TNF-TNFR2 Interactions Are Critical for the Development of Intestinal Graft-Versus-Host Disease in MHC Class II-Disparate (C57BL/6J→C57BL/6J × bm12)F₁ Mice

Mice

C57BL/6J and B6.C-H-2^bm12 (bm12) and C57BL/6-Tnfr2^tm1Mwm (B6.TNFR2^−/−) were obtained from The Jackson Laboratory (Bar Harbor, ME) and maintained in a specific pathogen-free environment. B6 males and bm12 females were bred to produce an F₁ strain (B6 × bm12F₁).

Transplantation

On the day of transplantation, the age- and sex-matched B6 × bm12F₁ (H-2^b/bm12) recipients were irradiated (900 cGy) and 2 h later were injected via the lateral tail vein with donor cells. Briefly, 4–5 × 10⁶ CD4⁺ T cell-enriched SpC and 2–3 × 10⁶ BMC from age- and sex-matched control B6 mice or B6.TNFR2^−/− mice were injected into the lateral tail vein of these lethally irradiated allogeneic B6 × bm12F₁ mice as previously described (1).This transplantation regimen has been documented previously to result in complete chimerism in this strain combination (10). In other experiments, 4 × 10⁶ splenic CD4⁺ T lymphocytes from B6.TNFR2^−/− and control B6 mice were transferred with 1–2 × 10⁶ T cell-depleted B6 BMC to irradiated MHC class II-disparate B6 × bm12F₁ mice. In other experiments, 4–5 × 10⁶ splenic CD4 T lymphocytes from control B6 mice were transferred with 1–2 × 10⁶ T cell-depleted B6 or B6.TNFR2^−/− BMC to irradiated MHC class II-disparate B6 × bm12F₁ mice. Recipients were maintained on acidified (pH 2) antibiotic (neomycin, 100 mg/L, and polymyxin B, 10 mg/L) H₂O for 7 days after transplantation. Splenic B6 or B6.TNFR2^−/−CD4⁺ T cells were enriched as previously described (1). Briefly, SpC were incubated with anti-CD8 Abs, anti-RBC, and B cell Abs (J11D.2) for 30 min at 4°C, then incubated with an adsorbed rabbit complement at 37°C for 30 min and then passed over a nylon wool column (11).

Histological scoring of intestinal GVHD

Individual recipients of either the control B6 or the experimental B6.TNFR2^−/− SpC were evaluated according to the scoring system used by Snover (12) and Bottomly et al. (13). The large intestine was isolated by removing all of the intestine distal to the appendix and 2-cm sections were cut, fixed in Formalin, and then stained with H&E. Peyer’s patches were removed from the small intestine and then 2-cm sections from the entire small intestine were cut from the duodenum to the ileum, fixed in Formalin, and stained with H&E. Specifically, intestinal GVHD was reported as grade I with evidence of increased apoptosis, grade II with evidence of cystically dilated crypts containing necrotic debris with individual crypt loss, grade III with loss of multiple crypts with preservation of surface epithelium, and grade IV with complete loss of epithelium (data not observed in this study). The pathologist reviewing the slides was not aware of the types of SpC that the mouse had received.

Preparation of lamina propria lymphocytes

Mucosal lymphocytes were isolated and prepared according to a modification of a previously published method (8, 14, 15). Large intestines and small intestines were removed and placed in 15-ml tubes containing HBSS without Ca²⁺ and Mg²⁺ (Life Technologies, Gaithersburg, MD). Each intestine was carefully cleaned and flushed of its fecal content. Peyer’s patches were removed from the small intestines. The small and the large intestines were opened longitudinally and then cut into six 0.5-cm pieces. To isolate the lamina propria lymphocytes (LPL), the intestinal tissue was vigorously pipetted, transferred to 17 × 100-mm polypropylene tubes, and shaken for 20 min at 37°C in complete RPMI 1640 supplemented with 1% FCS containing dispase at 1.5 mg/ml (Sigma-Aldrich, St. Louis, MO). The tissue suspensions were collected and passed through nylon mesh, and the cells were pelleted by centrifugation at 600 × g. The cells were layered over Ficoll-Hypaque and the lymphocytes were removed from the interface, resuspended in staining buffer (0.5 L HBSS, 0.5 L PBS, 2% BSA, and 0.02% sodium azide), and counted using a Coulter counter (8).

Isolation of SpC

SpC were harvested by surgically removing the spleen and then mincing and disrupting the spleen over a funnel covered with nylon mesh and washing repeatedly with HBSS into a 50-ml conical tube (11). After the suspension was centrifuged at 600 × g for 10 min, the cell pellet was resuspended in staining buffer (0.5 L HBSS, 0.5 L PBS, 2% BSA, and 0.02% sodium azide) and cells were counted using a Coulter counter.

Flow cytometric analysis of splenic and intestinal lymphocytes

LPL were obtained from individual mice and in some experiments LPL from two to three mice from the same experimental group were pooled before flow cytometric analysis. Since we had six to nine mice in each experimental group, we were able to prepare two to three pools from each experimental group for analysis. These pooled lymphocytes were suspended in staining buffer (0.5 L HBSS, 0.5 L PBS, 2% BSA, and 0.02% sodium azide). mAb used to characterize LPL populations included fluorescein-conjugated anti-CD3ε clone 145-2C11, biotinylated anti-CD4 clone RM4-4, FITC anti-CD8α clone 53-6.7, FITC anti-CD45RB clone 16A, and PE anti-CD4 Ab (all from BD PharMingen, San Diego, CA). mAb were added to cell suspensions at 4°C and incubated for 20 min. After staining with the primary Ab, samples were washed in PBS at 4°C. Avidin fluorescein was added as a secondary staining reagent for the biotinylated mAb. In some experiments, the cells were doubled stained with PE anti-CD4 Ab and FITC anti-CD45 Ab. After mAb staining and washing, all samples were fixed in PBS containing 1% paraformaldehyde and stored at 4°C until flow cytometric analysis as previously described (11). Flow cytometric analysis was performed on a FACscan flow cytometer (BD Biosciences, Mountain View, CA).

Multiprobe RNase protection assay

A multiprobe RNase protection assay system (Riboquant; BD PharMingen) that generated a series of a cytokine templates (IL-4, IL-10, Il-13, IL-15, IL-9, IL-2, IFN-γ, and GAPDH standard), each of distinct length and each representing a sequence in a distinct cytokine mRNA species, was used in analyzing small and large intestine LPL RNA from B6 × bm12F₁ mice that had received the control B6 CD4⁺ T lymphocytes and from mice that had received the B6.TNFR2^−/−CD4⁺ T lymphocytes. The mouse cytokine templates utilized a T7 polymerase-directed synthesis of a high specific activity ³²P-labeled antisense RNA probe set. The probe set was hybridized in excess to target small and large intestinal LPL RNA in solution, after which free probe and other ssRNA were digested with RNases (8). The remaining RNase-protected probes were purified, resolved on polyacrylamide gels, and quantified by phosphor imaging.

Statistical analysis

The results were analyzed by the Mann Whitney U rank sum nonparametric test.

Absence of TNFR2 interactions on donor CD4⁺ T cells ameliorates MHC class II-disparate intestinal GVHD induced by the transfer of B6 CD4⁺ SpC into irradiated B6 × bm12F₁ recipients

Previous publications have reported that intestinal GVHD is present after transfer of B6 SpC into lethally irradiated MHC class II-disparate B6 × bm12F₁ recipients (10) and that TNF-TNFR2 interactions promote MHC class II-stimulated T cell alloresponses (9). To assess whether the wasting disease induced by B6→B6 × bm12F₁ GVHD is dependent upon TNF-TNFR2 interactions on donor T cells, weights were assessed 11–14 days after transplantation. In the initial experiments, B6 CD4⁺ T cells and BMC and B6.TNFR2^−/−CD4⁺ T cells and BMC were transferred to B6 × bm12F₁ mice. A significant weight difference was noted 14 days after transplantation in recipients of B6.TNFR2^−/− donor cells (15.3 +/− 1.22 g; n = 3) vs recipients of control B6 donor cells (11.95 +/− 1.5 g; n = 3). The initial weights of the animals were not different.

To assess whether the weight difference was associated with TNFR2 deficiency on the donor T cells or donor BMC, B6 CD4⁺ T cells or B6.TNFR2^−/−CD4⁺ T cells were transferred with B6 T cell-depleted BMC. A significant weight difference was noted between the recipients of the B6.TNFR2^−/−CD4⁺ SpC (20.33 ± 2.67 g (n = 6)) in which no significant decline from pretransplant weight was observed vs recipients of B6 CD4⁺ T cells (mean weight, 14.64 ± 3.53 g (n = 7)) in which an average of 10.82% weight loss was observed. (Fig. 1⇓).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Day 14 weights of the B6 × bm12F₁ recipients of B6 CD4⁺ donor T cells were significantly lower than weights of recipients of B6.TNFR2^−/−CD4⁺ donor T cells. Briefly, 4 × 10⁶ splenic CD4⁺ T lymphocytes from B6.TNFR2^−/− (n = 7) or control B6 mice (n = 7) were transferred with 2–3 × 10⁶ B6 BMC to irradiated MHC class II-disparate B6 × bm12F₁ mice. Weights were assessed at the indicated time point.

Furthermore, there were histological differences in the intestines of recipients of the B6.TNFR2^−/− donor CD4⁺ T cells vs recipients of the B6 donor CD4⁺ T cells. The intestinal GVHD in B6 × bm12F₁ recipients of control B6 donor CD4⁺ T cells was manifested by evidence of apoptosis, cystically dilated crypts containing necrotic debris with individual crypt loss, loss of multiple crypts with preservation of surface epithelium, and complete loss of epithelium in the large intestine (Fig. 2⇓). More of the large intestines isolated from the B6 × bm12F₁ recipients of the B6 CD4⁺ T cells had evidence of grade II or higher GVHD (7 of 10) than recipients of the B6.TNFR2^−/−CD4⁺ T cells (2 of 9; Fig. 3⇓). Histological changes in the B6→B6 × bm12F₁ mice were clearly related to allogeneic responses as the large intestines of the syngeneic controls had no GVHD changes. These results indicate that TNF-TNFR2 interactions on alloreactive donor CD4⁺ T cells are important for the development of both intestinal GVHD and the wasting disease associated with GVHD in B6 →B6 × bm12F₁ mice.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Histological appearance of colons from B6 × bm12F₁ mice that had received either the control B6 CD4⁺ T cells or B6.TNFR2^−/− CD4⁺ T cells. Histological features of grade I GVHD, apoptosis of crypt cells (filled arrow), were observed in the colonic tissue (A) obtained 9–15 days after transplantation in recipients of both the control B6 and the TNFR2^−/−CD4⁺ T cells. Colonic tissue obtained 10–15 days after transplant of the control B6 CD4⁺ T cells displayed (B) cystically dilated crypt containing necrotic debris (filled arrow), partially destroyed crypt (arrowhead), and individual crypt loss (open arrow; grade 2 GVHD) as well as areas of (C) multiple crypts loss (filled arrow) with preservation of surface epithelium (grade 3 GVHD). Note normal crypt for comparison (open arrow).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Semiquantitative scoring of intestinal GVHD. Colonic tissue was harvested 9–15 days after transplantation from 10 recipients of control B6 CD4⁺ splenic T cells, 9 recipients of B6.TNFR2^−/−CD4⁺ splenic T cells, and 3 recipients of syngeneic donor cells (two B6→B6, one B6 × bm12F₁). The colonic tissue was fixed, sectioned, stained with H&E, and scored as detailed in Materials and Methods.

Absence of TNFR2 interactions on T cell-depleted BMC does not ameliorate MHC class II-disparate intestinal GVHD induced by the transfer of B6⁺CD4⁺ SpC and BMC into irradiated bm12 × B6F₁ recipients

To determine whether lack of TNFR2-TNF interactions on non-T cell-derived donor BMC were important for the development of GVHD, 4–5 × 10⁶ splenic CD4 T lymphocytes from control B6 mice were transferred with 1–2 × 10⁶ T cell-depleted B6 or TNFR2^−/− BMC to irradiated MHC class II-disparate B6 × bm12F₁ mice. Twelve days after transplant, weights of recipients of B6 CD4⁺ T cells and either TNFR2^−/− T cell-depleted BMC or B6 T cell-depleted BMC were assessed. The weights were similar for the TNFR2^−/− BMC recipients (mean, 18.3 g (n = 3) (15.9% decrease) vs control B6 recipients (21.65 g (n = 2) (7.25% decrease). Intestinal histopathology also was similar in both groups (data not shown).

Fewer T cells were isolated from the small intestine in recipients of the B6.TNFR2^−/−CD4⁺ T cells than from the recipients of the B6 CD4⁺ T cells

In addition to the analysis of colonic GVHD detailed in Fig. 3⇑, there was evidence of loss of villous architecture and gland drop-out in the small intestines of B6 →B6 × bm12F₁ GVHD mice (Fig. 4⇓). To assess whether recipients of the B6.TNFR2^−/−CD4⁺ T cells had lesser degrees of T cell infiltrates in the intestine than in the recipients of the control B6 CD4⁺ T cells, intestinal lymphocytes were removed from the small intestines of B6 × bm12F₁ recipients. Of interest, recipients of the B6 CD4⁺ T cells had a large variance in intestinal lymphocyte counts (3.3 ± 3.7 × 10⁶) while the recipients of the TNFR2^−/−CD4⁺ T cells were consistently low (0.67 ± 0.59 × 10⁶). The recipients of the syngeneic donor cells had counts similar to those of the recipients of TNFR2^−/− donor cells (0.6 ± 0.57). To determine the absolute number of CD3 and CD4⁺ T cells, these counts were multiplied by the percentage of cells that stained positive with anti-CD3 and anti-CD4 Abs. As demonstrated by the data displayed in Table I⇓, significantly fewer CD3⁺ and CD4⁺ T cells were obtained from the B6 × bm12F₁ recipients of the B6.TNFR2^−/−CD4⁺ T cells than from recipients of the B6 CD4⁺ T cells ,whereas recipients of syngeneic donor cells had even fewer CD3 and CD4⁺ T cells isolated. These data indicate that TNF-TNFR2 interactions are critical for the infiltration and expansion of the CD4⁺ T cells in the small intestine.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Photomicrograph illustrating the severity of the small intestinal disease in the B6 × bm12F₁ recipients of the control B6 CD4⁺ T cells. The small intestine was harvested from mice that received control B6 CD4⁺ T cells and B6 BMC. This is a representative photomicrograph of a section displaying loss of villous architecture and gland drop-out (A). In contrast, recipients of TNFR2^−/−CD4⁺ T cells demonstrated normal villous architecture and no gland drop-out (B).

Fewer CD45RB^low CD3⁺ T cells are observed in the small intestines of the recipients of B6.TNFR2^−/−CD4⁺ T cells

The level of expression of the CD45RB surface marker on T cells has been reported to be associated with prior T cell activation (memory T cells) (14, 15, 16, 17). Specifically, high levels of the CD45RB surface marker has been observed on T cells that have no prior exposure to Ags (naive), whereas those T cells with low levels of this marker on their surface have prior exposure to Ags (activated or memory cells). Intestinal LPL were isolated as described in Materials and Methods. FACS analysis revealed that >75% of these intestinal LPL were CD4⁺ T cells. The percentage of CD45RB^lowCD4⁺ T cells in B6 × bm12F₁ recipients of the B6 CD4⁺ T cells and of the B6.TNFR2^−/−CD4⁺ T cells was examined to determine whether the lack of B6.TNFR2^−/− on donor CD4⁺ T cells affected the percentage of memory T cells in the lamina propria of GVHD small intestines. A lower percentage and absolute number of CD4⁺CD45RB^low memory T cells was present in the recipients of the B6.TNFR2^−/−CD4⁺ T cells (Fig. 5⇓, A and B), suggesting that TNFR2-TNF interactions on donor CD4⁺ T cells may be necessary for alloantigen-induced activation and/or memory T cell development in this model of intestinal GVHD.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

A decreased percentage of CD45RB^low T cells was observed in small intestinal lymphocytes from B6 × bm12F₁ recipients of the TNFR2^−/− donor CD4⁺ splenic T cells. Small intestinal lymphocytes were harvested between day 8 and 14 after transplantation from B6 × bm12F₁ recipients of the TNFR2^−/− donor CD4⁺ T cells (n = 4) or B6 donor CD4⁺ T cells (n = 4). The histogram on the left of A displays CD45RB staining of LPL from a representative mouse that had received the B6 CD4⁺ donor T cells, and the histogram on the right displays CD45RB staining of LPL from a representative mouse that had received the TNFR2^−/−CD4⁺ donor T cells. In B are displayed the percentage and the absolute number of intestinal lymphocytes (mean ± SEM) expressing low levels of CD45RB (memory).

There were no differences in splenic CD3⁺ T cell number or CD45RB expression in recipients of the B6.TNFR2^−/−CD4⁺ donor T cells vs control B6 CD4⁺ donor T cells

To determine whether TNF-TNFR2 interactions were deterring the expansion or differentiation of donor T cells at extraintestinal sites, such as the spleen, total SpC counts were measured from mice that had received the B6.TNFR2^−/−CD4⁺ T cells and mice that had received the control B6 SpC. Eleven to 14 days after transplant, similar number of cells (5.5 × 10⁶ (n = 12 vs (6.47 × 10⁶ (n = 12; Fig. 6⇓) and of CD3⁺ T lymphocytes (Table II⇓) were obtained from the spleens isolated from the recipients of the control B6 or B6. TNFR2^−/−CD4⁺ donor T cells (p = 0.56). These results suggest that TNF-TNFR2 interactions on donor CD4⁺ T cells are not affecting the migration and expansion of donor T cell in the spleens of lethally irradiated B6 × bm12F₁ mice.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Similar number of T cells are observed in the spleens isolated from the B6 × bm12F₁ recipients of B6 or B6.TNFR2^−/− donor CD4⁺ T cell. Briefly, 4 × 10⁶ splenic CD4⁺ T cells from TNFR2^−/− (n = 11) or control B6 mice (n = 11) were transferred with 2–3 × 10⁶ control B6 BMC to irradiated MHC class II-disparate B6 × bm12F₁ mice. Cells from recipient spleens were obtained on days 11–14 after BMT and counted by Coulter counter.

CD45RB surface expression on splenic CD4⁺ T cells was also not affected by lack of TNFR2-TNF interactions on donor CD4⁺ T cells. The majority of the splenic CD4⁺ T cells exhibited the CD45RB^low memory phenotype. In the spleen, there were similar numbers of CD4⁺ memory CD45RB^low T cells in recipients of control or TNFR2-deficient donor CD4⁺ T cells (data not shown). These results suggest that TNF is critical for the development of memory T cells in the intestine during GVHD but not in the spleen.

A significant reduction in intestinal IL-2 and IFN-α mRNA expression was observed in recipients of the TNFR2-deficient CD4⁺ T cells

To assess the level of in vivo cytokine expression during intestinal GVHD, mRNA isolated from small intestinal sections extracted from recipients of the TNFR2^−/−CD4⁺ T cells or from recipients of B6 CD4⁺ T cells was probed with a multiprobe RNase protection assay system. After hybridization of the radioactive-labeled probe set to small intestinal mRNA, the radioactive counts for each cytokine mRNA were quantitated by phosphor imaging and then standardized to the control L32 RNA counts. Significantly lower levels of IL-2 (p = 0.019) and IFN-γ mRNA (p = 0.007) were observed in small intestine mRNA isolated from recipients of the TNFR2^−/−CD4⁺ T cells than from controls (Fig. 7⇓).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

Quantification of intestinal cytokine mRNA in GVHD mice. Quantification of cytokine mRNA in intestines from allogeneic bone marrow transplant recipients that received either the TNFR2^−/− (n = 6) or the control B6 CD4⁺ T cell (n = 8) was performed as detailed in Materials and Methods. Cytokine mRNA levels were normalized to L32 mRNA levels in each mouse.