Expression of β
2 integrins on T cells following viral infections has been described in lymphatic organs [
7,
8,
11] but knowledge about integrin expression on T cells following respiratory infections is scarce. Previously, we have described a population of CD8
+ T cells in the lung expressing the β
2 integrin CD11c following RSV infection [
14]. To characterize these cells in more detail, we have performed a time course analysis of CD3
+ cells displaying CD11c, thus excluding cells of myeloid origin. In mediastinal lymph nodes, where CD11c
+ cell numbers increase during inflammation [
17,
18], CD8α expressing dendritic cells [
19,
20] could be mistaken for CD11c
+ CD8
+ T cells if analyzed based on CD11c and CD8α only. Following RSV infection, CD11c expression on lung CD8
+ T cells increased with up to 50% of CD8 T cells positive for CD11c on dpi8, while, in contrast to LCMV infection, percentages of CD11c
+ CTL in spleen and blood were low. CD11c up-regulation on mucosal T cells has previously only been described in the intestine where a subpopulation of intraepithelial T cells expresses CD11c constitutively [
10]. In the lung, expression of CD11c on T cells was restricted to CD8
+ T cells and was paralleled by increases in CD11b. In contrast to CD11b and CD11c, expression of F4/80 was hardly detectable on pulmonary T cells following RSV infection. Expression of F4/80 may be dependent on both mouse strain and organ analyzed: F4/80
+ T cells seem to be more prevalent in C57Bl/6 than in BALB/c mice [
11] and were detectable in significant amounts only on splenic CD8
+ T cells, as shown here and by Lin and colleagues [
11]. In contrast to viral infections, allergic airway inflammation elicited by sensitization to OVA, resulted only in a small population of CD11c
+ CD8
+ T cells in the lung, demonstrating that up-regulation of CD11c is not a consequence of T cell activation and migration per se or of an inflammatory environment, but that it is dependent on the underlying pathology. Interestingly, reactivation of antigen-specific CD8
+ memory T cells by OVA results in CD11c up-regulation when cells are primed with an OVA-expressing virus [
21]. This suggests that MHC class I-restricted antigen presentation as well as co-stimulatory factors induced by viral infection, e.g. IFNα, favor expression of CD11c on CD8
+ T cells. Several studies have demonstrated up-regulation of CD11a [
22], expression of CD11b [
7,
8,
23] or CD11c [
11] on T cells following viral infections. Here, we demonstrate that following RSV infection a population of CTL co-expresses CD11a
hi, CD11b and CD11c. Furthermore CD11c mRNA was detectable in resting as well as activated CD8
+ T cells. Contamination by CD11c
+ non-T cells cannot be excluded completely but is very unlikely to account for CD11c mRNA detection in CD8 T cells since no major difference in CD11c mRNA quantity was detected between CD8
+ and CD11c
+ cell populations. CD11c mRNA seems to be expressed constitutively in CD8
+ T cells while expression of CD11c protein on the cell surface requires activation. This notion is supported further by the finding that activation markers were up regulated on CD11c
+ T cells both in the lung and in the draining MLN. Interestingly, CD11b mRNA is also expressed in resting as well as activated human T cells but CD11b becomes detectable on the cell surface only following stimulation of T cells [
24]. Taken together these observations make it tempting to speculate that there may be a common mechanism of post-transcriptional regulation of β
2 integrin expression following T cell activation.
Assessing functional properties of CD11c
+ CD8
+ T cells, we found that these were more efficient than CD11c
- CD8
+ T cells in IFNγ production, target cell lysis
in vitro and induction of viral clearance
in vivo. This is likely due to the fact that the CD11c
+ CTL population contains the majority of activated T cells specific for the RSV M2
82–90-peptide, the immuno-dominant CTL epitope of RSV in BALB/c mice. Thus, we show an association of adhesion molecule expression and high cytotoxic activity of pulmonary CD8
+ T cells during a respiratory viral infection, parallel to findings in lymphoid organs in LCMV infection [
8,
11]. A subset of CD11c
- CD8
+ T cells expressed high levels of the activation marker CD69 and low levels of CD62L. Cytotoxic effects induced by CD11c
- CD8
+ cells could be due to this subpopulation of activated CD11c
- CTL. We cannot exclude this, since we did not assess cytotoxic effects of isolated subpopulations of CD11c
- CD8
+ T cells. We believe though, that contamination with CD11c
+ CTL is a more likely explanation for cytotoxic activity observed when CD11c
- cells were used as effectors. In contrast to a study implicating CD11c on human T cell lines in target cell binding [
25] we could not detect an inhibitory effect of anti-CD11c-mAb on the cytotoxic activity of T cells. In addition to target cell binding, β
2 integrins are involved in migration of leukocytes and a role for CD11b in recruitment of T cells into inflamed areas has been shown by antibody blocking experiments [
9]. In our model, the percentage of CD11c
+ cells within the CD8
+ T cell population differed between the organs assessed, being lowest in MLN and spleen, higher in the blood, where a short-lived increase of these cells was noted during the acute phase of infection, and highest in the lung. This indicates that CD11c marks effector T cells, which are recruited to the site of infection. An analogous pattern of tissue distribution has been described for VLA-4
hi T cells following intra-cerebral LCMV infection [
8]. Interestingly, following adoptive transfer we observed efficient recruitment of CD11c
+ CD8
+ T cells to the lung in naïve mice, while only a small percentage of these cells were recruited to the spleen. This preferential recruitment of CD11c
+ CD8
+ T cells to the lung was even more pronounced in RSV infected recipients. Pre-treatment of transferred cells with anti-CD11c-mAb did not reveal a unique function of CD11c for homing of activated CTL to the lung. These observations suggest that CD11c is a marker for activated CTL, which are recruited preferentially to the lung, while the CD11c molecule itself may not be directly involved in the process of T cell recruitment to the lung. Functional redundancy of β
2 integrins in target cell binding or migration into effector sites is a possible explanation of our results. CD11a has been shown to be important for target cell binding [
26], recruitment of activated lymphocytes into the lung [
5,
27] and generation of an effective cytotoxic T cell response during primary RSV infection [
28].