Factors affecting human T cell engraftment, trafficking, and associated xenogeneic graft-vs-host disease in NOD/SCID beta2mnull mice

Abstract

Objective: Graft-vs-host disease (GVHD) is the major cause of morbidity and mortality after allogeneic hematopoietic stem cell transplantation. Models of immunodeficient mice that consistently and efficiently reconstitute with xenoreactive human T cells would be a valuable tool for the in vivo study of GVHD, as well as other human immune responses.

Materials and methods: We developed a consistent and sensitive model of human GVHD by retro-orbitally injecting purified human T cells into sublethally irradiated nonobese diabetic/severe combined immunodeficient (NOD/SCID)-beta2m(null) recipients. In addition, we characterized for the first time the trafficking patterns and expansion profiles of xenoreactive human T cells in NOD/SCID-beta2m(null) recipients using in vivo bioluminescence imaging.

Results: All NOD/SCID-beta2m(null) mice conditioned with 300 cGy total body irradiation and injected with 1 x 10(7) human T cells exhibited human T-cell engraftment, activation, and expansion, with infiltration of multiple target tissues and a subsequent >20% loss of pretransplantation body weight. Importantly, histological examination of the GVHD target tissues revealed changes consistent with human GVHD. Furthermore, we also showed by in vivo bioluminescence imaging that development of lethal GVHD in the NOD/SCID-beta2m(null) recipients was dependent upon the initial retention and early expansion of human T cells in the retro-orbital sinus cavity.

Conclusion: Our NOD/SCID-beta2m(null) mouse model provides a system to study the pathophysiology of acute GVHD induced by human T cells and aids in development of more effective therapies for human GVHD.

Figure 1

Human T cell engraftment and development of lethal X-GVHD in NOD/SCID-β2m^null mice. NOD/SCID-β2m^null mice (n = 59) were sublethally irradiated with 250 cGy of TBI and injected through the lateral tail vein (i.v.) or retro-orbital (r.o.) venous plexus with 5 × 10⁶ or 10 × 10⁶ naive purified huT cells the following day. (A) Human T cell engraftment. Peripheral blood samples collected from mice that developed (GVHD; open circles) or did not develop (No GVHD; closed circles) lethal GVHD were labeled with mAbs specific for human and murine CD45 and analyzed by flow cytometry. The proportion of human cells in the peripheral blood was calculated as follows: % huCD45⁺ = [huCD45⁺ / (huCD45⁺ + mCD45⁺)] × 100%. Maximum huT cell engraftment is shown for each mouse and typically occurred between 2 and 3 weeks after huT injection. (B) Kaplan-Meier survival analysis. Lethal X-GVHD was only observed when 10 × 10⁶ huT cell were injected r.o. Results are pooled from four separate experiments with different donors.

Figure 2

Characterization of mice that developed or did not develop lethal X-GVHD. NOD/SCID-β2m^null mice (n=34) were sublethally irradiated with 250cGy of TBI and injected r.o. with 10 × 10⁶ naive purified huT cells. Mice that developed lethal X-GVHD (n = 20; open bars) were then compared to animals that failed to develop lethal X-GVHD (n=14; closed bars). (A) Percent change in pretransplant body weight. (B) Kinetics of huT cell engraftment in the peripheral blood. The proportion of human cells in the peripheral blood was determined by flow cytometry as described in the legend to Fig. 1. (C) CD4⁺/CD8⁺ huT cell ratio. The percentage of human CD45⁺CD3⁺ T cells in the peripheral blood that expressed CD4 or CD8 was determined by flow cytometry. (D) Expression of T cell activation markers. The percentage of human CD45⁺CD3⁺ T cells in the peripheral blood that expressed CD25, CD30, or CD69 at 10 or 20 days after huT cell injection was determined by flow cytometry. Normal human PBMCs were used as controls in all analyses to establish consistent gating criteria for CD25, CD30, and CD69. (E) Human IFN-γ serum levels. Serum was obtained from peripheral blood samples collected at the indicated times after huT cell injection and human IFN-γ levels were determined using a cytometric bead assay (BD Biosciences). Values of p indicate statistically significant differences between the two groups. NS indicates differences were not statistically significant.

Figure 3

In vivo BLI of huT cells following i.v. or r.o. administration. NOD/SCID-β2m^null mice were sublethally irradiated with 250 cGy of TBI and injected through the lateral tail vein (i.v.; n = 5) or retro-orbital (r.o.; n = 5) venous plexus with 10 × 10⁶ CBRluc/egfp transduced huT (huT^CBRluc/EGFP) cells the following day. (A) Kaplan-Meier survival analysis. Lethal X-GVHD was only observed when huT^CBRluc/EGFP cells were injected r.o. (B) Percent change in pretransplant body weight. (C) Kinetics of huT cell engraftment in the peripheral blood. The proportion of human cells in the peripheral blood was determined by flow cytometry as described in the legend to Fig. 1. (D–G) BLI of luciferase activity. The trafficking and expansion of huT^CBRluc/EGFP cells following i.v. or r.o. injection was assessed in the dorsal (D) and ventral positions (E) by BLI (1h: 1 hour, 1d: 1 day, 4d: 4 days, 7d: 7 days, 14d: 14 days, 21d: 21 days post huT injection). One representative animal for each group is shown over time. Photon flux is indicated in the color scale bar. Overall quantifications of emitted photons in the dorsal (F) and ventral (G) positions are shown graphically.

Figure 4

Human T cell infiltration and phenotype in tissues of NOD/SCID-β2m^null mice that develop lethal GVHD. NOD/SCID-β2m^null mice (n=34) were sublethally irradiated with 250cGy of TBI and injected r.o. with 10 × 10⁶ naive purified huT cells. When possible, moribund animals with X-GVHD were euthanized and portions of blood, spleen, liver, lung, kidney and bone marrow (BM) were analyzed by flow cytometry. (A) Representative FACS analysis of blood, spleen, and liver harvested from a NOD/SCID-β2m^null mouse that developed lethal X-GVHD. In all FACS analyses, normal human PBMCs were used as controls to establish consistent gating criteria. (B) Human T cell subset composition. The percentage of human CD45⁺CD3⁺ T cells in the indicated tissues that expressed CD4 or CD8 was determined by flow cytometry. (C) Expression of T cell activation markers. The percentage of human CD45⁺CD3⁺ T cells in the indicated tissues that expressed CD25, CD30, or CD69 was determined by flow cytometry

Figure 5

Retention of X-GVHD-inducing potential following ex vivo activation of human T cells. NOD/SCID-β2m^null mice were sublethally irradiated with 300 cGy of TBI and left untreated (control; n = 4) or injected r.o. with 10 × 10⁶ naive (n = 8; open bars) 4-day activated (n = 9; closed bars), or 8-day activated (n = 5; hatched bars) huT cells. Activated T cells were prepared by incubating human PBMCs with CD3/CD28 beads at a ratio of 3 beads per cell in the presence of IL-2 (50 U/mL) for 4 or 8 days. (A) Kaplan-Meier survival analysis. Mice injected with naive or activated huT cells displayed similar overall survivals that were significantly decreased compared to the untreated control. (B) Kinetics of huT cell engraftment in the peripheral blood. The proportion of huT cells in the peripheral blood was determined by flow cytometry as described in the legend to Fig. 1. (C) Percent change in pretransplant body weight. (D) CD4⁺/CD8⁺ huT cell ratio. The percentage of human CD45⁺CD3⁺ T cells in the peripheral blood that expressed CD4 or CD8 was determined by flow cytometry. (E) Human IFN-γ serum levels. Serum was obtained from peripheral blood samples collected at the indicated times after huT cell injection and human IFN-γ levels were determined using a cytometric bead assay (BD Biosciences). (F) Kinetics of CD25 expression on naive and activated huT cells. The percentage of human CD45⁺CD3⁺ T cells in the peripheral blood that expressed CD25 was determined by flow cytometry. Normal human PBMCs were used as controls in all analyses to establish consistent gating criteria for CD25. Values of p in (C–F) indicate statistically significant differences between the naive and activated huT cell groups. NS indicates differences were not statistically significant.

Figure 6

Tissue infiltration of naive and activated human T cells in NOD/SCID-β2m^null mice. NOD/SCID-β2m^null mice were sublethally irradiated with 300 cGy of TBI and left untreated (control; n = 4) or injected r.o. with 10 × 10⁶ naive (n = 8; open bars) 4-day activated (n = 9; closed bars), or 8-day activated (n = 5; hatched bars) huT cells. Activated T cells were prepared by incubating human PBMCs with CD3/CD28 beads at a ratio of 3 beads per cell in the presence of IL-2 (50 U/mL) for 4 or 8 days. Moribund animals with X-GVHD were euthanized and portions of blood, spleen, liver, lung, and kidney were analyzed by flow cytometry. (A) Human T cell infiltration of different tissues. The percentage of human CD45⁺CD3⁺ T cells in the indicated tissues was determined by flow cytometry. (B) CD4⁺/CD8⁺ huT cell ratio. The percentage of human CD45⁺CD3⁺ T cells in the peripheral blood that expressed CD4 or CD8 was determined by flow cytometry. (C,D) Expression of T cell activation markers. The percentage of human CD45⁺CD3⁺ T cells in the indicated tissues that expressed CD25 and CD30 was determined by flow cytometry. Normal human PBMCs were used as controls in all analyses to establish consistent gating criteria for CD25 and CD30. Values of p in indicate statistically significant differences between the naive and activated huT cell groups. NS indicates differences were not statistically significant.

Figure 7

Histopathologic and immunohistochemical analysis of X-GVHD in NOD/SCID-β2m^null mice. NOD/SCID-β2m^null mice were sublethally irradiated with 300 cGy of TBI and injected r.o. with 10 × 10⁶ naive or 10 × 10⁶ activated huT cells. Moribund animals with X-GVHD were euthanized and portions of skin, liver, small intestine, colon, lung, kidney, salivary gland and spleen were saved for histopathologic and immunohistochemical analyses. (A) Liver from a mouse that did not show clinical or histological signs of X-GVHD following injection of huT. This liver shows an unremarkable portal tract and healthy hepatocytes. (B,C) X-GVHD in the liver characterized by (B) dense lymphocytic infiltrates in the portal tract and hepatocyte apoptosis, and (C) endotheliitis (head arrows). (D) X-GVHD in the lung showing perivascular and interstitial lymphocytic infiltration. (E) X-GVHD in the skin showing apoptosis and basal vacuolar damage of the keratinocytes. (F) X-GVHD in the colon showing apoptosis in the mucosa. (G) X-GVHD in the kidney showing prominent perivascular lymphocytic infiltration. (H) An example of immunohistochemical stains showing huT cells in X-GVHD expressing human CD45 in the spleen, (I,J) portal lymphocytes in liver X-GVHD consisting of huT lymphocytes expressing either human CD4 (I) or human CD8 (J). Slides in panels B–G were obtained from mice that received activated huT cells. Slides in panels A and H–J were obtained from mice that received naive huT cells. There were no significant differences in the pathologic damage originated by naive or activated huT cells. Black arrows indicate apoptotic cells.

Figure 8

X-GVHD score. NOD/SCID-β2m^null mice were sublethally irradiated with 300 cGy of TBI and injected r.o. with 10 × 10⁶ naive (n = 8; open bars) 4-day activated (n = 9; closed bars), or 8-day activated (n = 5; hatched bars) huT cells. Moribund animals with X-GVHD were euthanized and portions of skin, liver, small intestine, colon, lung, kidney, and salivary gland were saved for histopathologic analysis. Tissues were scored for pathologic damage as described in the Materials and Methods. NOD/SCID-β2m^null mice that received 250 cGy of TBI and were injected r.o. with 10 × 10⁶ naive huT cells but did not develop lethal X-GVHD served as no X-GVHD controls. Data represent mean ± SD.