Reactivation of Latent Human Cytomegalovirus in ... - Journal of Virology

JOURNAL OF VIROLOGY, Aug. 2001, p. 7543–7554 0022-538X/01/$04.00⫹0 DOI: 10.1128/JVI.75.16.7543–7554.2001 Copyright © 2001, American Society for Microbiology. All Rights Reserved.

Vol. 75, No. 16

Reactivation of Latent Human Cytomegalovirus in CD14⫹ Monocytes Is Differentiation Dependent ¨ DERBERG-NAUCLE ´ R,1,2* DANIEL N. STREBLOW,1 KENNETH N. FISH,1 CECILIA SO JUSTINE ALLAN-YORKE,3 PATRICIA P. SMITH,1 AND JAY A. NELSON1* Oregon Health Sciences University, Portland, Oregon 972011; Karolinska Institute, Department for Biosciences at Novum, Huddinge, Sweden2; and Institut National de la Sante´ et de la Recherche Me´dı`cale 395, 31024 Toulouse Cedex, France3 Received 29 January 2001/Received 4 May 2001

We have previously demonstrated reactivation of latent human cytomegalovirus (HCMV) in myeloid lineage cells obtained from healthy donors. Virus was obtained from allogenically stimulated monocyte-derived macrophages (Allo-MDM), but not from macrophages differentiated by mitogenic stimulation (ConA-MDM). In the present study, the cellular and cytokine components essential for HCMV replication and reactivation were examined in Allo-MDM. The importance of both CD4ⴙ and CD8ⴙ T cells in the generation of HCMVpermissive Allo-MDM was demonstrated by negative selection or blocking experiments using antibodies directed against both HLA class I and HLA class II molecules. Interestingly, contact of monocytes with CD4 or CD8 T cells was not essential for reactivation of HCMV, since virus was observed in macrophages derived from CD14ⴙ monocytes stimulated by supernatants produced by allogeneic stimulation of peripheral blood mononuclear cells. Examination of the cytokines produced in Allo-MDM and ConA-MDM cultures indicated a significant difference in the kinetics of production and quantity of these factors. Further examination of the cytokines essential for the generation of HCMV-permissive Allo-MDM identified gamma interferon (IFN-␥) but not interleukin-1 or -2, tumor necrosis factor alpha, or granulocyte-macrophage colony-stimulating factor as critical components in the generation of these macrophages. In addition, although IFN-␥ was crucial for reactivation of latent HCMV, addition of IFN-␥ to unstimulated macrophage cultures was insufficient to reactivate virus. Thus, this study characterizes two distinct monocyte-derived cell types which can be distinguished by their ability to reactivate and support HCMV replication and identifies the critical importance of IFN-␥ in the reactivation of HCMV. established to examine mechanisms of HCMV replication in vitro (19, 23, 28, 30, 55). In these studies, the ability of the virus to replicate in monocyte-derived macrophages (MDM) was dependent on the state of cellular differentiation. Infection of unstimulated monocytes resulted in either a lack of viral gene expression or replication restricted to immediate-early gene products (19, 30, 49). The block in HCMV expression in unstimulated monocytes was not at the level of virus entry and fusion with the cell, but rather at the level of transcriptional or posttranscriptional events (13, 19–21, 39). Differentiation of monocytes into macrophages resulting in fully permissive HCMV infection can be achieved by a number of different methods. One of the better-characterized MDM systems is based on concanavalin A (ConA) stimulation of autologous peripheral blood mononuclear cells (PBMC) for a defined period of time to allow macrophage differentiation (19). These HCMV-permissive macrophages can be maintained for prolonged periods without the addition of cytokines. We previously identified the specific cell-cell interactions and cytokines which were essential for ConA-mediated differentiation of HCMV-permissive macrophages in this system. HCMV replication in ConA-stimulated MDM cultures was dependent on the presence of CD8-positive T lymphocytes and the production of gamma interferon (IFN-␥) and tumor necrosis factor alpha (TNF-␣) (43). Although extensive studies have been performed to obtain HCMV from latently infected monocytes, reactivation of virus has not been demonstrated in ConA-MDM or other macrophage in vitro systems. However, reactivation of latent HCMV

Human cytomegalovirus (HCMV) infection remains a major cause of morbidity and mortality in transplant patients and AIDS patients. As with other members of the herpesvirus group, HCMV primary infection results in life-long persistence of the virus in the host, and reactivation frequently occurs in immunocompromised individuals. Reactivation of HCMV and severe disease development are common in bone marrow and solid organ transplant patients and have also been associated with complications following transplantation, such as acute graft-versus-host disease and acute rejection. Early epidemiological studies demonstrated transmission of HCMV by blood products, bone marrow grafts, and solid organs (5–8, 29, 50). Analysis of separated peripheral blood cell populations derived from individuals with HCMV disease (25, 41, 54) or asymptomatically infected individuals (9, 48) identified monocytes as the predominant infected cell type. Further examination of organ tissues by double-label immunohistochemistry with antibodies directed against viral antigens and cellular markers (14, 40) identified macrophages as a major source of virus early in the course of HCMV disease. Several primary monocyte-macrophage systems have been

* Corresponding author. Mailing address for Jay A. Nelson: Dept of Microbiology and Immunology, Oregon Health Sciences University, Portland, OR 97201-3098. Phone: (503) 494-7769. Fax: (503) 494-2441. E-mail: [email protected]. Mailing address for Cecilia So ¨derbergNaucle´r: Karolinska Institute, Department for Biosciences at Novum, S-141 57 Huddinge, Sweden. Phone: 46 8 608 9118. Fax: 46 8 774 5538. E-mail: [email protected]. 7543

7544

¨ DERBERG-NAUCLE ´ R ET AL. SO

was recently achieved in allogeneically stimulated monocytederived macrophages (Allo-MDM) from healthy blood donors. These results provided the first evidence that HCMV establishes a true latent infection in myeloid lineage cells, which can be reactivated upon allogeneic stimulation (42). The reactivation of HCMV in Allo-MDM but not in ConA-MDM suggests that the differentiation pathway of MDM mediated by antigenspecific recognition of activated T cells during an allogeneic reaction differs from the ConA-induced differentiation of MDM. In this study, we examined the cellular and cytokine components which were essential for HCMV replication and reactivation of latent virus in Allo-MDM. Our results indicate that the initial stimulus to induce monocyte differentiation is critical in the generation of HCMV-permissive macrophages. The reactivation of latent HCMV was dependent on the production of IFN-␥ early in the differentiation process. These studies provide further evidence for the importance of IFN-␥ in the pathogenesis of HCMV infection. MATERIALS AND METHODS Establishment of allogeneically stimulated PBMC cultures. PBMC were isolated from blood samples from 22 (16 donors for analyses of HCMV replication after in vitro infection and 6 donors for studies of reactivation of latent HCMV) healthy blood donors by density gradient centrifugation on Histopaque (Sigma Chemical Co.) as previously described (12). The PBMC were resuspended in Iscove’s complete medium containing penicillin (100 IU/ml), streptomycin (100 ␮g/ml; both from Gibco Laboratories), and 10% human AB serum (Sigma). Equal numbers of cells from two different blood donors (1.8 ⫻ 107 cells/ml) were mixed before plating on Primaria dishes (Becton Dickinson). After 48 h of culture at 37°C in 5% CO2, the majority of nonadherent cells were removed, and the cultures were replenished with complete 60/30 medium (60% AIM V and 30% Iscoves; Gibco) with 10% AB-negative human serum (from HCMV- and human immunodeficiency virus [HIV]-seronegative donors) (Sigma). The cultures were washed and fed with 50% spent medium–50% fresh medium every 3 to 4 days and kept in culture for up to 90 days poststimulation. Control cell cultures were established by stimulation of PBMC from individual donors with ConA as previously described (11). Day 1 poststimulation is defined as the day after the initial PBMC isolation and allogeneic or ConA stimulation. Induction of monocytes with allogeneically conditioned supernatants. PBMC from either HCMV-seropositive or -seronegative donors were isolated as described above. PBMC were resuspended in Iscove’s complete medium (2.0 ⫻ 107 cells/ml), and cells from a single donor were plated on 60-mm Primaria dishes. Cells were allowed to adhere for 2 h, after which the nonadherent cells were removed by extensive washing. Conditioned medium was added to the adherent monocytes from parallel Allo-MDM cultures at 1, 3, 5, and 7 days postisolation. Allo-conditioned medium was produced from two HCMV-seronegative donors. Cell-free conditioned medium was produced by pelleting cells at 3,000 rpm for 10 min, and the supernatants were passed through a 0.5-␮m syringe filter (Nalgene, Rochester, N.Y.). After day 7, the cultures were washed, fed fresh medium every 3 days, and kept in culture for 16 or 21 days postisolation. HCMV infection of MDM cultures. A recent patient isolate of HCMV was used to infect primary cultures of MDM. This isolate (PO) was obtained from a transplant patient with HCMV disease and passaged through human fibroblasts (HF). Cell-free virus stocks were prepared from supernatants of HF cultures, frozen, and stored until use at ⫺70°C. Virus used for infections was not passaged any more than 15 times in HF cells. MDM cultures were infected with virus obtained from supernatants of infected HF cells at a multiplicity of infection (MOI) of 10 at 7 to 10 days post-allogeneic stimulation or stimulation with ConA. For mock infection, cells were exposed to medium from uninfected HF cultures. The cultures were fed every third day and collected for viral titer assays and immunocytochemistry at different time points after infection. Negative selection of PBMC prior to allogeneic stimulation. In order to obtain CD4⫹ or CD8⫹ T-cell-depleted Allo-MDM cultures, the Mini MACS system (Miltenyi Biotec, Bergish Gladbach, Germany) was used for negative selection of the respective cell type. Freshly isolated PBMC were stained with monoclonal antibodies directed against CD4⫹ T cells (anti-human Leu-3a), CD8⫹ T cells (anti-human Leu-2a; both from Becton Dickinson), or isotype control serum

J. VIROL. (mouse immunoglobulin G1 [IgG1] Fc; R&D Systems, Minneapolis, Minn.). Cells (108 in 500 ␮l of serum-free Iscove’s medium) were incubated with a titered excess of the respective antibody at 4°C for 45 min. The cells were washed twice in cold phosphate-buffered saline (PBS) and resuspended in 250 ␮l of MACS buffer (PBS containing 5 mM EDTA and 0.5% bovine serum albumin) and incubated with 160 ␮l of MACS beads conjugated with rat anti-mouse IgG1 antibodies for 20 min at 4°C. Each MACS column was washed with 15 ml of MACS buffer before addition of the respective sample. PBMC coupled to MACS beads were eliminated from the samples by retention in the column in a magnetic field. Each column was washed with 4 ml of MACS buffer, and the collected cells were washed twice in serum-free medium and resuspended in complete 60/30 medium. Following depletion, the cells were allogeneically stimulated as described above. Small aliquots of each sample before and after negative selection were analyzed by flow cytometry to ensure satisfactory purity (⬍3% contamination) of each sample before the establishment of each Allo-MDM culture. Blocking of HLA class I and HLA class II molecules in Allo-MDM cultures. To block the interaction between T cells and monocytes, monoclonal antibodies directed against constant regions of HLA A, B, and C or HLA-DR (both from Immunotech, Westbrook, Maine) or isotype controls (mouse IgG2a or mouse IgG2b; both from R&D Systems) at a concentration of 35 ␮g/ml were incubated with 7 ⫻ 107 cells in Iscove’s complete medium for 1 h at 4°C before allogeneic stimulation. Thereafter, nonadherent cells and antibodies in the cultures were removed by three washes in serum-free medium, and the Allo-MDM cultures were cultured in complete 60/30 medium for up to 30 days. Quantitation of cytokines in Allo-MDM and ConA-MDM culture. The production of cytokines by allogeneically and mitogenically stimulated PBMC cultures were determined by a cytokine-specific enzyme-linked immunosorbent assay (ELISA). Allogeneic and ConA stimulation of PBMC was performed as described above. Briefly, PBMC (6 ⫻ 107 total) were mixed from histoincompatible donor pairs consisting of one seropositive donor and one seronegative donor. ConA-MDM cultures were established by adding 5 ␮g of ConA per ml to 6 ⫻ 107 PBMC from single donors. Cells were plated onto 60-mm dishes, and after 24 h nonadherent cells were removed by extensive washing. Cell culture supernatants were collected at 6, 12, 24, 36, and 48 h and at 3, 5, and 8 days poststimulation and analyzed by immunoassays for interleukin-1␤ (IL-1␤), IL-2, IL-3, IL-4, IL-6, IL-7, IL-8, IL-10, IL-12, IL-13, transforming growth factor beta (TGF-␤), TNF-␣, granulocyte colony-stimulating factor (G-CSF), macrophage CSF (M-CSF), GM-CSF, and IFN-␥ (R&D Systems). Neutralization of cytokines in Allo-MDM cultures. For neutralization experiments, polyclonal neutralizing goat antibodies against human TNF-␣, IL-1␣, IL-2, TGF-␤, GM-CSF, and IFN-␥ (all from R&D Systems) were used to block production of the respective lymphokine in Allo-MDM cultures. According to the manufacturer’s specifications, saturating concentrations of neutralizing antibodies were added to the cultures at the same time as allogeneic stimulation and were present in the cultures for 48 h poststimulation. Thereafter, nonadherent cells and antibodies in the cultures were removed by three washes in serum-free medium, and the Allo-MDM cultures were cultured in complete 60/30 medium for up to 30 days. Immunocytochemistry. HCMV-infected and mock-infected MDM or dendritic cell cultures grown in eight-well chamber slides or in Primaria 96-well plates were collected at different time points after infection. The cells were washed in PBS, fixed in phosphate-buffered 1% paraformaldehyde or methanolacetone (1:1) for 10 min at room temperature, and permeabilized with 0.3% Triton X-100 in PBS. Cells were blocked with 10% normal goat serum or 10% human AB serum in PBS for 30 min at room temperature and thereafter incubated with antibodies against different HCMV gene products (the major immediately-early [IE] protein [rabbit anti-MIE [45]) or gB (mouse anti-gB [UL55] (a kind gift from William Britt, University of Alabama, Birmingham [3])) in a 1:100 dilution for 1 to 6 h at room temperature. Cells were washed three times in PBS, and binding of the primary antibody was detected with a fluorescein isothiocyanate-tetramethyl (FITC)-conjugated goat anti-mouse or goat anti-rabbit Ig antibody for 1 to 2 h at room temperature. Double immunocytochemistry for cell surface markers was performed on live cells before fixation and staining for the HCMV IE antigen was performed. Stained cells were washed in PBS and mounted in Slowfade antifade kit (Molecular Probes Inc., Eugene, Oreg.) to ensure minimal fluorescence fading. Fluorescence-positive cells were visualized on an upright or inverted Leitz fluorescent microscope, and the number of infected cells was counted. Virus titer assays. At different days postinfection, supernatants from MDM and dendritic cell cultures were collected, and cells were harvested by scraping adherent cells into Dulbecco’s modified Eagle’s medium (DMEM) containing 2% fetal bovine serum (FBS), 2 mM L-glutamine, 100 IU of penicillin per ml and 100 ␮g of streptomycin per ml. Supernatants or sonicated MDM or dendritic

VOL. 75, 2001

DIFFERENTIATION-DEPENDENT REACTIVATION OF HCMV

cells were plated onto monolayers of subconfluent HF cells. After an initial 24 h of virus adherence at 37°C, cells were washed twice in medium and overlaid with DMEM containing 10% FBS, 2 mM L-glutamine, 100 IU of of penicillin per ml, 100 ␮g of streptomycin per ml, and 0.5% autoclaved SeaKem agarose (Sigma). The cultures were incubated for 14 days, with feeding every fourth day. The cells were fixed with 25% formaldehyde in PBS for 15 min and stained with a 0.05% solution of methylene blue, and plaques were counted (51). Detection of HCMV replication in Allo-MDM. Allo-MDM cultures were established by mixing PBMC from two healthy blood donors as described above. Samples were collected at days 1, 10, 17, 26, 36, 46, 54, 60, 70, 80, and 90 for detection of HCMV gene products by immunofluorescence or PCR and/or for virus recovery by plaquing on HF cells as previously described (42). Expression of HCMV proteins was detected in fixed cells (11). For PCR analysis, cell samples were collected at the indicated time points by scraping, and DNA and RNA were prepared with the Qiagen blood and cell culture DNA kit and RNeasy kit, respectively, according to the manufacturer’s instructions. HCMVspecific primer pairs were used in nested reverse transcription (RT)-PCRs detecting IE and pp150 RNA and with controls as previously described (41). PCR products were visualized by direct gel analysis on a 1% agarose gel. Flow cytometric analyses of PBMC. A fluorescence-activated cell analyzer (FACSCalibur; Becton Dickinson) was used for all analyses of cell surface expression on PBMC before and after negative selection using monoclonal antibodies directed against CD4 and CD8 (Becton Dickinson). Mean fluorescence values were obtained from histograms displaying the log fluorescence of FITC (FL1) of the samples which were generated against the background staining of cells stained with an isotype control antibody (mouse IgG1, IgG2a, or IgG2b) and the secondary antibody conjugated with FITC. Histograms displaying the log fluorescence of FITC (FL1) of the PBMC samples were generated before and after negative selection of CD4 and CD8. The percent positive cells was estimated by setting the level for positive cells not to include the background staining with a nonspecific isotype control antibody.

RESULTS Enhanced viral growth in Allo-MDM compared to ConAMDM. We have previously reported that HCMV can be reactivated from Allo-MDM but not macrophages derived from ConA stimulation of PBMC (ConA-MDM). The differential ability of HCMV to reactivate in Allo-MDM versus ConAMDM suggested that the method of monocyte stimulation greatly influenced the ability of HCMV to replicate in cell types derived from CD14⫹ monocytes. Therefore, we compared characteristics of viral replication in the two different macrophage subsets. To examine the kinetics of HCMV replication in Allo-MDM and ConA-MDM, in vitro infected cultures were monitored for production of infectious virus at a variety of time points. The kinetics of HCMV replication in Allo-MDM was rapid, and large quantities of virus were found in both the cellular and supernatant fractions (Fig. 1). These characteristics are similar to viral replication in fibroblasts, which are the prototypic cell for growing HCMV in vitro. In contrast, viral replication in the ConA-MDM was delayed and exhibited lower levels of viral production, and virus was only found associated with the cellular fraction. The importance of allogeneic stimulation for unrestricted HCMV replication was supported by the lack of viral replication in cultures derived by mixing PBMC from HLA-identical twins (Fig. 1). The frequency of HCMV-infected Allo-MDM was assessed by the detection of the IE as well as the early-late glycoprotein B (gB) antigen by immunofluorescence. In contrast to small numbers (⬍10%) of cells expressing viral antigens in the ConA-MDM cultures, greater than 50% of the Allo-MDM expressed both IE and gB antigens at 12 days postinfection (Fig. 1A). In addition, the kinetics of expression of the HCMV structural protein gB correlated with the rapid production of virus within Allo-MDM (Fig. 1B). These results indicate that the alloge-

7545

neically driven differentiation process, which results in the specific differentiation of a macrophage phenotype, ensures a more efficient replication of HCMV in comparison to mitogenically differentiated MDM. T-cell-produced cytokines IFN-␥ and IL-2 mediate the generation of HCMV-permissive Allo-MDM. To identify the cellular elements within the PBMC population which were important for the development of HCMV-permissive AlloMDM, CD4⫹ or CD8⫹ T cells were depleted from PBMC by a negative selection technique. For these experiments, the respective cell type was eliminated from the PBMC of each donor prior to the establishment of the Allo-MDM cultures. Flow cytometric analysis was performed on cells before and after negative selection to ensure that the residual cell phenotype was less than 3% (data not shown). Depletion of either CD4⫹ or CD8⫹ T cells from the PBMC before challenge with virus resulted in a 60 to 73% reduction in the number of Allo-MDM expressing IE proteins as well as a 1,000- to 10,000fold decrease in the production of virus (Fig. 2A and B). The addition of neutralizing antibodies directed against HLA class I or HLA class II to PBMC prior to the allogeneic reaction also resulted in an 80 to 90% reduction of IE-expressing cells as well as in a 1,000- to 10,000-fold decrease in virus production (Fig. 2C and D). These experiments indicate that the generation of HCMV-permissive Allo-MDM by an allogeneic reaction involves activation of both CD4⫹ and CD8⫹ T cells. Since cytokines produced by stimulated PBMC are crucial for the monocyte differentiation process, we examined potential differences in cytokines expressed during the development of ConA-MDM and Allo-MDM. Previously, we have demonstrated that IFN-␥ and TNF-␣ were critical cytokines necessary for the development of ConA-MDM (43). Therefore, we examined the expression and kinetics of these and other cytokines to compare their production in Allo-MDM and ConAMDM cultures. For these experiments, PBMC from six histocompatibly different donors either were not stimulated or were stimulated by either allogeneically matching different pairs within the group or stimulating cells from individuals mitogenically. Following adherence of activated PBMC, nonadherent cells were removed from the cultures at 24 h poststimulation, and supernatants were collected from the cultures at the indicated intervals. Interestingly, analysis of the supernatants for the cytokines IFN-␥, TNF-␣, IL-1, IL-2, IL-3, IL-4, IL-6, IL-7, IL-8, IL-10, IL-12, IL-13, GM-CSF, M-CSF, GCSF, and TGF-␤ indicated that Allo-MDM and ConA-MDM displayed differences not only in the kinetic appearance but also in the production of the proteins (Fig. 3). Whereas IFN-␥, TNF-␣, IL-2, IL-3, IL-10, M-CSF, G-CSF, and GM-CSF were produced in both the ConA-MDM and Allo-MDM cultures, the trend for production of these cytokines was delayed in the allogeneically stimulated cultures, with greater amounts of IFN-␥, GM-CSF, G-CSF, IL-2, and IL-3 produced by these cells in comparison to the mitogenically stimulated cells (Fig. 3A to H). IL-7 (Fig. 3J) and IL-8 (not shown) are also produced in both systems but with similar kinetics. Conversely, IL-4 and IL-12 (not shown) were undetectable in either ConAMDM or Allo-MDM cultures. Surprisingly, IL-1␤ and TGF-␤ were not expressed in allogeneically stimulated cells but were produced in significant amounts in the mitogenically stimulated cultures (Fig. 3K and L). IL-13, which can be used in

7546


J. VIROL.

FIG. 1. Expression of HCMV antigens IE and gB and virus growth in Allo-MDM and ConA-MDM. Parallel cultures of Allo-MDM and ConA-MDM were established as described in Materials and Methods, and the cultures (n ⫽ 4 with duplicates) were infected with HCMV. Cells were fixed at different time points after infection, and the expression of HCMV proteins was detected by double-label immunocytochemistry. HCMV IE proteins were observed earlier in Allo-MDM than in ConA-MDM. In addition, a sixfold increase in the number of HCMV-positive cells was detected in Allo-MDM compared to ConA-MDM (A). HCMV infection of fibroblasts, Allo-MDM, or PBMC from HLA-identical twins was performed at an MOI of 1 and with ConA-MDM at an MOI of 10 (B). Infectious virus was detected at 3 days postinfection in HF as well as Allo-MDM, whereas similar levels of infectious virus were detected in ConA-MDM first at 7 days postinfection. Virus was not produced in macrophage cultures, which were established by mixing PBMC from two HLA-identical twins. Similar amounts of HCMV were produced in the supernatants of infected Allo-MDM as well as HF cells, but not ConA-MDM. Bars represent the standard deviation of the mean.

combination with GM-CSF to differentiate monocytes into dendritic cells, was expressed in Allo-MDM cultures and not ConA-MDM (Fig. 3I). These observations highlight the differential expression of cytokines during the ConA-MDM and Allo-MDM differentiation process, which may be important in the genesis of these cell types. In order to determine the importance of these cytokines in mediating the formation of HCMV-permissive Allo-MDM, polyclonal antibodies with neutralizing activity for IL-1, IL-2, TNF-␣, TGF-␤, and IFN-␥ were added separately to AlloMDM cultures. Neutralization of IL-2 and IFN-␥ but not IL-1,

TNF-␣, or TGF-␤ within Allo-MDM cultures resulted in a 45 and 65% reduction, respectively, in the number of cells expressing the IE antigen, as well as in a 10,000-fold reduction in the production of infectious virus (Fig. 4). Interestingly, IFN-␥ is also important for the development of the HCMV-permissive ConA-MDM (42). While neutralization of TNF-␣ decreased virus titers in infected ConA-MDM, TNF-␣ did not demonstrate an effect on HCMV replication in Allo-MDM. IL-2 was not required for productive infection of ConA-MDM, although this cytokine was necessary for the development of HCMV-permissive Allo-MDM.

VOL. 75, 2001


7547

FIG. 2. Allogeneically induced differentiation of HCMV-permissive Allo-MDM is dependent on the presence of CD4⫹ and CD8⫹ T lymphocytes and HLA class I and II molecules. Allo-MDM cultures were established by negative selection of either CD4⫹ or CD8⫹ T lymphocytes from the respective donors before allogeneic stimulation (n ⫽ 6 with duplicates). The expression of the HCMV-encoded IE antigen in Allo-MDM was determined by immunofluorescence staining at day 7 postinfection. Depletion of CD4⫹ cells inhibited IE expression by 55 to 60%, while depletion of CD8⫹ cells reduced viral expression by 60 to 80% (A). Depletion of CD4⫹ and CD8⫹ T cells also inhibited the production of virus in Allo-MDM cultures by 1,000-fold and 10,000-fold, respectively (B). Blocking of HLA class I or HLA class II molecules with neutralizing antibody prior to allogeneic stimulation of PBMC also inhibited HCMV IE expression (C) and viral production (D), which correlated with the CD4⫹ and CD8⫹ depletion experiments.

Reactivation of latent HCMV in Allo-MDM is dependent on IFN-␥. While the above studies indicate that IFN-␥ and IL-2 are critical for HCMV in vitro infection of Allo-MDM, reactivation of virus in the Allo-MDM may require these cytokines or additional ones. To examine the cytokines necessary for the reactivation of latent HCMV, allogeneically stimulated cell cultures were established by mixing PBMC from histoincompatible donor pairs. Donors were tested for HCMV exposure by ELISA for serum antibodies and by PCR for the presence of HCMV DNA in PBMC to ensure the presence of latent HCMV genomes in at least one of the donors before stimulation. Polyclonal antibodies with neutralizing activity for IL-1,

IL-2, TGF-␤, TNF-␣, IFN-␥, or GM-CSF were added separately to Allo-MDM cultures. While HCMV IE proteins were detected at day 14 to 21 postinfection without the addition of neutralizing antibodies, the late HCMV gB antigen was detected in adherent cells between days 21 and 35 poststimulation in three of three allogeneically stimulated cultures (data not shown). At day 52 post-allogeneic stimulation, IE antigenpositive cells were quantified (Fig. 5). Neutralization of IFN-␥ but not IL-1, IL-2, TNF-␣, TGF-␤, or GM-CSF during the first 48 h poststimulation resulted in an 80 to 95% reduction in the number of HCMV-positive Allo-MDM (Fig. 5). These results indicate that reactivation of latent HCMV in Allo-MDM is

7548


J. VIROL.

VOL. 75, 2001


7549

FIG. 3. Allo-MDM and ConA-MDM display differences in kinetics and production of cytokines. To compare cytokine production in AlloMDM and ConA-MDM cultures, PBMC from six histoincompatible donors were stimulated by either allogeneically matching different pairs within the group or stimulating cells from individuals mitogenically. Following adherence of activated PBMC, nonadherent cells were removed from the cultures at 24 h poststimulation, and supernatants were collected from the cultures at the indicated intervals. Parallel cultures of Allo-MDM (red), ConA-MDM (blue), and unstimulated PBMC (green) were established as described in Materials and Methods. Cytokine-specific ELISA analysis was used to determine cytokine expression kinetics in cell-free culture supernatants collected at 6, 12, 24, 36, and 48 h and at 3, 5, and 8 days postisolation. Cytokines analyzed include IFN-␥ (A), TNF-␣ (B), IL-2 (C), IL-3 (D), GM-CSF (E), G-CSF (F), M-CSF (G), IL-10 (H), IL-13 (I), IL-7 (J), TGF-␤ (K), IL-1␤ (L), and IL-6 (M).

dependent on IFN-␥ production at early stages of monocyte differentiation. Reactivation of HCMV was not observed in ConA-MDM or control cultures obtained from the same individual donors. These results indicate the importance of a specific monocyte differentiation pathway which allows HCMV reactivation and unrestricted replication in macrophages.

Allo-MDM-conditioned medium induces reactivation of HCMV in monocytes. While the above experiments indicate that IFN-␥ is necessary for the production of virus in MDM, stimulation of cells with this cytokine alone was insufficient to reactivate virus. Therefore, to determine whether cytokines or other soluble factors produced by allogeneic stimulation of

7550


J. VIROL.

FIG. 4. Allogeneically induced differentiation of HCMV-permissive MDM is dependent on the presence of IL-2 and IFN-␥. HCMV infection was inhibited in Allo-MDM that were established by neutralization of IL-2 and IFN-␥, whereas an effect was not observed by neutralization of IL-1, TNF-␣, or TGF-␤ before establishment of the respective Allo-MDM cultures (n ⫽ 6 with duplicates). (A) Percent HCMV IE-expressing cells in the Allo-MDM cultures. (B) Production of cell-associated HCMV. All viral titer assays were performed on Allo-MDM collected by scraping at 14 days postinfection, and the expression of IE was determined in Allo-MDM fixed at 7 days postinfection.

PBMC were capable of generating Allo-MDM from monocytes, conditioned culture medium obtained from allogeneically stimulated cells was added sequentially over time to CD14⫹ cells. In these experiments, growth medium obtained from parallel Allo-MDM cultures generated from two HCMVseronegative donors at 1, 3, 5, and 7 days postisolation was added to CD14⫹ monocytes obtained from an HCMV-seropositive donor at the same intervals. Treatment of monocyte cultures with Allo-MDM-conditioned medium was sufficient to induce monocytes to differentiate into MDM with morphology similar to Allo-MDM and distinct from untreated cells (Fig. 6A, B, and C). Reactivation of HCMV in the monocytes treated with Allo-MDM-conditioned medium was also detected by the presence of the HCMV-specific proteins IE antigen and gB at 16 and 21 days poststimulation (Fig. 6D and E). These observations indicate that while IFN-␥ alone is insufficient to generate macrophages which reactivate HCMV, cytokines or other factors produced during allogeneic stimulation are capable of inducing the HCMV-permissive Allo-MDM phenotype. These results also indicate that Allo-MDM originate from CD14⫹ monocytes. DISCUSSION In this study, we describe two distinct monocyte-derived cell types which can be distinguished by their ability to reactivate and support HCMV replication. While HCMV infection of both ConA-MDM and Allo-MDM resulted in productive infection, viral infection of monocyte-derived dendritic cells (MDDC) was nonproductive and exhibited limited gene expression. Allo-MDM, in contrast to ConA-MDM, exhibited a

high frequency of infected cells as well as high titers of virus in both the cellular and supernatant fractions. More importantly, Allo-MDM but not ConA-MDM were capable of reactivating virus. These results indicate that the stimulus which initiates monocyte differentiation is a critical determinant in the generation of HCMV-permissive macrophages. Reactivation of latent HCMV in monocyte lineage cells is dependent on a specific macrophage differentiation stimulus. The ability of Allo-MDM but not ConA-MDM to reactivate HCMV would suggest that a specific monocyte-macrophage differentiation pathway was induced by the allogeneic reaction between T cells and monocytes. A model is presented in Fig. 7. Previous work from our group has shown that monocyte contact with ConA-stimulated CD8⫹ T cells and production of cytokines was required for generation of HCMV-permissive macrophages (42). In the present study, activation of both CD4⫹ and CD8⫹ T-cell populations was required for generation of HCMV-permissive Allo-MDM. These results indicate that the cytokines produced or the sequence of cytokine induction during the allogeneic activation of T cells may be different from those cytokines produced during mitogeneic stimulation. We have previously reported that production of TNF-␣ and IFN-␥ by CD8⫹ T cells was essential for HCMV replication in ConA-MDM (43). Although high levels of both IFN-␥ and TNF-␣ can be produced by both CD4⫹ and CD8⫹ allogeneically stimulated T cells (32, 51, 52), activated CD8⫹ T cells appeared to be a major producer of cytokines which were important for replication of HCMV in macrophages. In this study, IFN-␥ and IL-2 but not TNF-␣ produced by T cells upon allogeneic stimulation were identified as critical cytokines for the development of HCMV-permissive Allo-MDM. These

VOL. 75, 2001


FIG. 5. Reactivation of latent HCMV in Allo-MDM is dependent on production of IFN-␥. Allo-MDM were established by mixing PBMC from two healthy blood donors (n ⫽ 3 with duplicates), and reactivation of HCMV was detected by expression of the HCMV proteins IE and gB. (A) Confocal microscopy analysis of a representative sample of an Allo-MDM culture which reactivated HCMV. The cells expressed both HCMV IE (red) and gB (green). Magnification, ⫻630. (B) Reactivation of HCMV was inhibited in Allo-MDM, which were established in the presence of neutralizing antibodies to IFN-␥. In contrast, reactivation of virus was not affected by the addition of neutralizing antibodies directed against IL-1, IL-2, TNF-␣, TGF-␤, or GM-CSF at the time of establishment of the Allo-MDM cultures. (B) Percent HCMV IE-expressing cells in the Allo-MDM cultures at 52 days poststimulation.

findings suggests that, similar to the classic activation pathway of CD8⫹ T cells, a primary interaction occurs between CD4⫹ T cells and monocytes, which is followed by production of IL-2. IL-2 may serve two roles in the development of Allo-MDM cultures. First, IL-2 is a direct activator of monocytes, increasing their survival, migration, and cytokine secretion (TNF-␣, IL-1␤, GM-CSF, IL-6, and G-CSF) (10, 31, 33, 46), and the

7551

presence of the cytokines TNF-␣, GM-CSF, and G-CSF in our Allo-MDM cultures may be dependent this early IL-2 stimulation. The effects of IL-2 on monocyte activation are potentiated by IFN-␥ (found in both Allo-MDM and ConA-MDM) but blocked by TGF-␤ (present in the ConA-MDM only). Our findings suggest that the early presence of TGF-␤ in ConAMDM makes these cells unresponsive to the effects of IL-2 and drives them down a unique differentiation pathway, as evidenced by their functional differences from the Allo-MDM. Second, IL-2 activates CD8⫹ T cells, which results in the increased production of IFN-␥. Since IFN-␥ was required for HCMV replication in both Allo-MDM and ConA-MDM, the difference in the two systems may be explained by the presence of an IL-2-driven proliferation of T cells in Allo-MDM cultures and/or the production of other cytokine profiles, as shown in Fig. 3. Cytokines are critical mediators of macrophage differentiation, and the initial exposure to different cytokines determines their final phenotype (15, 35). Previously, we reported that the Allo-MDM cells, which are capable of HCMV reactivation, expressed both dendritic cell (CD1a and CD83) and macrophage (CD14 and CD68) phenotypic markers. GM-CSF and IL-13 are commonly used to generate dendritic cells from CD14⫹ monocytes. Our current findings that Allo-MDM cultures produce substantial levels of GM-CSF and IL-13 may explain the acquisition of dendritic cell characteristics. The presence of other cytokines such as M-CSF, which are critical for the maintenance of monocyte-macrophages, may explain the presence of macrophage markers (CD14 and CD68) (35). In addition to the effects of IL-13 on macrophage differentiation, this cytokine has been shown to enhance HCMV replication in fibroblasts, which suggests that this cytokine may be important in HCMV reactivation from latency (15). In our preliminary experiments, we have found that the kinetics of IL-13 expression in the Allo-MDM cultures corresponds to the earliest time that HCMV protein expression (IE and gB) can be detected in these cells (days 3 to 5; data not shown). Further experiments must be done to elucidate the role of this and other cytokines in the development of MDM capable of CMV reactivation and replication. Role of IFN-␥ in macrophage differentiation and viral replication. The observation that IFN-␥ is important for the generation of HCMV-permissive Allo-MDM contrasts with the reported antiviral effects of these cytokines in vivo. Probably the most important antiviral contribution of IFN-␥ is immunologic control of viral infection. For example, IFN-␥ was observed to restore CMV antigen presentation in infected animals, which implies a role for IFN-␥ in T-cell-mediated control of the virus (17, 18). In addition, IFN-␥ is also important for induction of NK cell-mediated cytotoxicity (reviewed in reference 4) and has been shown to mediate clearance of CMV from the salivary glands of infected animals (26, 27, 34, 36, 47). These findings may explain why IFN-␥-depleted mice demonstrate increased CMV titers in infected organs (16). Thus, IFN-␥ appears to play a dual role in the life cycle of HCMV. At one level the cytokine is crucial for the differentiation process necessary to reactivate virus in Allo-MDM. In contrast, production of IFN-␥ is critical for induction of the cellular immune response, which controls viral infection. Several groups have shown the antiviral effects of IFN-␥ (26, 27, 34, 36,

7552


J. VIROL.

FIG. 6. Reactivation of HCMV in monocytes stimulated with Allo-MDM-conditioned medium. Monocytes from single donors were isolated by adherence to plastic dishes for 2 h, after which nonadherent cells were removed by extensive washing. Supernatants from the enriched monocyte cultures were replaced at 1, 3, 5, and 7 days postisolation with Allo-MDM-conditioned medium obtained from allogeneic Allo-MDM cultures at parallel time points. (A) Allo-MDM generated from allogeneic stimulation of PBMC. Enriched monocyte cultures treated with canditioned medium are shown at 7 days poststimulation (B) in comparison to untreated adhered monocytes at the same time (C). HCMV reactivation in conditioned-medium-stimulated monocytes is demonstrated in panel D by immunofluorescence at 16 days poststimulation, with HCMV gB in green and IE in red. A graphical representation of reactivation in conditioned-medium-treated CD14⫹ monocytes derived from HCMVseropositive donors is shown in panel E. Reactivation was determined by positive fluorescent staining for the HMCV proteins IE and gB.

47); however, at the transcriptional level, this block can be overcome by treating cells with TNF-␣ (38, 44). Interestingly, the Allo-MDM cultures contain substantial levels of TNF-␣, which may activate the CMV IE promoter even in the presence of IFN-␥, which is necessary for Allo-MDM differentiation. This study is the first clear demonstration that IFN-␥ is crucial for the generation of HCMV-permissive macrophages which are capable of reactivating virus. Reactivation in these cells may depend on T-cell activation and production of cytokines leading to a specific macrophage phenotype capable of reactivating latent virus. Since immunosuppressed individuals often suffer from numerous of other infections, the virus may be frequently activated by T-cellproduced cytokines during an immune response against pathogens. If the immune system is not able to achieve control over

the virus, disease rather than persistence develops in these patients. This scenario may explain why reactivation of HCMV is common in transplant patients following bacterial infections (53) and in HIV-infected individuals, who often experience opportunistic infections. Furthermore, HCMV is often transmitted during blood transfusion (2) and is associated with the development of acute graft-versus-host disease in bone marrow transplant patients and acute rejection in organ transplant patients (1, 22, 24). All of these situations involve allogeneic immune processes, which may be the primary cause of viral transmission and disease in these patients. The above observations suggest that the ability of reactivated HCMV to establish a viremic state in the host is dependent on the kinetics of viral replication, ability to block IFN-␥ signaling, and subsequent immune response induced by IFN-␥ production. This

VOL. 75, 2001


7553

C.S.-N. is a scholar of the Wenner-Gren Foundation, Sweden. This work was supported by grants from the Public Health Service, the National Institutes of Health (AI 21640 to J.A.N.), Activated Cell Systems, LLC (J.A.N.), and the Swedish Medical Research Council (K98-06X-12615-01, C.S.-N.). D.N.S. is supported by a National Research Service Award. REFERENCES

FIG. 7. Model for cellular and cytokine factors necessary for generation of ConA-MDM and Allo-MDM. This model describes the cellular and cytokine components involved in the generation of ConAMDM and Allo-MDM. While CD8 T-cell contact and IFN-␥ and TNF-␣ are critical in the formation of ConA-MDM, both CD8 and CD4 T-cell contact and IFN-␥ and IL-2 are important in the generation of Allo-MDM, which can reactivate and replicate latent HCMV.

race to produce virus and an immune response would suggest that low-level virus production in macrophages might be controlled by IFN-␥-induced responses, whereas higher levels of virus would be more difficult to regulate in light of the fact that the cytokine induces more cells to produce HCMV. In support of this hypothesis, a recent study demonstrated that IFN-␥ reversibly inhibited reactivation of latent murine CMV (37). This effect was partly explained by an inhibition of low levels of murine CMV replication in infected tissues. Thus, IFN-␥ plays an important role in the regulation of virus replication, which could explain why immunosuppressed individuals often suffer from severe HCMV infections. In summary, our observations provide the first evidence that a specific monocyte activation pathway is crucial for HCMV replication and reactivation of latent HCMV. We demonstrate the importance of an immunological activation of T cells and the production of IFN-␥ for successful reactivation and replication of latent HCMV in monocyte-derived macrophages. This experimental cell system provides an important tool to examine HCMV immune surveillance and can be used to explore new therapeutic approaches to prevent viral reactivation of HCMV. ACKNOWLEDGMENTS We thank Ashlee Moses for helpful discussion and Andrew Townsend for graphic work.

1. Bostrom, L., O. Ringden, N. Jacobsen, F. Zwaan, and B. Nilsson. 1990. A European multicenter study of chronic graft-versus-host disease: the role of cytomegalovirus serology in recipients and donors-acute graft-versus-host disease, and splenectomy. Transplantation 49:1100–1105. 2. Bowden, R. A. 1995. Transfusion-transmitted cytomegalovirus infection. Hematol. Oncol. Clin. North Am. 9:155–166. 3. Britt, W. J., and L. G. Vugler. 1992. Oligomerization of the human cytomegalovirus major envelope glycoprotein complex gB (gp55–116). J. Virol. 66: 6747–6754. 4. Burns, G., C. Begley, I. Mackay, R. Triglia, and J. Werkmeister. 1985. Supernatural killer cells. Immunol. Today 6:370–373. 5. Chou, S., D. Y. Kim, and D. J. Norman. 1987. Transmission of cytomegalovirus by pretransplant leukocyte transfusions in renal transplant candidates. J. Infect. Dis. 155:565–567. 6. Chou, S. W. 1986. Acquisition of donor strains of cytomegalovirus by renaltransplant recipients. N. Engl. J. Med. 314:1418–1423. 7. Chou, S. W. 1987. Cytomegalovirus infection and reinfection transmitted by heart transplantation. J. Infect. Dis. 155:1054–1056. 8. Chou, S. W. 1989. Reactivation and recombination of multiple cytomegalovirus strains from individual organ donors. J. Infect. Dis. 160:11–15. 9. Dankner, W. M., J. A. McCutchan, D. D. Richman, K. Hirata, and S. A. Spector. 1990. Localization of human cytomegalovirus in peripheral blood leukocytes by in situ hybridization. J. Infect. Dis. 161:31–36. 10. Epling-Burnette, P. K., S. Wei, D. K. Blanchard, E. Spranzi, and J. Y. Djeu. 1993. Coinduction of granulocyte-macrophage colony-stimulating factor release and lymphokine-activated killer cell susceptibility in monocytes by interleukin-2 via interleukin-2 receptor beta. Blood 81:3130–3137. 11. Fish, K. N., W. Britt, and J. A. Nelson. 1996. A novel mechanism for persistence of human cytomegalovirus in macrophages. J. Virol. 70:1855– 1862. 12. Fish, K. N., A. S. Depto, A. V. Moses, W. Britt, and J. A. Nelson. 1995. Growth kinetics of human cytomegalovirus are altered in monocyte-derived macrophages. J. Virol. 69:3737–3743. 13. Fish, K. N., S. Stenglein, C. Ibanez, and J. A. Nelson. 1995. Cytomegalovirus persistence in macrophages and endothelial cells. Scand. J. Infect. Dis. Suppl. 99:34–40. 14. Gnann, J., Jr., J. Ahlmen, C. Svalander, L. Olding, M. Oldstone, and J. Nelson. 1988. Inflammatory cells in transplanted kidneys are infected by human cytomegalovirus. Am. J. Pathol. 132:239–248. 15. Hatch, W., A. R. Freedman, D. M. Boldt-Houle, J. E. Groopman, and E. F. Terwilliger. 1997. Differential effects of interleukin-13 on cytomegalovirus and human immunodeficiency virus infection in human alveolar macrophages. Blood 89:3443–3450. 16. Heise, M. T., and H. W. Virgin. 1995. The T-cell-independent role of gamma interferon and tumor necrosis factor alpha in macrophage activation during murine cytomegalovirus and herpes simplex virus infections. J. Virol. 69: 904–909. 17. Hengel, H., C. Esslinger, J. Pool, E. Goulmy, and U. H. Koszinowski. 1995. Cytokines restore MHC class I complex formation and control antigen presentation in human cytomegalovirus-infected cells. J. Gen. Virol. 76:2987– 2997. 18. Hengel, H., P. Lucin, S. Jonjic, T. Ruppert, and U. H. Koszinowski. 1994. Restoration of cytomegalovirus antigen presentation by gamma interferon combats viral escape. J. Virol. 68:289–297. 19. Ibanez, C., R. Schrier, P. Ghazal, C. Wiley, and J. A. Nelson. 1991. Human cytomegalovirus productively infects primary differentiated macrophages. J. Virol. 65:6581–6588. 20. Kondo, K., H. Kaneshima, and E. S. Mocarski. 1994. Human cytomegalovirus latent infection of granulocyte-macrophage progenitors. Proc. Natl. Acad. Sci. USA 91:11879–11883. 21. Kondo, K., J. Xu, and E. S. Mocarski. 1996. Human cytomegalovirus latent gene expression in granulocyte-macrophage progenitors in culture and in seropositive individuals. Proc. Natl. Acad. Sci. USA 93:11137–11142. 22. Koskinen, P. K. 1993. The association of the induction of vascular cell adhesion molecule-1 with cytomegalovirus antigenemia in human heart allografts. Transplantation 56:1103–1108. 23. Lathey, J. L., and S. A. Spector. 1991. Unrestricted replication of human cytomegalovirus in hydrocortisone- treated macrophages. J. Virol. 65:6371– 6375. 24. Lautenschlager, I., K. Hockerstedt, E. Taskinen, and E. von Willebrand. 1996. Expression of adhesion molecules and their ligands in liver allografts during cytomegalovirus (CMV) infection and acute rejection. Transpl. Int. 9:S213–S215.

7554


25. Link, H., K. Battmer, and C. Stumme. 1993. Cytomegalovirus infection in leucocytes after bone marrow transplantation demonstrated by mRNA in situ hybridization. Br. J. Haematol. 85:573–577. 26. Lucin, P., S. Jonjic, M. Messerle, B. Polic, H. Hengel, and U. H. Koszinowski. 1994. Late phase inhibition of murine cytomegalovirus replication by synergistic action of interferon-gamma and tumour necrosis factor. J. Gen. Virol. 75:101–110. 27. Lucin, P., I. Pavic, B. Polic, S. Jonjic, and U. H. Koszinowski. 1992. Gamma interferon-dependent clearance of cytomegalovirus infection in salivary glands. J. Virol. 66:1977–1984. 28. Maciejewski, J. P., E. E. Bruening, R. E. Donahue, S. E. Sellers, C. Carter, N. S. Young, and S. St Jeor. 1993. Infection of mononucleated phagocytes with human cytomegalovirus. Virology 195:327–336. 29. Meyers, J. D., N. Flournoy, and E. D. Thomas. 1986. Risk factors for cytomegalovirus infection after human marrow transplantation. J. Infect. Dis. 153:478–488. 30. Minton, E. J., C. Tysoe, J. H. Sinclair, and J. G. Sissons. 1994. Human cytomegalovirus infection of the monocyte/macrophage lineage in bone marrow. J. Virol. 68:4017–4021. 31. Misago, M., J. Tsukada, R. Ogawa, M. Kikuchi, T. Hanamura, S. Chiba, S. Oda, I. Morimoto, and S. Eto. 1993. Enhancing effects of IL-2 on M-CSF production by human peripheral blood monocytes. Int. J. Hematol. 58:43–51. 32. Molteni, M., S. Della Bella, B. Mascagni, C. Coppola, V. De Micheli, C. Zulian, S. Birindelli, M. Vanoli, and R. Scorza. 1994. Secretion of cytokines upon allogeneic stimulation: effect of aging. J. Biol. Regul. Homeost. Agents 8:41–47. 33. Musso, T., I. Espinoza-Delgado, K. Pulkki, G. L. Gusella, D. L. Longo, and L. Varesio. 1992. IL-2 induces IL-6 production in human monocytes. J. Immunol. 148:795–800. 34. Orange, J., B. Wang, C. Terhorst, and C. A. Biron. 1995. Requirement for natural killer cell-produced interferon gamma in defense against murine cytomegalovirus infection and enhancement of this defense pathway by interleukin 12 administration. J. Exp. Med. 182:1045–1056. 35. Palucka, K. A., N. Taquet, F. Sanchez-Chapuis, and J. C. Gluckman. 1998. Dendritic cells as the terminal stage of monocyte differentiation. J. Immunol. 160:4587–4595. 36. Pavic, I., B. Polic, I. Crnkovic, P. Lucin, S. Jonjic, and U. H. Koszinowski. 1993. Participation of endogenous tumour necrosis factor alpha in host resistance to cytomegalovirus infection. J. Gen. Virol. 74:2215–2223. 37. Presti, R. M., J. L. Pollock, A. J. Dal Canto, A. K. O’Guin, and H. W. Virgin IV. 1998. Interferon gamma regulates acute and latent murine cytomegalovirus infection and chronic disease of the great vessels. J. Exp. Med. 188: 577–588. 38. Ritter, T., C. Brandt, S. Prosch, A. Vergopoulos, K. Vogt, J. Kolls, and H. D. Volk. 2000. Stimulatory and inhibitory action of cytokines on the regulation of hCMV-IE promoter activity in human endothelial cells. Cytokine 12: 1163–1170. 39. Sinclair, J. H., J. Baillie, L. A. Bryant, J. A. Taylor-Wiedeman, and J. G. Sissons. 1992. Repression of human cytomegalovirus major immediate early gene expression in a monocytic cell line. J. Gen. Virol. 73:433–435. 40. Sinzger, C., H. Muntefering, T. Loning, H. Stoss, B. Plachter, and G. Jahn. 1993. Cell types infected in human cytomegalovirus placentitis identified by

J. VIROL.

41. 42. 43.

44.

45. 46.

47. 48. 49. 50. 51.

52.

53.

54.

55.

immunohistochemical double staining. Virchows Arch. A Pathol. Anat. Histopathol. 423:249–256. Soderberg, C., S. Larsson, S. Bergstedt-Lindqvist, and E. Moller. 1993. Definition of a subset of human peripheral blood mononuclear cells that are permissive to human cytomegalovirus infection. J. Virol. 67:3166–3175. So ¨derberg-Naucle´r, C., K. Fish, and J. A. Nelson. 1997. Reactivation of human cytomegalovirus in a novel dendritic cell phenotype from healthy donors. Cell 91:119–126. So ¨derberg-Naucle´r, C., K. N. Fish, and J. A. Nelson. 1997. IFN-␥ and TNF-␣ specifically induce the formation of cytomegalovirus permissive monocytederived macrophages which are refractory to the antiviral activity of these cytokines. J. Clin. Investig. 100:3154–3163. Stein, J., H. D. Volk, C. Liebenthal, D. H. Kruger, and S. Prosch. 1993. Tumour necrosis factor alpha stimulates the activity of the human cytomegalovirus major immediate early enhancer/promoter in immature monocytic cells. J. Gen. Virol. 74:2333–2338. Stenberg, R., A. S. Depto, J. Fortney, and J. A. Nelson. 1989. Regulated expression of early and late RNAs and proteins from the human cytomegalovirus immediate-early gene region. J. Virol. 63:2699–2708. Strieter, R. M., D. G. Remick, J. P. d. Lynch, R. N. Spengler, and S. L. Kunkel. 1989. Interleukin-2-induced tumor necrosis factor-alpha (TNF-alpha) gene expression in human alveolar macrophages and blood monocytes. Am. Rev. Respir. Dis. 139:335–342. Tay, C. H., and R. M. Welsh. 1997. Distinct organ-dependent mechanisms for the control of murine cytomegalovirus infection by natural killer cells. J. Virol. 71:267–275. Taylor-Wiedeman, J., J. G. Sissons, L. K. Borysiewicz, and J. H. Sinclair. 1991. Monocytes are a major site of persistence of human cytomegalovirus in peripheral blood mononuclear cells. J. Gen. Virol. 72:2059–2064. Taylor-Wiedeman, J., P. Sissons, and J. Sinclair. 1994. Induction of endogenous human cytomegalovirus gene expression after differentiation of monocytes from healthy carriers. J. Virol. 68:1597–1604. Tegtmeier, G. E. 1986. Cytomegalovirus infection as a complication of blood transfusion. Semin. Liver Dis. 6:82–95. Toungouz, M., C. Denys, D. de Groote, M. Andrien, and E. Dupont. 1996. Optimal control of interferon-gamma and tumor necrosis factor-alpha by interleukin-10 produced in response to one HLA-DR mismatch during the primary mixed lymphocyte reaction. Transplantation 61:497–502. Toungouz, M., C. Denys, and E. Dupont. 1996. Blockade of proliferation and tumor necrosis factor-alpha production occurring during mixed lymphocyte reaction by interferon-gamma-specific natural antibodies contained in intravenous immunoglobulins. Transplantation 62:1292–1296. van den Berg, A. P., I. J. Klompmaker, E. B. Haagsma, P. M. Peeters, L. Meerman, R. Verwer, T. H. The, and M. J. Slooff. 1996. Evidence for an increased rate of bacterial infections in liver transplant patients with cytomegalovirus infection. Clin. Transplant. 10:224–231. von Laer, D., A. Serr, U. Meyer-Konig, G. Kirste, F. T. Hufert, and O. Haller. 1995. Human cytomegalovirus immediate early and late transcripts are expressed in all major leukocyte populations in vivo. J. Infect. Dis. 172:365– 370. Weinshenker, B. G., S. Wilton, and G. P. Rice. 1988. Phorbol ester-induced differentiation permits productive human cytomegalovirus infection in a monocytic cell line. J. Immunol. 140:1625–1631.