Serum and EMA assay
Anti-endomysium antibodies (EMA) were determined by indirect immunofluorescence on pig intestine. 132 serum samples from non-pregnant donors were provided by the immunology laboratory of the Manchester Royal Infirmary and stored at -20°C. EMA-positive sera were reassayed blind at 1:30, 1:100, 1:300 and 1:1000.
A commercial ELISA (Celikey; Pharmacia Diagnostics) was used to determine anti-tTG IgA levels in patient sera. Results are reported as positive (OD ratio >1.4), borderline (OoralD ratio 1–1.4) and negative (OD ratio <1). The presence of tTG reactive IgA was confirmed by western blotting (not shown).
Sections of normal term placenta were dewaxed and incubated in methanol containing 0.15% hydrogen peroxide for 30 min to quench endogenous peroxidase activity, then microwaved in 0.01 M sodium citrate buffer pH 6.0, 10 min to achieve antigen recovery. Sections were incubated with protein block for 30 min then with autoimmune serum (tTG-positive or -negative; 1:10) overnight at 4°C. Mouse monoclonal anti-tTG (CUB 7402; 1:100, Labvision) was used as a positive control.
Sections were rinsed in TBS (x3) then incubated with either goat anti-mouse immunoglobulin-HRP or rabbit anti-human IgA-HRP (1:40, 30 min; Dako P0447 and P0216 respectively). After washing, sections were incubated with avidin-peroxidase (5 μg/ml in 0.125 M TBS, 1 h), developed in 3, 3' diaminobenzidine tetrahydrochloride (DAB) (10 mg/ml DAB/TBS, 0.0045% hydrogen peroxide) for 3 min, then nuclei counterstained with 0.25% methyl green.
The stained area (for syncytioplasm) or the linear staining (for the syncytial microvillous membrane) was obtained from a set of 20–25 serial images. Significance of difference between control and autoimmune staining were estimated using the Mann-Whitney test. Staining intensity was graded in a range + (weak) to +++ (intense).
Pre-absorption study: blockade of immunoperoxidase staining of human term placental tissue with human recombinant tTg
Serum positive for anti tTG antibodies (1:20 dilution in PBS) or mouse monoclonal anti-tTG (CUB 7204) at 1:400 dilution was incubated with recombinant human tTG (rhtTG) in different concentrations (5 μg, 2.5 μg, 1.25 μg) at 37°C for two hours prior to incubation with wax sections as described earlier.
For this experiment, mouse monoclonal anti tTG CUB 7402 (1:400 dilution) and PBS were used as positive and negative controls respectively. Goat anti mouse-HRP or rabbit anti human IgA-HRP were used in 1:40 dilutions as secondary antibodies. Controls were also carried out in which serum (1:20) or monoclonal anti-tTG were preincubated with recombinant human tTG (2.5 μg, 37°C 2 h) prior to staining. Full inhibition was achieved in both instances.
Direct binding of IgA to the placental villus
Placental tissue was fine-dissected to produce 2–3 mm pieces with intact villous architecture including stem, intermediate and terminal branches, fixed in 4% PFA/PBS for 2 h, then washed in 1% BSA/PBS (4 × 30 min). Positive or negative serum (1:10 in 5% BSA/PBS) and mouse monoclonal anti-tTG (1:200 in 5% BSA/PBS) were used as primary antibodies. Rabbit anti-human IgA-FITC (1:40; Dako F0204) and rabbit anti-mouse IgG-FITC (1:40; Dako F0261) secondary antibodies were used. The specimens were incubated with primary and secondary antibodies, washed in PBS (×4) in foil-wrapped eppendorf tubes over 30 min on a roller-mixer, snap frozen in OCT and cryosectioned. Negative controls contained either no primary antibody or anti-β-actin antibody (1:5000 in 5% BSA/PBS). The actin antibody produced staining only in the immediate vicinity (within ~50 μm) of cut surfaces (not shown), proving that fully epithelialised placental tissue was neither permeable to immunoglobulin nor leaky.
In situtTG activity and its inhibition by autoantibody
Syncytial tTG activity was assayed in situ by the incorporation of a fluorescent substrate into target proteins at the surface of intact placental tissue. Tissue (2–3 mm pieces) was incubated with the acyl donor hexapeptide biotinyl-TVQQEL[28, 29] (0.5 mM in serum-free F12/DMEM, 37°, 30 min). After the reaction, specimens were washed in F12/DMEM (2 × 2 min), fixed in 4% PFA (30 min), washed again (4 × 30 min), snap frozen in OCT and cryosectioned. Sections were incubated with streptavidin-FITC (1:200), washed and mounted using Vectashield containing propidium iodide. For the tTG activity inhibition assay, placental tissue pieces were directly incubated with serum (1:20 in 5% BSA/PBS) or monoclonal anti-tTG antibody (1:200) prior to addition of the acyl donor substrate.