The ability of HRSV, HMPV and HPIV3 to re-infect symptomatically throughout life without the need for significant antigenic change has led to the widely held speculation that these viruses, especially HRSV, can suppress or subvert the host adaptive immune response, resulting in incomplete and inefficient long-term immunity. A number of studies have addressed virus-specific effects on APC and T lymphocyte responses in vitro
, with varied and inconsistent conclusions. The first such studies reported that exposure of adult human peripheral blood mononuclear cells (PBMC) to HRSV, IAV, and Sendai virus suppressed proliferation in response to the non-specific mitogen phytohemagglutinin (PHA), an effect that was attributed to the expression of CD54/CD11a/CD18 (ICAM-1/LFA-1) and the interaction between APC and T cells 
. In 1992, Preston et al. 
showed that exposure of human cord blood mononuclear cells to HRSV resulted in a reduction in proliferation in response to PHA. The same study showed that exposure of adult PBMC to HRSV reduced the proliferation response to Epstein-Barr virus antigen, although this effect was not seen with all of the tested HRSV strains. This effect was attributed to secreted IFN-α 
. In another study, HPIV3 was shown to reduce proliferation of adult human PBMC in response to CD3-specific antibodies, an effect that was attributed to increased production of IL-10 
More recent studies have used increasingly more defined conditions. Bartz et al.
generated human immature DC in vitro
from cord blood CD34+ stem cells and showed that in presence of the toxic shock syndrome toxin-1 superantigen, HRSV caused increased DC apoptosis and reduced expression of IFN-γ and increased expression of IL-4 without affecting proliferation. Rothoeft et al. 
reported reduced T cell proliferation in response to HRSV stimulated CD34-derived DC, as well as reduced IFN-γ expression. De Graaf et al. 
showed that exposure of adult human MDDC to HRSV reduced SEB-mediated proliferation and cytokine production by naïve allogeneic CD4+ T cells, which was attributed to a soluble MDDC factor that was not identified but was not type I IFN. Chi et al. 
also showed that exposure of adult MDDC to HRSV resulted in reduced proliferation of autologous CD4+ T cells in response to SEB or cytomegalovirus antigen, whereas HPIV3 and IAV controls resulted in less and no inhibition, respectively. The inhibitory effect of HRSV was partially mediated by type I and type III IFN 
. Using human adult MDDC and allogeneic CD4+ T lymphocytes, Guerrero-Plata et al. 
showed that HRSV inhibited the proliferation of allogeneic CD4+ T lymphocytes to a greater extent than HMPV. However, Jones et al. 
did not observe HRSV-mediated inhibition of T cell proliferation under comparable conditions. Schlender et al. 
showed that exposure to HRSV F protein expressed on the surface of epithelial cells inhibited proliferation of T lymphocytes in response to mitogen, but several of the co-culture studies described above provided evidence that ruled out contact inhibition in the observed inhibitory effects 
. Thus, from these previous studies, there was inconsistency in the effects of HRSV on DC and on T cells and, in cases where inhibition or polarization was observed, there were differences in the proposed mechanism.
To address the question of virus-specific suppression of CD4 T cell function, we examined four viruses (HRSV, HMPV, HPIV3, and IAV) side-by-side, whereas the majority of the studies noted above examined a single virus, usually HRSV. While logistically more difficult, comparing a greater number of viruses provided for discrimination between effects that were unique to a particular virus versus those that were common to all. Also, analyzing more viruses and thus obtaining more comparisons for each individual donor was useful, given the heterogeneity of responses from an outbred human population. The “down-side” of our approach is that these studies were time consuming and laborious, and the donor-to-donor variability in human populations, and the relatively large panel of viruses studied, would have necessitated a high number of studies to reach statistical significance for the more nuanced differences between viruses, and forced us to interpret trends in differences.
We also used a more careful method of preparing virus. Viruses were grown in Vero cells, which do not produce type I IFN, and purified by sedimentation in sucrose gradients. We avoided the use of high input MOI of virus, especially with HRSV, which was used at an MOI of 10-20 or more in some studies 
, because HRSV is physically unstable, typically is contaminated with co-purified cellular membrane fragments, and has a high particle-to-PFU ratio 
. Thus, the use of a high MOI could result in a large, disproportionate dose of viral antigen and cell contaminants. In contrast to a number of studies that used allogeneic (unmatched) cells, which results in T cell proliferation due to an MHC incompatibility that is not relevant to viral infection, we used autologous cells. Also, rather than relying on a single time point, we evaluated kinetics and magnitude of T cell proliferation and cytokine production. Another difference is that we (i) investigated responses specific to each of these viruses, which primarily represented stimulation of memory T cells from prior natural infection, and which were dependent on antigen processing and presentation by the DC, and (ii), using SEB as a model, addressed the effect of viral infection of DC on secondary antigenic responses.
With regard to antigen-specific responses, CD4+ T cell proliferation in response to MDDC exposed to rHRSV was less than that to rHPIV3 and IAV and greater than to rHMPV, but the differences were not significant. In addition, proliferation was greater in response to MDDC exposed to live versus UV-inactivated virus, indicating that, although we cannot rule out the possibility of viral interference with CD4+ T cell proliferation, the net effect of exposure to live virus was stimulatory rather than inhibitory. The increased percentage of proliferating cells found in IAV cultures might reflect the presence of more IAV-specific CD4+ T cells at the beginning of the culture compared to the frequency of T cells specific for the other viruses. However, the magnitude of T cell proliferation induced by the four viruses correlated well with the extent of MDDC maturation we observed previously 
, as well as in additional MDDC maturation studies (data not shown) in which IAV was slightly but not significantly stronger in inducing MDDC maturation than rHPIV3, rHMPV, and rHRSV. Thus, the trend of increasing T cell proliferation responses in the order: rHMPV < rHRSV < rHPIV3 < IAV might reflect the relative potency of each of these viruses to induce MDDC maturation 
. The sub-maximal nature of MDDC maturation might represent insufficient stimulation rather than virus-mediated inhibition, since a secondary LPS stimulus further maturation to a similar final extent 
By day 7, the IFN-γ and TNF-α cytokine expression profiles were similar among the four viruses, with approximately the same proportion of IFN-γ single-positive or IFN-γ/TNF-α double-positive CD4+ T cells. This shows that MDDC stimulated by all four viruses induced the same cytokines at similar levels with no signs of inhibition or Th2- or Th17-biased responses.
The functionality of a protective T cell response depends on the quality of the cytokine producing cell. Several recent studies have shown that CD4+ T cells which are double-positive for IFN-γ and TNF-α produce these cytokines at a higher level compared to single-positive cells 
. This shows that these CD4+ T cells are more strongly activated, and more likely to provide stronger helper effects to CD8+ cells, leading to better protection 
. In the present study, we observed the same effect of increased cytokine production from double-positive cells as compared to single-positive cells. Thus, the proportion and number of IFN-γ/TNF-α positive cells was similar among the different viruses, providing no evidence of a deficit specific to any particular virus.
The CD4+ T cell recall response to all viruses was Th1-biased as characterized by the production of IFN-γ and the low IL-4 and IL-17 production by proliferating CD4+ T cells. This is offered with the caveat that the MDDC were generated in vitro in the presence of GM-CSF and IL-4, with the potential to bias the T cell response, in particular by stimulating or inhibiting the Th2 pathway. While there have been many reports of plasticity among CD4 T cell subsets, the data still points to a fair level of rigidity among Th1 and Th2 subsets, compared, for example, to Th17 and Treg subsets. This suggests that our in vitro observations reflect the level of Th1 responses to each of these viruses in vivo.
We also investigated whether any of the virus stimulated MDDC inhibited T cell proliferation and cytokine production to SEB, as a model of secondary infection that is independent of virus-mediated differential effects on antigen uptake and presentation pathways. Indeed, compared to their UV-inactivated counterparts, rHRSV, rHPIV3 and IAV inhibited the CD4 T cell response to SEB. However, this effect was transient, most pronounced on day 4, and showed little or no difference between live and UV-inactivated HRSV by day 6. The inhibitory effect was relatively less for live rHRSV and rHMPV, whereas live IAV and rHPIV3 induced a markedly stronger inhibition of proliferation and IFN-γ and TNF-α production by day 4 of co-culture.
The transient inhibition by live viruses in the SEB assay might be explained by the anti-proliferative effect of type I interferon on CD4+ T cells 
. We detected higher type I interferon expression levels in MDDC matured by rHPIV3 and influenza virus compared to the other viruses (
, and data not shown), suggesting a possible role in inhibition of proliferation. We tested this hypothesis in the present study by treating the MDDC/T cell co-cultures with IFN-β in concentrations similar to those in supernatants of rHPIV3-exposed MDDC 
and found a reduced proliferation in response to SEB, comparable to that of SEB-treated MDDC/T cell co-cultures with rHPIV3-MDDC. This supports the previous suggestion of Preston et al. 
and Chi et al. 
that type I IFN plays a role in inhibiting T cell responses to HRSV, and illustrates the importance of the local cytokine environment 
in the modulation of memory T cell proliferation.
In summary, each of these common human respiratory pathogens can affect the ability of MDDC to activate CD4+ T cells. The more biologically relevant response, namely the proliferation of virus-specific memory T cells, was somewhat less for the paramyxoviruses compared to IAV. While this might make a contribution to a trend of increased ease of re-infection, the differences in proliferation did not rise to the level of statistical significance. The modestly reduced paramyxovirus-specific proliferative responses correlated with reduced levels of DC maturation observed in previous studies, an effect that appeared to reflect a lower level of stimulation rather than virus-mediated inhibition. There was no obvious virus-specific bias to T cell polarization, and cytokine production was not significantly different between viruses. The non-specific proliferation response to SEB was lower for IAV than for HRSV and the other viruses. These results suggest that rHRSV-infected MDDC do not strongly and specifically inhibit proliferation of CD4+ T cells. Thus, HRSV-specific effects on DC/T cell interactions are unlikely to account for the ability of HRSV to cause repeat infections during life.