Absence of Vasoactive intestinal peptide expression in hematopoietic cells enhances Th1 polarization and anti-viral immunity in mice

Abstract

Vasoactive intestinal peptide (VIP) induces regulatory dendritic cells (DC) in vitro that inhibit cellular immune responses. We tested the role of physiological levels of VIP on immune responses to murine cytomegalovirus (mCMV) using VIP-knockout (VIP-KO) mice and radiation chimeras engrafted with syngenic VIP-KO hematopoietic cells. VIP-KO mice and had less weight loss and better survival following mCMV infection compared with wild-type littermates (WT). MCMV-infected VIP-KO mice had lower viral loads, faster clearance of virus, with increased numbers of IFN-γ⁺ NK and NKT cells, and enhanced cytolytic activity of NK cells. Adaptive anti-viral cellular immunity was increased in mCMV-infected VIP-KO mice compared with WT mice, with more Th1/Tc1 polarized T-cells, fewer IL-10⁺ T-cells, and more mCMV-M45 epitope peptide-MHC class I-tetramer⁺ CD8⁺ T-cells (tetramer⁺ CD8 T-cells). MCMV-immune VIP-KO mice had enhanced ability to clear mCMV-peptide pulsed target cells in vivo. Enhanced anti-viral immunity was also seen in WT transplant recipients engrafted with VIP-KO hematopoietic cells, indicating that VIP synthesized by neuronal cells did not suppress immune responses. Following mCMV infection there was a marked up-regulation of MHC class II (MHC-II) and CD80 co-stimulatory molecule expression on DC from VIP-KO mice compared with DC from WT mice, while PD-1 and PD-L1 expression were up-regulated in activated CD8⁺ T-cells and DC, respectively, in WT mice but not in VIP-KO mice. Since the absence of VIP in immune cells increased innate and adaptive anti-viral immunity by altering co-stimulatory and co-inhibitory pathways, selective targeting of VIP-signaling represents an attractive therapeutic target to enhance anti-viral immunity.

Introduction

Vasoactive intestinal peptide (VIP) is a multifunctional endogenous polypeptide that modulates both innate and adaptive immunity at multiple levels of immune cell differentiation and activation(1). VIP is secreted by neurons (in both the central and peripheral nervous systems) (2) and by B-cells, T-cells, accessory cells and other non-lymphoid cells (3-6). VIP and the closely related neuropeptide pituitary adenylyl cyclase-activating polypeptide (PACAP) bind to three known receptors: VPAC1, VPAC2, and PAC1. T-cells and dendritic cells (DC) express VPAC1 and VPAC2, but not PAC1(1). PAC1 is mainly expressed on neuron and endocrine cells in the brain and pituitary and adrenal glands, and selectively binds PACAP (7). Even though VIP and PACAP signal through the same receptors, PACAP does not fully compensate for the loss of VIP in VIP-KO mice (8). VIP-KO mice lack compensatory increase in PACAP peptide expression and expression of the VPAC1 and VPAC2 VIP receptors are diminished in brains of VIP-KO mice (8).

In adaptive immune responses, VIP polarizes CD4⁺ T-cells to an immunosuppressive Th2 response while suppressing the Th1 responses (9). T-cell activation and differentiation induce VPAC2 expression, while VPAC1 is down-regulated following stimulation of human blood T-cells with anti-CD3 monoclonal antibody plus PMA (10). VIP also acts on APC and regulates their function. Through the VPAC1 receptor, VIP leads to the development of bone marrow-derived tolerogenic DCs in vitro and in vivo (11). In a mouse model of allogeneic bone marrow transplantation, DC that were differentiated in the presence of VIP, and then transplanted along with bone marrow cells and splenic T-cells, induced the generation of regulatory T-cells and protected mice from acute graft versus host disease (12). Th2 polarization of immune responses by VIP-differentiated DC is likely achieved through VIP down regulation of co-stimulatory signals on antigen presenting cells (APC) and inhibition of IL-1, TNF-α, IL-6, and IL-12 production (13). VIP suppresses the expression of the pattern recognition receptors toll-like receptor (TLR) 2 and TLR4 on APC (14, 15) and inhibits TLR3-signaling (16). Conversely, activation of APC through binding of ligands to TLR2, TLR4, and TLR7 down-regulate VPAC2 expression (17).

Given the manifold effects of VIP on innate and adaptive immune responses, we explored the role of VIP in anti-viral responses to cytomegalovirus (CMV). Opportunistic CMV infection causes significant morbidity and transplant-related mortality in allogeneic BMT patients, and the pathogenesis of mouse cytomegalovirus (mCMV) infection in mice is similar to that in human CMV (hCMV) infection (18, 19). MCMV and hCMV exhibit 70% sequence similarity, comparable to the global level of DNA sequence homology between their natural hosts (20) and are predicted to contain approximately 170 and 165 open reading frames (ORFs), respectively (21, 22). The large number of homogeneous ORFs indicates that the two viruses are related, although immune evasive strategies of mCMV infection are quite different from those seen following hCMV infection (20) suggesting specific adaptation of a common ancestor virus to the immune environments of mice and humans (23). Furthermore, mice and humans have similar specific immune responses to their respective CMV (21, 24-26), with coordinated activities of innate and adaptive immune cells including DC, macrophages, natural killer (NK) cells, T-cells and B-cells (27-32). While cellular and humoral immune response to mCMV are robust an effective in clearing the virus, mCMV infection also leads to immuno-suppressive effects including expression of m144, a MHC class-I (MHC-I) decoy that binds to NK cells and inhibits anti-viral cytotoxicity (33, 34), and induction of a “paralyzed” DC phenotype, characterized by down-regulation of MHC-I and -II, co-stimulatory molecules, and pro-inflammatory cytokines (32). Hence, we were interested in whether interference with VIP-signaling could enhance immune responses to mCMV infection. Previous studies have explored the effect of VIP on inflammation and allogeneic immunity using supra-physiological, pharmacological administration of purified VIP peptide agonist (3, 9). To study the immuno-modulatory effects and anti-viral immunity of physiological levels of VIP, we used VIP-KO mice (35) and VIP-KO hematopoietic chimeras (36). We hypothesized that mice lacking VIP expression would show an increased response to viral infection due to a lack of immunosuppressive counter-regulatory activity from DCs. We challenged VIP-KO mice and radiation chimeras engrafted with VIP-KO hematopoietic cells with two sources of mCMV antigen: a Listeria monocytogenes vaccine that expresses an immuno-dominant CMV peptide (Lm-MCMV vaccine)(37, 38), and an infectious strain of mCMV (37, 39). Our results demonstrate that VIP-KO mice and recipients engrafted with VIP-KO hematopoietic cells have augmented cellular immune responses to mCMV antigen, and improved survival after viral infection. The kinetics of antigen-specific primary and secondary immune responses were accelerated in VIP-KO mice and in mice reconstituted with VIP-KO hematopoietic cells, supporting the role of VIP in immune counter-regulatory pathways.

Materials And Methods

Mice

B6 strain (H-2K^b, CD45.2, CD90.2) vasoactive intestinal peptide/peptide histidine isoleucine (VIP/PHI) knockout (KO) mice (VIP-KO) have been previously described (35). Both male and female VIP KO mice were used in experiments, using syngenic siblings as wild-type (WT) controls. Congenic strains of B6 mice were purchased from Jackson Laboratory (Bar Harbor, Maine) (H-2K^b, CD45.1, CD90.2) or were bred at the Emory University Animal Care Facility (Atlanta, GA) (H-2K^b, CD45.1/CD45.2). All mice were 8-10 weeks old. Procedures conformed to the Guide for the Care and Use of Laboratory Animals, and were approved by the Emory University Institutional Animal Care and Use Committee (IACUC). According to IACUC guidelines, any mouse that lost ≥ 25% bodyweight was euthanized and recorded as dying on the following day for statistical analysis.

Donor Cell Preparation for transplantation

Bone marrow transplantation was performed to create chimeric mice with hematopoietic cells from VIP-KO donors or WT donors (control). Femora, tibia, and spleens were obtained from VIP-KO or WT mice. Bone marrow cells were harvested by flushing the specimens with sterile RPMI-1640 containing 1% heat-inactivated fetal calf serum (RPMI/FCS). T-cells were purified from splenocytes by negative selection using a cocktail of biotinylated non-T-cell antibodies (anti-CD11b, B220, DX5, and Ter119), streptavidin microbeads and immuno-magnetic separation (MACS, Miltenyi Biotech, Auburn, CA). The average purity of CD3⁺ T-cells was 95%. Lineage^- (CD3, CD4, CD8, Gr-1, CD11b, I-A^b, DX5, B220, TER119 and CD19) c-kit⁺ sca-1⁺ hematopoietic stem cells (HSC) and lineage^- (CD3, DX5, IgM, TER119 and CD19) CD11c⁺ DC from donor BM were purified using a Becton Dickinson FACS Aria cell sorter (36). Purity of FACS-purified HSC and DC averaged 93% and 97%, respectively.

Radiation Chimeras and Stem Cell Transplantation

On day -1, 8-10 week old male B6 CD45.1 congenic mice were irradiated with two fractions of 5.5 Gy for a total of 11Gy (40). On day 0, irradiated mice received 5 × 10⁶ TCD-BM cells plus 3 × 10⁵ MACS purified splenic T-cells via tail vein injection. Some experiments used an alternate approach, transplanting a combination of 5 × 10³ HSC, 5 × 10⁴ DC, plus 3 × 10⁵ T-cells. Mice were monitored for signs of severe infection including fur texture, posture, activity, skin integrity, and weight loss. Each transplant group was followed for at least 100 days (41). Donor cell chimerism in peripheral blood was determined 2 months after transplantation, and was typically ≥ 95%. Chimeric mice were then used in vaccination and mCMV infection studies.

Virus and Immunization

The Smith strain of mCMV passaged in vivo in salivary glands and frozen in aliquots in liquid nitrogen (37, 39). WT and VIP-KO mice, as well as chimeric mice with hematopoietic cells from WT and VIP-KO donors, were given either 5 × 10⁴ (LD10; low dose) or 1 × 10⁵ (LD 50; high dose) plaque-forming unit (PFU) mCMV by intraperitoneal injection and then monitored for signs of illness including hunched posture, decreased activity, and weight loss. Mice were vaccinated intraperitoneally with 1 × 10⁶ colony-forming unit (CFU) Lm-MCMV, a Listeria monocytogenes which has been rendered non-pathogenic by knock-out of bacterial genes associated with virulence (42) and engineered to express the mCMV H-2D^b immuno-dominant peptide M45 aa-985~993- HGIRNASFI (43). The vaccine was prepared and supplied by Cerus Corporation (Concord, CA) (37, 38).

Analysis of Peripheral Blood and Spleen Samples

Blood and spleen samples were obtained on 3, 7, 10, 14, 17 and 21 days after vaccination or following mCMV infection. Leukocytes, red blood cells and platelets were counted using a Beckman Coulter automated counter. Blood and spleen samples were depleted of red blood cells by ammonium chloride lysis and washed twice. NK, NK-T, and T-cell subsets were enumerated using CD3 PE/PE-Cy7/FITC, CD4 PE-Alexa610/PE-Alexa700, CD8 PE-Cy7/Per-CP, CD62L FITC/APC, CD25 APC-Cy7, CD44 PE-Cy5, CD69 PE-Cy7, PD-1 PE, and NK1.1 PE (Pharmingen). Cells were stained with monoclonal antibodies specific for congenic markers CD45.2, CD45.1, CD90.1 and CD90.2 to determine donor chimerism. APC labeled mCMV M45 aa-985~993- peptide-HGIRNASFI-H-2D^b tetramer was obtained from the Emory Tetramer Core Facility. All samples were analyzed on a FACS Canto (Beckon Dickinson, San Jose, CA) and list mode files were analyzed using FlowJo software (Tree Star, Inc. 2007). Samples for flow cytometric analysis of mCMV-M45 epitope peptide-MHC-I tetramer⁺ CD8⁺ T-cells (tetramer⁺ CD8 T-cells) were gated for lymphocytes in the area of FSC and SSC, and setting a gate for tetramer⁺ T-cells such that 0.01% of control (non-immune) CD8⁺ T-cells were positive (37, 39). Flow cytometric analyses of the Treg-associated molecule PD-1 (44), the co-stimulatory molecule ICOS, the adhesion molecule CD62L (45), activation markers CD25 and CD69 (36, 46), intracellular cytokines (IFN-γ, TNF-α, IL4 and IL-10), and DC markers (I-A^b, CD80, and PD-L1) were analyzed as previously described (36).

In vivo Killing Assay

Naive splenocytes were harvested from CD45.1⁺/CD45.2⁺ heterozygous C57BL/6 mice and pulsed with 3 μM mCMV M45 aa-985~993- HGIRNASFI peptide in RPMI 1640 containing 3% FBS for 90 min at 37°C, and washed three times with ice-cold media. MCMV peptide-pulsed target splenocytes and non-pulsed splenocytes from CD45.1⁺ B6 congenic mice were mixed together in equal parts 40 × 10⁶ total target cells per mouse were injected i.v. into CD45.2⁺ VIP-KO or WT C57BL/6 mice that had been infected 9 days earlier with low dose (LD10) mCMV, or injected into non-infected WT control mice Sixteen hours following injection of target cells, recipients were sacrificed, splenocytes harvested, and the numbers of mCMV peptide-pulsed CD45.1⁺/CD45.2⁺ and non-pulsed CD45.1⁺ target cells quantified by FACS analysis. Immune mediated killing of mCMV peptide pulsed targets was calculated by first dividing the percentage of peptide-pulsed or non-pulsed targets recovered from the spleen of mCMV-immune mice with the mean percentage of the corresponding population of peptide-pulsed or non-pulsed targets from non-immune mice (ratio of immune killing). The specific anti-viral in vivo lytic activity for individual mice were calculated by the formula: (1- (ratio of immune killing mCMV-peptide pulsed-target cells/ ratio of immune killing non- pulsed target cells)) × 100.

In vitro Measurements of Immune Responses to mCMV Peptide

WT mice, VIP-KO mice, and mice engrafted with either WT or VIP-KO donor cells were infected with low dose mCMV and splenocytes were harvested 15 days later. Splenic DC and T-cells were purified by FACS and MACS, respectively (36). DC were plated at 2 × 10⁵ cells/mL in 12-well plates and centrifuged (300 × g for 30 min) with 3 μM mCMV peptide (37). After centrifugation, DC were washed 3 times with PBS, resuspended in complete medium, and incubated with 2 × 10⁶ T-cells at 37°C for 3 or 7 days (47). Cells were treated with Golgi Stop (Pharmingen, San Jose, CA) during the last 6 hours of culture. Cells were then harvested from culture plates and stained with fluorescently-labeled antibodies against DC and T-cell lineage markers (36), permeabilized, and stained with antibodies against IL-10 and IFN-γ, and analyzed by flow cytometry as previously described, using isotype-matched control antibodies to set the gates for distinguishing positive intracellular staining (36). Harvested culture media was stored at −20°C until use for cytokine analysis by ELISA (OptEIA ELISA sets for IL-10 and IFN-γ; BD Biosciences). ELISA plates were read using a SpectraMax 340PC spectrophotometer (Molecular Devices, Sunnyvale, CA)(36).

NK Cell Lytic Activity

Lytic activity of NK cells was analyzed as previously described (48) Briefly, YAC-1 cells, a sensitive target for NK cells, were labeled with 37 MBq of Na⁵¹CrO₄ at 37°C for 90 min and washed twice with warm RPMI 1640 medium. The labeled target cells (1×10⁴) were co-cultured with effector splenocytes (containing NK cells) at various ratios of effectors: targets (100:1, 50:1, and 25:1) in a final volume of 0.2 ml fresh medium in 96-well round bottom microplates. The plates were incubated for 4 hours at 37°C with 5% CO₂. The amount of ⁵¹Cr released in 0.1 ml supernatant was measured by a well-type gamma counter (Beta Liquid Scintillation Counter, EG&G Wallac, Perkin-Elmer, Ontario, Canada). Specific cytotoxicity was calculated as: % ⁵¹Cr release = 100 × (cpm experimental−cpm spontaneous release)/(cpm maximum release–cpm spontaneous release).

Determination Of Liver Viral Load

Viral load was analyzed as previously described (39). Briefly, livers were collected from CMV-infected recipients, homogenized, and centrifuged. Serially diluted supernatants were added to 3T3 confluent monolayers in 24-well tissue culture plates and incubated for 90 minutes at 37°C and 5% CO₂, then over layered with 1 mL 2.5% methylcellulose in DMEM and returned to the incubator. After 4 days, the methylcellulose was removed and the 3T3 confluent monolayers were stained with methylene blue. MCMV plaques were directly counted under a light microscope (Nikon, Melville, New York) PFUs were calculated.

Statistical Analyses

The data were analyzed using SPSS version 18 for MAC. In this study each treatment group (or time point) had 4-5 mice, and every experiment was repeated at least 2 times. Data are presented as mean ± SD of all evaluable samples if not specified. Survival differences among groups were calculated with the Kaplan-Meier log-rank test in a pair-wise fashion. Differences in tetramer response, cytokine levels, and T-cell numbers were compared using a 2-tailed Student’s t-test. A p-value of less than 0.05 was considered significant.

RESULTS

VIP-KO mice were resistant to mCMV infection

We first compared the hematological and immunological phenotypes of VIP-KO mice. We found no significant differences comparing blood from naïve WT and VIP-KO mice in the numbers of total leukocytes, CD4, CD8, αβ TCR T-cells, γδ T-cells, B-cells, myeloid leukocytes, and DCs in blood (Supplementary Figure 1). VIP-KO and WT mice were infected with a non-lethal dose of mCMV (5 × 10⁴ PFU) and sacrificed 3, 10 and 17 days later, VIP-KO mice had significantly less virus in their liver, a target for mCMV infection (37, 49), with more rapid clearance of virus than mCMV infected WT mice (p< 0.001; Figure 1). To test whether VIP-KO mice had better survival following mCMV infection, VIP-KO and WT mice were infected intraperitoneally with either 1 × 10⁵ PFU /mouse (high-dose) or 5 × 10⁴ PFU/mouse (low dose) mCMV. All WT mice given high-dose mCMV died by day 10 post-infection compared with 65% survival of the VIP-KO mice (p< 0.001, Figure 2A). Following low-dose mCMV infection both WT and VIP-KO mice had transient lethargy and weight-loss, with recovery to baseline values by day 20 post-infection, with 100% of WT mice and 92% of VIP-KO mice surviving to day 100 post-infection (Figure 2A, B). In a parallel experiment, serial measurements of CD4 and CD8 T-cells following mCMV infection showed that VIP-KO mice had more CD4⁺ and CD8⁺ T-cells in their blood and spleen compared with WT mice (Figure 2C-F).

Figure 1. Mice lacking VIP had lower levels of virus in the liver following mCMV infection.

VIP-KO (12-15 mice per time-point, from 3 replicate experiments) and WT mice (12-15 mice per time-point, from 3 replicate experiments) were infected (day 0) with low dose 5 × 10⁴ PFU mCMV. Livers were collected, weighed, and lysates prepared at days 3, 7, 10, 14 and 17 days post-mCMV infection. Day 0 control livers were from uninfected mice (n=3). Mean liver viral load was measured by plaque assay of a defined quantity of liver lysate on 3T3 cell monolayers, and the number of pfu/liver calculated. *** Signifies p<0.001, denoting a significant difference between VIP-KO and WT mice.

Figure 2. Mice lacking VIP had better survival and greater expansion of blood T-cells following mCMV infection.

VIP-KO and WT mice were infected (day 0) with low dose 5 × 10⁴ PFU or high dose 1 × 10⁵ PFU mCMV. Survival was recorded every day and body weight was recorded twice weekly. Peripheral blood and spleen were collected baseline, prior to mCMV infection, and 3, 7, 10, 14 and 17 days post-infection. Blood cells and splenocytes were stained with fluorescently conjugated monoclonal antibodies to CD45.2, CD3, CD4, and CD8 and analyzed by flow cytometry, and absolute numbers of cells per mL blood or per spleen were calculated. Data from 12-15 mice per group were pooled from 3 replicate experiments. A and B. Survival and body weight change of WT and VIP-KO mice that received graded doses of 5 × 10⁴, or 1 × 10⁵ PFU mCMV. C and D: Total numbers of CD4⁺ and CD8⁺ T-cells in blood following low dose mCMV infection. E and F: Total numbers of CD4⁺ and CD8⁺ T-cells in the spleen following low dose mCMV infection. * Signifies p<0.05, ** signifies p<0.01, and *** signifies p<0.001, denoting a significant difference between VIP-KO and WT mice.

Innate and adaptive anti-viral responses were enhanced in the absence of VIP

VIP-KO mice had significantly higher percentages (Figure 3A) and absolute numbers of antigen-specific tetramer⁺ CD8 T-cells in the blood (Figure 3B) and spleen (Figure 3C) following low-dose mCMV infection than WT mice. The highest frequency of tetramer⁺ CD8 T-cells in the blood was seen on day +10 post-infection with 9.1% ± 0.8% of blood CD8⁺ T-cells in VIP KO mice vs. 4.8% ± 0.7% of blood CD8⁺ T-cells in WT mice (p<0.001; Figure 3A). Since lethality was 100% in WT mice receiving high-dose mCMV compared with 35% mortality among VIP-KO mice (p<0.001), a longitudinal comparison of the numbers of antigen specific T-cells in WT vs. VIP KO mice could not be performed, but analysis at day 3 showed that VIP-KO mice had greater numbers of tetramer⁺ CD8 T-cells (295/mL ± 40/mL) compared with WT mice (124/mL ± 38/mL, p<0.001). Enhanced innate anti-viral immunity among VIP-KO mice was evidenced by higher levels of NK-mediated cytotoxicity against YAC1 targets in VIP-KO splenocytes harvested 3 days post-infection (Figure 3D). Using mCMV-peptide-pulsed and non-pulsed congenic splenocytes as targets in an in vivo cytotoxicity assay in immune mice (previously infected with low dose mCMV), the specific lysis of mCMV-peptide-pulsed targets was significantly enhanced in VIP-KO mice compared with WT mice (Figure 4A, B). Significantly, VIP-KO mice had similar baseline-numbers but more IFN-γ-expressing NK, NKT cells, and Th1/Tc1 polarized (IFN-γ⁺ and TNF-α⁺) T-cells on days 3-17 post-infection compared with WT mice (Supplementary Figure 2 A-H).

Figure 3. Mice lacking VIP had larger increases of antigen-specific T-cells following mCMV infection.

VIP-KO and WT mice were infected (day 0) with low dose 5 × 10⁴ PFU or high dose 1 × 10⁵ PFU mCMV. Peripheral blood and spleen were collected at baseline, prior to infection and 3, 7, 10, 14 and 17 days post-mCMV infection. Blood cells and splenocytes were stained with fluorescently conjugated monoclonal antibodies to CD45.2, CD3, CD4, CD8 and mCMV M45-epitope peptide specific MHC-I tetramer reagents, analyzed by flow cytometry, and absolute numbers of cells per mL blood and per spleen were calculated. NK cell killing activity were measured by Cr⁵¹ releasing assay using YAC-1 pulsed Cr⁵¹. A. Percentages of tetramer⁺ CD8 T-cells in blood and spleen following low-dose mCMV infection. Dot graphs showed concatenated list mode files from analysis of 4 mice per group 10 days following mCMV infection (mean ± SD), and are representative of 5 replicate experiments. B. Absolute numbers of tetramer⁺ CD8 T-cells/mL in blood following low-dose mCMV infection (data are from 5 replicate experiments involving a total of 12-20 mice per time point). C. Absolute numbers of tetramer⁺ CD8 T-cells in the spleen following low-dose mCMV infection (data are from 5 replicate experiments involving a total of 12-20 mice per time point). D. NK cells mediated cytolytic activity (data are from 3 replicate experiments with 12 mice per time point). ** Signifies p<0.01 and *** signifies p<0.001, denoting a significant difference between VIP-KO and wild-type mice.

Figure 4. VIP-KO mice had increased cytolytic activity against M45 peptide-pulsed targets following mCMV infection.

A mixture of peptide-pulsed targets (CD45.1⁺ CD45.2⁺) and non-pulsed targets (CD45.2^- CD45.1⁺) were adoptively transferred to VIP-KO and WT mice 9 days after infection with low-dose mCMV. Target cells were harvested from the recipient spleens 16 hours after iv injection, and peptide-pulsed targets and non-pulsed targets were differentiated by flow cytometry following staining for CD45 congenic markers. A. A representative flow cytometry analysis plot of splenocytes from recipient mice showing mean percentages of peptide-pulsed target cells and non-pulsed target cells. Dot graphs showed concatenated list mode files from analysis of 5 mice per group. B. Calculated mean specific cytolytic activity, n=5, representative of 2 replicate experiments.

The absence of VIP expression in donor hematopoietic cells enhanced anti-viral immunity in radiation hematopoietic chimeras

Since VIP is expressed in multiple cell lineages (2-6) we tested whether mice lacking VIP expression only in their hematopoietic cells had the same level of enhanced anti-viral immunity as we observed in VIP-KO mice. We used VIP-KO mice as donors of hematopoietic cells and created radiation chimeras with syngenic BMT in which recipients had >95% donor cell engraftment (36). The day 59 survival of mice transplanted with VIP-KO 3 × 10³ FACS purified HSC, 5 × 10⁴ FACS purified DC and 3 × 10⁵ MACS purified T-cells (75% ± 10%) was similar to the survival seen among mice transplanted with WT HSC, DC and T-cells (80% ± 9%). To explore the effect of VIP expression in hematopoietic cells on primary and secondary immune responses, VIP-KO→WT and WT→WT syngeneic transplant recipients were primed with PBS or the Lm-MCMV vaccine (containing mCMV immunodominant M45 epitope peptide aa 985~993) followed by infection 21 days later with low dose mCMV (Figure 5A, B). Peripheral blood samples obtained prior to Lm-MCMV vaccination (day 59 post-transplant), after vaccination, and following mCMV infection (day 80 post-transplant) were analyzed for the numbers of tetramer⁺ CD8 T-cells. Non-immunized WT and VIP-KO chimeric mice had minimal numbers of mCMV-peptide tetramer⁺ CD8⁺ T-cells in their blood at baseline (Figure 5A). Following primary mCMV infection, recipients engrafted with VIP-KO hematopoietic cells had significantly more mCMV-peptide tetramer⁺ CD8⁺ T-cells in their blood compared with WT mice (Figure 5A). Vaccination with Lm-MCMV led to a larger increase in blood mCMV tetramer⁺ T-cells in the VIP-KO→WT chimeras compared with WT→WT chimeras (Figure 5B) indicating that mCMV peptide presentation alone in VIP-KO mice (in the absence of viral infection) was sufficient to result in enhanced expansion of antigen-specific T-cells. Subsequent infection of the Lm-MCMV vaccinated mice with low dose mCMV led to an accelerated anamnestic response in VIP-KO→WT chimeras compared with mice engrafted with WT BM (Figure 5B). Since both T-cells and accessory cells can secrete VIP (4-6), we further explored the role of VIP synthesis by different immune cell subsets by creating radiation chimeras engrafted with the combination of donor DC & HSC from VIP-KO mice and donor T-cells from WT mice. Mice transplanted with the heterogeneous combination of VIP-KO HSC & DC and WT T-cells did not show the enhanced immune responses seen in mice engrafted with the homogeneous combination of VIP-KO HSC, DC and T-cells (Figure 5B) indicating that VIP production by donor T-cells was sufficient to attenuate anti-viral cellular immunity.

Figure 5. Radiation chimeras engrafted with hematopoietic cells from VIP-KO donors had enhanced primary and secondary antigen specific cellular immune responses following Lm-mCMV vaccination and mCMV infection.

Syngeneic bone marrow chimeric mice were generated by transplanting lethally irradiated H-2K^b recipients with 3 × 10³ HSC, 5 × 10⁴ DC and 3 × 10⁵ T-cells from either VIP-KO or WT H-2K^b donor mice. 59 days post-transplant, mice were vaccinated with 1 × 10⁶ CFU Lm-MCMV or PBS, and then 80 days post-transplant, mice were infected with low dose 5 × 10⁴ PFU mCMV. Blood samples were collected at day 59, 62, 66, 80, 83, 87 and 101 post-transplantation and analyzed by flow cytometry for tetramer⁺ CD8 T-cells. Data represent mean values of 12-16, per time-point, pooled from 3 replicate experiments. A. Mice were treated first with PBS then infected with mCMV. B. Primary and secondary immune responses in mice following vaccination with Lm-MCMV and then infection with low-dose mCMV. *** Signifies p<0.001 comparing tetramer⁺ T-cell levels between mice transplanted with VIP-KO hematopoietic cells and WT hematopoietic cells.

Absence of VIP augmented anti-viral CD8+ T-cell proliferation and Th1/Tc1 polarization in vitro

To study the effect of VIP on anti-viral immunity in vitro, we analyzed cultures of T-cells and mCMV-peptide-pulsed DC for tetramer⁺ CD8 T-cells and for Th1 & Th2 cytokines. DC and T-cells were purified from WT or VIP-KO mice (36), the DC were pulsed with mCMV peptide, and then mixed with T-cells. The numbers of tetramer⁺ CD8 T-cells generated over 10 days of culture were measured by flow cytometry. Significantly greater numbers of antigen-specific tetramer⁺ CD8 T-cells were detected after 3 days in cultures of T-cells with DC that had been isolated from mCMV-immune VIP-KO mice compared with similar cells isolated from mCMV-immune WT mice (Figure 6A). To rule out an effect of VIP synthesized by non-hematopoietic cells on in vitro immune responses to mCMV peptides, donor-derived T-cells and DC were recovered from syngeneic transplants recipients of VIP-KO→WT or WT→WT radiation chimeras. Homogeneous cultures of DC and T-cells recovered from VIP-KO→WT radiation chimera generated more tetramer⁺ CD8 T-cells than cultures of DC and T-cells from WT→WT radiation chimeras (Figure 6B), indicating the absence of VIP synthesis by hematopoietic cells in radiation chimeras programed T-cells and DC towards enhanced cellular immune responses. Supernatants from cultures of T-cells and mCMV-peptide-pulsed DC from WT mice had higher levels of IL-10, and lower levels of IFN-γ compared with supernatants from cultures of T-cells and mCMV-peptide-pulsed DC from VIP-KO mice (Figure 6C, D). To determine whether synthesis of VIP by T-cells was sufficient to down-regulate immune responses to mCMV, we cultured WT T-cells and VIP-KO DC isolated from radiation chimeras originally transplanted with the heterogeneous combination of WT T-cells plus VIP-KO DC and VIP-KO HSC. In contrast to the larger numbers of tetramer⁺ CD8 T-cells seen in homogeneous cultures of T-cells and DC from VIP-KO mice, heterogeneous cultures of WT T-cells plus VIP-KO DC generated fewer tetramer⁺ CD8 T-cells, similar to cultures of WT T-cells and WT DC, indicating that VIP synthesis by T-cells acts as a dominant negative regulatory mechanism in anti-viral cellular immunity in vitro (Figure 6 B).

Figure 6. The generation of antigen-specific anti-viral T-cells and Th1 polarization was increased in cultures of DC and T-cells from VIP-KO mice compared with cells from WT mice.

DC and T-cells were isolated from spleens of VIP-KO and WT mice, and from radiation chimeric mice that received homogeneous grafts from VIP-KO or WT (3 × 10³) HSC, (5 × 10⁴) DC and (1 × 10⁶) T-cells, and heterogeneous grafts from the combination of VIP-KO HSC and DC and WT T-cells 15 days following infection with 5 × 10⁴ PFU (low dose) mCMV (15-18 per group were studied). FACS-purified DC from these mice were incubated with 3μM mCMV peptide for 30 minutes, washed, and then triple-wells co-cultured with T-cells from the same groups. On day 3 and day 7 of culture, antigen-specific T-cells were measured by FACS using mCMV-M45 epitope peptide-MHC-I tetramer reagent. A and B: the absolute numbers of tetramer⁺ CD8 T-cells per mL in cultures of cells from non-transplanted (A) and radiation chimeric mice (B). Day 0 data were obtained using cells from non-infected mice. Culture media from day 3 cultures of cells from radiation chimeric mice were assayed for IL-10 (C) and IFN-γ (D) by ELISA. * Signifies p<0.05, ** p<0.01, *** p<0.001 comparing VIP-KO mice and WT groups. Means ± SE from pooled results of 3 repeat experiments. The experiment was repeated 3 times.

VIP-KO mice had higher levels of co-stimulatory molecule and MHC-II expression on DC and less PD-1/PD-L1 expression compared with WT mice following mCMV infection

To explore the mechanism by which the absence of VIP enhanced anti-viral immunity, we studied the expression of co-stimulatory molecules and PD-1/PD-L1 expression in WT and VIP-KO mice following mCMV infection. Prior to mCMV infection, baseline levels of MHC-II, CD80, and PD-L1 expression on DCs, and PD-1 expression on CD4 and CD8 T-cells were similar comparing WT with VIP-KO mice (Figure 7). VIP-KO mice had a marked up-regulation of CD80 and MHC-II expression on cDC and pDC 3 days after mCMV infection compared with the corresponding DC subsets from mCMV-infected WT mice. Of note, the absence of VIP expression had a significant impact on the up-regulation of co-inhibitory molecules and ligands that normally follows mCMV infection: PD-L1 expression was up-regulated 3 days after mCMV infection in DC from WT but not VIP-KO mice, while WT CD8⁺ T-cells showed a striking up-regulation of PD-1 expression on day 10 after mCMV infection that was not seen in CD8⁺ T-cells from VIP-KO mice (Figure 7).

Figure 7. Higher levels of CD80 and MHC-II expression on DC and lower levels of PD-1 and PD-1 expression on CD8⁺ T-cells and DC from VIP-KO mice following mCMV infection.

Splenocytes were isolated from VIP-KO and WT mice at baseline and 3, 10 and 17 days after infection with with 5 × 10⁴ PFU mCMV. Expression patterns of CD80 (A), MHC-II (B) and PD-L1 (C) on conventional DC (cDC, lineage^-, CD11c^hi, B220^-) and plasmacytoid DC (pDC, lineage^-, CD11c^lo, B220⁺) and the percentages of CD8⁺ T-cells expressing PD-1 (D) were analyzed by flow cytometry. Histograms depict analysis of concatenated list mode files from 4 mice per group at each time-point, and are representative of 3 replicate experiments. Dashed lines represent the staining profile using an isotype-matched control antibody; filled lines represent specific staining.

Discussion

In this study we explored the immuno-regulatory effect of VIP in immune responses to mCMV infection, hypothesizing that the absence of VIP would increase innate and adaptive immune responses to viral infection. Our data using VIP-KO mice demonstrates that the absence of physiological levels of VIP in hematopoietic cells led to striking enhancement of innate and adaptive anti-viral cellular immune responses. VIP-KO mice had less mortality and faster viral clearance compared with WT mice. The increased expansion of tetramer⁺ CD8 T-cells and increased cytolytic activity of NK cells seen in VIP-KO mice are likely responsible for their greater resistance to mCMV infection (50). While we used the M45 epitope peptide to measure mCMV specific T-cells, and T-cells recognizing this epitope have been shown to be relative ineffective in clearing virus infected cells due to m152/gp40-mediated immune interference (51), the enhanced killing of M45 epitope-containing peptide-pulsed-target cells supports the contribution of M45 reactive T-cells to functional anti-viral cytotoxic activity in vivo.

To clarify the effect of various physiological sources of VIP (hematopoietic versus neuronal), we used C57BL/6 radiation chimeras engrafted with syngeneic VIP-KO or WT hematopoietic cells following myeloablative radiation. Recipients of VIP-KO hematopoietic grafts showed accelerated kinetics of cellular immune responses to primary mCMV infection and LmCMV vaccination as well as greater amnestic responses following Lm-mCMV vaccination and mCMV infection compared with recipients of wild-type grafts. These data indicate that VIP produced by hematopoietic cells has a dominant negative effect on anti-viral cellular immune responses, and that VIP synthesis by non-hematopoietic neuronal cells does not significantly affect anti-viral immune responses in this system.

Immune cells in VIP-KO mice had more Th1 polarization (52, 53), less Th2 polarization, and higher MHC-II expression (47) than those of WT mice following mCMV infection, consistent with the reports that VIP is a negative regulator of Th1 immune responses (3, 54). A simple in vitro model of T-cells co-cultured with mCMV-peptide pulsed DC recapitulated the in vivo immunology of VIP KO mice. Co-cultures of DC and T-cells from VIP-KO mice had higher levels of IFN-γ⁺ CD4⁺ and CD8⁺ T-cells and more antigen-specific anti-viral CD8⁺ T cells compared with cultures of WT DC and WT T-cells. Conditioned media from cultures of WT T-cells and WT DC had higher levels of IL-10, and lower levels of IFN-γ, compared with culture media from VIP-KO T-cells VIP-KO DC, consistent with other reports (55). Heterogeneous co-cultures of VIP-KO DC and WT T-cells had the same (lower) numbers of antigen-specific anti-viral CD8⁺ T cells as cultures of WT DC and WT T-cells, confirming that T-cells making VIP are sufficient to polarize Th2 immunity and suppress Th1 immunity, and that VIP made by T-cells is a dominant negative regulator of anti-viral immune responses (56, 57).

The mechanisms for the enhanced antiviral cellular immunity and greater Th1/TC1 immune polarization seen in VIP-KO mice following mCMV infection appears to be due to a profound shift in the pattern of co-stimulatory and co-inhibitory molecule expression on DC and CD8⁺ T-cells. The higher levels of MHC-II and CD80 on cultured VIP-KO DC compared with WT DC are consistent with previous reports that mature DC activate Th1 immune responses (36, 58) and that supra-physiological levels of VIP induces tolerogenic DC that express lower levels of co-stimulatory molecules (12). Another possible mechanism is that VIP-signaling interferes with the ability of the mCMV protein m138 to target CD80 expression on DC (59). An important new finding in this study is that VIP modulates the expression of the PD-1 and PD-L1 co-inhibitory molecules that regulate immune polarization and survival of T-cells. PD-L1-PD-1 interactions are known to regulate the initial priming of naive T cells by mCMV-infected APC, and are distinct from the role that PD-1 signaling plays in T cell “exhaustion” described for several persistent/ chronic viral infections in humans and mice (60), including human CMV (61). Following viral infection, up-regulation of the PD-L1/L2 – PD-1 pathway has been associated with immunosuppression (62) due to cell-cycle arrest, and death of T-cells, either through the direct engagement of a death pathway or indirectly by down-regulating survival signals and growth factors (61). PD-L1/L2 expression on DC is associated with reduced expression of CD40, CD80, and CD86 and increased IL-10 production (63). We found that DC from mice transplanted with VIP-KO cells had dramatically reduced PD-L1 expression on DC and PD-1 expression on activated memory CD8⁺ T-cells that were associated with increased quantitative and qualitative antiviral T cell responses following mCMV infection. Our studies indicate that physiological levels of VIP contribute to the up-regulation of PD-L1/PD-1 expression seen in WT mice following mCMV infection. This work further suggests that induction of VIP may be part of the active suppression of adaptive immune responses that occur following mCMV infection.

In summary, these data indicate that VIP synthesis by hematopoietic cells is a key factor in regulating the development of protective Th1 immune responses following vaccination or infection with mCMV. The absence of VIP synthesis by hematopoietic cells leads to lower levels of counter-regulatory co-inhibitory molecules and changes in serum cytokines consistent with global Th1 immune polarization. The increased anti-viral immunity seen in the absence of VIP suggests that VIP antagonists may be of clinical benefit for patients with viral infection.

Abbreviations

APC

Antigen presenting cells

BM

Bone marrow

DC

Dendritic cell

mCMV

murine cytomegalovirus

VIP

Vasoactive Intestinal Peptide

MACS

Magnetic cell activator sorter

FACS

Fluorescence-activated cell sorting

KO

knockout

WT

wild-type

PACAP

pituitary adenylyl cyclase-activating polypeptide

Tetramer⁺ CD8 T-cells

CMV-M45 epitope peptide-MHC class I tetramer⁺ CD8⁺ T-cells

PD-1

Programmed Death-1

PD-L1

Programmed Death Ligand-1