Prevention of allograft tolerance by bacterial infection with Listeria monocytogenes

Abstract

Exposure to certain viruses and parasites has been shown to prevent the induction of transplantation tolerance in mice via the generation of cross-reactive memory T cell responses or the induction of bystander activation. Bacterial infections are common in the perioperative period of solid organ allograft recipients in the clinic, and correlations between bacterial infections and acute allograft rejection have been reported. However, whether bacterial infections at the time of transplantation have any effect on the generation of transplantation tolerance remains to be established. We used the Gram-positive intracellular bacterium Listeria monocytogenes (LM) as a model pathogen because its effects on immune responses are well described. Perioperative LM infection prevented cardiac and skin allograft acceptance induced by anti-CD154 and donor-specific transfusion in mice. LM-mediated rejection was not due to the generation of cross-reactive T cells and was largely independent of signaling via MyD88, an adaptor for most TLRs, IL-1, and IL-18. Instead, transplant rejection following LM infection was dependent on the expression of the phagosome-lysing pore former listeriolysin O and on type I IFN receptor signaling. Our results indicate that bacterial exposure at the time of transplantation can antagonize tolerogenic regimens by enhancing alloantigen-specific immune responses independently of the generation of cross-reactive memory T cells.

Figure 1. LM infection prevents anti-CD154/DST-mediated allograft acceptance

(A) LM infection prevents anti-CD154/DST-mediated cardiac allograft acceptance. B6 recipients of B/c heterotopic heart grafts were left untreated (n=5) or were treated with anti-CD154 (1mg/mouse on days 0, 7 and 14 post-transplantation) and DST (5×10⁶ splenocytes, i.v. on the day of transplantation) in the absence (n=6) or presence (n=8) of an infection with LM (10⁵ cfu i.p. on the day of transplantation). Recipients of syngeneic hearts were untreated but received the same dose of LM as the anti-CD154/DST groups (n=4). p<0.01 for anti-CD154/DST vs no treatment; p<0.001 for anti-CD154/DST vs anti-CD154/DST+LM. (B) Histology and immunohistochemistry of cardiac allografts isolated at the time of rejection from an untreated mouse and from mice treated with anti-CD154/DST and anti-CD154/DST+LM, as well as from a syngeneic graft harvested on day 30 from a mouse infected with LM. Cellular infiltration of CD4⁺ and CD8⁺ cells is observed, as well as alloantibody deposition, in rejecting but not syngeneic grafts. This result is representative of 3 allografts analyzed/group (magnification 200X). (C) LM infection prevents anti-CD154/DST-mediated skin allograft acceptance. B6 mice were transplanted with B/c or B6 skin and treated as indicated (no treatment, n=5; anti-CD154/DST, n=8; anti-CD154/DST+LM, n=5; syngeneic+LM, n=5). Skin grafts were considered rejected when fully scabbed and necrotic. p<0.001 for anti-CD154/DST vs no treatment and vs anti-CD154/DST+LM.

Figure 2. LM infection restores B and T cell alloreactivity in anti-CD154/DST-treated recipients

(A) Restoration of allo-IgG responses. Serum from transplanted mice was collected at the indicated time points and concentrations of allo-reactive IgM and IgG antibodies was determined by flow cytometry. Results represent the mean and standard deviation of 4 determinations per group. p<0.05 for Allo+anti-CD154/DST+LM vs no treatment, vs Syn+LM and vs Allo+anti-CD154/DST. (B) Restoration of primed allo-specific IFN-γ-producing cells. Splenocytes from transplanted or naïve mice treated as indicated were isolated 2-3 weeks post transplantation and stimulated with syngeneic (B6), donor (B/c) or third party (C3H/HEJ) irradiated splenocytes. The frequency of IFN-γ-producing cells was analyzed as indicated in the Materials and Methods. Results represent the mean and standard deviation of 4 determinations per group. p<0.001 for Allo+anti-CD154/DST+LM vs Allo+anti-CD154/DST, vs naïve+LM and vs naïve; not significant vs no treatment. (C) LM infection results in increased APC capacity to prime T cells. B6 mice were infected with LM and either untreated or treated with anti-CD154/DST. The splenocytes were isolated 48h after infection and depleted of T cells using anti-Th1.2 mAb and rabbit complement. T cell-depleted splenocytes either unpulsed or pulsed with OVA peptide were used to stimulate CFSE-labeled OVA-specific B6/RAG1^-/-/OTII-Tg T cells that had been enriched by negative selection over magnetic beads. Cultures were harvested after 5 days and analyzed by flow cytometry. Results represent CFSE fluorescence intensity of CD4⁺/Vα2⁺ cells and are representative of 3 independent experiments. p<0.01 for LM-infected+OVA and LM-infected/anti-CD154/DST+OVA vs naïve+OVA.

Figure 3. CD4⁺ or CD8⁺ T cells can mediate allograft rejection in LM-infected recipients

(A) B6/CD8^-/- mice were transplanted with BALB/c hearts and treated as described for Figure 1 (n=5 mice/group). p,0.01 for CD8^-/-+anti-CD154/DST vs no treatment and vs CD8^-/-+anti-CD154/DST+LM. (B) B6/CD4^-/- mice were used as recipients of B/c hearts. Mice were left untreated, or infected with LM. In one group, CD8⁺ cells were depleted with anti-CD8 mAb (n=4-5 mice/group). p<0.01 for wildtype (WT) vs CD4^-/-, vs CD4^-/-+anti-CD8+LM and vs CD4^-/-+LM; p<0.05 for CD4^-/-+LM vs CD4^-/- and vs CD4^-/-+anti-CD8+LM.

Figure 4. The ability of LM infection to prevent transplantation tolerance is not due to the stimulation of cross-reactive T cells

B6 mice were infected with LM only on day -45 (n=5), or on days -14 and +0 (n=4), or only on day +0 (n=7) relative to transplantation with B/c hearts and initiation of the anti-CD154/DST treatments. p<0.001 for no treatment vs anti-CD154/DST, vs anti-CD154/DST+LM-D-45 and vs anti-CD154/DST+LM-D-14+0.

Figure 5. Mechanisms of acute rejection by LM

(A) B6/MyD88^-/- recipients of B/c/MyD88^-/- hearts were treated as in Figure 1 (untreated, n=3; anti-CD154/DST, n=6; anti-CD154/DST+LM, n=10). p<0.01 for MyD88^-/- vs MyD88^-/-+anti-CD154/DST+LM and vs MyD88^-/-+anti-CD154/DST; p<0.01 for MyD88+anti-CD154/DST vs MyD88^-/-+anti-CD154/DST+LM. (B) B6 recipients of B/c cardiac allografts were treated as in Figure 1. Some groups received ampicillin (Amp, 25 mg/mouse for 5 days) or cephazolin (10 mg/mouse) starting on the day of LM infection (n=6) or 2 days later (n=6). p<0.001 for WT+anti-CD154/DST vs WT+anti-CD154/DST+LM and vs WT+anti-CD154/DST+LM+antibiotic-D2. (C) LM (10⁵ cfu, n=7), ActA-deficient LM (LM-Act^-, 10⁷ cfu, n=5), or LLO-deficient LM (LM-LLO^-, 5×10⁷ cfu, n=9) were used on the day of transplantation to infect B6 recipients of B/c hearts treated with anti-CD154/DST as for Figure 1. p<0.001 for LM-LLO^- vs LM-WT and vs LM-ActA^-. (D) Number of LM-WT and LM-LLO^- bacteria recovered from the spleen 2 days after infection (n=4/group), as described in Materials and Methods. p=0.2096, not significant.

Figure 6. Prevention of tolerance by LM infection is mediated by Type I IFNs

(A) In vivo production of IFN-β by B6, B6/MyD88^-/- and B6/IFNαR1^-/- mice infected with LM (10⁵ cfu). Animals were bled at the indicated time points and the concentration of serum IFN-β was measured by ELISA. *p<0.05 between serum from LM-infected B6 mice at 48h and all other time points and with serum from LM-infected IFNαR1^-/- mice at 48h. (B) Prevention of tolerance by LM infection is dependent on Type I IFNs. B6/IFNαR1^-/- mice were used as recipients of B/c heart (Left Panel) or skin (Right Panel) allografts. Mice were left untreated or were treated with anti-CD154/DST or with anti-CD154/DST/LM (n=4-6 mice in all groups). (Left Panel) p<0.05 for IFNαR1^-/-+anti-CD154/DST+LM vs IFNαR1^-/- no treatment and WT+anti-CD154/DST+LM. (Right Panel) p<0.01 for IFNRαR1^-/-+no treatment vs IFNRαR1^-/-+anti-CD154/DST and +anti-CD154/DST+LM. (C) IFN-β is sufficient to prevent anti-CD154/DST-mediated graft survival. B6 mice were transplanted with B/c skin and left untreated or treated with anti-CD154/DST. Two groups of mice received i.p. injections of IFN-β (10⁴U on day 0, n=5 or 2×10⁴ U/mouse on days 0 and 2 post-transplant, n=5). P<0.05 between IFN-β-treated and anti-CD154/DST alone groups.

Figure 6. Prevention of tolerance by LM infection is mediated by Type I IFNs

(A) In vivo production of IFN-β by B6, B6/MyD88^-/- and B6/IFNαR1^-/- mice infected with LM (10⁵ cfu). Animals were bled at the indicated time points and the concentration of serum IFN-β was measured by ELISA. *p<0.05 between serum from LM-infected B6 mice at 48h and all other time points and with serum from LM-infected IFNαR1^-/- mice at 48h. (B) Prevention of tolerance by LM infection is dependent on Type I IFNs. B6/IFNαR1^-/- mice were used as recipients of B/c heart (Left Panel) or skin (Right Panel) allografts. Mice were left untreated or were treated with anti-CD154/DST or with anti-CD154/DST/LM (n=4-6 mice in all groups). (Left Panel) p<0.05 for IFNαR1^-/-+anti-CD154/DST+LM vs IFNαR1^-/- no treatment and WT+anti-CD154/DST+LM. (Right Panel) p<0.01 for IFNRαR1^-/-+no treatment vs IFNRαR1^-/-+anti-CD154/DST and +anti-CD154/DST+LM. (C) IFN-β is sufficient to prevent anti-CD154/DST-mediated graft survival. B6 mice were transplanted with B/c skin and left untreated or treated with anti-CD154/DST. Two groups of mice received i.p. injections of IFN-β (10⁴U on day 0, n=5 or 2×10⁴ U/mouse on days 0 and 2 post-transplant, n=5). P<0.05 between IFN-β-treated and anti-CD154/DST alone groups.