Preemptive HMG-CoA reductase inhibition provides graft-versus-host disease protection by Th-2 polarization while sparing graft-versus-leukemia activity

Abstract

We investigated whether atorvastatin (AT) was capable of protecting animals from acute graft-versus-host disease (aGVHD) across major histocompatibility complex (MHC) mismatch barriers. AT treatment of the donor induced a Th-2 cytokine profile in the adoptively transferred T cells and reduced their in vivo expansion, which translated into significantly reduced aGVHD lethality. Host treatment down-regulated costimulatory molecules and MHC class II expression on recipient antigen-presenting cells (APCs) and enhanced the protective statin effect, without impacting graft-versus-leukemia (GVL) activity. The AT effect was partially reversed in STAT6(-/-) donors and abrogated by L-mevalonate, indicating the relevance of STAT6 signaling and the L-mevalonate pathway for AT-mediated aGVHD protection. AT reduced prenylation levels of GTPases, abolished T-bet expression, and increased c-MAF and GATA-3 protein in vivo. Thus, AT has significant protective impact on aGVHD lethality by Th-2 polarization and inhibition of an uncontrolled Th-1 response while maintaining GVL activity, which is of great clinical relevance given the modest toxicity profile of AT.

Figures

Figure 1

Donor pretreatment with atorvastatin improves survival after major MHC mismatch BMT by reducing alloreactive T-cell expansion in vivo. (A) Balb/c mice were given 5 × 106 TCD-BM cells and 1.2 × 106 CD4+/CD8+ (4:1) T cells (both H-2Kq) after lethal irradiation with 800 cGy. Donor animals were fed with either PBS or AT with or without L-mevalonate by oral garvage for 10 days prior to transplantation. Survival of mice receiving TCD-BM (▵, n = 15), with T cells from donors treated with PBS (□, n = 15), atorvastatin 1 mg/kg (○, AT1, n = 15), atorvastatin 10 mg/kg (■, AT10, n = 15), AT1 plus L-mevalonate (◇, n = 10), AT10 plus L-mevalonate (●, n = 10). Percentage survival of Balb/c recipients is significantly higher when T cells are derived from donors treated with atorvastatin 1 mg/kg or 10 mg/kg as compared with PBS (○ versus □, P < .001; ■ versus □, P < .001). Survival data from 2 independent experiments including 8 and 7 recipients, respectively, are combined. (B) Expansion of luciferase-labeled T cells was quantified in emitted photons over total body area at serial time points after BMT. BLI signal intensity of mice receiving TCD-BM (▵, n = 15) with T cells from donors treated with PBS (□, n = 15), atorvastatin 1 mg/kg (○, AT1, n = 15), atorvastatin 10 mg/kg (■, AT10, n = 15), AT1 plus L-mevalonate (◇, n = 10), or AT10 plus L-mevalonate (●, n = 10). Signal intensity is significantly higher in animals receiving T cells from PBS-treated as compared with AT1- or AT10-treated animals (□ vs ○, P = .007; □ vs ■, P = .003). (C) Single time points depicting the expansion of luciferase transgenic (luc+) donor T cells in representative Balb/c mice receiving T cells from donors treated as indicated on top of each column for days 4, 8, and 12 after BMT. (D) Cytokine levels were measured in the supernatant of cocultures combining irradiated (30 Gy) APCs (Stimulator, H-2Kd) with T cells (Responder, H-2Kq) of each cell type, 2 × 105 cells/well in flat-bottomed, 96-well plates derived from donors that were fed for 10 days with PBS or atorvastatin 10 mg/kg (AT) as indicated in the respective histogram (*P < .05). ELISA was performed as described in detail in “ELISA-based cytokine analysis.”

Figure 2

In vitro statin treatment reduces T-cell expansion in response to alloantigen in vitro and in vivo. (A) In vitro expansion of CFSE- (A) or luciferase-labeled (B) CD4+/CD8+ T cells (4:1, FVB/N, 2 × 105 T cells/ flat-bottomed, 96-well plate, 200 μL per well) in the presence of CD3/CD28 stimulation (2 μg/mL or 5 μg/mL each) for 48 hours in complete media (cRPMI) and the indicated statins (10 μM). Afterward, cells were washed and analyzed by FACS for CFSE dilution. cRPMI indicates no statin; AT, atorvastatin; SI, simvastatin; FLU, fluvastatin; LO, lovastatin; Unstim, no CD3/CD28 mAb stimulation. T cells expand upon in vitro stimulation (upper panel). AT, SI, FLU, and LO reduce T-cell expansion significantly as measured by levels of CFSE dilution. (B) The same experimental in vitro conditions as described in panel A were employed. Expansion of luciferase transgenic T cells was quantified in photons/second/cm2. All statins reduce in vitro T-cell expansion significantly as compared with cRPMI (P < .01). (C) Survival of mice receiving TCD-BM (●, n = 10) with T cells exposed to CD3/CD28 stimulation as described in panel A in cRPMI alone (▵, n = 10) or with FLU (○, n = 10) or AT (■, n = 10). Survival of Balb/c recipients was significantly higher when T cells were exposed to AT or FLU as compared with cRPMI (■ vs ▵, P = .003; and ○ vs ▵, P = .004). All Balb/c recipients (H-2d) underwent transplantation as described in “aGVHD model.” Data from 2 independent experiments are combined. (D) Expansion of luciferase-labeled (luc+) T cells was quantified in emitted photons over total body area at serial time points after BMT in mice receiving TCD-BM (●, n = 10) with T cells exposed to cRPMI (▵, n = 10), fluvastatin (FLU, ○, n = 10), or atorvastatin (AT, ■, n = 10). Signal intensity is significantly higher in animals receiving T cells exposed to cRPMI only as compared with FLU- and AT-treated animals (▵ vs ○, P < .001; and ▵ vs ■, P < .001). Data from 2 independent experiments are combined.

Figure 3

AT and FLU recipient treatment reduced expression of costimulatory molecules and MHC class II on antigen-presenting cells in liver and lymphoid organs. (A) Representative flow cytometric analysis of the cell surface expression of CD80, CD86, CD40, and MHC class II on CD11c+ APCs isolated from Balb/c recipients after 10 days of feeding with phosphate-buffered saline (PBS, heavy black line), atorvastatin (AT, red line), or fluvastatin (FLU, blue line) prior to transplantation (day 0). The thin black line represents the isotype control. (B) Recipient-type APCs were isolated on day 2 after transplantation from the indicated organs. The mean fluorescence intensity of the H-2KdCD11c gated population is quantified for the indicated surface molecules (PBS vs AT, *P < .05). (C) IL-4 and IL-10 expression within the CD4+ T-cell population derived from the spleen prior to transplantation is shown. AT and FLU pretreatment as described in panel A increases the frequency of CD4+IL-4+IL-10+ T cells as well as CD4+IL-4−IL-10+ Tr-1 regulatory cells. One representative FACS analysis of 3 independent experiments is shown. (D) Balb/c mice were given 5 × 106 TCD-BM cells and 1.2 × 106 CD4+/CD8+ (4:1) T cells (both H-2q) after lethal irradiation with 800 cGy. Recipient animals were fed with either PBS or the indicated statin by oral gavage for 10 days prior to transplantation. Survival of mice receiving TCD-BM (●, n = 10) plus T cells, and being pretreated with PBS (▵, n = 10), fluvastatin (○, n = 10), or atorvastatin (■, n = 10). Survival of Balb/c recipients is significantly higher when recipients are treated with FLU or AT as compared with PBS (○ vs ▵, P = .004; ■ vs ▵, P = .006). (E) Expansion of luciferase-labeled (luc+) T cells was quantified in emitted photons over total body area at serial time points after BMT in mice receiving TCD-BM (●, n = 10) and being pretreated with PBS (▵, n = 10), fluvastatin (FLU, ○, n = 10), or atorvastatin (AT, ■, n = 10). Signal intensity is significantly higher in animals pretreated with PBS only as compared with FLU- and AT-treated animals (▵ vs ○, P = .003; ▵ vs ■, P < .003). Data from 2 independent experiments are combined.

Figure 4

Combined donor and recipient feeding has an additive protective effect and protection is predicted by the kinetics of Th-2 biased donor T cells in secondary lymphoid organs and the liver. (A) Survival rate of mice receiving TCD-BM alone (●, n = 10). Survival rate of recipients of TCD-BM with T cells when donor and recipient were treated with PBS (▵, n = 10) or AT (AT(D + R), ○, n = 10). Recipients of TCD-BM with T cells when the donor (□, n = 10) or the recipient (▾, n = 10) was treated with AT. Survival percentage of Balb/c recipients when both recipient and donor were fed with AT (65%), when only the T-cell/BM donors were fed with AT (43%) and when only the recipient was fed with AT (40%). (B) Serum was collected from recipients on the indicated days after BMT and IFN-γ and TNF levels were determined by ELISA (*P < .05, **P < .01). (C) Adoptively transferred CD4+H-2Kq+ donor T cells were isolated from the indicated organs and analyzed for their intracellular cytokine profile on days 3, 15, and 25. The y-axis depicts the percentage of IL-4–, IL-10–, TNF-, or IFN-γ–positive CD4+ T cells within the total CD4+H-2Kq+ donor population. Data presented are derived from 2 independent experiments (3 mice per time point, values are mean plus or minus standard deviation [SD]).

Figure 5

STAT6 deficiency allows only for partial protection by AT while L-mevalonate reverses the protective statin effect. C57BL/6 mice (H-2b, Th-1.1) were given 5 × 106 TCD-BM cells and 2 × 106 CD4+/CD8+ (4:1) T cells (both H-2d, Thy-1.2) after lethal irradiation with 800 cGy. Donor animals were fed with either PBS or AT for 10 days prior to transplantation. (A) Survival of mice receiving TCD-BM (▵, n = 15) with T cells from PBS-treated donors (□, n = 15), atorvastatin-treated (AT, 10 mg/kg) wt donors (■, n = 15), and STAT6-deficient donors treated with PBS (●, n = 15) or AT (○, n = 15). Percentage survival of Balb/c recipients is significantly higher when T cells are derived from wt donors as compared with STAT6-deficient donors when both are treated with AT (■ vs ○, P = .02). Data from 2 independent experiments are combined. (B) Survival of mice receiving TCD-BM (▵, n = 15) with T cells from PBS-treated donors (□, n = 15), atorvastatin-treated (AT, 10 mg/kg) donors (■, n = 15), AT with L-mevalonate pretreatment (○, n = 15), and PBS with L-mevalonate (●, n = 15). Percentage survival of Balb/c recipients is significantly higher when T cells are derived from AT as compared with AT with L-mevalonate–treated donors (■ vs ○, P = .004). Data from 2 independent experiments are combined. (C) In vivo expansion of CFSE-labeled Balb/c (H-2Kd, Thy-1.2+) CD4 T cells in C57BL/6 recipients (H-2Kd Thy-1.1+) on day 3 after BMT is depicted. Percentages of T cells having undergone cell division are indicated in the left upper corner of each histogram. Proliferation of T cells in the group pretreated with AT is significantly reduced as compared with PBS, STAT6−/− with AT, and AT with L-mevalonate (P < .05). (D) At 10 days after transplantation, mice from the indicated group were killed and sections of small bowel, large bowel, and liver were stained with H&E. Slides were analyzed by a pathologist blinded to the treatment groups (N.K.) for evidence of pathologic damage. Histopathologic scoring was performed as described in “GVHD histopathology scoring” with a range from 0 (absence of GVHD signs) to 4 (maximal GVHD damage). Data are pooled from 3 experiments, representing 6 animals per group. (E) Histopathologic evaluation of representative liver samples collected 10 days after transplantation is shown for the PBS group (i) and the groups where donor T cells are derived from AT-treated wt (ii) or STAT6−/− (iii) donors. Maximal liver tissue damage and strong lymphocytic infiltration surrounding the bile duct region as indicated by the solid arrow (i), minimal tissue damage (ii), intermediate liver damage, and low degree of lymphocytic infiltration surrounding the bile duct (iii) (dashed arrow). Magnification is ×400.

Figure 6

Differential impact of AT on RAS, RAP-1, and Rho-B prenylation, T-bet, GATA3, and c-MAF expression. (A) CD4+ T cells (H-2Kq) were exposed to irradiated allogeneic APCs (H-2Kd) of each cell type, 2 × 106 T cells/flat-bottomed, 6-well plate for 48 hours in the presence or absence of AT (10 μM), harvested and analyzed by intracellular cytokine staining with anti–IL-10, –IL-4, –TNF-α, or –IFN-γ. Mean fluorescence for the representative cytokines of CD4+H-2Kq+ gated cells is presented. (B) Cells cultured under the conditions described in panel A were then washed and responder T cells (H-2Kq) were re-isolated by MACS with anti–H-2Kq-FITC and anti-FITC beads. Protein samples from in vitro–treated T cells were subject to Western blot analysis. AT is effective at inhibiting the prenylation of Ras, Rap-1, and Rho-B in the allogeneic activation culture. Increased amounts of nonprenylated proteins are found when AT is present (solid arrow). (C) T cells were harvested from AT-treated donors on day 0 (left panel) or on day 3 (right panel) after BMT from the recipients' secondary lymphoid organs and were analyzed by Western blot for the indicated total or phosphorylated proteins. (D) Mean fluorescence for the representative pSTATs is analyzed by phospho-flow analysis of CD4+H-2Kq+ gated cells, presented for the respective day prior to or after BMT. MFI for pSTAT4 was 11.4 (± 0.7) versus 47.4 (± 3.4) for the AT versus the PBS group, respectively, prior to transfer into the irradiated host (P < .05). MFI for pSTAT4 was 97.2 (± 2.1) versus 541.4 (± 11.3) for the AT versus the PBS group, respectively, on day 3 after BMT (P < .01). (E) Donor type T cells derived from AT-treated donors are presented for the respective day prior to or after BMT. Mean fluorescence for c-MAF, T-bet, and GATA-3 within CD4+H-2Kq+ gated cells is presented. MFI for T-bet was 12.4 (± 0.8) versus 98.4 (± 6.3) for the AT versus the PBS group, respectively, prior to transfer into the irradiated host (P < .01). MFI for T-bet was 137.2 (± 4.1) versus 21.4 (± 5.2) for the PBS versus the AT group, respectively, on day 3 after BMT (P < .01).

Figure 7

GVL effector function of cytolytic T cells is preserved in the presence of atorvastatin treatment of donor and recipient. Tumor growth and GVL effect are visualized by bioluminescence imaging (A), recipient survival (B), emitted photons over time (C), and FACS analysis (D). All Balb/c recipients (H-2Kd) were given 105 A20 leukemia cells (H-2Kd) on day 0 and BMT was performed as described in “Tumor models.” (A) Representative animals of each group display similar luc+A20 tumor cell localization on day 4 (top row) after BMT. Tumor elimination in animals receiving T cells on days 8 (middle row) and 16 (bottom row). Mice underwent transplantation with TCD-BM after PBS (first column) or AT (second column) treatment, TCD-BM plus T cells exposed to PBS (third column) or AT (fourth column). Tumor infiltration is found in the following locations: ST, sternum; HU, humerus; FE, femur; SPL, spleen. (B) Survival of BALB/c mice receiving TCD-BM alone (●, n = 10), TCD-BM plus A20 leukemia cells after pretreatment of donor and recipient with PBS (▵, n = 10) or AT (□, n = 10), TCD-BM plus A20 leukemia cells plus T cells after pretreatment of donor and recipient with PBS (*, n = 10) or AT (○, n = 10). Survival: ○ versus *, P < .001. Data from 2 independent experiments are combined. (C) Expansion of luc+A20 tumor cell as measured in photons over total body area (photons/second/mouse). Animals rejecting the A20 leukemia cells demonstrate a durable signal loss as depicted until day 112 after BMT. Data from 2 independent experiments are combined. (D) Bone marrow samples with yfp+ A20 cells of recipients on days 7 and 15 after transplantation with TCD-BM plus A20 leukemia cells after pretreatment of donor and recipient with PBS or AT, TCD-BM plus A20 leukemia cells plus T cells after pretreatment of donor and recipient with PBS or AT as indicated for the respective column. (E) In vitro cytolytic effector function of CD8+ T cells (FVB/N) derived from animals treated with PBS or AT. Target cells were A20 or Yac1 tumor cells and chromium release assay was performed as described in “51Cr release cytotoxicity assay.”