Methionine aminopeptidase-2 blockade reduces chronic collagen-induced arthritis: potential role for angiogenesis inhibition

Abstract

The enzyme methionine aminopeptidase-2 (MetAP-2) is thought to play an important function in human endothelial cell proliferation, and as such provides a valuable target in both inflammation and cancer. Rheumatoid arthritis (RA) is a chronic inflammatory disease associated with increased synovial vascularity, and hence is a potential therapeutic target for angiogenesis inhibitors. We examined the use of PPI-2458, a selective non-reversible inhibitor of MetAP-2, in disease models of RA, namely acute and chronic collagen-induced arthritis (CIA) in mice. Whilst acute CIA is a monophasic disease, CIA induced with murine collagen type II manifests as a chronic relapsing arthritis and mimics more closely the disease course of RA. Our study showed PPI-2458 was able to reduce clinical signs of arthritis in both acute and chronic CIA models. This reduction in arthritis was paralleled by decreased joint inflammation and destruction. Detailed mechanism of action studies demonstrated that PPI-2458 inhibited human endothelial cell proliferation and angiogenesis in vitro, without affecting production of inflammatory cytokines. Furthermore, we also investigated release of inflammatory cytokines and chemokines from human RA synovial cell cultures, and observed no effect of PPI-2458 on spontaneous expression of cytokines and chemokines, or indeed on the angiogenic molecule vascular endothelial growth factor (VEGF). These results highlight MetAP-2 as a good candidate for therapeutic intervention in RA.

Figure 1

Structure of PPI-2458.

Figure 2

PPI-2458 reduces acute CIA. Following onset of arthritis induced by bovine type II collagen, mice were treated with PPI-2458 (0.5, 1.5, or 5 mg/kg per day) or vehicle alone (n = 8 per group). (a) Change in clinical score for each animal from day 1 of arthritis and (b) paw swelling were assessed over a 10-day period. Data are expressed mean ± standard error. Statistical analysis was carried out using a two-way analysis of variance versus vehicle-treated mice: ** P < 0.01, ***P < 0.001. CIA, collagen-induced arthritis.

Figure 3

PPI-2458 reduces chronic CIA. Following onset of arthritis induced by murine type II collagen, mice were treated with either PPI-2458 (1.5 mg/kg every other day), soluble tumour necrosis factor receptor (sTNF-R)II (5 mg/kg every other day) or vehicle alone, or were left untreated (as indicated for the relevant experiments). (a) Clinical score was assessed over 4 weeks. Data are expressed mean ± standard error. Statistical analysis was carried out using a two-way analysis of variance versus vehicle-treated mice: ***P < 0.001. Each treatment group contained at least six mice. (b) Area under curve analysis, with data represented as box-and-whiskers plots. Statistical analysis was carried out using a one-way analysis of variance versus vehicle-treated mice: *P < 0.05, **P < 0.01. (c) Clinical score was assessed over 10 weeks. Data are expressed mean ± standard error. Statistical analysis was carried out using a two-way ANOVA versus vehicle-treated mice: ***P < 0.001. Each treatment group contained at least eight mice. (d) Area under curve analysis, with data represented as box-and-whiskers plots. Statistical analysis was carried out using a one-way analysis of variance versus vehicle-treated mice: *P < 0.05, **P < 0.01.

Figure 4

PPI-2458 reduces joint destruction in chronic collagen-induced arthritis. Following onset of arthritis, mice were treated with either PPI-2458 (1.5 mg/kg every other day), soluble tumour necrosis factor receptor (sTNF-R)II (5 mg/kg every other day), or vehicle alone for 28 days. Hematoxylin and eosin sections were scored in a blinded manner for pannus formation, synovitis, bone and cartilage erosion. Representative images, with bar equivalent to 20 μm, are illustrated.

Figure 5

PPI-2458 reduces proliferation of HUVEC. Human umbilical vein endothelial cells (HUVEC) were plated at 3 × 10³ in 30 mm² wells and stimulated with either (a) vascular endothelial growth factor (VEGF; 10 ng/ml) or (b) fibroblast growth factor (FGF)-2 (50 ng/ml) in 10% foetal calf serum, or were left untreated (serum-free medium). Fumagillin (10 nmol/l) and PPI-2458 (0.01 to 10 nmol/l) or 0.001% dimethyl sulfoxide as a vehicle control were added where indicated for 24 hours. Cells were pulsed with 10 μCi per well ³H-thymidine for a further 24 hours. Data are expressed as means ± standard deviation, representative of three experiments, and were analysed by one-way analysis of variance versus response in the absence of inhibitor: **P < 0.01.

Figure 6

PPI-2458 reduces VEGF and FGF-2 mediated angiogenesis. Angiogenesis in response to either (a,b,d) 2 ng/ml vascular endothelial growth factor (VEGF) or (c) 50 ng/ml fibroblast growth factor (FGF)-2 was assessed after 11 days. Fumagillin (10 nmol/l) and PPI-2458 (1 nmol/l or 10 nmol/l), or 0.001% dimethyl sulfoxide as a vehicle control were added where indicated. CD31 expression was measured using alkaline phosphatase-mediated conversion of (panel a) p-nitrophenol phosphate followed by colorimetric assay (panels b and c) BCIP/NBT (5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium) followed by image analysis. Data are expressed as means ± standatd deviation, representative of three experiments, and were analysed by one-way analysis of variance versus response in the absence of inhibitor: **P < 0.01. Included in panel d are representative images showing morphology of the formed tubes stained for CD31 at day 11 (objective magnification 40×).

Figure 7

PPI-2458 does not affect HUVEC inflammatory responses. Human umbilical vein endothelial cells (HUVEC) were plated at 3 × 10⁴ in 75 mm² wells and stimulated for 24 hours with either (a) IL-1β (10 ng/ml), (b) tumour necrosis factor (TNF)-α (10 ng/ml), or (c) lipopolysaccharide (LPS; 500 ng/ml). Fumagillin (10 nmol/l) and PPI-2458 (0.01 to 10 nmol/l), or 0.001% dimethyl sulfoxide as a vehicle control were added where indicated. Release of monocyte chemoattractant protein (MCP)-1 was measured by enzyme-linked immunosorbent assay (ELISA). MCP-1 was measured by ELISA. Data are expressed as mean ± standard deviation, representative of at least three experiments.

Figure 8

PPI-2458 does not affect RA synovial cell inflammatory responses. Rheumatoid arthritis (RA) synovial cells were plated at 5 × 10⁵ in 75 mm² wells and incubated for 24 hours with either fumagillin (10 nmol/l), PPI-2458 (0.01 to 10 nmol/l), or 0.001% dimethyl sulfoxide as a vehicle control. Release of (a) vascular endothelial growth factor (VEGF), (b) IL-6, (c) transforming growth factor (TNF)-α and (d) monocyte chemoattractant protein (MCP)-1 were measured using enzyme-linked immunosorbent assay.