A critical role for the retinoic acid signaling pathway in the pathophysiology of gastrointestinal graft-versus-host disease

Abstract

Damage to the gastrointestinal tract during graft-versus-host disease (GVHD) is one of the major causes of morbidity and mortality in allogeneic hematopoietic stem cell transplant (HSCT) recipients. In the current study, we identified a critical role for the retinoic acid (RA) signaling pathway in the induction and propagation of gastrointestinal GVHD. The administration of exogenous RA significantly increased expression of the gut-homing molecules, CCR9 and α4β7, on donor T cells in mesenteric lymph nodes, and augmented the accumulation of proinflammatory CD4(+) and CD8(+) T cells within the gut mucosa, leading to a selective exacerbation of colonic GVHD and increased overall mortality. Conversely, depletion of RA in recipient mice by vitamin A deprivation resulted in a dramatic reduction of gut-homing molecule expression on donor T cells after HSCT. Significantly, absence of the RA receptor-α on donor T cells markedly attenuated the ability of these cells to cause lethal GVHD. This observation was attributable to a significant reduction in pathological damage within the colon. These findings identify an organ-specific role for RA in GVHD and provide evidence that blockade of the RA signaling pathway may represent a novel strategy for mitigating the severity of colonic GVHD.

Figures

Figure 1

Administration of exogenous RA selectively intensifies GVHD of the colon. Lethally irradiated Balb/c mice received a transplant of 10 × 106 B6 BM alone (○) or together with B6 spleen cells (adjusted to yield a dose of 0.6 × 106 T cells). Cohorts of mice that received adjunctive spleen cells were then treated by intraperitoneal injection with either DMSO (▪) or RA (●)(450 μg/mouse) every other day beginning on the day of transplantation until day 20. (A) Overall survival. Data are cumulative results of 3 experiments (n = 6 for BM control group, n = 12 for both DMSO- and RA-treated groups). (B) Mean percent initial body weight of recipient mice transplanted as in panel A over the first 35 days. (C) Lethally irradiated Balb/c mice were transplanted with B6 BM and spleen cells as above with the exception that the spleen cell dose was adjusted to yield a dose of 0.3 × 106 T cells. Groups were then treated with either DMSO or RA. Mice were killed 5 weeks after transplantation. Pathological damage in the tongue, liver, lung, and colon was examined using a semiquantitative scoring system as described in the Material and Methods section. Data are presented as the mean (± standard error of the mean) and are the cumulative results from 3 experiments (n = 12 for both groups). Statistics: **P ≤ .01. (D) Histology of tongue, liver, lung, and colon from representative Balb/c recipients transplanted with B6 BM and spleen cells and treated with either DMSO or RA. Tongues in DMSO- and RA-treated mice reveal lymphocytic infiltration into the epidermis. Livers in the 2 groups show mononuclear cell infiltration in the portal triads, while lungs demonstrate both perivascular cuffing attributable to mononuclear cells and associated interstitial inflammation. Colon in DMSO-treated animals demonstrates loss of mucin and lamina propria inflammation but preservation of crypt cell architecture. Conversely, RA-treated mice show extensive loss of crypts with goblet cell depletion and inflammation extending into the muscle. Original magnification is 200× for liver, colon, and tongue images and 100× for lung images.

Figure 2

RA enhances the expression of α4β7 and CCR9 on donor T cells in mesenteric lymph nodes. Lethally irradiated Balb/c mice were transplanted with B6 BM and spleen cells and treated with either DMSO or RA. Mice from both groups were killed at day 7 after BMT. The expression of integrin and chemokine receptors on donor T cells isolated from different secondary lymphoid tissues of recipient mice was examined by gating on donor-derived H-2Kb+ CD4+ or H-2Kb+ CD8+ cells. (A) Representative contour plots for CCR9 and α4β7 expression on donor T cells isolated from MLNs. (B) Mean percent positive (± SEM) donor T cells expressing CCR9 and α4β7 isolated from MLNs (n = 6-8 per group). Statistics: ***P ≤ .001. (C) Recipient mice were treated with either DMSO or escalating doses of RA, the expression of CCR9 and α4β7 on donor T cells from the MLNs of recipient mice 7 days after BMT is depicted (n = 3 per group). Data are from 1 of 2 representative experiments. (D) Representative contour plots for CXCR3 and CCR5 expression on donor T cells isolated from spleen. (E) Mean percent positive (± SEM) donor T cells expressing CXCR3 and CCR5 isolated from spleen (n = 6-8 per group). (F, G) Mean percent positive (± SEM) donor T cells expressing CCR10 (F) and gut-homing molecules (G) isolated from peripheral lymph nodes (n = 6-8 per group). Data are combined results from two independent experiments. Statistics: *P ≤ .05, **P < .01.

Figure 3

Effect of exogenous RA administration on effector and regulatory T-cell infiltration into the colon. Lethally irradiated Balb/c mice were transplanted with 10 × 106 B6 BM together with B6 spleen cells (adjusted to yield a T-cell dose of 0.6 × 106). Cohorts of mice were then treated with either DMSO (n = 6 to 8/group, black bars) or RA (n = 6 to 8/group, white bars). (A) Absolute cell number of CD4+IFN-γ+ or (B) CD4+IL-17+ T cells in the spleen, liver, lung, and colon of animals treated with either DMSO or RA and then killed 19 to 21 days after transplantation. (C) Absolute cell number of CD8+CD103+ cells in the colon of recipient mice. Data are presented as the mean (± standard error of the mean [SEM]) and are the cumulative results from 2 independent experiments. (D) Lethally irradiated Balb/c mice were transplanted with 10 × 106 Foxp3EGFP BM and spleen cells (adjusted to yield a T-cell dose of 0.6 × 106). Cohorts of mice were then treated with either DMSO (n = 8) or RA (n = 8). Mice in both groups were killed 19 to 21 days after transplantation. The absolute cell number of Foxp3+ (GFP+) Tregs in the spleen, liver, and colon of mice is depicted. Data are presented as the mean (± SEM) and are the cumulative results from 2 independent experiments. (E) Percentage of CD4+ and CD8+ T cells that expressed either α4β7 in the colon or CCR9 in the small intestines of transplant recipients that were treated with either DMSO or RA. Animals were killed 19 to 21 days post transplantation. Data are presented as the mean (± SEM) and are the cumulative results from 2 independent experiments. Statistics: *P ≤ .05, **P < .01, ***P < .001.

Figure 4

Vitamin A deprivation in recipient mice impairs the acquisition of gut-homing molecule expression on donor T cells after BMT. Lethally irradiated Balb/c Rag mice fed with either VAS or VAD diets were transplanted with B6 BM and CellTrace-violet–labeled splenic T cells (adjusted to yield a T-cell dose of 1 × 106). Mice from both groups were killed on day 3 after BMT. (A) Representative contour plots of CCR9 and α4β7 expression on donor T cells isolated from the MLNs of recipient mice after gating on H-2Kb+ CD4+ or H-2Kb+ CD8+ cells. (B) The mean percentage positive donor T cells (± SEM) expressing CCR9 and α4β7 from 4 mice per group. Data are from 1 of 2 representative experiments. Statistics: ***P ≤ .001. (C) Representative contour plots of CCR10 expression on donor T cells isolated from the PLNs of recipient mice. (D) The mean percentage positive donor T cells (± SEM) expressing CCR10 from 4 mice per group. Data are from 1 of 2 representative experiments. Statistics: *P ≤ .05, **P < .01. (E) Representative example of CellTrace-violet-labeled donor H-2Kb+ CD4+ or CD8+ T-cell proliferation in the spleen of VAS and VAD recipients 3 days after transplantation. (F) Mean (± SEM) percentage of proliferating (CellTrace-violet low) splenic donor T cells (n = 4 mice per group). One of 2 representative experiments is shown.

Figure 5

RAR-α signaling is important in mediating the effects of RA on T cells. B6 spleen cells (adjusted to 1 × 105 T cells/well) were cocultured with 5 × 104 Balb/c dendritic cells in the presence or absence of 1 μM AM80, an RAR-α agonist. Four days later, cells were harvested for flow cytometric analysis. (A) Representative contour plots of α4β7 and CCR9 expression on CD4+ and CD8+ T cells. (B) Mean (± standard error of the mean [SEM]) percentage of CD4+ and CD8+ T cells that expressed α4β7 and CCR9. Data are cumulative results from 4 independent experiments. Statistics: *P ≤ .05, ***P < .001. (C, D) B6 spleen cells (adjusted to 1 × 105 T cells/well) were cocultured with 5 × 104 Balb/c dendritic cells in the presence of 10 nM exogenous RA. An RAR-α antagonist, BMS195614 (1 μM), was added to some cultures 1 hour prior to the addition of RA. Cells were analyzed after 4 days in culture. Representative contour plots of CCR9, α4β7, CD62L, and CD25 expression on CD4+ and CD8+ T cells is depicted. Data are from 1 of 4 representative experiments with similar results. (E) Cells were cultured as in panels C and D with the exception of 3H-thymidine was added for the last 18 hours to assess proliferation. Data are presented as mean counts per minute ± SEM and are representative of 1 of 2 experiments with similar results. Statistics: *P ≤ .05

Figure 6

Endogenous RA production and RAR signaling on donor T cells are enhanced after allogeneic BMT. (A) Lethally irradiated Balb/c mice were transplanted with Balb/c BM (10 × 106) and spleen cells (0.4–0.5 × 106, syngeneic) or B6 BM and an equivalent number of B6 spleen cells (allogeneic). Cohorts of animals were killed weekly, and colon tissue was analyzed for RA levels by mass spectrometric analysis as described in the section titled “Quantification of RA in tissues by mass spectrometry.” Data are derived from 9 to 10 mice per group and are cumulative results from 2 independent experiments. Statistics: *P ≤ .05. (B) Lethally irradiated Balb/c mice were transplanted with B6 BM and spleen cells to induce GVHD. RNA was extracted from sort-purified CD4+ and CD8+ T cells isolated from spleens of allogeneic transplant recipients at the indicated time points after BMT, and gene expression of RAR-α was analyzed by real-time q-PCR as described in the section titled “Real-time q-PCR.” Data are normalized for 18S ribosomal RNA and presented as fold increase over RAR-α expression in sort-purified original naive CD4+ or CD8+ donor T cells used for transplantation. Data are derived from 2 independent experiments and are presented as the mean ± standard error of the mean.

Figure 7

Recipients of RAR-α–deficient T cells are protected from lethal GVHD yet are able to mount a GVL response. Lethally irradiated Balb/c mice were transplanted with BM cells (5 × 106) from either RAR-α─/─ (□) or wild type (WT) littermate controls (○) or BM and spleen cells (adjusted to yield a T-cell dose of 0.7−1.0 × 106 T cells) from either RAR-α─/─ (▪) or WT littermate controls (●). (A) Overall survival. Data are cumulative results from 3 independent experiments (n = 3 for BM control groups, n = 14 for each BM + T-cell group). (B) Lethally irradiated Balb/c mice were transplanted with BM (5 × 106) and spleen cells (adjusted to yield a T-cell dose of 2.1 × 106) from either RAR-α─/─ (n = 9) or WT littermate controls (n = 5). Mice were killed at days 10 to 12 after BMT and pathological damage in the liver, lung, and colon was examined. Data are derived from 2 independent experiments and are presented as the mean (± standard error of the mean [SEM]). Representative photomicrographs of the colon from mice transplanted with WT vs RAR-α─/─ marrow grafts are shown. Colon from recipients of WT grafts show loss of mucin and destruction of crypt architecture, while animals reconstituted with grafts from RAR-α─/─ mice have normal-appearing crypts. Statistics: *P ≤ .05, ***P ≤ .001. (C,D) The absolute number of donor-derived CD4+ (C) and CD8+ cells (D) in the liver, spleen, lung, and colon is shown for each group. Data are derived from 2 independent experiments and are presented as the mean (± SEM). (E) Lethally irradiated Balb/c mice were administered 0.6 × 106 A20 cells transfected with firefly luciferase (A20-luc) and then transplanted with WT BM alone (n = 5), WT BM plus spleen cells (adjusted to yield a dose of 1.2 × 106 T cells; n = 5), or RAR-α─/─ BM plus spleen cells (adjusted to yield an equivalent T cell dose; n = 5). Overall survival and whole body distribution of A20-luc cells using in vivo BLI is depicted.