TNFRSF13B polymorphisms counter microbial adaptation to enteric IgA
Jeffrey L. Platt, Mayara Garcia de Mattos Barbosa, Daniel Huynh, Adam R. Lefferts, Juhi Katta, Cyra Kharas, Peter Freddolino, Christine M. Bassis, Christiane Wobus, Raif Geha, Richard Bram, Gabriel Nunez, Nobuhiko Kamada, Marilia Cascalho
Jeffrey L. Platt, Mayara Garcia de Mattos Barbosa, Daniel Huynh, Adam R. Lefferts, Juhi Katta, Cyra Kharas, Peter Freddolino, Christine M. Bassis, Christiane Wobus, Raif Geha, Richard Bram, Gabriel Nunez, Nobuhiko Kamada, Marilia Cascalho
View: Text | PDF
Research Article Immunology Microbiology

TNFRSF13B polymorphisms counter microbial adaptation to enteric IgA

Abstract

TNFRSF13B encodes the transmembrane activator and CAML interactor (TACI) receptor, which drives plasma cell differentiation. Although TNFRSF13B supports host defense, dominant-negative TNFRSF13B alleles are common in humans and other species and only rarely associate with disease. We reasoned that the high frequency of disruptive TNFRSF13B alleles reflects balancing selection, the loss of function conferring advantage in some settings. Testing that concept, we investigated how a common human dominant-negative variant, TNFRSF13B A181E, imparts resistance to enteric pathogens. Mice engineered to express mono- or biallelic A144E variants of tnrsf13B, corresponding to A181E, exhibited a striking resistance to pathogenicity and transmission of Citrobacter rodentium, a murine pathogen that models enterohemorrhagic Escherichia coli, and resistance was principally owed to natural IgA deficiency in the intestine. In WT mice with gut IgA and in mutant mice reconstituted with enteric IgA obtained from WT mice, IgA induces LEE expression of encoded virulence genes, which confer pathogenicity and transmission. Taken together, our results show that C. rodentium and most likely other enteric organisms appropriated binding of otherwise protective antibodies to signal induction of the virulence program. Additionally, the high prevalence of TNFRSF13B dominant-negative variants reflects balancing selection.

Authors

Jeffrey L. Platt, Mayara Garcia de Mattos Barbosa, Daniel Huynh, Adam R. Lefferts, Juhi Katta, Cyra Kharas, Peter Freddolino, Christine M. Bassis, Christiane Wobus, Raif Geha, Richard Bram, Gabriel Nunez, Nobuhiko Kamada, Marilia Cascalho

×

Figure 4

IgA sequences from sorted C. rodentium–bound IgA+ B cells of C57BL/6 and Tnfrsf13b mutant mice have distinct properties and determine C. rodentium spreading.

Options: View larger image (or click on image) Download as PowerPoint
IgA sequences from sorted C. rodentium–bound IgA+ B cells of C57BL/6 and...
Single IgA+ GFP-C. rodentium–bound B cells were sorted from the Peyer’s patches of infected C57BL/6, A144E/WT, or A144E/A144E mice 14 days following primary infection (A–C) (See also Supplemental Figure 5 and Supplemental Tables 1–6). IgA H+L sequences were obtained from cDNA by PCR followed by Sanger sequencing or by next generation sequencing of barcoded single cell barcoded cDNA libraries. (A) Graphs compare the lengths of the N regions (number of nts) of CDR3 regions (in amino acids) of HC IgA sequences obtained from C57BL/6 A144E/WT or A144E/A144E mice. The Kruskal-Wallis test followed by Dunn’s multiple comparisons test yielded P < 0.0001 in comparison of the N region lengths in mutant mice with those in WT mice. (B) Graphs depict the frequencies (%) of mutated nts in VH or Vκ exons relative to their closest germline in HC sequences obtained from C57BL/6, A144E/WT, or A144E/A144E mice. The Kruskal-Wallis test followed by Dunn’s multiple comparisons test yielded P < 0.0001 in comparison of the N region lengths in mutant mice with those in WT mice, and yielded P = 0.0094 or P = 0.0195 in comparisons of the frequencies of VH gene mutations in A144E/A144E mutants or A144E/WT mice, to those in WT mice, respectively. (C) Contingency analysis of frequency of germline C. rodentium binding IgA, VH, and Vκ sequences obtained from A144E/WT or A144E/A144E mice. The χ2 test yielded a P = 0.016, indicating that germline sequences are rarer in A144E/A144E mice compared with A144E/WT mice. (D) Schematic of experimental protocol. Founder mice are either C57BL/6, Tnfrsf13b mutant mice, or PIgR-KO mice. Two founder mice were cohoused with 3 C57BL/6 mice/cage. Only founders were inoculated with low virulence 108 C. rodentium. CFUs were counted on not primarily infected mice 7 days after infection. (E) Graph depicts CFU/g feces in mice not primarily infected according to the strain of the inoculated founders (depicted in the x axis). Comparisons were by Kruskal-Wallis tests followed by Dunn’s multiple comparison tests. The Kruskal-Wallis test yielded P < 0.0001; Dunn’s multiple comparison tests to WT yielded *P = 0.0111, **P = 0.0060, ***P = 0.0004, ****P < 0.0001.