Proteasome inhibition reduces superantigen-mediated T cell activation and the severity of psoriasis in a SCID-hu model (original) (raw)
T cell activation–induced NF-κB DNA binding in human T cells is suppressed by the proteasome inhibitor PS-519. Since activated T cells are involved in maintaining psoriatic skin lesions and many T cell activating signals are transduced via activation of the nuclear transcription factor κB, we assessed the effects of PS-519 on superantigen- and cytokine-stimulated cells. PBMCs were stimulated with TSST-1 (100 ng/ml), and highly purified human T cells were stimulated with PHA (1 μg/ml), IL-2 (100 U/ml), and TGF-β (50 ng/ml). Total cell extracts were isolated 4 hours after culture initiation as outlined in Methods and analyzed for NF-κB–binding activity using a 32P-labeled oligonucleotide probe. As we reported earlier, we found, in a time-dependent manner, the appearance of a protein-DNA complex, which was first observable after 30 minutes and reached a maximum after 4 hours (41). The PHA/IL-2/TGF-β–induced protein-DNA complex was characterized by using subunit-specific antibodies against p50 and p65 subunits of NF-κB. These subunits form prototypic heterodimers that are most frequently observed upon stimulation of various cell types with diverse conditions. Almost all of the protein-DNA complex newly induced by PHA/IL-2/TGF-β treatment was supershifted upon the addition of specific antibodies against the p50 and p65 subunits of NF-κB to the DNA-binding reaction. A nonspecific antibody was ineffective (41). This suggests that PHA/IL-2/TGF-β treatment induced complexes predominantly containing p50 and p65 subunits.
The appearance of the NF-κB DNA-binding complex was reduced in a dose-dependent manner using PS-519. T cells or PBMCs were pretreated for 1 hour with PS-519 (1–10 μg/ml). Thereafter, cells were stimulated with PHA/IL-2/TGF-β or TSST-1 for 4 hours. As expected, PS-519 suppressed NF-κB DNA-binding activity in stimulated T cells and PBMCs (Figure 1).
The PHA/IL-2/TGF-β–induced NF-κB DNA complex is suppressed by the proteasome inhibitor PS-519. Human T cells were stimulated with PHA/IL-2/TGF-β for 4 hours. PS-519 (1–10 μg/ml) suppressed NF-κB DNA-binding activity. The NF-κB DNA complex is indicated by a filled arrowhead. The open arrowhead shows the position of the unbound DNA probe. Binding mixture without cell extract was applied on lanes 1 and 10. The antioxidant _N_-acetylcysteine (NAC; 25 mM), which is known to inhibit IκB kinase, served as control (lane 4). The results are representative of three electrophoretic mobility shift assays from three independent donors.
PS-519 inhibits TSST-1–induced T cell proliferation. Superantigens such as TSST-1 belong to the strongest T cell–activating compounds because they stimulate a large proportion of T cells, as compared with nominal antigens, via their interaction with the T cell receptor Vβ-chain. Additionally, they have been shown to be involved in the induction of psoriasis in animal models of the disease and in human studies (12, 13, 42, 43). Therefore, we analyzed the effects of PS-519 on T cell proliferation and activation in superantigen-stimulated cells. TSST-1 induced a 28-fold increase in 3H-thymidine incorporation, which was inhibited in a dose-dependent manner by PS-519 and reached a maximum at 2.5 μg/ml. Stimulation indexes were: unstimulated cells, 1; TSST-1–stimulated cells, 28.13 ± 3.30; cells stimulated with TSST-1 and PS-519 (2.5 μg/ml), 1.65 ± 1.81 (Figure 2). Proliferation of unstimulated cells cultured in the presence of PS-519 (0.25–2.5 μg/ml) remained unchanged (Figure 2). Using trypan blue exclusion tests and propidium iodide uptake, we excluded toxic effects of PS-519 as a possible cause for reduced T cell proliferation (data not shown).
PS-519 inhibits TSST-1–induced T cell proliferation. PBMCs (2 × 106/ml) obtained from five healthy volunteers were stimulated with TSST-1 (100 ng/ml) in the absence or presence of PS-519 (0.25–2.5 μg/ml) for 4 days and thereafter pulsed with 3H-thymidine. Incorporation of 3H-thymidine into DNA was calculated using a liquid scintillation counter. Stimulation index was calculated by the ratio: decays per minute of experimental group/decays per minute of control group. Open symbols represent resting PBMCs, filled symbols TSST-1–stimulated PBMCs. A significant reduction in proliferation was observed starting at 0.25 μg/ml PS-519 (*P < 0.001). Values represent mean ± SD of five healthy donors.
PS-519 inhibits TSST-1–induced expression of T cell activation molecules. Since superantigens are potent activators of T cells and activated T cells are involved in maintaining the psoriatic skin lesions, we investigated the effects of the proteasome inhibitor PS-519 on superantigen-mediated T cell activation. For this purpose we chose the very early activation marker CD69, the early activation marker CD25, and the late activation marker HLA-DR, which are all under the transcriptional control of NF-κB (15, 25, 44, 45). In TSST-1–stimulated cells, CD69 expression increased about fourfold at day 1 (unstimulated cells 9.62% ± 5.24% versus TSST-1–stimulated cells 43.89% ± 10.14%) and remained unchanged for 5 days. PS-519 significantly and dose-dependently reduced CD69 expression by T cells up to 40% (Figure 3a). CD25 expression by T cells continuously increased, reaching a plateau at day 5 of TSST-1 stimulation (day 5: unstimulated cells 8.32% ± 2.29% versus TSST-1–stimulated cells 53.34% ± 2.99%) and remained stable until day 9 (data not shown). Again, PS-519 induced a dose-dependent inhibition of this activation marker with a complete suppression of CD25 upregulation at 2.5 μg/ml (Figure 3b). HLA-DR expression by TSST-1–stimulated T cells peaked at day 5 (unstimulated cells 21.79% ± 3.41% versus TSST-1–stimulated cells 51.79% ± 12.26%) with a decline at day 9 (data not shown). HLA-DR expression by T cells treated in the presence of PS-519 was significantly reduced even as compared with unstimulated cells (5.81% ± 4.40%; P < 0.01; Figure 3c).
PS-519 inhibits TSST-1–induced expression of T cell activation molecules. PBMCs were stimulated with TSST-1 (100 ng/ml) in the presence or absence of PS-519 (0.25–2.5 μg/ml). CD69+ CD3+ (a), CD25+ CD3+ (b), and HLA-DR+ CD3+ (c) surface expression was measured at days 1, 3, 5, 7, and 9 (days 7 and 9 not shown) by flow cytometry. Appropriate isotype Ig’s served as controls to set gates for positive and negative staining. For CD69 expression, significant reduction was observed on day 1 starting at 1.0 μg/ml (P < 0.05), and on days 3 and 5 starting at 0.5 μg/ml PS-519 (P < 0.001 and P < 0.05, respectively). For CD25 expression, significant reduction was observed on day 3 starting at 1.0 μg/ml (P < 0.001), and on day 5 starting at 0.25 μg/ml PS-519 (P < 0.001). For HLA-DR expression, significant reduction was observed on day 1 at 2.5 μg/ml (P < 0.05), on day 3 starting at 1.0 μg/ml (P < 0.05), and on day 5 starting at 0.25 μg/ml PS-519 (P < 0.001). Data represent means of five experiments ± SD.
PS-519 inhibits TSST-1–induced expression of T cell homing molecules. Recirculation of activated T cells into the skin is of key importance for both immunosurveillance and T cell–mediated inflammatory disorders such as psoriasis. We therefore examined the effects of PS-519 on TSST-1–induced upregulation of molecules known to be expressed by skin-seeking T cells. Expression of CLA and the closely related moiety sialyl LewisX (sLeX = CD15s) is induced by superantigens (2–4, 46). As described, we observed a maximum induction of both antigens at day 5 of TSST-1 stimulation (CLA: control cells 14.57% ± 2.52% versus TSST-1–stimulated cells 35.67% ± 5.63%; CD15s: control cells 15.31% ± 1.99% versus TSST-1–stimulated cells 50.19% ± 2.51%). In PS-519–treated cells, the TSST-1–induced upregulation of both skin-homing molecules was completely inhibited in a dose-dependent manner, and, even in concentrations as low as 0.5 μg/ml, CLA and CD15s expression remained below control levels (CLA: PS-519–treated [2.5 μg/ml] and TSST-1–treated cells 9.61% ± 1.72%; CD15s: 9.61% ± 1.72%; Figure 4, a and b). CLA and CD15s expression correlated with in vitro E-selectin binding, which was also completely inhibited by PS-519 treatment in concentrations above 1 μg/ml (Figure 4c). In vitro binding of T cells to P-selectin remained unchanged in TSST-1– and TSST-1+PS-519–treated cells as compared with control cells (data not shown). CD54 expression was also inhibited by PS-519, but to a lesser extent as compared with suppression of CLA or CD15s (data not shown).
PS-519 inhibits TSST-1–induced expression of T cell adhesion molecules. PBMCs were stimulated with TSST-1 (100 ng/ml) in the presence or absence of PS-519 (0.25–2.5 μg/ml). CLA+CD3+ (a) and CD15s+CD3+ (b) surface expression and binding of CD3+ cells to E-selectin (c) was measured at days 1, 3, 5, 7, and 9 (days 1 and 9 not shown) by flow cytometry. Appropriate isotype Ig’s served as controls to set gates for positive and negative staining. Staining in the absence of fusion proteins and in the presence of anti-CD3–FITC together with secondary anti-human IgG-phycoerythrin demonstrated absence of unspecific binding reactivity. For CLA expression, significant reduction was observed on day 3 at 2.5 μg/ml (P < 0.05), and on days 5 and 7 starting at 0.25 μg/ml PS-519 (P < 0.001). For CD15s expression, significant reduction was observed on days 5 and 7 starting at 0.25 μg/ml PS-519 (P < 0.001). For E-selectin binding, significant reduction was observed on day 5 starting at 0.25 μg/ml (P < 0.001), and on day 7 starting at 0.5 μg/ml PS-519 (P < 0.05). Data represent means of five independent experiments ± SD.
Interestingly, lower concentrations of PS-519 were necessary for inhibition of T cell proliferation, of T cell activation markers, and of cell adhesion molecules as compared with inhibition of NF-κB DNA binding (Figures 1–4). This can be explained by the fact that threshold levels of active NF-κB are required for T cell activation and optimal cellular adhesion molecule expression. Therefore, a small decrease in the level of activated NF-κB can lead to a profound change in the level of transcription (27, 30).
PS-519 is effective in the treatment of psoriasis in a xenogeneic transplantation model. Having demonstrated reduced superantigen-mediated T cell activation in vitro, we employed PS-519 as a therapeutic tool in a SCID-hu xenogeneic transplantation model that proved to be a useful tool in the evaluation of antipsoriatic drugs (47, 48). Only grafts from lesional psoriatic skin treated with the vehicle retained their increased epidermal thickness (414 ± 54 μm; Figures 5a and 6) and papillomatosis. The percentage of proliferating basal keratinocytes expressing the proliferation marker Ki-67 was 10% ± 4% (Figures 5b and 6). A dense leukocytic infiltrate was present in the upper dermis (75 ± 12 leukocytes per high power field; Figure 5c). In contrast, grafts treated with PS-519 exhibited a markedly reduced epidermal thickness (110 ± 43 μm) along with fewer Ki-67–positive basal keratinocytes (1% ± 1%). The remaining inflammatory infiltrate localized in the upper dermis was reduced (41 ± 7 leukocytes per high power field; Figure 5, a–c). These changes paralleled the findings in grafts of mice that received dexamethasone treatment (Figure 5, a–c, and Figure 6).
PS-519 suppresses hallmarks of psoriasis in a xenogeneic transplantation model. Grafts from lesional psoriatic skin were transplanted onto SCID mice as outlined in Methods. After 2 weeks, mice were treated with PS-519 (1 mg/kg body weight), dexamethasone (0.2 mg/kg body weight), or vehicle for 4 weeks. Subsequently, epidermal thickness (a), proliferation of basal keratinocytes measured by Ki-67 reactivity (b), and leukocytic infiltration (c) were determined by a blinded investigator. All parameters showed a marked reduction (*P < 0.05, **P < 0.001). Data represent means of four independent experiments ± SD.
PS-519 suppresses hallmarks of psoriasis in a xenogeneic transplantation model. Grafted skin in PS-519–treated mice (1 mg/kg body weight) showed normalization of epidermal architecture, loss of papillomatosis, and marked reduction of acanthosis (c, hematoxylin-and-eosin stain) as compared with vehicle-treated mice (a, hematoxylin-and-eosin stain). In PS-519–treated (d, Ki-67 stain) as compared with vehicle-treated mice (b, Ki-67 stain), proliferation of basal keratinocytes was markedly reduced. Treatment with dexamethasone (0.2 mg/kg body weight; e and f) was as effective as PS-519 treatment. The sections show one representative of four experiments. ×200.
20S proteasome activity is reduced in PS-519–treated animals. To determine the pharmacodynamics of PS-519 from these transplanted mice, blood was collected 2 hours after treatment at days 4, 8, 14, and 28 from PS-519– and vehicle-treated animals. Results indicate an 85.7% ± 8.6% (mean ± SD) reduction in proteasome activity in the PS-519–treated animals as compared with the control (Figure 7). Already after 4 days of treatment, 20S proteasome activity was reduced by 81.6% ± 1.7%, indicating that the effects were achieved quite early after the beginning of treatment. In the vehicle-treated mice, 20S proteasome activity remained constant during the whole experimental period (data not shown). No effect on 20S proteasome activity was seen in the dexamethasone-treated group (data not shown). These data support the critical role the proteasome plays to suppress the activation of NF-κB, thereby decreasing the transcription of many genes encoding proinflammatory proteins, including the T cell surface markers.
20S proteasome activity is reduced in PS-519 as compared with vehicle treated mice. Peripheral blood from vehicle- and PS-519–treated mice was drawn 2 hours after the final injection at days 4, 8, 14, and 28. Thereafter, 20S proteasome activity was determined as described above. PS-519–treated animals showed an 85.7% ± 8.6% (mean ± SD) inhibition of the 20S proteasome activity as compared with vehicle-treated mice at day 28 (**P < 0.0001). Already after 4 days, 81.7% ± 1.6% inhibition as compared with controls was achieved (*P < 0.01). The 20S proteasome activity in vehicle-treated mice remained unchanged when measured at 4, 8, and 14 days as compared with day 28 (data not shown). The 20S proteasome activity is given in pmol/s/mg protein. Data represent mean ± SD; n = 6, days 4–14; n = 8, day 28.