Investigation of the binding profile and specificity of monoclonal IgA to microbiota communities under steady state and inflammatory conditions

Kabbert, Johanna; Pabst, Oliver (Thesis advisor); Blank, Lars M. (Thesis advisor)

Aachen : RWTH Aachen University (2021)
Dissertation / PhD Thesis

Dissertation, RWTH Aachen University, 2021


Immunoglobulin A (IgA) is the dominant antibody isotype secreted into the intestinal lumen and plays a key role in gut homeostasis. IgA binds to luminal pathogens, toxins and members of the gut microbiota, thereby contributing to intestinal host immune-defense as well as maintaining a stable microbiota composition. In addition to IgA, in the human gut low levels of IgG are also present, which markedly increase during inflammation. Considering the vast changes of microbial consortia in the gut, it remains elusive how the host can generate and regulate relevant antibody-responses in this highly dynamic setting. Similarly, the specificity of human IgA and IgG to the microbiota and the mechanisms underlying antibody-microbiota interactions are still largely unknown. Here, we examined the binding capacity and specificity to the intestinal microbiota of a large set of monoclonal antibodies (mAbs) generated from intestinal IgA+ and IgG+ plasma cells (PCs) derived from human healthy donors (HD) or Crohn’s disease (CD) patients by bacterial flow cytometry. Our study revealed that a high frequency of IgA mAbs from both healthy and inflamed gut exhibited high microbiota-binding. Interestingly, we found similarly high microbiota-binding capacities among human IgG mAbs. To further determine the bacterial-binding profiles of microbiota-reactive human IgA mAbs, we sorted IgA-bound and unbound fecal bacteria and characterized their composition by 16S rRNA sequencing. Some bacterial taxa were commonly targeted by several mAbs, whereas others were selectively bound by single mAbs only. Notably, all high microbiota-reactive HD and CD derived IgA mAbs bound phylogenetically distant bacteria but showed in addition to common bacterial specificities unique binding profiles. We refer to this phenomenon as "cross-species reactivity". Importantly, microbiota cross-species reactivity did not correlate with polyreactivity (i.e. binding to structurally non-related antigens) but was crucially dependent on accumulated somatic mutations. Accordingly, cross-species reactive IgA mAbs carried frequent somatic mutations and the majority of mAbs showed substantial loss of microbiota binding in their germ-line configuration. In addition, germ-line reversion did not cause these mAbs to gain polyreactivity. This strongly suggests that microbiota-reactive mAbs do not derive from originally polyreactive antibodies. We therefore propose that "early" polyreactive antibodies may become supplanted by highly mutated, affinity-matured antibodies during aging. The high numbers of somatic mutations suggest that cross-species reactive antibodies may be the result of ongoing selection during multiple rounds of affinity-maturation and most likely depend on T cell interactions. Notably, cross-species reactive antibodies did not enrich all members of a targeted species, implying that cross-species reactive IgA may target distinct genetic sub-strains within bacterial species or bacteria in a particular growth-state. Collectively, our data suggest that a system of affinity-matured, cross-species reactive antibodies is one dominant mechanism of IgA-microbiota interactions in the human gut. We propose that cross-species reactivity of IgA facilitates broad but specific binding to phylogenetically distant gut bacteria to efficiently interact with the microbiota in both the healthy and inflamed gut. The continuous exposure to varied but structurally-similar antigens might enable the selection for cross-species reactive IgA responses. Prospectively, identifying the microbiota-targeting profiles of IgA derived from healthy or inflamed gut may allow to develop anti-microbial mAbs as diagnostic or therapeutic tools. Another aspect of intestinal immunity relates to the generation of IgA-producing PCs. However, how IgA responses to the microbiota are generated under steady state conditions is largely unknown. Also, it is uncertain which inductive compartments of the gut associated lymphoid tissue predominantly contribute to the generation and differentiation of intestinal IgA-secreting PCs. Additionally, there is still limited knowledge about the contribution of newly generated IgA+ PCs to the prevailing intestinal PC pool in homeostasis. To investigate these questions, we have established an in vivo mouse model to monitor the differentiation and migration kinetics of activated B cells. In this model, the local injection of 4-Hydroxy-Tamoxifen into a single intestinal Peyer’s patch (PP) leads to the irreversible eYFP-labeling of activated B cells, which allows to track activated B cells and all their progeny. Notably, in steady state and in the absence of neo-antigen challenge, we found a surprisingly high frequency of activated eYFP+ B cells in PPs. Our data further demonstrate that a marked frequency of eYFP+ B cells originating from a single PP re-circulate through different PPs, with only transient homing to the mesenteric lymph nodes. This suggests that B cells activated in a single PP may undergo affinity-maturation in multiple PPs leading to the generation of highly specific IgA+ PCs. Moreover, eYFP+ B cells generated in a single PP had the capacity to transition into IgA+ PCs and contributed to the PC pool in the small intestinal lamina propria up to 90 days after induction. In the future, our established model will allow to study the clonal overlap of PC-repertoires in different gut compartments and to identify functional differences in IgA+ PCs generated under healthy or inflammatory conditions. Therefore, our method of local B cell labeling in single PPs, may provide further understanding of the underlying mechanisms leading to cross-species reactive IgA responses to the microbiota.


  • Department of Biology [160000]
  • Chair of Applied Microbiology [161710]
  • [526000-2]