EAE is less severe in tPA deficient mice.
tPA−/− mice developed less severe EAE disease than WT mice, with lower clinical scores during the plateau phase between days 17 and 23 (d20 ± 3) (Fig. 1A). Although no significant difference was observed between tPA−/− and WT mice in the incidence of the disease (P = 0.1194, Fig. 1B) nor for the day of disease onset (tPA−/−: 15.13 ± 3.54 vs WT: 14.6 ± 2.01, P = 0.8159, data not shown), peak score and cumulative clinical score were significantly lower in tPA−/− than in WT mice (tPA−/− peak score: 1.16 ± 1.38 vs WT: 2.07 ± 1.56, P = 0.0164; cumulative clinical score: 10.17 ± 17.12 vs 20.38 ± 20.23, P = 0.0163, Fig. 1C and D). In addition, severity index ([25]), was significantly lower in tPA−/− than in WT mice (tPA−/− severity index: 0.3440 ± 0.4986 vs WT: 0.6652 ± 0.5522, P = 0.0075, Fig. 1E).
tPA enhances T-cell response by a proteolytic mechanism.
Given that immune cell infiltration is a cardinal feature of EAE, we analysed T cell subpopulations in the spinal cords of EAE mice during the plateau phase of the disease (d20 ± 3) by flow cytometry and immunofluorescence. We found that the absolute number of CD4+ T cells was lower in tPA−/− than in WT mice (6940 ± 2110 vs 27110 ± 8920 cells; P = 0.0238, Fig. 2A), an observation consistent with the clinical scores observed in tPA/− EAE mice (Fig. 1). Neither CD8+ (WT: 61291 ± 27112 cells vs tPA−/−: 42721 ± 8640 cells, P = 0.5476, Fig. 2B) nor regulatory FoxP3+ T cell (Treg) (WT: 2396 ± 2322 cells vs tPA−/−: 394.6 ± 444 cells; P = 0.1667, Fig. 2C) were significantly altered. In accordance with the above data, CD4+ T cells were the only CD3+ T cell subset that showed changes in the spinal cord of tPA−/− mice (tPA−/−: 13.87%±4.22 vs WT: 29.85%±9.83, P = 0.0476, Fig. 2D). Concordantly, the histological analysis of the spinal cords (Fig. 2E) further showed that the average density of CD4+ T cells within infiltrated area was lower in tPA−/− than in WT EAE mice (absolute number of WT: 514.4 ± 235.1 vs tPA−/− 245.3 ± 87.95, P = 0.0495, Fig. 2E-G and Suppl. Figure 1A-C).
To determine the mechanisms involved in the decrease of CD4+ T cell number in tPA/−EAE mice, we assessed their functional response in vitro after activation with anti-CD3ε/CD28. CD4+ T cells from tPA−/− naive mice proliferated less than CD4 + T cells from WT mice (proliferation index: tPA−/−: 2.03 ± 0.95 vs WT: 6.18 ± 3.08; P = 0.0317; Fig. 3A and B). Activation level measured by mean fluorescence intensity (MFI) of CD25 was also reduced in CD4+ T cells from tPA−/− as compared with WT CD4+ T cells (51.42%±21.91 of WT; P = 0.0286, Fig. 3D). Importantly, the addition of exogenous tPA rescued the proliferation and activation of tPA/− CD4 + T cells at the same level as WT CD4 + T cells (P = 0.8413 and P = 0.3143, respectively; Fig. 3A, B and D). As in the case of the in vivo analysis, CD8+ T cell proliferation was not affected in tPA−/− (proliferation index WT: 11.65 ± 11.01 vs tPA−/−: 6.57 ± 7.48, P = 0.3095) although their activation was reduced (59.34%±28.96 of WT, P = 0.0286; Fig. 3A, C and E).
We next examined the effect of tPA on T cell response in vitro. tPA significantly increased proliferation and activation of anti-CD3ε/CD28 activated CD4+ and CD8+ T cells in a dose-dependent manner (Fig. 4A-E). At the dose of 10 µg/mL of tPA, proliferation of CD4+ and CD8+ T cells were increased to reach 163.40%±33.85 (P = 0.0004) and 170.90%±35.60 (P = 0.0012), respectively. No effect of tPA was observed on Treg cells (P = 0.9999, Fig. 4F).
Next, we investigated the mechanisms underlying tPA actions on T cells. Since the primary action of tPA is to activate plasminogen into plasmin, we interrogated the different elements of the tPA/plasminogen/plasmin cascade. Plasminogen treatment neither increased CD4+ T cell proliferation nor potentiated the proliferative action of tPA on CD4+ T cells (P = 0.400 and P = 0.200; Fig. 4J). However, aprotinin, a specific inhibitor of plasmin, did reverse tPA-mediated activation (for CD4+ T cells, P = 0.0190; for CD8+, P = 0.0109; Fig. 4H-I). In addition, the inactivation of the catalytic activity of tPA (tPA-GGACK) abolished tPA stimulatory effect on CD4+ and CD8+ T cell proliferation (P = 0.0252 and 0.0162, respectively; Fig. 4H-I). In addition, ε-ACA, an inhibitor of plasmin generation from plasminogen inhibited the stimulatory effect of tPA on CD4+ and CD8+ T cell proliferation (P = 0.0056 and P = 0.0040, respectively; Fig. 4H, I). This indicates that tPA increases T cell proliferation via the proteolytic activation of plasminogen into plasmin.
We next explored if tPA-mediated effects on T cell proliferation was related to its previously demonstrated proteolytic action on NMDAR [27], as this receptor was previously reported to be expressed on T cells [28, 29] Glunomab®, a monoclonal antibody that blocks the interaction between tPA and NMDAR [18], did not alter the proliferative effect of tPA on CD4+ T cells (Supplementary Fig. 2). This result excluded that tPA may act on T cell proliferation by acting on NMDAR.
Furthermore, since cytokines are key mediators of T cell-driven autoimmunity, we analysed the impact of tPA on the late cytokine pattern of activated T cells. tPA induced an increase of IL-6 and IL-10 secretion by activated splenocytes at four days of culture (respectively 233.9%±125.5, P = 0.0289 and 499.4%±487.6, P = 0.0029, Fig. 4G). In addition, concerning IL6, this effect was not observed with tPA-GGACK and was inhibited in presence of ε-ACA or aprotinin (P = 0.0167, 0.0167 and 0.0333 respectively, Fig. 4K). ε-ACA also reverted the activation of IL-10 secretion by tPA (P = 0.0167, Fig. 4L). Together, these data indicate that tPA increases T cell proliferation via the generation of plasmin to increase their proliferation, activation, and secretion of cytokines.
tPA enhances myeloid cell maturation by a mechanism dependent of proteolysis.
Our next step was to analyse whether the distribution of myeloid cells is altered in the spinal cord of tPA−/− EAE mice. We found that absolute numbers of CD11c+/CD11b+ (dendritic cells, DCs) and CD45high/CD11c−/CD11bhigh (activated microglia and infiltrated macrophages, Mɸ) were lower in the spinal cords of tPA−/− EAE mice as compared to WT EAE mice, (DCs: 15474 ± 5169 vs 72283 ± 20405; P = 0.0238; microglia/Mɸ: 14980 ± 9426 vs 119228 ± 14980, P = 0.0238; Fig. 5A and B).
We then analysed whether tPA may affect the proportion and the phenotype of DCs and Mɸ. First, splenocytes extracted from EAE mice at the peak of the clinical course were treated with different concentrations of exogenous tPA (0.2, 2 and 20 µg/ml) during 24 hours. None of the tPA concentrations modified the percentage of antigen presenting DCs (CD11c+ MHC-II+) and Mɸ (F4/80+ MHC-II+) populations (Fig. 6A and B). However, tPA (2 µg/mL) induced a significant increase of MHC-II+ MFI in both cell populations compared to control conditions (DCs: 120.75%±17.10; Mɸ: 121.69%±21.12, P < 0.001, Fig. 6C and D). Interestingly, the same tPA dose promoted the polarization on the MHC-II+ expressing antigen presenting cell (APC) subsets towards a more pro-inflammatory and immunogenic phenotype, with a significant increase in the percentage of MHC-II+ CD80+ CD86+-APCs (DCs: 122.45 ± 27.79, P = 0.002; Mɸ: 128.04 ± 25.49; P = 0.004, Fig. 6E and F) and a decrease in the percentage of MHC-II+ CD80− CD86− tolerogenic APCs (DCs: 87.09%±14.29 Mɸ: 86.71%±12.08; P = 0.005 and P < 0.001, respectively; Fig. 6G and H).
Then, we aimed at analysing the receptors and/or tPA functional domains enrolled in APC maturation. The inactivation of the catalytic activity of tPA (tPA-GGACK) partially reversed MHCII+ upregulation on APCs (tPA: DCs: 121.12%±14.59 Mɸ: 113.96%±15.26 with respectively P = 0.040 and P = 0,042 versus control condition; tPA GGACK: DCs: 108.30%±14.43; Mɸ: 105.21%±16.10, with respectively P = 0.156 and P = 0.440 versus control condition and P = 0.072; P = 0.186 versus tPA; Fig. 7A and B). Nonetheless, none of the tPA-mediated effects on APC polarization were modified by the addition of Glunomab® (Suppl. Figure 3). These data indicate that the action of tPA on APCs is partly due to its proteolytic effect and is not mediated via interaction with NMDAR.
On the other hand, it has been shown that tPA can act via non-proteolytic “growth factor-like” effects [30], some of them mediated by its EGF-like domain [4, 14] The addition of the EGFR blocking agent AG1478 decreased MHC-II+ expression on MHC-II+ APCs as compared to the condition with tPA (DCs: 55.49% ±20.64; Mɸ: 77.76%±18.82; P < 0.001, Fig. 7C and D). These data indicate that tPA promotes APC maturation partly by the activation of EGFR.
tPA-mediated APC polarization is accompanied by a higher MOG-specific T cell response .
In order to determine whether the stimulatory effect of tPA on APC maturation may modify T cell functions, splenocytes extracted from EAE mice at the peak of the clinical course were cultured in presence/absence of tPA, with or without ex-vivo reactivation with MOG. In absence of MOG reactivation, tPA by itself did not induce CD4+ or CD8+ T cell proliferation (CD4+: 107.22%±29.02; CD8+: 124.73%±66.40 data not shown). Condition with MOG reactivation showed a higher percentage of proliferation than control conditions for both CD4+ and CD8+ T cells (CD4+: 852.79%±525.09; CD8+: 1525.90%±1096.05, both P < 0.05, data not shown). Interestingly, tPA potentiated MOGinduced CD4+ T cell proliferation (132.43%±17.77; P < 0.05, Fig. 7E). In line with the previous results about APC maturation, the stimulatory effect of tPA on MOG-activated CD4+T cell proliferation was abrogated when EGFR activity was blocked by AG1478 (81.88%±40.36; P < 0.05 Fig. 7E), while AG1478 did not modify MOG-induced APC proliferation in the absence of tPA (Fig. 7E). Remarkably, the stimulatory effect of tPA was not observed in MOG-activated CD8+ T cells (108.26%±15.42; P > 0.05, Fig. 7F), in accordance with what observed in the context of CD3/CD28 polyclonal activation of CD8+ T cells extracted from tPA−/− mice (Fig. 3C).
In sum, our data indicated that the effect of tPA on APC maturation and pro-inflammatory polarization resulted in amplified MOG-induced CD4+ T cell response. This effect may explain the deleterious role of tPA in EAE.