ALBERT

Description: Background : Recipients of hematopoietic stem cell transplantation (HSCT) can be severely immunocompromised, predisposing to opportunistic infections including cytomegalovirus (CMV). While adoptive transfer of ex vivo expanded donor antiviral T-cells can prevent viral disease, this approach is not frequently used clinically because cell culture and selection procedures are lengthy, labor-intensive, and expensive. We describe an alternative approach. A single inoculation of a strongly immunogenic, but safe, live-attenuated, Listeria-based vaccine (Lm-MCMV) can rapidly drive extensive in vivo expansion of anti-murine CMV (MCMV) CD8+ T-cells during immune reconstitution following HSCT. Methods : The Lm-MCMV vaccine is derived from a genetically defined Listeria monocytogenes ΔactAΔinlB vaccine strain (Brockstedt et al. PNAS in press) that is attenuated by 4-logs in a mouse virulence assay, as compared to wild-type Listeria. Lm ΔactAΔinlB was engineered to express and secrete the MCMV H-2b immunodominant peptide HGIRNASFI within an ovalbumin scaffold. C57BL/6 (H-2b; CD45.2) HSCT recipients were conditioned with 11 Gy irradiation on day -1, and injected with 5x106 T cell depleted (TCD) C57BL/6 bone marrow cells on day 0. Selected mice simultaneously received 3x107 splenocytes from PepBoy (H-2b; CD45.1) donors that were either naive or immunized with 1x107 colony forming units (cfu) Lm-MCMV on day -7. On day +21, selected groups were vaccinated with 1x107 cfu Lm-MCMV, and sacrificed up to 14 days later. MCMV-specific CD8+ T-cells were quantified by intracellular cytokine and tetramer staining. Residual viable Lm-MCMV vaccine in the spleen, liver, and brain of BMT recipients was measured by plating on BHI agar media. Results : No mortality occurred following vaccination. Viable Lm-MCMV, which peaked the day after immunization in the liver (avg = 6.1x104 cfu) and spleen (avg = 1.4x105 cfu) of HSCT recipients, were completely cleared within 5 days of inoculation. Viable Listeria were not isolated from brain. Without vaccination at day +21, none of the HSCT recipients had detectable anti-MCMV CD8+ T cells. In contrast, immunization with Lm-MCMV at day +21 lead to marked antiviral T-cell expansion in all groups. HGIRNASFI-specific CD8+ T-cells averaged 2.4% of total CD8+ T-cells in vaccinated TCD marrow recipients. Levels were significantly higher in recipients of TCD marrow plus naive (7.1%) or immune (19.0%) donor splenocytes. Conclusions : Immunization of BMT mice with live-attenuated Lm-MCMV was safe. All HSCT recipients survived immunization and rapidly cleared viable Listeria. A single Lm-MCMV immunization during the period of immune reconstitution following HSCT was also effective, driving extensive antiviral T-cell expansion. In HSCT recipients of TCD marrow alone, vaccination produced levels of HGIRNASFI-specific CD8+ T-cells (2.4%) comparable to that seen with adoptive immunotherapy. Co-transplantation of naive or immune donor splenocytes lead to significantly higher levels of antiviral CD8+ T-cells following vaccination (7.1% and 19%, respectively). This approach could represent a broadly applicable alternative to adoptive immunotherapy. Future studies will focus on optimizing vaccination strategies, including use in allogeneic transplantation.

Description: CMV infection is reported to increase the incidence and severity of chronic GvHD and clinical data have shown that preemptive antiviral therapy decreased the risk of extensive chronic GvHD. Using mouse model of Allogeneic BMT, we investigated the mechanism for the association of MCMV infection and GvHD. We hypothesize that MCMV infection leads to generalized immune activation and increases the number of donor derived allo-reactive T cells and GvHD activity. Methods: A parenteral to F1 mouse BMT model was used to study anti-CMV immunity and GvHD. Low dose splenocytes (3x106) from MCMV immunized C57BL/6 donors (H-2b, Thy1.2+, CD45.1+) were transplanted with 5x106 T cell depleted bone marrow (TCD BM) from congeneic mice (H-2b, Thy1.1+, CD45.2+) into lethally irradiated (11Gy) CB6F1 recipients (C57BL/6 x Balb/C, H-2b/d, Thy1.2+, CD45.2+). Previous work has established this as a dose that protects against CMV without immediate lethality from GvHD. Non-GvHD control mice received a dose of Amotosalen treated splenocytes (10x106) and 5x106 TCD BM that protects against CMV without GvHD. Recipient mice were infected i.p. with a supralethal dose (2.5x104 pfu) of MCMV 7 days post transplant. Clinical GvHD was monitored twice weekly by weight loss, hair loss, ruffled fur, diarrhea, and decreased activity. T cell chimerism in recipient spleen and thymus, and MCMV peptide specific tetramer+CD8+ T cells were determined by flow cytometry. Liver and lung viral loads were determined by counting PFU in tissue homogenates plated onto 3T3 confluent monolayers. Results: During the acute phase of MCMV infection (day 3 to 14 post infection), recipient mice that received 3x106 untreated donor splenocytes developed GvHD characterized by weight loss and higher mortality than the non-GvHD control mice. Although both GvHD+ and control mice effectively cleared the virus from their liver, delayed viral clearance from the lung was found in non-GvHD recipients. Viral clearance was associated with expansion of donor spleen-derived MCMV peptide specific tetramer+ CD8+ T cells. The kinetics of donor T-cell expansion varied significantly between the two treatment groups, with GvHD+ recipients showed rapid early expansion of donor derived T-cells followed by the development of GvHD with subsequent lymphopenia. Recipients of Amotosalen-treated splenocytes had more gradual expansion of total and 400-fold expansion of antigen specific T-cells with sustained lymphoid reconstitution. GvHD+ recipients of untreated splenocytes had complete chimerism comprised of 〉80% of CD8+ donor T cells whereas non-GvHD controls had significant expansion of host derived T cells following MCMV infection and lacked allo-reactive of donor- spleen-derived T cells. Thymic function was inhibited among animals that developed GvHD and preserved among control mice throughout the infectious phase. Very delayed CMV infection (on day 60 after transplant) in mice with established chronic GvHD exacerbated GvHD and was associated with delayed lung viral clearance. Conclusion: After CMV infection there is extensive expansion of allo-reactive T cells in GvHD+ mice with associated damage to the microenvironment in the spleen and thymus. Amelioration of the immuno-suppressive effect of CMV infection (in clinical transplantation) will likely require more effective prophylaxis and treatment of GvHD in allotransplant recipients.

Description: Background: Graft versus host disease (GvHD) and infections by opportunistic pathogens, such as cytomegalovirus (CMV), are causes of significant morbidity and mortality in recipients of allogeneic bone marrow transplants (BMT). Thus, there is a need for methods of graft engineering that inhibit the alloreactive T cells responsible for lethal GvHD without compromising the activity of antiviral T cells. In a murine MHC-mismatched BMT model, we have previously demonstrated that adoptive transfer of polyclonal T cells from donors immunized to murine CMV (MCMV) can decrease viral load but cause lethal GvHD. However, pretreatment of these cells with the psoralen, amotosalen hydrochloride, and ultraviolet A (UVA) light prevents GvHD without compromising antiviral response. We have previously hypothesized that these effects were due to differential sensitivity of naïve and memory T cells to amotosalen/UVA. Recent investigations have demonstrated that CD4 T cells are most directly responsible for lethal GvHD in this model. This observation suggested an alternative hypothesis, equally consistent with previous data, that the observed in vivo effects of amotosalen/UVA treatment are due to differential effects on CD4 and CD8 T cells. The assessment of this new hypothesis is the focus of the current studies. Methods: Two models of T cell activation/proliferation were utilized to test the effects of amotosalen/UVA treatment on CD4 and CD8 cells: stimulation of B6.PL (H-2b) cells with concavalinA, and primary mixed lymphocyte reaction (MLR) between MHC-mismatched B6.PL (H-2b) and BALB/c (H-2d). Responder cells were pretreated with 2nM amotosalen and varying doses of UVA light (0–5 minutes). Proliferation of CD4 and CD8 cells was measured by flow cytometric analysis of CFSE-labeled responder cells. Results: In both systems, CD4 proliferation was effectively eliminated by immediately prior treatment with amotosalen and UVA doses of 1 minute or higher. CD8 proliferation was eliminated by amotosalen and UVA doses of 2 minutes and higher. Both the amount of division on a per cell basis and the overall number of cells that initiated division followed similar trends. Conclusions: These data demonstrate that both CD4 and CD8 T cells are sensitive to treatment with amotosalen/UVA and suggest a subtle difference in sensitivity of CD4 and CD8 populations. Since division of both CD4 or CD8 cells is inhibited at doses of amotosalen/UVA that prevent GvHD but allow anti-viral activity in vivo, it is unlikely that the observed differential sensitivity of T-cell subsets is sufficient to explain the in vivo effects of amotosalen/UVA treatment in this model. Using similar methodologies, ongoing studies are assessing the hypothesis that amotosalen/UVA has differential effects on naïve and mature T cells.