The regulatory T cell effector molecule fibrinogen-like protein 2 is necessary for the development of rapamycin-induced tolerance to fully MHC-mismatched murine cardiac allografts

Abstract

Therapies that promote tolerance in solid organ transplantation will improve patient outcomes by eliminating the need for long-term immunosuppression. To investigate mechanisms of rapamycin-induced tolerance, C3H/HeJ mice were heterotopically transplanted with MHC-mismatched hearts from BALB/cJ mice and were monitored for rejection after a short course of rapamycin treatment. Mice that had received rapamycin developed tolerance with indefinite graft survival, whereas untreated mice all rejected their grafts within 9 days. In vitro, splenic mononuclear cells from tolerant mice maintained primary CD4(+) and CD8(+) immune responses to donor antigens consistent with a mechanism that involves active suppression of immune responses. Furthermore, infection with lymphocytic choriomeningitis virus strain WE led to loss of tolerance suggesting that tolerance could be overcome by infection. Rapamycin-induced, donor-specific tolerance was associated with an expansion of regulatory T (Treg) cells in both the spleen and allograft and elevated plasma levels of fibrinogen-like protein 2 (FGL2). Depletion of Treg cells with anti-CD25 (PC61) and treatment with anti-FGL2 antibody both prevented tolerance induction. Tolerant allografts were populated with Treg cells that co-expressed FGL2 and FoxP3, whereas rejecting allografts and syngeneic grafts were nearly devoid of dual-staining cells. We examined the utility of an immunoregulatory gene panel to discriminate between tolerance and rejection. We observed that Treg-associated genes (foxp3, lag3, tgf-β and fgl2) had increased expression and pro-inflammatory genes (ifn-γ and gzmb) had decreased expression in tolerant compared with rejecting allografts. Taken together, these data strongly suggest that Treg cells expressing FGL2 mediate rapamycin-induced tolerance. Furthermore, a gene biomarker panel that includes fgl2 can distinguish between rejecting and tolerant grafts.

Keywords: fgl2; regulatory T cell; tolerance; transplantation.

Figure 1

Rapamycin treatment leads to indefinite heart allograft survival (tolerance) in a donor-specific manner. (a) Survival of BALB/cJ hearts transplanted into C3H/HeJ recipient mice. Recipient groups included non-treated (○: mean survival time = 9·0 days, n = 5), rapamycin-treated (•: survival time > 100 days, n = 12), cyclosporin A-treated (▵: mean survival time = 29·5 days, n = 6), and syngeneic (C3H/HeJ → C3H/HeJ) transplants (□: survival time > 100 days, n = 5). *P < 0·05 versus non-treated. (b) Histopathology of heart allografts (H&E stain; magnification 100×). (i) Allografts from non-treated recipients [post-operative day (POD) 7], black arrow indicates vasculitis. (ii) Allograft from rapamycin-treated recipient (POD 100), black arrow indicates normal appearing vessel. (iii) Syngeneic control graft (POD 100). (iv) Non-transplanted control heart. No evidence of vasculitis (black arrows) is present in either control. (c) Survival curves of donor BALB/cJ (▪: survival time > 30 days, n = 5) and third-party C57BL/6J (♦: mean survival time = 14·3 days, n = 4) skin grafts after heart transplantation. *P < 0·05 versus third party. Rapa, rapamycin. CsA, cyclosporin.

Figure 2

Lymphocytes from tolerant mice maintain primary immune responses to donor antigens in vitro. (a) Mixed lymphocyte reaction (MLR) assay. Splenic mononuclear cells (SMNC) were isolated from non-transplanted (□), post-operative day (POD) 100 non-treated and rejecting (), and POD 100 rapamycin-treated and tolerant (▪) C3H/HeJ mice and co-cultured with irradiated SMNC from syngeneic (C3H/HeJ), donor (BALB/cJ) or third-party (C57BL/6J) mice. *P < 0·05 versus rejecting group. (b) CFSE proliferation profiles of (i) non-transplanted, (ii) POD 30 tolerant, and (iii) POD 30 rejecting C3H/HeJ SMNC stimulated with irradiated donor SMNC. (c) Quantification of CFSE proliferation data. Proliferation of SMNC from the different groups is shown as a fold change compared with SMNC from non-transplanted mice (mean ± SEM). *P < 0·05 versus rejecting group. (d, e) Cytotoxic T lymphocyte (CTL) assay. SMNC were isolated from non-transplanted controls (♢), POD 100 tolerant () and POD 100 rejecting (▪) C3H/HeJ graft recipients and co-cultured with donor A20 (d) or third-party EL4 (e) target cells. *P < 0·05 versus rejecting group. (f) Sera from syngeneic, tolerant and rejecting recipients were analysed for the presence of donor-specific antibody (DSA) by flow cytometry. Data are expressed as mean ± SEM. *P < 0·05 versus syngeneic control. Data for (a), (c) and (f) have three or four mice per group. Data for (b), (d) and (e) are representative of three independent experiments performed in triplicate. Rapa, rapamycin.

Figure 3

Infection with lymphocytic choriomeningitis virus (LCMV) leads to rejection of tolerant allografts. Hearts were isolated from non-transplanted (Non-Tx) mice and mice that had received syngeneic grafts and allografts 6 days post-infection with 2 × 10⁶ plaque-forming units of LCMV. Mice that had received allografts were tolerant to the grafts following a short course of rapamycin. (a–c) Haematoxylin & eosin staining (magnification 200×) revealed significant inflammatory cell infiltrates and myocyte necrosis at day 6 post-LCMV infection in allografts consistent with the histological picture of acute cellular rejection and loss of tolerance. Non-transplanted hearts and syngeneic grafts did not show any signs of rejection or inflammation. (d–f) Anti-LCMV immunostaining (black arrow) (magnification 200×) showed near equivalent distribution of nucleocapsid protein in non-transplanted hearts, syngeneic grafts and allografts. (g) Survival of heart grafts is shown as a Kaplan–Meir survival curve post-LCMV infection. The images and survival curve are representative of three mice per group.

Figure 4

Regulatory T (Treg) cells are increased in tolerant mice. (a) Flow cytometry plots of the splenic Treg population (FoxP3⁺) as a percentage of CD4⁺ splenic mononuclear cells (SMNC) from (i) non-transplanted (Non-Tx), (ii) post-operative day (POD) 30 rejecting, (iii) POD 30 tolerant, (iv) rapamycin-only non-transplanted (Rapa only), and (v) POD 30 syngeneic C3H/HeJ recipients are shown. Graphs are representative of three independent experiments. (b) Quantification of flow cytometry plots. FoxP3⁺ cells are shown as a percentage of CD4⁺ SMNC from the above groups with three or four mice per group (mean ± SEM). *P < 0·05 versus tolerant group. (c) FoxP3 immunohistochemical staining (magnification 400×) in (i) non-transplanted hearts, (ii) POD 100 syngeneic grafts, (iii) POD 7 rejecting and (iv) POD 100 tolerant heart allografts. (d) Morphometric analysis of allografts. Cells positive for FoxP3 are shown as a percentage of CD3⁺ graft-infiltrating cells from syngeneic (), tolerant (□) and rejecting (▪) groups at different time-points after transplantation. A minimum of three mice per group at each time-point. Data are expressed as the mean ± SEM. *P < 0·05 versus rejecting group or syngeneic group. ND, not determined.

Figure 5

CD4⁺ CD25⁺ FoxP3⁺ regulatory T (Treg) cells are necessary for rapamycin-induced graft tolerance. (a) CD4⁺ CD25⁺ FoxP3⁺ Treg cells are shown as percentages of CD4⁺ splenic mononuclear cells (SMNC) from rapamycin treated (Rapa), anti-CD25 treated (Rapa + anti-CD25), and IgG isotype control treated (Rapa + Isotype) recipients. Mice received an intraperitoneal injection of 250 µg of anti-CD25 antibody (PC61) or 250 µg of isotype control antibody 2 days before transplantation and on day 0 and 3, 6, 9, 12, and 15 days post-transplantation concurrently with rapamycin. Minimum of three mice per group. *P < 0·05 compared with Rapa + anti-CD25 group. (b) Histology of grafts from (i) anti-CD25 and (iii) isotype control treated mice (haematoxylin & eosin stain, magnification 300×). FoxP3 immunohistochemistry of grafts from (ii) anti-CD25 and (iv) isotype control treated mice (magnification 100×). (c) Morphometric analysis of allografts. Cells stained positive for FoxP3 are shown as percentage of CD3⁺ allograft-infiltrating cells from anti-CD25 antibody (Rapa + anti-CD25) and isotype control (Rapa + Isotype) treated groups. Minimum of three mice per group. Graph shows mean ± SEM. *P < 0·001 by two-tailed t-test. (d) Graft survival in anti-CD25 antibody (○, mean survival time = 52·2 days, n = 6) and isotype control treated (•: survival time > 100 days, n = 3) recipients. *P < 0·05. Rapa, rapamycin.

Figure 6

Plasma levels of fibrinogen-like protein 2 (FGL2) after allotransplantation. Plasma levels were measured at different time-points after transplantation in tolerant (▪) and rejecting (□) recipients. n = 3 mice per group at each time-point. Graphs show mean level (ng/ml) ± SEM. *P < 0·05.

Figure 7

Inhibition of fibrinogen-like protein 2 (FGL2) reverses the tolerizing effects of rapamycin. (a) Proliferation of splenic mononuclear cells (SMNC) from post-operative day (POD) 30 tolerant mice with either (i) anti-FGL2 or (ii) isotype control by CFSE proliferation. Graphs are representative of three independent experiments. (b) Quantification of CFSE proliferation data. Graph shows proliferation of anti-FGL2-treated SMNC normalized to isotype control-treated SMNC (mean ± SEM). Three independent experiments were performed. *P < 0·05 by two-tailed t-test. (c) Allograft survival in mice following rapamycin treatment and administration of either anti-FGL2 (o: mean survival time = 47·7 days, n = 6) or isotype control (▲: survival time > 100 days, n = 3). *P < 0·05. Mice received an intraperitoneal injection 250 μg of anti-FGL2 antibody or isotype control antibody 2 days before transplantation and on day 0 and 3, 6, 9, 12, and 15 days post-transplantation concurrently with rapamycin. (d) Graft histology following antibody administration. (i) anti-FGL2, black arrow indicates severe vasculitis. (ii) isotype control, black arrow indicates normal appearing blood vessel. (Haematoxylin & eosin stain; magnification 100×).

Figure 8

Anti-fibrinogen-like protein 2 (FGL2) antibody does not reduce numbers of regulatory T (Treg) cells in allografts. (a) FoxP3 immunohistochemistry of grafts from (i) isotype control and (ii) anti-FGL2 treated mice (magnification 200×). (b) Morphometric analysis of allografts. Cells stained positive for FoxP3 are shown as percentage of CD3⁺ allograft-infiltrating cells from anti-FGL2 antibody (Rapa + anti-FGL2) (n = 6) and isotype control (Rapa + Isotype) (n = 3) treated groups. Graph shows mean ± SEM. P = 0·71 by two-tailed t-test.

Figure 9

Co-expression of fibrinogen-like protein 2 (FGL2) and FoxP3 in regulatory T (Treg) cells in tolerant allografts. (a, b) Transplanted hearts were harvested from (a) rejecting mice or from (b) tolerant C3H mice at post-operative day (POD) 100 and subsequently immunostained for FoxP3 (red) and FGL2 (green) (magnification 200×). Nuclei were visualized with DAPI (blue). Tolerant mice had significantly increased numbers of FoxP3⁺ Treg (white arrow). Whereas FoxP3⁺ Treg from tolerant mice largely expressed FGL2, FoxP3⁺ Treg in rejecting mice did not express FGL2. Inset shows a FGL2⁻ Treg in a rejecting allograft and a FGL2⁺ Treg in a tolerant allograft (magnification 1000×). (c–e) Morphometric analysis of the immunostained sections was performed using a Definiens analysis assessing the (c) number of FoxP3⁺/μm², (d) FGL2⁺/μm² and (e) FoxP3⁺ FGL2⁺/μm². Cardiac myocytes were excluded from analysis using size exclusion. Lymphocytes were defined based on size of 10 µm or less. The morphometric analysis of heart allografts is from six rejecting mice, seven tolerant mice and three syngeneic mice with four serial sections taken at multiple levels of the heart. Data are expressed as the mean ± SEM. Statistical significance was assessed using a Student's t-test.

Figure 10

Differentially expressed regulatory T (Treg) -related genes in cardiac allografts serve as putative biomarkers of tolerance. Graphs display differentially expressed genes between tolerant (□) and rejecting (•) grafts from a panel of 22 Treg-related genes (see Supporting information, Table S1) as assessed by multiplex RT-PCR. The expression of a gene was normalized to the housekeeping gene hypoxanthine phosphoribosyltransferase and expression was then calculated as a ratio compared with the expression in non-transplanted hearts. Three allografts were used for each time-point for both tolerant and rejecting groups. Graph shows mean ± SEM. *P < 0·05 versus rejecting group at the same time-point.