Human adenoviruses (HAdV)

Adenovirus-Specific CD4 T Cell Clones Recognizing Endogenous Antigen Inhibit Viral Replication In Vitro through Cognate Interaction 1 Bianca Heemskerk, 2 * Tamara van Vreeswijk,* Louise A. Veltrop-Duits,* Claudia C. Sombroek,* Kees Franken, Renate M. Verhoosel, Pieter S. Hiemstra, Daphne van Leeuwen, Maaike E. Ressing, René E. M. Toes, Maarten J. D. van Tol,* and Marco W. Schilham 3 * Human adenovirus (HAdV) infection is a frequent and potentially severe complication following allogeneic stem cell transplantation in children. Because treatment with antiviral drugs is often ineffective, adoptive transfer of donor-derived HAdV-specific T cells able to control viral replication of HAdV of multiple serotypes may be an option for therapy. In healthy donors, predominantly HAdV-specific T cells expressing CD4 are detected. In this study, a preclinical in vitro model was used to measure the antiviral effect of HAdV-specific CD4 T cells. CD4 HAdV-specific T cell clones restricted by HLA class II molecules were generated and most of these clones recognized conserved peptides derived from the hexon protein. These cross-reactive T cell clones were able to control viral replication of multiple serotypes of HAdV in EBV-transformed B cells (B-LCL), melanoma cells (MJS) and primary bronchial epithelial cells through cognate interaction. The HAdV-specific CD4 T cell clones were able to specifically lyse infected target cells using a perforin-dependent mechanism. Antigenic peptides were also presented to the CD4 T cell clones when derived from endogenously produced hexon protein. Together, these results show that cross-reactive HAdVspecific CD4 T cells can control replication of HAdV in vitro and provide a rationale for the use of HAdV-specific T cells in adoptive immunotherapy protocols for control of life-threatening HAdV-infections in immunocompromised patients. The Journal of Immunology, 2006, 177: 8851 8859. *Department of Pediatrics, Department of Immunohematology and Bloodtransfusion, Department of Pulmonology, Department of Medical Microbiology, and Department of Rheumatology, Leiden University Medical Center, Leiden, The Netherlands Received for publication June 7, 2006. Accepted for publication September 26, 2006. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported by Dutch Cancer Foundation Grant RUL-2001-2492 and European Community Grant QLK2-CT-2002-01432. 2 Current address: Surgery Branch, National Cancer Institute, National Institutes of Health, 10 Center Drive, Bethesda, MD 20892-1201. 3 Address correspondence and reprint requests to Dr. Marco W. Schilham, Department of Pediatrics, Leiden University Medical Center, P. O. Box 9600, 2300 RC Leiden, The Netherlands. E-mail address: M.W.Schilham@lumc.nl 4 Abbreviations used in this paper: HAdV, human adenovirus; SCT, stem cell transplant; PBEC, primary bronchial epithelial cell; MB, methylene blue; egfp, enhanced GFP; MOI, multiplicity of infection; TCID 50, tissue culture-infective dose of 50%; CRA, chromium release assay; CMA, concanamycin A. Human adenoviruses (HAdV) 4 rarely cause severe clinical symptoms in healthy children and adults, because infections in immunocompetent individuals are usually selflimiting. However, HAdV may cause life-threatening complications in immunocompromised patients (1, 2). In recent years, the incidence of HAdV infections in pediatric stem cell transplant (SCT) recipients has increased remarkably (3 7). Recipients of a T cell-depleted or CD34 -enriched allogeneic stem cell graft, i.e., patients with a non-hla-identical donor, receiving serotherapy with anti-cd52 mab (Campath) or anti-thymocyte globulin, have a higher risk of developing HAdV infection and dissemination due to the delayed immune reconstitution in these children after SCT (7 9). Clinical symptoms include gastroenteritis, hemorrhagic cystitis, hepatitis, pneumonia, encephalitis, and multiorgan failure. Dissemination of the infection, documented by a rise of HAdV DNA loads in plasma, is associated with a fatal outcome (7, 10 14). Currently, 51 serotypes of adenovirus have been identified, distributed among six species (A F) based on their differential hemagglutination with erythrocytes and their DNA homology (15, 16). Species A, B, and C serotypes are most frequently isolated from pediatric immunocompromised hosts and are the major cause of disease (1, 7, 17 19). Treatment of adenoviral infections with antiviral medication, such as cidofovir and ribavirin, has not been unequivocally effective (reviewed in (20)). Therefore, new strategies to treat disseminated HAdV infections are needed. A potential approach is adoptive immunotherapy by infusion of HAdV-specific lymphocytes. This strategy has already been successfully pursued for other viral infections or reactivations such as CMV or EBV (21 23). Furthermore, case reports have suggested that donor lymphocyte infusions or tapering of immunosuppression may contribute to clearance of an HAdV infection, illustrating the potential role of T cells in the immune response to HAdV (8, 24, 25). Few reports on HAdV-specific immune responses in healthy donors have been published until recently (19, 26, 27). Current reports on the occurrence and frequency of HAdV-specific T cells have shown that HAdV-specific T cells are predominantly CD4, although HAdV-specific CD8 T cells have been described as well (28 33). Furthermore, in a previous study, we have shown the presence of CD4 HAdV-specific T cells in the blood of patients recovering from HAdV infection or viremia (7), suggesting that these cells may be functionally involved in clearance of the virus. As mainly CD4 HAdV-specific T cells can be detected in healthy donors and patients after SCT, the question arises whether these CD4 T cells can exert direct antiviral functions, and if so, by which mechanisms. To study this, we generated HAdV-specific Copyright 2006 by The American Association of Immunologists, Inc. 0022-1767/06/$02.00

8852 HAdV-SPECIFIC CD4 T CELLS INHIBIT VIRAL REPLICATION CD4 T cell clones as described previously (31) and determined the HLA restriction, protein, and peptide specificity of these clones. To investigate the antiviral activity of these T cells, inhibition of viral replication in HAdV-infected B-LCL, MJS, or primary bronchial epithelial cells (PBEC) was determined in a newly developed in vitro inhibition assay. In this assay, the viral titer was determined in cell lysates of HAdV-infected cells after 3 days of culture in the presence or absence of a HAdV-specific CD4 T cell clone. These CD4 T cells were able to inhibit viral replication, and this inhibition required cognate MHC-peptide interaction. T cell clones lysed target cells in a perforin-mediated manner, which could be the mechanism of inhibition. In addition, it was established that the hexon epitopes can be presented on HLA class II after intracellular synthesis by stable transduction of the hexon gene into target cells. Together, these results imply that HAdVspecific CD4 T cells are a promising tool for immunotherapy in immunocompromised patients at risk of developing severe HAdV infection. Materials and Methods Cells and viruses The human EC line HEp2 was used to propagate HAdV. Cells were maintained in RPMI 1640 medium with glutamax (Invitrogen Life Technologies) supplemented with 10% heat-inactivated FCS (Invitrogen Life Technologies), 100 U/ml penicillin, and 100 g/ml streptomycin (RPMI 1640 medium/10% FCS). HAdV strains of serotypes 2 and 5 (species C), 12 (species A), and 35 (species B) from the National Institute of Public Health and the Environment (Bilthoven, The Netherlands) were grown on HEp2 cells and harvested when cytopathological effect was present. Virus was released from the cells by two freeze-thaw cycles and purified by CsCl density-gradient centrifugation. Virus stocks were titrated using the plaque assay on 293 cells at the Department of Molecular Cell Biology (Leiden, The Netherlands). Inactivation of HAdV was performed by incubation of virus with 1.3 M methylene blue (MB) followed by irradiation with visible light for 1 h (31, 34). B-LCL were generated from PBMC of HLAtyped donors by incubation with supernatant of the marmoset B cell line B95-8 in the presence of cyclosporin A. Cell lines were cultured in supplemented RPMI 1640 medium with 10% FCS. MJS (Meljuso, HLA class II-positive melanoma) cells were cultured in RPMI 1640 medium/10% FCS. To generate MJS cells expressing fragments of the hexon protein, the Moloney murine leukemia virus-based retroviral vector LZRS and packaging cells -NX-A were used (35). Two retroviral constructs were produced that transcribed HAdV hexon C fragment (504 743 aa) and hexon D fragment (761 952 aa), respectively, as a bicistronic messenger RNA with enhanced GFP (egfp). Replication-defective retrovirus was used to transduce MJS cells, and the cells were isolated for high egfp expression by FACS. The two cell lines were called MJS-Hex-C and MJS-Hex-D, respectively. PBEC were obtained from resected lung tissue obtained from patients who underwent surgery for lung cancer by enzymatic digestion as described previously (36). Cells were subcultured in a 1/1 mixture of DMEM (Invitrogen Life Technologies) and bronchial epithelial growth medium (Clonetics) supplemented with 0.4% (w/v) bovine pituitary extract, 0.5 ng/ml epidermal growth factor, 5 g/ml insulin, 0.1 ng/ml retinoic acid, 10 g/ml transferrin, 1 M hydrocortisone, 6.5 ng/ml T3, 0.5 g/ml epinephrine (all from Clonetics), 1.5 g/ml BSA (Sigma-Aldrich), 1 mm HEPES (Invitrogen Life Technologies), 20 U/ml penicillin, and 20 g/ml streptomycin (BioWhittaker). HAdV-specific T cell clones PBMC from healthy blood bank donors (Sanquin) were isolated using Ficoll gradient centrifugation. Monocytes were depleted by adherence to plastic for 2 h, after which the nonadherent peripheral blood lymphocytes were collected and used as responder cells. Peripheral blood lymphocytes (2 10 6 per well) were added in a 24-well plate (Corning) and stimulated with 1 10 5 irradiated (30 Gy) autologous PBMC preincubated with MBinactivated HAdV5 at a multiplicity of infection (MOI) of 10. Cells were cultured for 12 days in RPMI 1640 medium/10% human AB medium, harvested, seeded at 2.5 10 5 cells per well, and restimulated with 1 10 5 HAdV5-infected stimulator PBMC as before. IL-2 (50 IU/ml; Chiron) was added at day 15 and replenished two or three times a week thereafter. At day 28, a limiting dilution assay was performed in which cells were plated at a density of 10, 1, or 0.3 cells per well in 96-well round-bottom plates (Corning) and stimulated with 1 g/ml PHA (Murex Biotech), 150 IU/ml IL-2, and 1 10 6 /ml irradiated allogeneic PBMC. After 2 weeks, wells with expanding cells were restimulated as before with PHA, IL-2, and allogeneic PBMC. Specificity of clones was determined by proliferation as described below, and the phenotype of the clones was assessed by flow cytometry (see below). HAdV-specific proliferation T cell clones (2 10 4 cells per well) were plated in 96-well round-bottom plates in triplicate. As stimulator cells, autologous EBV-transformed B cells (B-LCL) were resuspended in RPMI 1640 medium with 0.5% BSA at 5 10 6 B-LCL/ml, irradiated at 40 Gy, and infected for 1 h at 37 C with HAdV5 at an MOI of 100, or uninfected as control. Hereafter, a total of 5 10 4 stimulator cells was added per well in RPMI 1640/10% AB medium. After 3 days at 37 C and 5% CO 2, cultures were pulsed with 1 Ci [ 3 H]thymidine (Amersham Biosciences) per well for 18 h. Plates were harvested (Skatron), and filters were subsequently counted in a Betaplate counter (Wallac). To determine which protein was recognized by a T cell clone, B-LCL were loaded with recombinant hexon protein (protein II, generated in four parts as IIA (1 273 aa), IIB (245 509 aa), IIC (479 743 aa), and IID (719 952 aa)) (37), the penton base protein (III) or an early protein, E1A, at a final concentration of 5 g/ml. For peptide recognition, the hexon protein was synthesized in overlapping peptides of 30 aa with a 15-aa overlap (37). Proliferation was determined against B-LCL loaded with pools of five to six peptides (5 g/ml), and specific peptides were identified from positive peptide pools by loading single peptides on B-LCL. The HLA restriction element was determined by the addition of blocking mabs against HLA class I (W6/32), HLA-DR (B811.2), HLA-DQ (SPVL3), and HLA-DP (B7.21) in the proliferation assay (Abs were a gift from Dr. A. Mulder, Department of Immunohematology and Blood Transfusion, Leiden University Medical Center, Leiden, The Netherlands) and using partially HLA-matched B-LCL as stimulators. Detection of cytokine production and mediators of apoptosis HAdV-specific T cell clones were tested for cytokine production by intracellular staining and flow cytometry. T cells (2 10 5 ) were stimulated with 1 10 5 irradiated autologous B-LCL either uninfected, infected with HAdV5 at an MOI of 100, or loaded with the specific peptide in roundbottom polystyrene 5-ml tubes in a total volume of 400 l. After 1 h, brefeldin A (Sigma-Aldrich) was added for 18 h to block exocytosis. Cells were stained on ice and analyzed using flow cytometry. Cells were washed with PBS containing 0.2% w/v NaN 3, fixed with freshly made 4% paraformaldehyde (Sigma-Aldrich) in PBS for 4 min, and washed twice with PBS/NaN 3. Cells were permeabilized by washing with PBS/NaN 3 containing 0.1% saponin and 0.5% BSA, and nonspecific binding was blocked with PBS/NaN 3 /saponin/bsa/10% FCS for 10 min. After washing with PBS/NaN 3 /saponin/bsa, Abs were added: anti-cd3-percp-cy5.5, anti- CD4-FITC, anti-ifn- -PE/anti-TNF- -PE/anti-IL-10-PE/anti-IL-4-PE (BD Biosciences), and anti-cd8-allophycocyanin (Immunotech) and incubated for 30 min. For detection of perforin and granzyme B intracellularly, cells were stained with anti-perforin-fitc (Hölzel Diagnostika) or with anti-granzyme B-PE (Sanquin) in combination with anti-cd3-percp-cy5.5 and anti-cd4- allophycocyanin as above. After washing, cells were either fixed in 0.5% paraformaldehyde or analyzed immediately on a FACSCalibur flow cytometer using the CellQuest software (BD Biosciences). The level of secreted IFN- in culture supernatant was measured with an ELISA kit (Sanquin) according to manufacturer s guidelines. HAdV replication inhibition assay To test whether HAdV-specific T cell clones were able to inhibit viral replication, we developed an in vitro inhibition assay. B-LCL, MJS, or PBEC (after 2 days of incubation with 100 U/ml IFN- ), were infected with HAdV5 at an MOI of 100 for B-LCL or 10 for MJS and PBEC. MJS cells and PBEC (1 10 5 ) were plated 1 day before infection in the wells of a 24-well plate. After 1 h of infection, B-LCL were washed three times with 10 ml of RPMI 1640 medium/10% FCS in tubes, and MJS/PBEC three times with 1 ml per well to remove free virus. For B-LCL, cells were plated at 1 10 5 in 500 l of RPMI 1640 medium/10% FCS per well of a 24-well plate (Corning). HAdV-specific T cell clones (at an E:T ratio of 10:1 unless indicated otherwise) were added to the three types of target cells in 500 l of RPMI 1640 medium/10% FCS, or medium alone as a control for viral replication. In preliminary experiments, cells were harvested daily from days 0 through 6 to determine the kinetics of viral replication in the B-LCL. In later experiments, cells were harvested at day 3

8853 after infection, which was found to be optimal. Cells were washed three times with RPMI 1640 medium/10% FCS to remove free virus, and virus was released from the cells by two freeze-thaw cycles. The virus titers of cell lysates were determined with the tissue culture-infective dose of 50% (TCID 50 ) assay. For this, lysates were diluted 10 0 -to10 8 -fold, and each dilution was plated in 10-fold in 96-well flat-bottom plates. HEp2 cells were added to the wells at 5 10 3 cells per well. After 7 days of incubation at 37 C and 5% CO 2, wells were scored for the presence of cytophatological effect, which indicates the presence of infectious virus in those wells. The viral titer of each lysate was calculated via the method of Reed and Munch, which determines the dilution at which 50% of wells are infected as described elsewhere (31). For detection of HAdV12 and HAdV35 viral replication, A549 cells were used to determine the viral titer, as these viruses do not grow optimally on HEp2 cells. In some experiments, a coculture of autologous B-LCL and HLA-DRmismatched B-LCL was performed. To stimulate the HAdV-specific T cell clone, autologous B-LCL were loaded with the specific peptide, while the mismatched B-LCL were infected with HAdV5. Viral titers were determined in cell lysates to investigate the bystander effect of soluble factors or non-hla-restricted recognition on inhibition of viral replication. Chromium release assay (CRA) To determine HAdV-specific cytotoxicity, CRA were performed on B-LCL or MJS cells. Target cells were infected with HAdV5 (infectious or MBinactivated) at an MOI of 100 as described above or loaded with peptide at 5 g/ml and cultured for 16 h unless otherwise indicated. Target cells were labeled with 51 Cr for 1.5 h at 37 C and washed three times with RPMI 1640 medium. Various numbers of effector cells were cocultured with 2.5 10 3 B-LCL or 2.5 10 3 MJS per well for 4hat37 C and 5% CO 2, and 51 Cr release was determined in the supernatant. Maximal release was determined by the addition of 2 N HCl to target cells, and spontaneous release was obtained by the addition of medium alone. Counts were measured in an automatic counter counter (Wallac). The percentage of specific release was calculated as: ((cpm experimental release cpm spontaneous)/ (cpm maximal cpm spontaneous)) 100. In some experiments, inhibitors of the perforin-mediated lysis were added. Concanamycin A (CMA; 50 nm; a gift from Dr. J. P. Medema, Department of Clinical Oncology, Leiden University Medical Center, Leiden, The Netherlands) or EGTA (at 2 mm; Boehringer Mannheim) was added to the effector cells 1 h before coculture with target cells, and CMA or EGTA were added during cocultivation. To determine whether antigenic peptides were transferred between cells during culture, MJS cells were mixed with MJS-Hex-D cells in a 1:1 ratio and cultured for 4 5 days. Then, the two cell populations were separated by FACS sorting based on egfp expression, incubated with peptide or medium during labeling with 51 Cr, and used as target cells. Table I. HAdV-specific CD4 T cell clones with hexon peptides and HLA restriction T Cell Clone Donor Hexon Protein Hexon Peptide HLA Restriction A1.1 1 - a - DR7 D1.23 1 - - DR7 C30.4 2 - - DR7 A4 0.3 4 3 IIC II44 (43) b DR17 A5 1.32 3 IID II64, 910-924 c DP4 M2.11 4 IID II57 DR17 K1.2 5 IID II64, 910-924 DP4 K3.1 5 IID II61 DR17 a -, Indicates none of the hexon, penton base, or E1A proteins were recognized. b Indicates that peptide II44 induced the highest proliferation; proliferation to II43 was observed to a lesser extent, suggesting that the minimal epitope resides in the N-terminal part of peptide II44. c Peptide II64 was recognized as well as the minimal epitope that is present in II64, the II 910-924 epitope described in Ref. 29. Results Characterization of HAdV-specific CD4 T cell clones HAdV-specific CD4 T cell clones, generated against HAdV5 (species C), have been obtained from several healthy donors as described recently (31). These clones exhibited a restricted or a broad (most clones) cross-reactivity pattern, recognizing serotypes from species C only or from A, B, and C, respectively. Because such broadly cross-reactive T cell clones are ideal candidates to boost HAdV-specific immunity in immunocompromised patients, the nature of the epitopes recognized by these clones was characterized. First, it was investigated which protein of HAdV was recognized using proliferation assays with recombinant hexon protein (generated in four parts, IIA IID), recombinant penton base, and early gene product E1A (37). Three of eight HAdV-specific T cell clones were unresponsive to any of these proteins, but the remaining five recognized the C-terminal part of the hexon protein (IIC or IID) (Table I and example in Fig. 1A), which is highly conserved between different species (38, 39) and might explain the broad cross-reactivity that we observed previously (31). To further investigate which peptides of the hexon protein were recognized, 63 overlapping peptides of 30 aa with a 15-aa overlap of the hexon protein (peptides II4 II66) were synthesized and tested in proliferation assays. An example is shown in Fig. 1B, and results are summarized in Table I. Two clones recognized II64, which contains a conserved epitope II 910 924 that has recently been described (29, 40). Proliferation was also observed against this minimal epitope, indicating that these two HAdV-specific clones specifically recognized this II 910 924 epitope contained within peptide II64. The other clones recognized three different hexon peptides. HLA restriction of these clones was determined by stimulation with HLA-DR- or HLA-DP-matched B-LCLs or the addition of blocking Abs. The T cell clones of which the protein specificity could not be determined were all HLA-DR7 restricted, while the other five clones that recognized various peptides in the hexon were HLA-DR17 or HLA-DP4 restricted (Table I). FIGURE 1. Most HAdV-specific CD4 T cell clones recognize peptides in hexon protein (protein II). A, T cells from clone A5 1.32 were tested in a 3-day proliferation assay against autologous B-LCL infected with HAdV5 or loaded with recombinant proteins from hexon (IIA IID), the penton base protein (III), or the early E1A protein. Proliferation, as measured by [ 3 H]thymidine incorporation, was only observed against HAdV and the recombinant IID protein. B, Clone A5 1.32 was further tested for proliferation against autologous B-LCL loaded with 30-aa peptides from the IID protein, and was found to proliferate exclusively against peptide II64. This peptide contains a minimal epitope 910-924 as described recently (29), which was found to be the epitope for this clone. Results are representative for five clones.

8854 HAdV-SPECIFIC CD4 T CELLS INHIBIT VIRAL REPLICATION FIGURE 2. HAdV-specific CD4 T cell clones secrete IFN- and TNF-. Autologous B-LCL were infected with HAdV5 at an MOI of 100 or uninfected (medium) as control. T cells from clone M2.11 were added for 16 h, and cytokine production was determined with intracellular cytokine staining using anti-ifn- -PE, anti-tnf- -PE, anti-il-2-pe, or anti- IL-4-PE in combination with anti-cd4-fitc. The percentage of CD4 T cells producing a specific cytokine is shown in the upper right quadrant. Results are also representative for the other clones tested (A1.1 and K3.1). When the clones were stimulated with HAdV5-infected B-LCL, 40% of the T cells produced IFN- and TNF- as determined by intracellular cytokine staining (Fig. 2), despite the fact that HAdV infection of B-LCL is inefficient and results only in 5 10% infected cells as visualized by immunohistochemistry (Ref. 41 and data not shown). IL-2 was produced in few cells (Fig. 2), whereas IL-4, IL-5, and IL-10 were not detected (Fig. 2 and data not shown). When T cell clones were stimulated with peptide-loaded B-LCL, 90% of T cells coexpressed IFN- and TNF- (data not shown). Inhibition of HAdV replication in vitro If HAdV-specific CD4 T cells are to be administered to patients as adoptive immunotherapy, these T cells should have antiviral functions to combat the infection in vivo. To investigate whether HAdV-specific CD4 T cell clones can exert antiviral activity, an in vitro inhibition assay was developed to determine the effect of these T cell clones on HAdV replication. HAdV replication in B-LCL was determined by measuring the virus titers in cell lysates daily from days 0 through 3 after infection. HAdV virus titers increased 100- to 1000-fold in 3 days (Fig. 3A). When an HAdVspecific T cell clone, A1.1, was added to autologous infected B- LCL at an E:T ratio of 10:1, this increase in viral titers was almost completely inhibited (Fig. 3A). The inhibitory effect of the CD4 T cell clones was dependent on the number of T cells present. At an E:T ratio of 1:1, viral replication was still inhibited significantly, whereas inhibition could no longer be observed at an E:T ratio of 0.1:1 (Fig. 3B). Strong reduction in viral titers was observed not only when autologous B-LCL were used, but also with HLA-DR- or HLA-DP-matched B-LCL as targets (Fig. 3C). However, viral titers were not reduced when mismatched B-LCL were applied as target, indicating that inhibition of viral replication is HLA restricted (Fig. 3C). B cells are not likely to be the physiological target cells for infection with HAdV. To test the ability of the HAdV-specific T cells to inhibit virus production in a more appropriate cell type, PBEC with the relevant HLA-DR genotype were used. To induce HLA-DR on the surface of the EC, PBEC were treated with IFN- for 2 days before infection. Virus production in infected EC was inhibited by the T cells at a ratio of 50:1 and less well at a ratio of 10:1 (Fig. 3D). The higher E:T ratio needed to inhibit virus production in the primary EC, compared with when B-LCL were used as targets, may partly be due to the fact that the fraction of infected cells in the case of B-LCL (5 10%) is lower than that of the EC ( 50%) (data not shown). The results indicated that the HAdVspecific T cells also inhibited virus production in physiological target cells infected with HAdV. FIGURE 3. Inhibition of viral replication in B-LCL and PBEC by HAdV-specific CD4 T cell clones. The antiviral activity of T cell clones was tested in a 3-day in vitro assay in which viral replication was assessed. A, Autologous B-LCL were infected with HAdV5 at an MOI of 100 and cultured for 3 days in the presence (f) or absence ( ) of T cell clone A1.1 at an E:T ratio of 10:1. Cell lysates were obtained on days 0, 1, 2, and 3, and the titer of infectious virus was determined with the TCID 50 assay on HEp2 cells. B, Autologous B-LCL were infected with HAdV5 at an MOI of 100 and cultured for 3 days in the presence of the T cell clone K3.1 at different E:T ratios of 10:1, 1:1, and 0.1:1. At day 3, cell lysates were harvested and viral titers were determined. C, Autologous B-LCL, HLA- DR- or HLA-DP-matched B-LCL, or HLA-mismatched B-LCL were infected with HAdV5 at an MOI of 100 and used in the in vitro assay at an E:T ratio of 10:1. At day 3, cell lysates were harvested and viral titers were determined. Two representative examples (left panel, clone A1.1; right panel, clone K3.1) are shown of four clones (A1.1, M2.11, K3.1, A51.32) tested. D, PBEC cells were incubated for 1 day with IFN-, then trypsinized and replated in the presence of IFN- at 1 10 5 cells per well in 24-well plates. The next day, IFN- was removed, and the cells were infected with HAdV for 1 h. Virus was removed and T cells were added at E:T ratios of 50:1 and 10:1. Three days later, cell lysates were harvested and viral titers determined as described above (left panel, clone A.1.1; right panel, clone K3.1). Cognate interaction required for inhibition of viral replication To elucidate potential mechanisms by which these HAdV-specific CD4 T cell clones exert their antiviral effect, we investigated whether cognate interaction between T cells and target cells was required or whether the inhibition of viral replication could be achieved by soluble factors produced by the T cells. Therefore, a coculture experiment was performed with autologous B-LCL loaded with the specific peptide together with HLA-mismatched B-LCL infected with HAdV5. In this setting, T cell clones were stimulated by the peptide-loaded autologous B-LCL, while the viral replication was assessed in the HLA-mismatched B-LCL. Any factor secreted by the peptide-stimulated T cells may then directly have its potential effect on the virus production in HAdV-infected HLA-mismatched B-LCL. As a control, the virally infected B- LCL was, similar to peptide-loaded B-LCL, of autologous origin. As expected, a 1000-fold reduction in titer was seen in the control experiment in which T cells were cocultured with peptide-loaded as well as infected autologous B-LCL (Fig. 4). However, when the virus was replicating in the mismatched B-LCL, viral titer was

FIGURE 4. Inhibition of viral replication is not mediated via soluble factors. In a coculture experiment, autologous B-LCL were loaded with the specific peptide ( P) to stimulate the HAdV-specific T cell clone K3.1. In the same well, HLA-DR-mismatched B-LCL were added, which were infected with HAdV5 at an MOI of 100. In control wells, HAdV-infected autologous B-LCL were added that could directly be recognized by the T cell clone. T cells were added at an E:T ratio of 10:1. Viral titers were determined in cell lysates at day 3. Similar results were obtained with clone M2.11. reduced only marginally (Fig. 4). This lack of inhibition was observed despite the fact that the T cells were strongly stimulated by the peptide-loaded autologous B-LCL as measured by IFN- production in the supernatant ( 50 ng/ml IFN- ). Together, these results indicate that the antiviral effect of HAdV-specific CD4 T cell clones does not seem to be mediated by soluble factors (such as IFN- ) or non-hla-restricted cell-cell contact. Instead, cognate interaction between the T cell and the target cell appears to be required to achieve an antiviral effect. 8855 Lysis of HAdV-infected target cells As HLA-restricted recognition of target cells as well as cell-cell contact appeared to be essential for the antiviral function of the T cell clones, we investigated whether these HAdV-specific CD4 T cell clones had cytolytic activity against infected target cells. Maximal lysis of HAdV5-infected B-LCL was usually low ( 30%) (Fig. 5A), but this observation is probably due to the fact that HAdV infection of B-LCL is rather inefficient as mentioned above. However, when B-LCL were loaded with the specific 30-aa peptide, lysis increased to 80% or more (Fig. 5B). A melanoma cell line, MJS, that is more easily infected ( 90% infected cells; data not shown) and expresses HLA-DR17, was used to further investigate lysis of HAdV-infected cells. Using these targets, lysis of HAdV5-infected MJS by the CD4 HAdV-specific T cell clones was as efficient as lysis obtained with the peptide-loaded B-LCL (Fig. 5C). Mechanisms of contact-dependent T cell-induced death of target cells include release of preformed vacuolar perforin and granzymes into the intercellular space and up-regulation of cell surface molecules such as FAS ligand and TRAIL (42 44). The presence of both perforin and granzyme B in the cytoplasm of HAdV-specific T cell clones was confirmed by intracellular staining, supporting their role in HAdV-specific lysis (Fig. 6). The perforin effector pathway can be inhibited by the action of CMA, which raises the ph in intracellular vacuoles and inactivates perforin, as well as by EGTA which binds intracellular calcium that is required for release of perforin (45). Lysis was almost completely abrogated by either of these inhibitors, indicating that these CD4 T cells lyse predominantly using the perforin pathway (Fig. 5D and data not shown). Loading of hexon peptide on MHC class II molecules To address the question whether recognition of infected cells requires protein synthesis, we compared T cell recognition of cells exposed to infectious virus or MB-inactivated virus. MJS cells pulsed with inactivated HAdV were recognized as well as cells infected with infectious virus (Fig. 7A). MB inactivation prohibits gene transcription (34), indicating that viral protein synthesis is not required, and that processing of input virions can be sufficient for presentation of T cell epitopes on HLA class II. Subsequently, it was investigated whether the antigenic peptides recognized by HAdV-specific cytotoxic CD4 T cell clones could also be processed from proteins synthesized in the cytosol. Two MJS-derived cell lines were generated, each containing a different fragment of 700-bp coding sequence of the hexon gene, MJS- Hex-C and MJS-Hex-D. The antigenic peptides recognized by the CD4 T cell clones M2.11 and K3.1 are encoded by fragment D and not by fragment C. The T cell clones lysed the MJS-Hex-D, whereas the MJS-Hex-C cells were not lysed, indicating that the FIGURE 5. HAdV-specific CD4 T cell clones lyse HAdV-infected or peptide-loaded targets using the perforin-pathway. A, HLA-DR-matched (DR7/DR17-positive) B-LCL were infected with HAdV5 at an MOI of 100 (F), loaded with peptides II61 ( ) and II64 (Œ) or uninfected ( ) as controls (clone A1.1 does not recognize any peptide in the hexon protein) and used as targets in a 4-h CRA. T cells from clone A1.1 were added at different E:T ratios, and the percentage of specific lysis is shown. B, The same targets as in A were used in a CRA with T cells from clone K3.1 at different E:T ratios. C, MJS cells (HLA-DR17) were infected with HAdV5 at an MOI of 10 (F), loaded with specific peptide II57 (f), or uninfected as control ( ) and used as targets in a CRA with T cells from clone M2.11 as effectors. D, T cells from clone M2.11 were preincubated with 50 nm CMA for 1 h before the addition of T cells to MJS target cells in medium containing CMA. Lysis results in A are representative of eight T cell clones tested against HAdV5, and in B of five clones that were tested against the specific peptide. CMA results (D) are representative for three T cell clones tested.

8856 HAdV-SPECIFIC CD4 T CELLS INHIBIT VIRAL REPLICATION FIGURE 6. HAdV-specific CD4 T cell clones express perforin and granzyme B. Intracellular FACS staining was performed on T cell clones. Cells were stained with anti-perforin-fitc or anti-granzyme B-PE (filled histograms) or unstained as control (open histograms). Shown is one example of three T cell clones (A1.1, M2.11, and K3.1) tested. FIGURE 7. Recognition of cells with hexon protein from exogenous as well as endogenous sources. A, MJS cells ( ) were infected with infectious (F) or MB-inactivated HAdV (E) for1hat37 C, then medium was added, and after overnight incubation, the cells were used as targets in a CRA. B, Untransduced MJS ( ), MJS-Hex-C ( ), and MJS-Hex-D (Œ) were used as target cells. Lack of recognition of MJS-Hex-C was excluded to be due to changes in surface molecules by loading with relevant peptide (f). C, To exclude that MJS-Hex-D presented Ag by endocytosis of Ag secreted by neighboring cells, the MJS-Hex-D cells (Œ) were mixed in a 1:1 ratio with untransduced MJS cells (MJS, ). After 5 days of coculture, both cell lines were separated again by FACS based on egfp expression and used as target cells in a CRA (MJS after coculture, and MJS-Hex-D after coculture, ). As a control for cellular integrity after the sorting procedure, MJS cells were pulsed with the relevant peptide (f). All experiments shown were performed with clone M2.11, and similar results were obtained using clone K3.1. FIGURE 8. Inhibition of viral replication of serotypes from different HAdV species. MJS were infected at an MOI of 10 with HAdV5 and HAdV2 (both species C), HAdV12 (species A) and HAdV35 (species B). Inhibition of viral replication of these serotypes was tested using the T cell clone K3.1 at an E:T ratio of 10:1. Viral titers were determined in cell lysates at day 3. The sequence of the peptide II61 (recognized by clone K3.1) present in the hexons of HAdV5, 2, 12, and 35, respectively, is depicted below the figure and illustrates its conserved nature. epitope was processed from cytosolic protein sources (Fig. 7B and data not shown). As a control, MJS-Hex-C cells pulsed with the relevant peptides were lysed, indicating that this cell line still expressed all molecules necessary for presentation of Ag. A possible explanation for the observed recognition of MJS- Hex-D cells could be that endogenously produced protein is released from the cells during culture and is taken up by neighboring cells and presented as if it is exogenously derived. To test this hypothesis, untransduced MJS cells were mixed in a 1:1 ratio with the MJS-Hex-D cells (which are egfp ). After 5 days of coculture, the two cell populations were separated by FACS on the basis of egfp expression and used as target cells in a cytotoxicity assay. Clones M2.11 and K3.1 did not recognize the egfp neg MJS cells, indicating that processing of hexon protein fragment released in any form from neighboring MJS-Hex-D cells had not occurred at a sufficiently high level to lead to recognition (Fig. 7C and data not shown). The same (untransduced and sorted MJS) cells pulsed with the relevant peptide were lysed, indicating that the experimental procedures had not rendered the cells resistant to cytotoxicity. These results indicate that recognition of the transduced MJS-Hex-D cells by the T cells is not due to endocytosis of secreted protein or cellular fragments from neighboring cells, suggesting that endogenously produced hexon protein is also presented directly on HLA class II molecules in infected MJS cells. Inhibition of viral replication of different serotypes by HAdV-specific T cell clones After SCT, patients can be infected with HAdV belonging to various serotypes. If adoptive immunotherapy is to be effective, T cells should be cross-reactive to multiple serotypes. The peptide recognized by the broadly cross-reactive HAdV-specific T cell clone, K3.1, is reasonably well conserved between several serotypes of HAdV (Fig. 8). To analyze whether the T cell clones are also able to inhibit viral replication of different serotypes, viral replication inhibition assays with multiple serotypes from different HAdV species were performed. As not all serotypes tested were able to replicate in B-LCL, the MJS melanoma cell line was used in these experiments. Clone K3.1 was able to inhibit viral replication of HAdV5 as well as HAdV2 (belonging to the same species C), HAdV12 (species A) and HAdV35 (species B) (all 1000-fold reduction of titer), confirming that T cell clones showing cross-reactivity in proliferation assays can exert crossreactive antiviral activity as well (Ref. 31 and Fig. 8). These results are promising for adoptive immunotherapy, as T cells generated against HAdV5 could also be effective in patients infected with different serotypes of HAdV. Discussion To develop clinical protocols for adoptive immunotherapy with T cells for children suffering from disseminated HAdV infections, we and others have studied the human immune response to HAdV (19, 28, 31, 32, 46, 47). In healthy donors the majority of T cells reactive against HAdV is CD4, although CD8 T cells are also observed. In general, CD4 T cells can mediate antiviral effects indirectly by providing help to B cells in the production of neutralizing Abs or by initiating and maintaining an HLA class I-restricted CD8 T cell response through the activation of dendritic cells (48). More directly, CD4 T cells can restrict viral growth by producing cytokines, e.g., IFN-, or by lysing infected target cells (49, 50). For HAdV, the relative contribution of CD4 and CD8 T cells and of Ab-producing B cells to in vivo protection against

infection are largely unknown. Murine models using HAdV provide only limited information because HAdV does not produce a permissive infection in murine cells. Studies in mice using murine adenovirus reveal that B cells are involved in protection from early disease, whereas T cells, either CD4 or CD8, are responsible for late viral clearance, even in the absence of perforin (51, 52). However, there are significant differences with adenovirus infections in humans; the main target cells in mice are endothelial cells vs EC in humans, T cells can cause immunopathology, and the course of infection is dependent on the genetic background of the mice (52). Therefore, conclusions about protective immune responses in murine models of adenovirus infection are of limited relevance to HAdV infections in humans. Because it has not been investigated whether and by which mechanism CD4 HAdV-specific T cells are capable of mediating antiviral activity, several CD4 HAdV-specific T cell clones were generated. The clones recognize peptides derived from structural viral proteins and appeared to have a Th1-like phenotype based on the secreted cytokines. Whether directly ex vivo isolated memory T cells have the same cytokine secretion pattern remains to be investigated. These clones were then tested in a newly developed in vitro assay in which the effect of T cells on viral replication was measured. In this assay, viral replication was determined in infected B-LCL, which have the advantage that they can be generated from PBMC of a large panel of donors expressing almost all HLA alleles. Although B cells may not be the main physiological target of HAdV, they do facilitate a productive infection in which a fraction (5 10%) of the cells is infected (53). The addition of a T cell clone at a ratio of 10:1 or 1:1 reduced viral replication by a factor of 1000 in 3 days. Although the addition of the T cell clone did not result in a lower number of B-LCL, the number of infected cells, as shown by intense staining with an anti-hexon Ab, was strongly decreased (data not shown). Virus production in melanoma cells and also in PBEC, which constitute a more physiological host cell type for HAdV, was also inhibited by the T cell clones. Because HAdV infects primarily EC, which under normal conditions do not express high levels of HLA class II molecules, the question arises how HAdV-specific CD4 T cells would be able to recognize their infected targets. It could be that the levels of HLA class II expression are high enough for activated T cells to be able to recognize the infected EC. In experiments with peptide-pulsed PBEC that were not pretreated with IFN-, low levels of IFN- were produced by T cells indicating that at least some recognition occurred (data not shown). Subsequently, the IFN- produced by the activated T cells could then induce class II up-regulation in the infected cells and the surrounding uninfected EC (54). Alternatively, class II expression on these cells might be up-regulated by the HAdV infection itself via production of IFN- or IFN- and other stress proteins or by interaction with IFN- -producing NK cells. Preliminary data from our group obtained in postmortem biopsies from patients who succumbed to HAdV infection showed that, in some cases, HAdV-infected cells were strongly positive for HLA class II (B. Heemskerk, unpublished data). For other viruses such as CMV and EBV, numerous reports on virus-specific CD8 T cells have been published (reviewed in Ref. 55). Nevertheless, in recent years, a growing number of reports on virus-specific CD4 T cells with lytic activity have been published for EBV, CMV, and HSV (49, 56, 57), which in some reports have been shown to inhibit viral outgrowth in vitro (49, 58, 59). Most reports indicate that perforin-mediated lysis is the main mechanism of human cytotoxic CD4 T cells, either directly ex vivo (60) or in cell lines or clones (49, 56, 57, 61), whereas some CD4 T cell clones have been shown to use the FAS/FAS-L pathway (58). 8857 In this respect, it is noteworthy that HAdV has developed immune evasion mechanisms. The early protein E3-gp19K has been reported to down-regulate MHC class I expression by retention of MHC molecules in the endoplasmic reticulum, which might result in reduced recognition of the infected cell by CD8 T cells and, conversely, in increased NK cell recognition (reviewed in Ref. 62). Another early E3 protein complex, the RID (receptor internalization and degradation) complex, has been reported to internalize FAS, TNF-RI, and TRAIL-R and induce their degradation, thereby circumventing FAS- or TRAIL-mediated lysis of infected cells (63). In view of these immune evasion strategies of HAdV, it is not surprising that HAdV-specific, HLA class II-restricted CD4 T cells are present that have antiviral activity against HAdV by mediating lysis of infected cells using the perforin pathway. Nevertheless, HAdV-specific CD8 T cells have been described (32, 33), indicating that these viral evasion mechanisms may not be absolute. The origin of the peptides presented in the HLA class II molecule and recognized by the CD4 T cells is a point of interest. When cells are infected with HAdV, hexon protein is available from the input virions, but hexon will also be produced intracellularly for the production of new virions. Classically, it is assumed that antigenic determinants presented on HLA class II originate from endocytosed Ags and are processed via the endosomal-lysosomal pathway. However, a growing number of recent studies describe exceptions to this assumption (64 69). The hexon epitopes described in this study were shown to be presented from exogenous sources, as illustrated by the recognition of target cells pulsed with inactivated virus or hexon protein fragments. In addition, cells retrovirally transduced with the relevant fragment of the hexon gene were recognized, indicating that processing and presentation of endogenously produced hexon protein can occur as well, although expression of full-length hexon protein could have given different results, for example, because of its localization to different cellular compartments. In cultures of the transduced cells, exogenous loading of the HLA class II molecules by peptides taken up from neighboring cells was not very likely as demonstrated in the 5-day coculture experiments, although it cannot be excluded that longer periods of coculture would have been necessary to observe such events. It was reported recently that, in cultures of B-LCL, transfer of BHRF1 Ag derived from the few cells entering lytic cycle to latently infected cells could take as long as 21 days (70). Thus, despite these limitations, our experiments suggest that the same functional HAdV epitope can be presented via both an exogenous as well as an endogenous pathway. During an in vivo infection, cells are most likely infected by only few virus particles. If CD4, HLA class II-restricted T cells are to be effective in the response to HAdV-infected cells, antigenic peptides presented directly after intracellular synthesis of hexon protein on HLA class II molecules should also be recognized. It is tempting to speculate that, if the reactivity of HLA class I-restricted CD8 T cells is hampered due to immune evasive strategies of a virus, HLA class II-restricted CD4 T cells can and will fulfill this function by expressing perforin-mediated cytotoxicity (71). The fact that the epitopes recognized by the cytotoxic CD4 T cells are presented on class II MHC molecules when derived from endogenous sources would be in agreement with this role-switching and raises an interesting question. Could there be a correlation between processing of epitopes from endogenous sources and cytotoxicity of CD4 T cells? The cross-reactive inhibition of viral replication of serotypes from different species of HAdV by the CD4 T cell clones, which were generated against HAdV5, is encouraging for the clinical

8858 HAdV-SPECIFIC CD4 T CELLS INHIBIT VIRAL REPLICATION application of these cells in immunocompromised patients suffering from disseminated HAdV infection of various serotypes. However, the observed cytolytic activity and cross-reactivity of HAdVspecific CD4 T cell clones may have implications for gene therapy trials with HAdV-based vectors as well. The longevity of gene expression might be reduced when HAdV-specific, cytotoxic CD4 T cells are present, and the cross-reactivity of HAdV-specific T cells may also reduce the applicability of switching to vectors based on different serotypes of HAdV. In conclusion, HAdVspecific CD4 T cell clones have antiviral activity, as they were able to inhibit viral replication in vitro, most likely dependent on perforin-mediated lysis. It is to be expected that, in the near future, generation of HAdV-specific T cell lines with antiviral activity may become a feasible option for immunotherapy in immunocompromised patients suffering from disseminated HAdV infections. Acknowledgments We thank Martijn Rabelink (Department of Molecular Cell Biology (Leiden University Medical Center) for virus purification; Dr. Jan-Wouter Drijfhout (Department of Immunohematology and Blood Transfusion, Leiden University Medical Center) for synthesizing the peptides; Prof. Rob Hoeben for fruitful discussions; and Prof. Kees Melief and Dr. Nicola Annels for critical reading of the manuscript. Disclosures The authors have no financial conflict of interest. References 1. Hierholzer, J. 1992. Adenoviruses in the immunocompromised host. Clin. Microbiol. Rev. 5: 262 274. 2. Ljungman, P. 1997. Respiratory virus infections in bone marrow transplant recipients: the European perspective. Am. J. Med. 102: 44 47. 3. Hale, G. A., H. E. Heslop, R. A. Krance, M. A. Brenner, D. Jayawardene, D. K. Srivastava, and C. C. Patrick. 1999. Adenovirus infection after pediatric bone marrow transplantation. Bone Marrow Transplant. 23: 277 282. 4. Baldwin, A., H. Kingman, M. Darville, A. B. Foot, D. Grier, J. M. Cornish, N. Goulden, A. Oakhill, D. H. Pamphilon, C. G. Steward, and D. I. Marks. 2000. Outcome and clinical course of 100 patients with adenovirus infection following bone marrow transplantation. Bone Marrow Transplant. 26: 1333 1338. 5. Runde, V., S. Ross, R. Trenschel, E. Lagemann, O. Basu, K. Renzing-Kohler, U. W. Schaefer, M. Roggendorf, and E. Holler. 2001. Adenoviral infection after allogeneic stem cell transplantation (SCT): report on 130 patients from a single SCT unit involved in a prospective multi center surveillance study. Bone Marrow Transplant. 28: 51 57. 6. Kojaoghlanian, T., P. Flomenberg, and M. S. Horwitz. 2003. The impact of adenovirus infection on the immunocompromised host. Rev. Med. Virol. 13: 155 171. 7. Heemskerk, B., A. C. Lankester, T. van Vreeswijk, M. F. Beersma, E. C. Claas, L. A. Veltrop-Duits, A. C. Kroes, J. M. Vossen, M. W. Schilham, and M. J. van Tol. 2005. Immune reconstitution and clearance of human adenovirus viremia in pediatric stem-cell recipients. J. Infect. Dis. 191: 520 530. 8. Chakrabarti, S., V. Mautner, H. Osman, K. E. Collingham, C. D. Fegan, P. E. Klapper, P. A. Moss, and D. W. Milligan. 2002. Adenovirus infections following allogeneic stem cell transplantation: incidence and outcome in relation to graft manipulation, immunosuppression, and immune recovery. Blood 100: 1619 1627. 9. van Tol, M. J., A. C. Kroes, J. Schinkel, W. Dinkelaar, E. C. Claas, C. M. Jol-Van Der Zijde, and J. M. Vossen. 2005. Adenovirus infection in paediatric stem cell transplant recipients: increased risk in young children with a delayed immune recovery. Bone Marrow Transplant. 36: 39 50. 10. Echavarria, M., M. Forman, M. J. van Tol, J. M. Vossen, P. Charache, and A. C. Kroes. 2001. Prediction of severe disseminated adenovirus infection by serum PCR. Lancet 358: 384 385. 11. Schilham, M. W., E. C. Claas, W. van Zaane, B. Heemskerk, J. M. Vossen, A. C. Lankester, R. E. Toes, M. Echavarria, A. C. Kroes, and M. J. van Tol. 2002. High levels of adenovirus DNA in serum correlate with fatal outcome of adenovirus infection in children after allogeneic stem-cell transplantation. Clin. Infect. Dis. 35: 526 532. 12. Lankester, A. C., M. J. van Tol, E. C. Claas, J. M. Vossen, and A. C. Kroes. 2002. Quantification of adenovirus DNA in plasma for management of infection in stem cell graft recipients. Clin. Infect. Dis. 34: 864 867. 13. Lion, T., R. Baumgartinger, F. Watzinger, S. Matthes-Martin, M. Suda, S. Preuner, B. Futterknecht, A. Lawitschka, C. Peters, U. Potschger, and H. Gadner. 2003. Molecular monitoring of adenovirus in peripheral blood after allogeneic bone marrow transplantation permits early diagnosis of disseminated disease. Blood 102: 1114 1120. 14. Leruez-Ville, M., V. Minard, F. Lacaille, A. Buzyn, E. Abachin, S. Blanche, F. Freymuth, and C. Rouzioux. 2004. Real-time blood plasma polymerase chain reaction for management of disseminated adenovirus infection. Clin. Infect. Dis. 38: 45 52. 15. Horwitz, M. S. 2001. Adenoviruses. In Fields Virology. D. M. Knipe and P. M. Howley, eds. Lippincott Williams & Wilkins, Philadelphia, pp. 2301 2326. 16. De Jong, J. C., A. G. Wermenbol, M. W. Verweij-Uijterwaal, K. W. Slaterus, P. Wertheim-Van Dillen, G. J. Van Doornum, S. H. Khoo, and J. C. Hierholzer. 1999. Adenoviruses from human immunodeficiency virus-infected individuals, including two strains that represent new candidate serotypes Ad50 and Ad51 of species B1 and D, respectively. J. Clin. Microbiol. 37: 3940 3945. 17. Flomenberg, P., J. Babbitt, W. R. Drobyski, R. C. Ash, D. R. Carrigan, G. V. Sedmak, T. McAuliffe, B. Camitta, M. M. Horowitz, N. Bunin, et al. 1994. Increasing incidence of adenovirus disease in bone marrow transplant recipients. J. Infect. Dis. 169: 775 781. 18. Carrigan, D. R. 1997. Adenovirus infections in immunocompromised patients. Am. J. Med. 102: 71 74. 19. Flomenberg, P., V. Piaskowski, R. L. Truitt, and J. T. Casper. 1995. Characterization of human proliferative T cell responses to adenovirus (see comments). J. Infect. Dis. 171: 1090 1096. 20. Ljungman, P. 2004. Treatment of adenovirus infections in the immunocompromised host. Eur. J. Clin. Microbiol. Infect. Dis. 23: 583 588. 21. Walter, E. A., P. D. Greenberg, M. J. Gilbert, R. J. Finch, K. S. Watanabe, E. D. Thomas, and S. R. Riddell. 1995. Reconstitution of cellular immunity against cytomegalovirus in recipients of allogeneic bone marrow by transfer of T-cell clones from the donor. N. Engl. J. Med. 333: 1038 1044. 22. Heslop, H. E., and C. M. Rooney. 1997. Adoptive cellular immunotherapy for EBV lymphoproliferative disease. Immunol. Rev. 157: 217 222. 23. Einsele, H., E. Roosnek, N. Rufer, C. Sinzger, S. Riegler, J. Loffler, U. Grigoleit, A. Moris, H. G. Rammensee, L. Kanz, et al. 2002. Infusion of cytomegalovirus (CMV)-specific T cells for the treatment of CMV infection not responding to antiviral chemotherapy. Blood 99: 3916 3922. 24. Hromas, R., K. Cornetta, E. Srour, C. Blanke, and E. R. Broun. 1994. Donor leukocyte infusion as therapy of life-threatening adenoviral infections after T cell depleted bone marrow transplantation. Blood 84: 1689 1690. 25. Chakrabarti, S., K. E. Collingham, C. D. Fegan, D. Pillay, and D. W. Milligan. 2000. Adenovirus infections following haematopoietic cell transplantation: is there a role for adoptive immunotherapy? Bone Marrow Transplant. 26: 305 307. 26. Smith, C. A., L. S. Woodruff, G. R. Kitchingman, and C. Rooney. 1996. Adenovirus-pulsed dendritic cells stimulate human virus-specific T-cell responses in vitro. J. Virol. 70: 6733 6740. 27. Souberbielle, B. E., and W. C. Russell. 1995. Human T cell proliferative responses to polypeptides from Adenovirus type 2. J. Infect. Dis. 172: 1421 1422. 28. Olive, M., L. C. Eisenlohr, and P. Flomenberg. 2001. Quantitative analysis of adenovirus-specific CD4 T-cell responses from healthy adults. Viral Immunol. 14: 403 413. 29. Olive, M., L. Eisenlohr, N. Flomenberg, S. Hsu, and P. Flomenberg. 2002. The adenovirus capsid protein hexon contains a highly conserved human CD4 T-cell epitope. Hum. Gene Ther. 13: 1167 1178. 30. Leen, A. M., U. Sili, B. Savoldo, A. M. Jewell, P. A. Piedra, M. K. Brenner, and C. M. Rooney. 2004. Fiber-modified adenoviruses generate subgroup cross-reactive, adenovirus-specific cytotoxic T lymphocytes for therapeutic applications. Blood 103: 1011 1019. 31. Heemskerk, B., L. A. Veltrop-Duits, T. van Vreeswijk, M. M. ten Dam, S. Heidt, R. E. Toes, M. J. van Tol, and M. W. Schilham. 2003. Extensive cross-reactivity of CD4 adenovirus-specific T cells: implications for immunotherapy and gene therapy. J. Virol. 77: 6562 6566. 32. Leen, A. M., U. Sili, E. F. Vanin, A. M. Jewell, W. Xie, D. Vignali, P. A. Piedra, M. K. Brenner, and C. M. Rooney. 2004. Conserved CTL epitopes on the adenovirus hexon protein expand subgroup cross-reactive and subgroup-specific CD8 T cells. Blood 104: 2432 2440. 33. Tang, J., M. Olive, R. Pulmanausahakul, M. Schnell, N. Flomenberg, L. Eisenlohr, and P. Flomenberg. 2006. Human CD8 cytotoxic T cell responses to adenovirus capsid proteins. Virology 350: 312 322. 34. Schagen, F. H., A. C. Moor, S. C. Cheong, S. J. Cramer, H. van Ormondt, A. J. van der Eb, T. M. Dubbelman, and R. C. Hoeben. 1999. Photodynamic treatment of adenoviral vectors with visible light: an easy and convenient method for viral inactivation. Gene Ther. 6: 873 881. 35. Kinsella, T. M., and G. P. Nolan. 1996. Episomal vectors rapidly and stably produce high-titer recombinant retrovirus. Hum. Gene Ther. 7: 1405 1413. 36. van Wetering, S., A. C. van der Linden, M. A. van Sterkenburg, K. F. Rabe, J. Schalkwijk, and P. S. Hiemstra. 2000. Regulation of secretory leukocyte proteinase inhibitor (SLPI) production by human bronchial epithelial cells: increase of cell-associated SLPI by neutrophil elastase. J. Investig. Med. 48: 359 366. 37. Veltrop-Duits, L. A., B. Heemskerk, C. C. Sombroek, T. van Vreeswijk, S. Gubbels, R. E. Toes, C. J. Melief, K. L. Franken, M. Havenga, M. J. van Tol, and M. W. Schilham. 2006. Human CD4 T cells stimulated by conserved adenovirus 5 hexon peptides recognize cells infected with different species of human adenovirus. Eur. J. Immunol. 36: 2410 2423. 38. Scott-Taylor, T. H., and G. W. Hammond. 1992. Conserved sequences of the adenovirus genome for detection of all human adenovirus types by hybridization. J. Clin. Microbiol. 30: 1703 1710. 39. Rux, J. J., and R. M. Burnett. 2000. Type-specific epitope locations revealed by X-ray crystallographic study of adenovirus type 5 hexon. Mol. Ther. 1: 18 30. 40. Tang, J., M. Olive, K. Champagne, N. Flomenberg, L. Eisenlohr, S. Hsu, and P. Flomenberg. 2004. Adenovirus hexon T-cell epitope is recognized by most adults and is restricted by HLA DP4, the most common class II allele. Gene Ther. 11: 1408 1415.

8859 41. Regn, S., S. Raffegerst, X. Chen, D. Schendel, H. J. Kolb, and M. Roskrow. 2001. Ex vivo generation of cytotoxic T lymphocytes specific for one or two distinct viruses for the prophylaxis of patients receiving an allogeneic bone marrow transplant. Bone Marrow Transplant. 27: 53 64. 42. Smyth, M. J., J. M. Kelly, V. R. Sutton, J. E. Davis, K. A. Browne, T. J. Sayers, and J. A. Trapani. 2001. Unlocking the secrets of cytotoxic granule proteins. J. Leukocyte Biol. 70: 18 29. 43. Shresta, S., C. T. Pham, D. A. Thomas, T. A. Graubert, and T. J. Ley. 1998. How do cytotoxic lymphocytes kill their targets? Curr. Opin. Immunol. 10: 581 587. 44. Kayagaki, N., N. Yamaguchi, M. Nakayama, A. Kawasaki, H. Akiba, K. Okumura, and H. Yagita. 1999. Involvement of TNF-related apoptosis-inducing ligand in human CD4 T cell-mediated cytotoxicity. J. Immunol. 162: 2639 2647. 45. Kataoka, T., N. Shinohara, H. Takayama, K. Takaku, S. Kondo, S. Yonehara, and K. Nagai. 1996. Concanamycin A, a powerful tool for characterization and estimation of contribution of perforin- and Fas-based lytic pathways in cell-mediated cytotoxicity. J. Immunol. 156: 3678 3686. 46. Sester, M., U. Sester, S. A. Salvador, G. Heine, S. Lipfert, M. Girndt, B. Gartner, and H. Kohler. 2002. Age-related decrease in adenovirus-specific T cell responses. J. Infect. Dis. 185: 1379 1387. 47. Smith, C. A., L. S. Woodruff, C. Rooney, and G. R. Kitchingman. 1998. Extensive cross-reactivity of adenovirus-specific cytotoxic T cells. Hum. Gene Ther. 9: 1419 1427. 48. Schoenberger, S. P., R. E. Toes, E. I. van der Voort, R. Offringa, and C. J. Melief. 1998. T-cell help for cytotoxic T lymphocytes is mediated by CD40-CD40L interactions. Nature 393: 480 483. 49. Adhikary, D., U. Behrends, A. Moosmann, K. Witter, G. W. Bornkamm, and J. Mautner. 2006. Control of Epstein-Barr virus infection in vitro by T helper cells specific for virion glycoproteins. J. Exp. Med. 203: 995 1006. 50. Brown, D. M., E. Roman, and S. L. Swain. 2004. CD4 T cell responses to influenza infection. Semin. Immunol. 16: 171 177. 51. Moore, M. L., E. L. McKissic, C. C. Brown, J. E. Wilkinson, and K. R. Spindler. 2004. Fatal disseminated mouse adenovirus type 1 infection in mice lacking B cells or Bruton s tyrosine kinase. J. Virol. 78: 5584 5590. 52. Moore, M. L., C. C. Brown, and K. R. Spindler. 2003. T cells cause acute immunopathology and are required for long-term survival in mouse adenovirus type 1-induced encephalomyelitis. J. Virol. 77: 10060 10070. 53. Lavery, D., S. M. Fu, T. Lufkin, and S. Chen-Kiang. 1987. Productive infection of cultured human lymphoid cells by adenovirus. J. Virol. 61: 1466 1472. 54. Boss, J. M., and P. E. Jensen. 2003. Transcriptional regulation of the MHC class II antigen presentation pathway. Curr. Opin. Immunol. 15: 105 111. 55. Van Lier, R. A., I. J. ten Berge, and L. E. Gamadia. 2003. Human CD8 T-cell differentiation in response to viruses. Nat. Rev. Immunol. 3: 931 939. 56. Khanolkar, A., H. Yagita, and M. J. Cannon. 2001. Preferential utilization of the perforin/granzyme pathway for lysis of Epstein-Barr virus-transformed lymphoblastoid cells by virus-specific CD4 T cells. Virology 287: 79 88. 57. Yasukawa, M., H. Ohminami, Y. Yakushijin, J. Arai, A. Hasegawa, Y. Ishida, and S. Fujita. 1999. Fas-independent cytotoxicity mediated by human CD4 CTL directed against herpes simplex virus-infected cells. J. Immunol. 162: 6100 6106. 58. Nikiforow, S., K. Bottomly, G. Miller, and C. Munz. 2003. Cytolytic CD4 -Tcell clones reactive to EBNA1 inhibit Epstein-Barr virus-induced B-cell proliferation. J. Virol. 77: 12088 12104. 59. Le Roy, E., M. Baron, W. Faigle, D. Clement, D. M. Lewinsohn, D. N. Streblow, J. A. Nelson, S. Amigorena, and J. L. Davignon. 2002. Infection of APC by human cytomegalovirus controlled through recognition of endogenous nuclear immediate early protein 1 by specific CD4 T lymphocytes. J. Immunol. 169: 1293 1301. 60. Appay, V., J. J. Zaunders, L. Papagno, J. Sutton, A. Jaramillo, A. Waters, P. Easterbrook, P. Grey, D. Smith, A. J. McMichael, et al. 2002. Characterization of CD4 CTLs ex vivo. J. Immunol. 168: 5954 5958. 61. Yanai, F., E. Ishii, K. Kojima, A. Hasegawa, T. Azuma, S. Hirose, N. Suga, A. Mitsudome, M. Zaitsu, Y. Ishida, et al. 2003. Essential roles of perforin in antigen-specific cytotoxicity mediated by human CD4 T lymphocytes: analysis using the combination of hereditary perforin-deficient effector cells and Fas-deficient target cells. J. Immunol. 170: 2205 2213. 62. Wold, W. S., K. Doronin, K. Toth, M. Kuppuswamy, D. L. Lichtenstein, and A. E. Tollefson. 1999. Immune responses to adenoviruses: viral evasion mechanisms and their implications for the clinic. Curr. Opin. Immunol. 11: 380 386. 63. Tollefson, A. E., T. W. Hermiston, D. L. Lichtenstein, C. F. Colle, R. A. Tripp, T. Dimitrov, K. Toth, C. E. Wells, P. C. Doherty, and W. S. Wold. 1998. Forced degradation of Fas inhibits apoptosis in adenovirus-infected cells. Nature 392: 726 730. 64. Hegde, N. R., C. Dunn, D. M. Lewinsohn, M. A. Jarvis, J. A. Nelson, and D. C. Johnson. 2005. Endogenous human cytomegalovirus gb is presented efficiently by MHC class II molecules to CD4 CTL. J. Exp. Med. 202: 1109 1119. 65. Mukherjee, P., A. Dani, S. Bhatia, N. Singh, A. Y. Rudensky, A. George, V. Bal, S. Mayor, and S. Rath. 2001. Efficient presentation of both cytosolic and endogenous transmembrane protein antigens on MHC class II is dependent on cytoplasmic proteolysis. J. Immunol. 167: 2632 2641. 66. Tewari, M. K., G. Sinnathamby, D. Rajagopal, and L. C. Eisenlohr. 2005. A cytosolic pathway for MHC class II-restricted antigen processing that is proteasome and TAP dependent. Nat. Immunol. 6: 287 294. 67. Lich, J. D., J. F. Elliott, and J. S. Blum. 2000. Cytoplasmic processing is a prerequisite for presentation of an endogenous antigen by major histocompatibility complex class II proteins. J. Exp. Med. 191: 1513 1524. 68. Dani, A., A. Chaudhry, P. Mukherjee, D. Rajagopal, S. Bhatia, A. George, V. Bal, S. Rath, and S. Mayor. 2004. The pathway for MHCII-mediated presentation of endogenous proteins involves peptide transport to the endo-lysosomal compartment. J. Cell Sci. 117: 4219 4230. 69. Paludan, C., D. Schmid, M. Landthaler, M. Vockerodt, D. Kube, T. Tuschl, and C. Munz. 2005. Endogenous MHC class II processing of a viral nuclear antigen after autophagy. Science 307: 593 596. 70. Landais, E., X. Saulquin, M. Bonneville, and E. Houssaint. 2005. Long-term MHC class II presentation of the EBV lytic protein BHRF1 by EBV latently infected b cells following capture of BHRF1 antigen. J. Immunol. 175: 7939 7946. 71. Heemskerk, M. H., H. M. Schoemaker, W. J. Spaan, and C. J. Boog. 1995. Predominance of MHC class II-restricted CD4 cytotoxic T cells against mouse hepatitis virus A59. Immunology 84: 521 527.