TLR-induced reorganization of the IgM-BCR complex regulates B-1 cell responses to infections

Neonatally-developing, self-reactive B-1 cells generate steady levels natural antibodies throughout life. They can, however, also rapidly respond to infections with increased local antibody production. The mechanisms regulating these two seemingly very distinct functions are poorly understood, but have been linked to expression of CD5, an inhibitor of BCR-signaling. Here we demonstrate that TLR-mediated activation of CD5+ B-1 cells induced the rapid reorganization of the IgM-BCR complex, leading to the eventual loss of CD5 expression, and a concomitant increase in BCR-downstream signaling, both in vitro and in vivo after infections with influenza virus and Salmonella typhimurium. Both, initial CD5 expression and TLR-mediated stimulation, were required for the differentiation of B-1 cells to IgM-producing plasmablasts after infections. Thus, TLR-mediated signals support participation of B-1 cells in immune defense via BCR-complex reorganization.


173
2C-D). Surface CD5 levels were decreased first, by 1.5 days of culture, while cd5 mRNA was not 174 reduced until 2 days after culture onset ( Fig. 2C-D). The stimulated cells began secreting IgM 175 before CD5 levels were reduced, but the increase in IgM secretion was more pronounced after at 176 least 2 days of stimulation compared to the earlier time points (Fig. 2E).

177
A number of control studies ensured that the reduced frequencies of CD5+ B-1 cells in the 178 cultures were due to a loss of surface expression and not to an expansion of small numbers (< 179 5%) CD5-cells that might have contaminated the cultures at onset. First, we separated by FACS 180 CD5+ and CD5-B-1 cells from the body cavities to very high purities, and then cultured either 181 100%, 99% or 95% sorted CD5+ cells to which we added 0%, 1% and 5% CD5-cells, respectively.

182
The percentage of CD5+ and CD5-cells after 3 days of culture was unaffected by the initial 183 composition of the culture wells (Fig. 2F). When cultures with 100% and 95% CD5+ at onset were 184 compared on day 3 of culture, there was no significant difference in either CD5 MFI or in the 185 percent of CD5+ and CD5-cells (Fig. 2G). Thus, small percentages of CD5-B-1 cells at culture with and without LPS stimulation. To ensure that the two populations were exposed to the same 190 culture conditions, CD5+ and CD5-B-1 cells were sorted, labeled with different proliferation dyes, 191 and cultured together (Fig. 2H). Compared to B-1 cells that expressed CD5 on day 0, CD5-cells 192 survived less well after stimulation (Fig. 2I)

212
Finally, we examined whether phosphatidylcholine (PTC)-binding B-1 cells can lose CD5 213 surface expression after TLR-stimulation. PTC is a well-known specificity of a large subset of indistinguishable from the proposed B-1 cell "sister" population, the CD5-"B-1b" cells.

225
To test this hypothesis, neonatal B-1 cell chimeras generated with FACS-purified CD5+

257
Neonatal chimeric mice generated with varying ratios of FACS-purified CD5+ and/or CD5-258 B-1 cells, as described above were orally infected with Salmonella typhimurium (Fig. 5A). Again primarily B-1b cells formed fewer IgM secreting cells, although more than the uninfected chimeras ASC did not increase significantly (Fig. 5F). Instead, we found OmpD-specific IgM secretion by 271 B-2 derived plasmablasts. Of note, the phenotype of B-2-derived plasmablasts is indistinguishable 272 from that of "B-1b" cells (CD19low CD45lo IgM+ CD43+ (Fig. 5F)), and thus only identifiable using 273 a lineage-marking approach, such as used here.

274
Together these findings demonstrate that in response to both bacterial and viral infections,

291
As expected, stimulation of CD5+ B-1 cells with anti-IgM failed to induce their proliferation 292 but did induce proliferation of B-2 cells. In contrast, stimulation with CpG (TLR9-ligands) induced stimulation was explained by the PLA data, which showed the maintenance and even increase in on B-1 cells inhibits BCR-signaling by reducing steady-state IgM-CD19 interactions and likely also 299 by initiating instead interactions between CD5 and CD19.

300
In contrast to direct stimulation of the IgM-BCR, CpG stimulation led to changes in the

362
As we showed that it is the CD5+ B-1 cell population that initially responded to infections 363 with influenza virus and with S. typhimurium, the latter previously identified as an exclusive "B-

372
The data presented here are inconsistent with models that regard B-1 cell responses as derived from TLR-activated CD5+ B-1 cells. Since no other clearly subset-defining differences 377 between B-1a and B-1b cells have been reported to-date, it may be more prudent to describe B-      bacteria were centrifuged for 20 minutes at 6,000-8,000 rcf at 4°C after concentration was 459 determined by spectrophotometer reading at OD600. Bacterial pellets were resuspended in mouse 460 drinking water to a concentration of 10 9 CFU/ml. Water was provided to mice ad lib.

549
After FACS sorting, cells were resuspended in RPMI and rested for at least two hours before 550 designated stimuli were added to culture media. Stimulated and unstimulated cells were cultured 551 for 5 minutes, and 24 and 48 hours prior to PLA. PLA was performed as previously described 57 .

552
In brief: For PLA-probes against specific targets, the following unlabeled Abs were used: anti-IgM