(A) SEA-DR1 and SEH-DR1 complexes are SDS stable. Various Enterotoxins in the presence or absence of HLA-DR1/peptide complexes were incubated in PBS with a final SDS concentration of 0.1%, for 5 min at room temperature and subjected to electrophoresis in a 12% polyacrylamide gradient gel. The gel was silver stained by standard protocols. There was no evidence of multimer formation (*) Lane 1: empty DR1. Lane 2: DR1/HA_306-318 complex. Lane 3: DR1/HA_Y308A complex. Lane 4: SEA+DR1/HA_Y308A Lane 5: SEA only Lane 6: SEA+DR1/HA_Y308A Lane 7: SEH only. (B) SEA-DR1 but not SEH-DR1 complexes can bind an additional DR1. A solution of DR1 was injected on a HA_306-318 peptide decorated BIAcore CM5 chip (*) and allowed to dissociate for 3 minutes. A second injection (**) of preformed SEA-DR1 (red) resulted in a 30% greater RU change than SEH-DR1 (blue), as measured by the difference between the solid and dashed arrows. Two consecutive injections of DR1 alone on a control flow cell (inset) showed the same fold increase in RUs as SEH-DR1. ▴, start of injection.

B cells treated as in Fig. 2 were stained using anti-HLA-DR primary and gold-labeled secondary antibodies, fixed and analyzed. Representative tomograms of 250 nm thick sections of these cells were generated and segmented to highlight the 3D distribution of DR1 labeled by 6 nm gold beads (gold spheres) on the B cell membrane (purple). B cells were untreated (A), treated with SEA+Zn²⁺ (B), SEA+EDTA (C) or SEH (D). Representative stretches of cell membrane with a higher density of DR1 in cells treated with SEA (E), or untreated B cells (F) also showed no aggregates. Forced crosslinking of DR1 (G) caused aggregation (capping) of DR1 in some areas (black arrow/asterisk) but not others (white arrowhead).

(A) Acceptor Photobleach experiments can quantitatively measure energy transfer. Images of B cells isolated from wtDR1 transgenic mice were stained with Donor (Cy3) and Acceptor (Cy5) conjugated anti-DR monoclonal antibodies (L243) at Donor∶Acceptor ratios of (1∶1) before (i, ii) and after (iii, iv) acceptor photobleaching, indicated by arrow. Donor channel, i, iii; acceptor channel, ii, iv. The photobleached area is boxed in the range indicator (to the right of each image). (B) FRET in the various regions of interest (ROI) on the cell surface (overlay, i) can be quantified from changes in the intensity in the Donor and Acceptor channels (ii). Acceptor photobleach (red) results in variable increase in Donor intensity (green; ROI 1, iii; ROI 1a, iv), while control ROIs 2 and 3 (v,vi) show little change in donor and acceptor intensities. (C) SEA induces clustering of HLA-DR1 on B cells. Energy Transfer (E%) values were plotted against acceptor Fluorescence ([A] pre-bleach) for untreated B cells (i), B cells treated with with SEH +/− EDTA (ii,iii), SEA in the presence of EDTA (iv) or B cells treated with 10 µg or 2.5 µg SEA (v,vi) for 45 minutes, here, at Donor∶Acceptor molar ratio 1∶4. Data from one of two experiments.

B cells were treated, stained and fixed as in Figure 3, and thin sections (80 nm) analyzed by transmission electron microscopy. (A) Images of untreated B cells (i) or B cells treated with 2.5 µg SEA (ii) showed varying degrees of clustering of HLA-DR1. arrow/asterisk, “clustered” gold beads; arrowhead, “unclustered” gold beads. (B) Nearest neighbor analysis of B cells treated with SEA (red) shows an increased number of closely apposed gold beads compared to B cells treated with no enterotoxin (yellow) or SEH (blue). SEA+EDTA (green) has an intermediate distribution, and these differences disappear when nearest neighbors are separated by large distances. ▴, separation at 75 nm. (C) B cells treated with SEA show a statistically significant increase in the fraction of beads that are within 75 nm of each other, when compared to cells that were untreated or treated with SEH or SEA+EDTA. Same color scheme as above. (D) SEA treatment does not mediate higher order aggregation of HLA-DR1. Equivalent stretches of B cell membranes of various groups were evaluated for number of instances of “higher order” clusters (clusters of >2 gold beads). There was an increase in the fraction of cells with higher order clusters on SEA treatment, but no dramatic change in the distribution profile between the groups. Data pooled from two experiments.