Induction of TGF-beta 1, not regulatory T cells, impairs antiviral immunity in the lung following bone marrow transplant

Abstract

Patients receiving hematopoietic stem cell transplantation or bone marrow transplantation (BMT) as therapy for various malignancies or autoimmune diseases have an increased risk for infectious complications posttransplant, especially in the lung. We have used BMT in mice and murine gammaherpesvirus, gammaHV-68, to study the efficacy of adaptive immune responses post-BMT. Five weeks posttransplant, mice have fully reconstituted their hematopoietic lineages in both the lung and periphery. When challenged with virus, however, BMT mice have a reduced ability to clear lytic virus from the lung. Defective viral control in BMT mice is not related to impaired leukocyte recruitment or defective APC function. Rather, BMT mice are characterized by defective CD4 cell proliferation, skewing of effector CD4 T cells from a Th1 to a Th17 phenotype, and an immunosuppressive lung environment at the time of infection that includes overexpression of TGF-beta1 and PGE(2) and increased numbers of regulatory T cells. Neither indomethacin treatment to block PG synthesis nor anti-CD25 depletion of regulatory T cells improved antiviral host defense post-BMT. Transplanting mice with transgenic bone marrow expressing a dominant-negative TGF-betaRII under the permissive CD4 promoter created mice in which effector CD4 and CD8 cells were unresponsive to TGF-beta1. Mice with TGF-beta1-nonresponsive effector T cells had restored antiviral immunity and improved Th1 responses post-BMT. Thus, our results indicate that overexpression of TGF-beta1 following myeloablative conditioning post-BMT results in impaired effector T cell responses to viral infection.

FIGURE 1

Syngeneic BMT mice have higher viral burden in the lungs at day 7 postinfection 5 wk following BMT. Control and syngeneic BMT mice (conditioned with 650, 900, or 1350 rad) were infected i.n. with 5 × 10⁴ PFU γHV-68 5 wk following BMT. A and B, At day 7 postinfection, left lungs were processed for RNA, and expression of lytic viral genes was measured using real-time RT-PCR. Expression of the viral capsid gene gB and viral DNA polymerase was significantly increased (p = 0.0085 and p = 0.0017, respectively) in the lungs of BMT mice conditioned with 1350 rad when compared with nontransplanted controls (n = at least 9 mice/group; data combined from two experiments). C, At day 7 postinfection, right lungs from BMT mice conditioned with 1350 rad and control mice were harvested for plaque assay (p = 0.0014; n = 10 control, 9 BMT; data combined from two experiments). D, Frozen sections of lungs from control and BMT mice (1350 rad) were prepared at day 7 postinfection and were stained with rabbit polyclonal antisera against γHV-68 or with nonimmune rabbit sera as control (original magnification × 100). The goat anti-rabbit secondary was linked to alkaline phosphatase.

FIGURE 2

BMT mice do not have a defect in inflammatory cell recruitment. A, Whole lungs from uninfected or infected (5 × 10⁴ PFU γHV-68 i.n., day 7 postinfection) nontransplanted control and syngeneic BMT mice were digested with collagenase, and total cells were enumerated. Total cell numbers between control and BMT mice did not differ significantly in the uninfected or infected groups (n = 4 mice/group; representative of at least three independent experiments). B, Seven days postinfection with γHV-68 (5 × 10⁴ PFU, i.n.), whole lungs from control and BMT mice were digested in collagenase and analyzed by flow cytometry for expression of cell surface molecules. Numbers of each subset were not statistically different between control and BMT mice (data in each group is representative of at least two independent experiments; n = at least three mice per group).

FIGURE 3

BMT BMDCs are efficient stimulators in an MLR. BMDCs from nontransplanted control or BMT mice 5 wk posttransplant were grown for 7 d in GM-CSF and analyzed by flow cytometry for expression of I-A^b (A), CD80 (B), and CD86 (C). BMT BMDCs expressed levels of CD80 and CD86 comparable to that of control cells and slightly increased levels of I-A^b (n = 2 mice/group). Similar results were found in a separate experiment where BMDCs were matured for 24 h with IL-4 (data not shown). In D, 2 × 10⁵ irradiated BMDCs from BMT or control mice were used as stimulators in an MLR using 1 × 10⁵ and 2 × 10⁵ BALB/c splenocytes as responders. BMT BMDCs were able to stimulate proliferation of responder cells at least as well as control cells; BMT BMDCs stimulated significantly more proliferation in the 2 × 10⁵ responder group (p = 0.0003). Error bars represent differences between triplicate or quadruplicate wells. Similar results were found when irradiated splenocytes from BMT or control mice were used as MLR stimulators (data not shown).

FIGURE 4

BMT cells are poor responders in an MLR. A total of 2 × 10⁵ BALB/c-irradiated splenocytes were used as stimulators in an MLR. In A, BMT whole spleen cell responders proliferated significantly less than control cells. In B, purified CD4 cells from BMT or control spleens were used as responders in an MLR. BMT CD4 cells responded significantly less than control cells when 2 × 10⁵ and 4 × 10⁵ responders were used. Error bars represent differences between triplicate or quadruplicate wells.

FIGURE 5

Syngeneic BMT mice overexpress PGE₂ and TGF-β1. A, Whole lungs were harvested and assayed for PGE₂ by ELISA at day 7 postinfection with 5 × 10⁴ PFU γHV-68. Lungs of BMT mice produced significantly more PGE₂ than nontransplanted controls (p = 0.0230; n = 5 mice/group; representative of two independent experiments). B, Lungs of uninfected BMT mice produced significantly more total TGF-β1 than control, as determined by ELISA (p = 0.0015; n = 5 mice/group; representative of two independent experiments). C, AECs were isolated from lungs of uninfected BMT, and control mice and were assayed for production of TGF-β1 by ELISA following 24-h culture in serum-free medium. BMT AECs produced significantly more TGF-β1 than control AECs (p < 0.0001; n = 6 mice/group; representative of two independent experiments).

FIGURE 6

BMT mice have elevated numbers of Tregs in the lung. Whole lungs from BMT and control mice, uninfected or infected (5 × 10⁴ PFU γHV-68, i.n., day 7), were digested with collagenase and analyzed by flow cytometry. Lungs from uninfected BMT mice had significantly more CD4⁺Foxp3⁺ cells than nontransplanted controls (p = 0.0002; n = 4 mice/group; representative of two independent experiments). At day 7 post-infection, BMT lungs had significantly higher numbers of CD4⁺Foxp3⁺ cells than nontransplant controls (p = 0.0104; n = 4 mice/group; representative of at least six independent experiments).

FIGURE 7

Allogeneic BMT mice show increased susceptibility to γHV-68 in absence of GVHD. A, Mice receiving syngeneic or allogeneic (C57BL/6→BALB/c) BMT were weighed twice per week as a measure of GVHD for 5 wk post-BMT (n = 5 syngeneic; 10 allogeneic per time point; data representative of two independent experiments). B and C, Left lungs were harvested from control and allogeneic BMT mice at day 7 post-infection with γHV-68, processed for RNA, and analyzed for expression of lytic viral genes. Viral gene expression was significantly increased in allogeneic BMT mice when compared with nontransplanted control mice (p = 0.0225 for viral capsid gene; p = 0.0367 for viral DNA polymerase; n = 3 control, 5 allogeneic BMT; data representative of two independent experiments).

FIGURE 8

Allogeneic BMT mice have increased TGF-β1 and increased Treg numbers in the lungs. A, TGF-β1 levels in the lungs of syngeneic and allogeneic BMT mice were determined by ELISA. Both BMT groups expressed similar increases in TGF-β1 levels (n = 3 control, 3 syngeneic, and 5 allogeneic). B, Control, syngeneic, and allogeneic BMT mice were infected with 5 × 10⁴ PFU γHV-68. At day 7 postinfection, right lungs were digested in collagenase and analyzed by flow cytometry for expression of CD4 and Foxp3. CD4⁺Foxp3⁺ cells were significantly increased in both syngeneic and allogeneic BMT mice compared to non-transplanted controls (n = 7 control, 7 syngeneic, 3 allogeneic).

FIGURE 9

Depletion of Tregs does not restore antiviral immunity in syngeneic BMT mice. A and B, Syngeneic BMT mice were treated with a single dose of either anti-CD25 or isotype control Ab at 4 wk post-BMT. At 5 wk post-BMT (1 wk following Ab treatment), mice were infected with 5 × 10⁴ PFU γHV-68. Left lungs were harvested for RNA at day 7 post-infection, and expression of lytic viral genes was determined by real-time RT-PCR. Expression of both the capsid gene gB and viral DNA polymerase was significantly increased in BMT mice treated with either isotype or anti-CD25 compared with isotype control-treated, nontransplanted animals (n = 5 mice/group). C, One week following Ab treatment, uninfected whole lungs were digested in collagenase and analyzed by flow cytometry for expression of CD4 and Foxp3. Numbers of CD4⁺Foxp3⁺ cells were significantly increased in the isotype-treated BMT mice but were at control levels in anti–CD25-treated mice (n = 5 mice/group). Similar results were found at day 7 postinfection with γHV-68 (data not shown).

FIGURE 10

Transplanting mice with CD4-DN-TGF-βRII bone marrow restores immunity to γHV-68. Control mice and mice transplanted with syngeneic wild-type or CD4-DN-TGF-βRII bone marrow were infected i.n. with 5 × 10⁴ PFU γHV-68 and analyzed at day 7 postinfection. A, Expression of the viral capsid gene gB was significantly increased in the lungs of wild-type BMT mice compared with nontransplanted control mice; however, there was no significant difference between control and CD4-DN-TGF-βRII BMT groups. B, Expression of viral DNA polymerase was significantly increased in wild-type BMT lungs when compared with nontransplanted control mice. There was no significant difference between the control and CD4-DN-TGF-βRII BMT mice (n = 4–5 mice/group; data representative of two independent experiments). C, Right lungs were digested in collagenase and analyzed using flow cytometry. Both wild-type and CD4-DN-TGF-βRII BMT mice had significantly higher numbers of Tregs compared with nontransplanted controls (n = 5 mice/group; data representative of two independent experiments).

FIGURE 11

Altered T cell differentiation in lungs of BMT mice in response to γHV-68. Syngeneic BMT and control mice were infected with 5 × 10⁴ PFU γHV-68. At day 7 postinfection, lungs were digested in collagenase. Cells were then stimulated with PMA and ionomycin and analyzed by flow cytometry by using Abs against CD4, IFN-γ, and IL-17a. A, BMT mice showed a significant decrease in numbers of CD4⁺IFN-γ⁺ cells compared with nontransplanted control mice (p = 0.0207). B, BMT lungs had a significant increase (p = 0.0002) in numbers of CD4⁺IL-17a⁺ cells compared with nontransplant controls (n = 5 mice/group; data representative of two independent experiments).

FIGURE 12

T cells from CD4-DN-TGF-βRII BMT mice express higher amounts of IFN-γ and lesser amounts of IL-17a. Splenocytes from syngeneic BMT and CD4-DN-TGF-βRII BMT mice were depleted of CD19-expressing cells via magnetic separation. Cells were then cultured with PMA and ionomycin for 24 h. Supernatants were harvested and assayed for expression of IFN-γ and IL-17a by ELISA. A, Syngeneic BMT cells expressed significantly less IFN-γ than CD4-DN-TGF-βRII BMT cells (p = 0.0004; n = 5 mice/group). B, Syngeneic BMT cells expressed significantly more IL-17a than CD4-DN-TGF-βRII BMT cells (p = 0.0006; n = 5 mice/group).