Results

Flow cytometric analysis of CD4+ T lymphocytes in peripheral lymphoid tissues and the lumbar spinal cord
Previously, we have shown that spinal cord CD4+ T lymphocytes contribute to L5Tx-induced maintenance of mechanical hypersensitivity [4]. To identify the phenotype of CD4+ T lymphocytes involved in the development of L5Tx-induced mechanical hypersensitivity, we examined CD4+ T lymphocytes in both peripheral lymphoid tissues and the lumbar spinal cord via flow cytometry with antibodies that identify the Th1 and Th2 subtypes. WT BALB/c mice were randomly assigned to naive (no surgery), L5Tx, and sham groups. Since 7 days post-surgery is the peak time of detecting spinal cord-infiltrating CD4+ T lymphocytes, samples were collected at day 7 post-surgery.
To examine CD4+ T lymphocytes in the periphery, spleens and lumbar LNs were collected and processed for intracellular flow cytometric analysis. CD4+ T lymphocytes were first identified (the CD45+CD3+CD4+ population), and then the expression of T-bet (transcription factor found in Th1, and not Th2, cells), GATA-3 (transcription factor found in Th2, and not Th1, cells), IFN-γ (one of the signature cytokines produced by Th1 cells), and IL-4 (one of the signature cytokines produced by Th2 cells) were examined within the CD4+ T lymphocytes (Figure 1A). Consistent with our previous findings regarding peripheral responses post-L5Tx [4], no significant differences in the percentages of T-bet+, GATA-3+, IFN-γ+, or IL-4+ CD4+ T lymphocytes were detected in either spleen or lumbar LNs when the naive, L5Tx and sham groups were compared (Figure 1B and 1C, one-way ANOVA, p>0.05 for all data sets within each graph). Since signs of splenic CD4+ T lymphocyte activation were previously observed at day 3 post-surgery [4], similar experiments were performed to examine splenic CD4+ T lymphocytes at day 3 post-surgery. However, no significant differences were detected in any of the above populations assessed (data not shown). Thus, the data indicate a lack of a significant Th1- or Th2-dominant response in the periphery after either L5Tx or sham surgery.
We then examined the phenotype of the lumbar spinal cord-infiltrating CD4+ T lymphocytes in the naive, L5Tx, and sham groups. At day 7 post-surgery, lumbar spinal cord mononuclear cells were harvested and analyzed as described above via intracellular flow cytometric analysis (Figure 2). First, we examined the expression of the transcription factors T-bet and GATA-3 within the lumbar spinal cord-infiltrating CD4+ T lymphocytes. Within each set of experiments, the number of T-bet+CD4+ T lymphocytes in the L5Tx group was consistently higher than that of the naive and sham groups. As a result, there was a significant increase in the number of T-bet+CD4+ T lymphocytes in the L5Tx group compared to both the naive and sham groups. A slight, but not significant, increase in T-bet+CD4+ T lymphocytes was also observed in sham animals compared to naive mice (Figure 2B, one-way ANOVA, p<0.05). On the other hand, GATA-3+ CD4+ T lymphocytes were below the level of detection. Thus, these data indicated that the L5Tx-induced lumbar spinal cord infiltrating CD4+ T lymphocytes were predominantly Th1 cells. We further examined the cytokine expression of the infiltrating CD4+ T lymphocytes by measuring the number of IFN-γ+, IL-4+, TNF-α+, and GM-CSF+ CD4+ T lymphocytes. TNF-α and GM-CSF are two other cytokines predominantly associated with Th1 rather than Th2 cells. There were L5Tx-induced significant increases in the number of IFN-γ+ and GM-CSF+ lumbar spinal cord infiltrating CD4+ T lymphocytes (Figure 2C and E, one-way ANOVA, p<0.05). L5Tx induced similar, but not significant, changes in the number of TNF-α+ CD4+ T lymphocytes (Figure 2D, one-way ANOVA, p>0.05; p=0.051, L5Tx vs. sham), while no significant changes in the number of IL-4+ CD4+ T lymphocytes were detected (data not shown). These data further support lumbar spinal cord-infiltrating Th1 CD4+ T lymphocytes mediating L5Txinduced maintenance of mechanical hypersensitivity, and also suggest that this mediation might involve multiple Th1 cytokines.

Lumbar spinal cord IFN-γ production in both WT and CD4 KO mice post-L5Tx
We further measured the total levels of IFN-γ in the lumbar spinal cord at selected times (day 0 (naive), 3, 7, and 14) after L5Tx or sham surgery in both WT BALB/c and BALB/c CD4 KO mice via ELISA. First, no significant differences between the L5Tx and sham treatments were detected within either of the genotypes (Figure 3, one-way ANOVA for WT mice, p=0.685, and for CD4 KO mice, p=0.601). Therefore, further statistical analyses were performed on all data using “time” and “genotype” as factors regardless of treatment. It was found that CD4 KO mice expressed significantly lower levels of IFN-γ in the lumbar spinal cord compared to WT mice at all-time points, while there were no significant changes in the total levels of IFN-γ detected within either genotype over time, (Figure 3, two-way ANOVA, pgenotype<0.001, ptime=0.146, pgenotype × time=0.243). In addition, IFN-γ was not detectable in the splenic tissue of any of the mice. These data support that a reduction in IFN-γ levels (which can be associated with a blunted Th1 response) may contribute to the observed reduced mechanical hypersensitivity in CD4 KO mice compared to WT mice following L5Tx [4].

Mechanical sensitivity of CD154 KO mice
In the periphery, Th1 cells can promote macrophage proinflammatory responses through the ligation of CD154 expressed by Th1 cells and CD40 expressed by macrophages [21]. Previously, we have reported the critical role of microglial CD40 in the development of L5Tx-induced mechanical hypersensitivity [13]. To further identify the infiltrating CD4+ T lymphocyte-mediated downstream responses that contribute to the development of sustained mechanical hypersensitivity following L5Tx, we investigated the potential involvement of the interaction between microglial CD40 and CD154 expressed by spinal cord-infiltrating CD4+ T lymphocytes. First, we confirmed that L5Txinduced spinal cord infiltrating CD4+ T lymphocytes express CD154 via IHC. Serial sections of L5 spinal cord segments from two WT BALB/c mice 7 days post-L5Tx were stained for CD4 and CD154. Consistent with what we have shown previously, the majority of CD4+ T lymphocytes were observed in the ipsilateral side of the dorsal horn [4]. CD154+CD4+ T lymphocytes were detected in the ipsilateral side of the dorsal horn region (see representative images in Figure 4A). Due to the limited numbers of infiltrating CD4+ T lymphocytes, we could not accurately quantify the CD154+CD4+ T lymphocytes via IHC. To further investigate the CD40–CD154 pathway, an in vivo study with CD154 KO mice was conducted. We tested the mechanical sensitivity of BALB/c CD154 KO mice and WT BALB/c mice following either L5Tx or sham surgery with von Frey filaments. No basal differences in mechanical sensitivity were observed between the CD154 KO mice and WT mice. Interestingly, L5Tx induced significant mechanical hypersensitivity starting at day 1 and up to 21 days post-surgery in both CD154 KO and WT mice compared to the corresponding sham groups, while no significant differences in mechanical sensitivity between CD154 KO and WT mice were detected (Figure 4B; Two-way repeated ANOVA, ptime<0.001, pgroup<0.001, and ptime × group<0.001). Thus, it is unlikely that Th1 CD4+ T lymphocytes mediate L5Tx-induced behavioral hypersensitivity via the CD40–CD154 pathway.

Lumbar spinal cord GFAP expression in WT and CD4 KO mice post-L5Tx
To explore whether lumbar spinal cord-infiltrating CD4+ T lymphocytes can contribute to the maintenance of L5Tx-induced mechanical hypersensitivity by regulating astrocytic responses, we evaluated L5 lumbar spinal cord GFAP expression (increase of GFAP expression is an indicator of astrocytic activation) at selected times (day 0 (naive), 1, 3, 7, 10, and 14) after either L5Tx or sham surgery in both WT and BALB/c CD4 KO mice via IHC (Figure 5). The basal levels of GFAP expression were similar between WT and CD4 KO mice. In WT mice, L5Tx induced significant increases in GFAP expression in both the ipsilateral and contralateral sides (no significant differences between sides) over time (Figure 5B; Two-way ANOVA, ptime<0.001, pgroup=0.049, and ptime × group<0.945). Consistent with what has been previously reported [22]; this increase is most significant after day 7 post-L5Tx. However, in CD4 KO mice, although L5Tx induced an increase in GFAP expression similar to WT mice up to day 10 post-L5Tx, the increase in GFAP expression was not observed in CD4 KO mice at day 14 post-L5Tx (Figure 5C; Two-way ANOVA, ptime<0.001, pgroup=0.084, and ptime × group<0.757). These data suggest a potential link between spinal cord-infiltrating CD4+ T lymphocytes and L5Tx-induced astrocytic responses, particularly those that occur during the later maintenance phase of neuropathic pain.