Exacerbation of a TNF-α–induced panniculitis by mPR3-ANCAs. Mice treated intravenously with either autologous mPR3-ANCA serum (antiserum pool 1) or serum from mock-immunized mice received TNF-α locally in the same skin area on 4 consecutive days. The contralateral side was treated either with PBS or PBS/0.1% BSA (TNF-α solvent) as a control. (A) The degree of cellular infiltrations in the subcutaneous adipose tissue was measured with digital image analysis software. The plotted data show the net focus size increase between TNF-α–treated and control-treated skin in 17 mice belonging to 4 independent experimental series, which are distinguished by squares, circles, triangles, and diamonds. Nine mice were treated with autologous mPR3-ANCA serum (left column), and 8 mice received serum from mock-immunized mice (right column). Horizontal lines indicate the median of the 2 experimental groups. The difference between mPR3-ANCA serum and mock immune serum–treated animals is statistically significant (P = .043, Mann-Whitney U test). (B) Histologic sections of biopsies from an mPR3-ANCA serum-treated mouse (i) and a mock immune serum-treated mouse (ii) are shown (original magnification × 200; inset × 400). Note the dense cellular infiltrate in the subcutaneous adipose tissue that is accompanied by an apparent local increase of the septal volume. The cellular infiltrate consisted of neutrophils, eosinophils without leukocytoclasia, and mononuclear cells. Staining with an antibody against the neutrophil marker Ly-6G confirms the presence of neutrophils in the inflammatory infiltrate (red-stained cells in the inset). Images were acquired using the same equipment described in the Figure 2B legend.