To investigate how the clustering of different B-cell subsets along PC1 might relate to selection for one or more specific gene segments, we next asked how use of specific VH gene segments contributed to PC1 of productive rearrangements (Fig

To investigate how the clustering of different B-cell subsets along PC1 might relate to selection for one or more specific gene segments, we next asked how use of specific VH gene segments contributed to PC1 of productive rearrangements (Fig. in FO subsets was significantly narrower (= 1.8 10?9), suggesting selection against extreme-length CDR3s during maturation from pre-B to FO cells. Meanwhile, MZ CDR3s were, on average, 12.7 2.8 aa long, or about one-half of an amino acid shorter than in either pre-B or FO cells (= Rabbit Polyclonal to CELSR3 3.1 10?8 and = 8.0 10?10, respectively), consistent with differences between MZ and FO cells in the rat model (11) and, again, significantly narrower than in pre-B cells (= 3.8 10?6). Open in a separate window Fig. 1. CDR3 length and charge distributions for productive and nonproductive rearrangements in pre-B, FO, and MZ Gastrodenol cells and source of differences in productive rearrangements. (= 3.1e?8) and between MZ and FO subsets (= 8.0 10?10). (= 3.2 10?8) and between pre-B and MZ subsets (= 8.4 10?7). ( 0.14 for pairwise comparisons), but the length distributions were wider for nonproductive rearrangements than for productive rearrangements (= 5.1 10?9 to 6.2 10?11 for pairwise comparisons between nonproductive and productive rearrangements within each subset; compare Fig. 1vs. = 4.6 10?24C5.7 10?17), consistent with previous reports (3, 14). Among productive rearrangements, CDR3s from FO and MZ cells were, on average, 0.13C0.15 unit more negative than those from pre-B cells (= 3.2 10?8 and 8.4 10?7, respectively; Fig. 1= 6.8 10?6C3.0 10?14), consistent with selection against extremes of charge in addition to selection for more negative charge. Among productive rearrangements, the distribution of charges in MZ and FO cells was narrower than in pre-B cells (= 2.4 10?11 and 3.5 10?3, respectively; Fig. 1and axis and PC2 as the axis makes it possible to visualize variability of samples across many variables in a single scatterplot. PCA of VDJ use showed a clear separation between pre-B, FO, and MZ subsets (Fig. 2test confirmed that all four of these subsets differed significantly from each other in VH use ( 2.0 10?4). These four subsets also differed in D use, albeit more weakly ( 8.8 10?3), but not in JH use. B-Cell Maturation Selects for Specific VH Gene Segments. In PCA of VH use, each principal component summarizes the use of multiple different VH gene segments. To investigate how the clustering of different B-cell subsets along PC1 might relate to selection for one or more specific gene segments, we next asked how use of specific VH gene segments contributed to PC1 of productive rearrangements (Fig. 2contribute the same three residues to CDR3 (the initial C-A-R). Thus, the observed pattern does not result from VH gene segments contribution to CDR3, consistent with our finding that VH gene segments do not explain the differences between subsets in CDR3 length and charge. We therefore tested for systematic differences in coding sequence upstream of CDR3. Indeed, alignment of all VH gene segments showed systematic bias in amino acid use in two regions: in framework 1 (FR1) at VH amino acid positions 10C13 and just distal to CDR1 at positions 37C39 (Fig. 4and Fig. S4). Analysis of nucleotide sequence revealed no additional regions. We investigated these two regions in detail. Four of the top five VH gene segments that were used more frequently in MZ cells had the amino acid sequence E-L-V-K at positions 10C13, compared with only one of the top five VH gene segments that were used more Gastrodenol frequently in pre-B cells (Fig. 4and and and ?and55). How could a loose elbow promote survival? B-cell selection and maturation depend on signaling through the antigen receptor (1). Amino acid substitution experiments have shown that Gastrodenol disrupting the ball-and-socket joint can markedly decrease signaling while having only a minimal effect on antigen binding (24). Antigen binding is usually thought to cause the elbow to bend, affecting the initiation of signaling (25, 26). However, precisely how.