Alloantibodies produced by recipient B cells and plasma cells pose a significant risk for rejection after solid organ transplantation, including intestinal transplantation (ITx). Dynamic turnover of recipient B cells and plasma cells in intestinal allografts and their roles in mediating rejection have not been investigated. 
We analyzed the chimerism, phenotype and clonotypes of B cells and plasma cells in the allograft and peripheral blood of ITx recipients using multicolor flow cytometry and high-throughput BCR sequencing.
Serial analyses of ileal biopsies demonstrated eventual replacement of lamina propria donor B cells by the recipient, over periods of >1 year in some patients (Fig.1). Recipient B cell replacement rates were often faster than those of T cells. In patients with donor T cell macrochimerism (>4%) in blood, replacement of donor cells by recipient cells tended to occur slowly, over a period of several months. In contrast, in patients without T cell macrochimerism and with early rejection, recipient B cells and T cells rapidly replaced those of the donor in the graft. High levels (>10%) of donor B cell chimerism in blood were only observed in multivisceral transplant (MVTx) recipients and peak donor B cell blood chimerism was generally higher than that of donor T cells within the same patient.
Phenotypic analyses of lamina propria B cells on sequential biopsies of a 2-year-old recipient of an allograft from a 6-month-old donor revealed the early presence of naïve donor and recipient B cells in the graft that over time were largely replaced by both class-switched and non-switched memory B cells of donor and recipient origin. Recipient-derived surface IgG⁺ and IgA⁺ B cells appeared in the long-term graft of a patient who had had DSA but not in a DSA^- patient. Even when recipient B cells had almost fully replaced donor-derived B cells, donor-derived plasma cells (CD138⁺CD38⁺) persisted long-term (POD>900), demonstrating the existence of long-lived plasma cells in human intestine. The presence of recipient plasma cells in the graft may correlate with rejection, as they were only detected during rejection episodes. (Fig.2)
IgH V region sequencing on sorted recipient B cells from ileal specimens of a MVTx patient on Days 29, 32, 43, 57, 78, 105 and 141 post-Tx revealed clonal overlap, especially at later time points. The fraction of recipient B cells co-expressing CD45RB and CD69 in the ileal biopsies increased with time in 3 out of 3 patients. (Fig.3) CD45RB⁺CD69⁺ B cells may belong to a resident memory pool that contributes to tissue-based immunity (Weisel et al, in preparation). 
In summary, the appearance of recipient plasma cells and IgG⁺/IgA⁺ B cells in the graft correlates with rejection and DSA development, respectively. We hypothesize that graft-repopulating recipient B cell clones may expand and acquire a B resident memory phenotype, promoting organ-specific alloimmunity.