7’s downstream signaling pathway leads to the activation of proinflammatory cytokines and IFNs; other RNA sensing PRRs, such as RIG-I, activate similar responses. Previously, it was shown that MyD88 signaling, not RIG-I signaling, plays a unique role in the CD4+ polarized response to IAV infection, but the mechanism of this was unknown. We demonstrate that the lack of TLR7 signaling significantly increases the accumulation of MDSCs at the site of infection. Because MDSCs have been previously shown to alter CD4+ T-cell polarization, we propose that MDSC recruitment is the mechanism by which TLR7 effects the resulting T-cell TH1/TH2 balance.
In a study by Seo et al.,
MyD88 was shown to be required for protection from IAV infection, as MyD88-/-
mice displayed increased morbidity and increased viral titers when infected with PR8 
mice were inhibited in their ability to recruit CD11b+
granulocytes, produce inflammatory cytokines, and Th1 cytokine production by CD4+ T cells following IAV infection when compared to B6 mice. A study by Koyama et al.
, conversely, showed no changes in viral titer when MyD88-/-
mice were infected with A/New Caledonia/20/99. They demonstrated that RIG-I and MyD88 were redundant in their ability to induce the early IFNα/β response in vivo
and in vitro 
. However, they found changes in the IgG istotypes in MyD88-/-
mice following IAV infection with increased IgG1 with a concomitant decrease in IgG2a on day D10 
. These studies demonstrated that MyD88 signaling is instrumental in the shaping of not only the innate but also the adaptive responses to IAV.
MyD88 is the downstream adaptor not only for many of the TLRs, but is also downstream of IL-1 receptor signaling 
. For this reason, it is difficult to ascertain if the phenotypes observed in MyD88-/-
mice are a result of dysfunctional IL-1 cytokine signaling or if they are due to inhibited viral recognition by a PRR. Like MyD88-/-
mice display an increase in morbidity and lung viral titers in response to IAV 
. Because of these similarities, it is possible that many of the effects observed in the MyD88-/-
model could be attributed to the inhibition of IL-1 signaling more so than the inhibition of TLR signaling. To better understand the effects of PRR specific recognition of IAV, independent of cytokine signaling inhibition, we utilized TLR7-/-
mice. We found that TLR7-/-
mice had increased morbidity, starting on day 7 p.i. (). In many viral infections, increased morbidity is caused by increased viral titers and/or cytokine storm 
. Previously, it was shown that absence of the inflammatory cytokines TNFα and IL-1 reduces the severity of H5N1 infection 
. In our study, viral titers did not increase in the absence of TLR7 (). Also, many of the inflammatory cytokines measured were slightly decreased rather than showing signs of a cytokine storm (). The only significant change in cytokines observed in the lungs was the decrease in IFNγ.
The relative decrease in inflammatory cytokines in the lungs did not have a detrimental effect the recruitment of myeloid cells (). This is contrary to what was observed in the MyD88-/-
models of IAV 
. Macrophage and neutrophil accumulation during IAV has been shown to play both a protective and pathogenic role depending on the magnitude of the cellular infiltrate. Depletion of neutrophils after a sub-lethal dose of IAV induces uncontrolled virus growth and lethality 
. In addition to hypercytokinemia, much of the pathology associated with highly pathogenic IAV infection is attributed to the recruitment of these myeloid cell types 
. Lin et al.
found that CCR2 dependent Gr1+
iMo were the largest population recruited to the lungs during PR8 infection, and were responsible for much of the pathology associated with IAV in mice, but not in the control of virus replication 
. In our study, we also show that the relatively large recruitment of Gr1+
cells in TLR7-/-
mice was likely the major cause of morbidity observed with little or no effect on lung viral titers.
cells are recruited to a site of inflammation in several models of inflammation 
. MDSCs have been described to be a Gr1+
immature myeloid cell population derived from monocytes migrating out of the blood to the site of infection 
. It has been previously shown that the presence of MDSCs skews the immune response towards a Th2 response 
. This is mainly attributed to their ability to induce IL-10 and reactive oxygen species 
Like other groups, the TLR7-/-
MDSCs not only produced IL-10, but also TNFα () 
. Interestingly, TNFα has been shown to amplify the ongoing Th1 or Th2 response rather than favoring one over the other 
. The ability of TLR7-/-
MDSCs to coproduce IL-10 and TNFα may be one mechanism that encourages the Th2 bias observed in TLR7-/-
mice during IAV infection. On day 10 p.i., the lungs of TLR7-/-
mice had increased numbers of IL-4 producing CD4+ T-cells present (). One month following infection with IAV, we found increased levels of IgG1 and decreased amounts of IgG2a in sera of TLR7-/-
mice (), suggesting that TLR7-/-
mice do in fact display a Th2 bias during IAV infection. We hypothesize that this is due to the increased recruitment of MDSCs to the site of infection.
Interestingly, it has been shown in different models that MyD88 is required for the expansion of MDSCs in vivo 
MDSCs, when compared to B6, not only show increased recruitment to the site of infection and secrete increased amounts of Th1 inhibiting IL-10 in vivo
, but also increase the amount of IL-4 production from OT-II T-cells (, ). It is possible that while MyD88 is required for MDSC expansion, through either IL-1 signaling or another TLR becoming activated, TLR7 stimulation during IAV infection may be involved in the suppression of MDSC activity. TLR7 signaling could accomplish this through an inhibitory feedback mechanism or by inducing MDSCs to differentiate into another monocytic cell type. It has been previously shown that activation of MDSCs through either TLR9 or TLR4 can further differentiate MDSCs into a myeloid cell that no longer has tumor suppressor activity or Th2 allergy inducing polarization 
. If this is true in our model, the absence of TLR7 signaling in MDSCs may be responsible for further aggravating the Th2 bias observed due to increased MDSCs cytokine secretion and recruitment.
In the sepsis model, Gr1+
MDSCs have been implicated for Th2 polarization observed 
. The accumulation of MDSCs was dependent on MyD88 expression, as Gr1+
cell recruitment to the spleen was inhibited during sepsis in MyD88-/-
cells in the spleens of wild type mice constitutively express TLR7, and TLR7’s expression is highly up regulated during sepsis. When mice were pre-treated with R-848 (a TLR7 ligand) before sepsis induction, the treated mice showed an increase in their ability to control bacterial load at the site of infection. This increase in pathogen control and inflammation demonstrate that TLR7 signaling will overcome the inhibitory phenotype predominant during sepsis 
During IAV infection, RNA released during cell lysis serve as ligands to activate TLR7+
myeloid cells, including macrophages, neutrophils, DCs, and MDSCs. Resident alveolar macrophages are among the first responders to infection 
. Neutrophils and macrophages are recruited on days 3 and 5 p.i. (). On day 5 p.i., MDSCs are recruited to the site of inflammation (). They then can become activated by the presence of inflammatory mediators like IL-6 and IL-1 
. In mice, virus replication is present through day 7, and is cleared by day 10 (), indicating that TLR7 ligands would be present at least through day 10. Our findings suggest that the MDSCs recruited to the lungs starting on day 5 would recognize the presence of RNA through TLR7 and become hindered in their capacity to inhibit the ongoing inflammatory response similar to TLR9 ligand mediated inhibition of MDSC tumor suppression 
(). It is possible that the sensing of RNA by TLR7 actually leads to further differentiation of MDSCs into a more classical Th1 inducing macrophage. MDSCs are known to have plasticity in their ability to further differentiate into a different type of myeloid derived cell, as has been previously shown with activation from TLR4 activation 
TLR7 inhibition of MDSC mediated Th1 suppression.
It would not be beneficial for the host to have numerous suppressor cells present before the acute infection is cleared. After day 7 p.i., when RNA is no longer in excess, MDSC activity would no longer be inhibited and would then be allowed to suppress the remaining inflammatory response. In a situation where TLR7 is not present, MDSCs would lose their ability to sense the presence of RNA during an ongoing infection. When the MDSCs are recruited on day 5, they would not suspend their suppressive activity in the presence of inflammation 
. This would lead to accumulation of MDSCs and the inhibition of a Th1 polarized immune response, further aggravating inflammation in an unbalanced manner ().
Together, our findings suggest that TLR7 signaling inhibits the suppressive activity of MDSCs during IAV infection in mice, resulting in a predominant Th1 response. It has yet to be determined what the long term effects of a Th2 polarized response observed in these TLR7-/-
mice are on the memory response. It would be interesting to see if this Th2 polarization increases or decreases the recall response to subsequent infections, and what the consequences of this would be on vaccine design. Because of TLR7 ability to promote a Th1 biased response, it is possible that the inclusion of TLR7 ligands may enhance the immunogenicity of influenza vaccines