Human tonsil implants xenotransplanted in SCID mice display broad lymphocytic diversity and cellular activation profile similar to those in the original lymphoid organ.

Background: Models consisting of human immune cells in suspension transferred to severe combined immune deficient (SCID) mice have been invaluable for studying immune response, autoimmunity, and lymphomagenesis. The dissemination of human cells within the mouse body hampers immune functionality with time and favorites the development of human graft vs. mouse host (GvH) disease. To circumvent these limitations we surgically implanted tonsil pieces subcutaneously in SCID animals (hu-ton-SCID mice). Recall humoral responses was elicited and animals did not suffer from signs of GvH disease. A detailed cell subset and cell activation analysis of implants has not yet been reported.

Methods: Implants from 86 hu-ton-SCID mice were evaluated by immunohistochemistry and flow cytometry analyses to assess human lymphoid cell subpopulation surviving with time after implantation, and to evaluate status of human cell activation.

Results: B cells persist over 3 months in implants. The proportion of class and type-specific Ig+ cells varied between implants, but as a whole IgG+ cells were more abundant than IgA+, and IgM+ cells, and kappa+ cells predominated over lambda+ cells. The mean proportions of these cells resemble those in the original tonsil. Fine analysis of CD19+ B cells demonstrated no expansion of activated (CD5+, CD23+, CD69+) B cells in implants compared with tonsils, and a decrease of CD19+CD77+ B cells corresponding to a centroblastic phenotype, which is consistent with the disappearance of follicular structure in implants. Double positive CD20+CD27+ memory B cells were detected in implants by immunohistochemistry. T cell CD4+CD8-/CD4-CD8+ ratios were about 4 in implants, that is similar to those in tonsils, and there was no expansion of CD3+CD4+CD8+ and of CD3+CD4-CD8- T-cell subpopulations. T cells activation markers (CD25, CD69) were similarly expressed in implants and tonsils, and implants contained cells with a memory T cell phenotype (CD45RO). Finally cells within implants depicted a low rate of proliferation when assessed by Ki-67 expression levels.

Conclusions: Compared with original tonsils, tonsil implants in hu-ton-SCID mice lose the germinal center architecture, which is correlated with the decrease of CD77+ B cells, but conserve T and B cell subpopulation diversity, notably memory cells. In addition, implant T and B cells are not differently activated when compared with those in original tonsils and do not proliferate extensively. These observations indicate indirectly absence of GvH reaction at the cellular level in this model. Collectively, the detailed implant cellular characterization in the hu-ton-SCID model provides a strong rationale for the use of this model in the study of human recall antibody response.