Altered B-Cell Expansion and Maturation in Draining Mesenteric Lymph Nodes of Inflamed Gut in Crohn’s Disease

Recent studies have investigated the role of B-cell responses in ulcerative colitis, which exclusively affects the colon, whereas data in Crohn’s disease (CD), which mainly affects the terminal ileum, are insufficient.1, 2, 3, 4 Granuloma formation within the thickened, inflamed mesentery of patients with CD, however, is associated with significantly worse outcome,5, 6, 7 and microstructural analysis has suggested increased numbers of B cells in CD mesentery.⁸ Although a healthy orchestrated mucosal immune response involves B-cell maturation and development of IgA-secreting plasma cells directed toward more invasive strains of bacteria, anticommensal IgG can be found in patients with CD, which usually are absent in healthy controls.9, 10, 11 Although a single-cell sequencing study in therapy-refractory CD found an intestinal immune cell signature in inflamed areas that included IgG⁺ plasmablasts, rather than IgA⁺ plasmablasts,¹² B-cell receptor (BCR) sequencing in CD has been limited to peripheral blood to date.¹³ Anticommensal IgG antibodies, however, can be transported across mucosal barriers by a neonatal fragmented crystallizable region (Fc) receptor and sustain inflammation in ulcerative colitis, with data lacking for CD.1, 2, 3^,14, 15, 16, 17 To investigate the local B-cell response in CD we therefore characterized paired samples of draining mesenteric lymph nodes (MLNs) of affected and adjacent healthy small intestinal segments (Figure 1A, Supplementary Table 1, Supplementary Methods, and Supplementary Figure 1). Fractions of CD45⁺ leukocytes were higher in affected areas (Figure 1B), macrophages were negligeable (Figure 1C), T cells were reduced (Figure 1D), and CD19⁺ B cell fractions were expanded in affected MLNs (Figure 1E). Further characterization of CD45⁺CD19⁺ cells¹⁸ within MLNs showed that IgD⁺CD27⁺ marginal zone B-cell fractions¹⁹ were comparable between affected and healthy MLNs (Figure 1F), IgD^-CD27^- double-negative B cells were more abundant within CD45⁺CD19⁺ B cells of affected MLNs (Figure 1G), CD38⁺ plasmablasts numbers were increased and CD38^- memory B cells were reduced within the CD45⁺CD19⁺CD27⁺IgD^- B-cell fraction (Figure 1H and I), overall suggesting ongoing antigenic stimulation within affected MLNs.^20,21

Figure 1.

B cells expand within affected draining MLNs. (A–I) Characterization within paired draining MLNs of affected and healthy small intestinal segments. (A) Schematic depiction of sample collection. (B–E) Eighteen patients were included for basic immunologic phenotyping for (B) CD45⁺ cells, (C) CD45⁺CD14⁺ cells, (D) CD45⁺CD3⁺ cells, and (E) CD45⁺CD19⁺ cells. (F–I) A more detailed B-cell subset analysis of 9 patients showing (F) IgD⁺CD27⁺ marginal zone B cells, (G) IgD^-CD27^- double-negative B cells of CD45⁺CD19⁺ B cells, (H) CD38⁺ cells, and (I) CD38^- cells of CD45⁺CD19⁺CD27⁺IgD^- B cells. (J and K) Differences in GC (J) abundance and (K) size per high-power field (HPF). (L–Q) Immunohistochemistry (IHC) staining: representative image of a draining MLN of (L) healthy and (M) affected small intestine stained for CD19. Purple arrow and yellow arrow indicate GCs. MLN of a (N) healthy small intestine and (O) affected segment stained for Ki67. MLN of a (P) healthy small intestine and (Q) affected segment stained for Bcl6. (B–K) A paired Student t test was used for analysis. ∗P < .05, ∗∗P < .01, and ∗∗∗P < .001.

B cells mature within germinal centers (GCs), which are specific microanatomic structures within secondary lymphoid organs such as MLNs.²² Immunohistochemistry staining for CD19 showed that GC abundance was unchanged between affected and healthy MLNs (Figure 1J), but GCs in affected areas were significantly larger (Figure 1K). In addition, MLNs from healthy areas displayed GCs that appeared to be immature (Figure 1L), whereas affected MLNs often showed GCs with dark and light zones (Figure 1M). Because B-cell isotype switching within GCs is a T-cell–dependent process,^23,24 we next performed immunohistochemistry staining for Ki67, a marker indicating cell proliferation,²⁵ and Bcl6, a master regulator for T-follicular helper cells and expressed in class-switching B cells.²³ Ki67 and Bcl6 staining marked individual cells scattered within the MLNs, but the majority of healthy MLNs showed a lack of positive GCs (Figure 1N and P), whereas Ki67 and Bcl6 were highly positive within GCs of affected MLNs (Figure 1O and Q).

Thus, BCR sequencing was performed to investigate differences in class switching between MLNs (Figure 2A) and showed decreased use of Immunoglobulin Heavy Constant Alpha (IGHA) and Immunoglobulin Heavy Constant Epsilon (IGHE), with a concomitant significant increase in Immunoglobulin Heavy Constant Gamma 1/2 (IGHG1/2) in affected MLNs (Figure 2B). Although isotype analysis can provide important information about class switching, analysis of the BCR variable region provides more information about affinity maturation. Affinity maturation, a process in which B cells increase their antigen affinity and avidity, occurs via somatic hypermutation (SHM) within the GC’s dark zone.²⁶ Thus, mutation rates within complementary determining regions (CDRs) and framework regions (FWRs) of the BCR were compared and analyzed for silent mutations (ie, mutations that do not result in a change in amino acid sequences), and substitution mutations (ie, mutations that result in a change in amino acid sequence, and, therefore, Ig structure) (Figure 2C). The mean SHM frequency was significantly higher in IGHA as well as Immunoglobulin Heavy Constant Mu (IGHM) B cells (Figure 2C and D). This increased rate of SHM in affected MLNs was driven by replacement mutations in the CDRs as well as framework regions of IGHA B cells, and CDR mutations in IGHM B cells (Figure 2C and D). In line with this, diversity analysis indicated an increased diversity of the BCR in IGHG1/2 B cells whereas diversity was unchanged in IGHM or IGHA B cells (Figure 2E and F).

Figure 2.

B-cell–receptor sequencing indicates pathologic B-cell maturation. (A) Schematic depiction of the BCR sequencing strategy performed in 24 patients. (B) BCR isotype use in MLNs. (C) Rate of SHM in CDRs or framework regions (FWRs). (D) Rate of SHM is depicted as the ratio of affected/healthy MLNs. (E) Diversity analysis on BCR sequences of the variable regions. Richness refers to the abundance of unique clones in a repertoire, Simpson’s index assesses the probability of 2 randomly sampled reads belonging to the same clone, Shannon’s index is a measure of evenness. (F) Depiction of individual results and (E) as log of affected/healthy. ∗P < .05, ∗∗P < .01, and ∗∗∗P < .001. freq., frequency; IGHA, Immunoglobulin Heavy Constant Alpha; IGHD, Immunoglobulin Heavy Constant Delta; IGHE, Immunoglobulin Heavy Constant Epsilon; IGHG, Immunoglobulin Heavy Constant Gamma; IGHM, Immunoglobulin Heavy Constant Mu.

Overall, our results indicate ongoing class switching within draining MLNs of affected intestinal segments, with a shift toward IGHG1/2 BCRs. The lack of high SHM rates within IGHG1/2 BCRs, the difference between IGHA and IGHG1/2 BCRs in single MLNs, and increased diversity in IGHG1/2 BCRs suggests that many antigens do not result in long-lasting immunologic stimulation, and IGHA and IGHG1/2 responses may target different pathogens/commensals.