HSV efficiently infects dendritic cells (DCs) in their immature state and induces down-regulation of costimulatory and adhesion molecules. As in mice, HSV infection of human DCs also leads to their rapid and progressive apoptosis, and we show that both early and late viral proteins contribute to its induction. Because topical HSV infection is confined to the epidermis, Langerhans cells are expected to be the major APCs in draining lymph nodes. However, recent observations in murine models show T cell activation to be mediated by nonepidermal DC subsets, suggesting cross-presentation of viral Ag. In this study we provide an explanation for this phenomenon, demonstrating that HSV-infected apoptotic DCs are readily phagocytosed by uninfected bystander DCs, which, in turn, stimulate virus-specific CD8+ T cell clones.
The outcome of primary or recurrent HSV infection of skin or mucosae is thought to depend on the complex interaction between virus immunoevasive mechanisms and the host immune system countermeasures. In an immunocompetent host, recurrent herpes immune control and clearance is mediated mainly by infiltrating CD4+ and CD8+ T cell effectors and their cytokines (1, 2, 3). CD4+ T cells are the initial infiltrating effectors in the recurrent lesion, followed by CD8+ T cells, which correlates with viral clearance from the skin (1, 4). However, HSV has developed several immune evasion mechanisms, including inhibition of complement and Ab binding, dendritic cell (DC)3 function, and MHC class I-mediated Ag presentation on infected cells. The latter occurs through binding of an immediate-early HSV gene product, ICP47, via its N terminus to the TAP (5, 6, 7, 8, 9). This ICP47-TAP complex blocks translocation of the MHC class I-processed peptide complex to the cell surface, thereby preventing CD8+ T cell recognition. However, the ICP47 effect is cell type dependent and can be reversed. IFN-γ secreted by CD4+ cells in the lesion partially reverses the MHC class I down-regulation and stimulates MHC class II expression, allowing targeting of HSV-infected epidermal cells by both CD4+ and CD8+ T cells (1, 10, 11). The persistence of HSV-specific CD8+ T cell clones in seropositive individuals, however, suggests their complementary role to CD4+ T cells in the control of recurrences (12).
To ensure early clearance of HSV in the periphery, prompt activation of T cell effectors by APCs, such as DCs, is required. In skin or mucosae, Langerhans cells or dermal/submucosal DCs acquire and process foreign Ags, leading to DC activation and migration to the draining lymph nodes where they present peptide Ags to CD4+ and CD8+ T cells (13). DC maturation is required for optimal activation of naive T cells and entails a set of phenotypic changes that can be induced directly by the pathogen via pattern recognition receptors or indirectly through danger signals, such as IL-1β or TNF-α that are secreted by the surrounding cells (14). These complex changes include responsiveness to chemokines, such as CCL19 and CCL21, that guide them into the lymphoid organs as well as up-regulation of surface expression of molecules that are involved in T cell activation, including MHC classes I and II, costimulatory and adhesion molecules (13).
However, HSV greatly impairs the function of initially infected DCs. DCs generated from PBMC express HSV entry receptors, HVEM (Hve-A) and nectin-2 (Hve-B), and are highly susceptible to HSV infection (15, 16). HSV-1 infection of immature monocyte-derived DCs results in asynchronous down-regulation of CD40, CD54 (ICAM-1), CD80, and CD86, but not of MHC classes I and II (16). We have recently reported that HSV-2 induces rapid cell death via apoptosis in murine bone marrow-derived DCs (17). In this study we extend these findings and show that both HSV-1 and -2 also induce apoptosis in human monocyte-derived DCs in a replication-dependent manner. Given that in natural HSV infection, cellular immune responses are strongly activated, our in vitro findings raise an important question: if HSV-infected DCs are being rapidly cleared by programmed cell death, how are HSV-specific T cells being activated? Recently, two studies (18, 19) in murine models of vaginal and cutaneous infection showed that although HSV infection was confined to the epidermis, Langerhans cells, the predominant epidermal DC population, were not responsible for CD4+ or CD8+ T cell stimulation in draining lymph nodes. Instead, CD4+ and CD8+ T cells were found to be stimulated by CD11b+ submucosal and CD8α+ DCs, respectively. Nevertheless, Langerhans cells appear to play a key role in HSV infection, because their depletion from mouse skin exacerbates HSV infection in the periphery (20). Together, these findings hint at cooperation between different DC subsets and suggest that cross-presentation of viral Ags from infected epidermal DCs via other DC subsets may be the mechanism responsible for the generation of a cellular antiviral immune response in HSV infection.
In our in vitro study of human DCs, we demonstrate that HSV infection of human monocyte-derived DCs results in apoptosis and subsequent phagocytosis by bystander DCs, which can stimulate human CD8+ T cells by cross-presentation, thereby providing a mechanism for such a two-DC model of Ag presentation.
Materials and Methods
Complete medium consisting of RPMI 1640 with 2.05 mM l-glutamine (Invitrogen Life Technologies) and 10% heat-inactivated FCS (CSL) supplemented with 200 U/ml GM-CSF and 250 U/ml IL-4 (Schering Plough) was used for culture and maintenance of DCs.
Generation of monocyte-derived DCs
PBMC from HSV-seropositive and -seronegative male and female donors enriched by countercurrent elutriation or selection using magnetic beads conjugated to anti-CD14 Ab (Miltenyi Biotec) were cultured over 5–7 days in complete culture medium, as previously described (16). These culture conditions result in a highly pure CD1a+CD11c+ DC population with no detectable CD14 or CD83 expression, indicating that the monocytes had been converted to immature DCs.
HSV-1 and HSV-2 stocks were prepared by infecting 80–90% confluent monolayers of African green monkey kidney cells (Vero). Infected cells were incubated in DMEM supplemented with 5% FCS and were harvested after 2 days and briefly sonicated using a cup horn sonicator (Branson) to release cell-bound virus. After centrifugation at 20,000 × g for 10 min to pellet cell debris, clarified supernatant was subjected to ultracentrifugation for 2 h at 24,000 rpm using a Beckman Coulter SW28 rotor to pellet the virus. Pellets were resuspended in complete medium and stored at −80°C until use. Virus titers were determined by plaque assay on Vero monolayers. Briefly, 10-fold dilutions of virus were adsorbed onto Vero cells for 1 h at 37°C. The inoculum was then removed and fresh medium containing 0.4% agarose was added. Cells were fixed and stained with a 0.1% crystal violet solution after 2 days to determine the number of plaques.
UV light inactivation of virus
HSV was inactivated by placing a 100-mm petri dish, containing 2 ml of virus suspension, ∼10 cm distant from a 30-W UV light source (UV-HSV-1) (21). The dish was kept on ice at all times and mixed by pipetting every 2–3 min. Inactivated virus was stored at −80°C until use.
Infection of DCs with HSV
Phosphonoacetic acid (PAA) treatment
PAA was dissolved in PBS and adjusted to pH 7.4 with NaOH, then made up to final concentration of 10 mg/ml. DCs were treated with and maintained in 400 μg/ml PAA throughout the course of infection, commencing at the virus adsorption stage. At this concentration, PAA completely blocks viral DNA synthesis in infected cells (2, 23).
Annexin V binding assay
Uninfected and HSV-infected DCs were collected at 6, 12, and 24 h postinfection (p.i.) and analyzed for the early signs of apoptosis using an annexin V and propidium iodide (PI) staining protocol (BD Pharmingen). Briefly, cells were washed twice with PBS and resuspended in incubation buffer containing 10 mM HEPES (pH 7.4), 140 mM NaCl, and 2.5 mM CaCl2. Aliquots containing 5 × 105 cells/tube were incubated with FITC-conjugated annexin V and 0.1 μg of PI. After a 15-min incubation in the dark at room temperature, 10,000 events/sample were acquired using a FACSCalibur flow cytometer (BD Biosicences) and were later analyzed with CellQuest software (BD Biosicences). Cells showing red fluorescence due to PI uptake had lost their membrane integrity and were considered necrotic or late apoptotic. Only those cells negative for PI that also stained for FITC-annexin V were considered apoptotic. Cells treated with 3% formaldehyde for 30 min on ice were used as a positive control (not shown). For concurrent labeling of viral Ags, cells were incubated with an anti-glycoprotein C-FITC mAb (Trinity Biotech), annexin V-Alexa Fluor 647 (Molecular Probes), and PI for 15 min at room temperature, washed twice in incubation buffer, and analyzed as described above.
Colorimetric detection of caspase-3 activity
Uninfected or HSV-infected DCs (2 × 106 cells/treatment) were collected at selected time points and washed with PBS by centrifugation at 300 × g for 10 min. Supernatant was removed, and dry cell pellets were stored at −80°C until use. Caspase-3 activity was determined using a commercially available colorimetric assay (R&D Systems) that relies on the cleavage of a fluorochrome-conjugated caspase-3-specific peptide by active caspase. The level of caspase-3 activity in cell lysate is directly proportional to the color reaction. The results were expressed as the fold increase in caspase-3 activity of HSV-infected cells over that of uninfected cells.
Intracellular caspase detection
Uninfected, UV-HSV-1-infected, and HSV-1-infected DCs, seeded at 1 × 106 cells/ml, were incubated with 500 ng/ml of the fluorescein-conjugated pan-caspase inhibitor, benzyloxycarbonyl-Val-Asp-fluormethylketone (z-VD-fmk; R&D Systems), for 30 min at 37°C at specific time points after infection. Cells were then washed in PBS to remove unbound inhibitor, fixed in 2% paraformaldehyde, and analyzed by flow cytometry.
Transmission electron microscopy
Uninfected and HSV-1-infected DCs at 12 h p.i. and DCs treated with 1 M d-sorbitol (1 × 106 cells for each treatment) were collected and prepared as previously described (24). Briefly, cells were fixed in modified Karnovsky’s fixative for 1 h, washed twice in 0.1 M MOPS buffer, postfixed in 2% buffered osmium tetroxide for 1 h, followed by 2% aqueous uranyl acetate (Fluka) for 1 h, dehydrated through graded ethanols, and embedded in Spurr resin (TAAB Laboratories). Polymerization occurred at 70°C for 10 h.
Ultrathin sections were cut using a Reichert-Jung Ultracut E microtome, collected on 400-mesh, thin-bar copper grids (TAAB Laboratories), and stained with 1% uranyl acetate in 50% ethanol and Reynold’s lead citrate. The sections were examined with a Philips BioTWIN transmission electron microscope at 80 kV.
To induce apoptosis, DCs were incubated with 1 M d-sorbitol (Sigma-Aldrich) in RPMI 1640 for 1 h at 37°C. Cells were then pelleted and incubated in complete medium for an extra hour at 37°C before being assessed for apoptosis.
CFSE labeling of DCs
DCs were washed in RPMI 1640 and resuspended at 2 × 106 cells/ml. Cells were then added to an equal volume of 3 μM CFDA, SE (Molecular Probes) and incubated for 8 min at room temperature. CFDA, SE is converted by intracellular esterases into amine-reactive fluorescent CFSE. An equal volume of prewarmed FCS was then added, and cells were incubated for 10 min at 37°C to stop the labeling reaction and allow for dye efflux. Cells were centrifuged at 400 × g for 5 min, and supernatant was removed. Cells were resuspended in complete medium and washed three times at 400 × g for 5 min each time.
PKH26 labeling of DCs
DCs were labeled with PKH26 (Sigma-Aldrich), a fluorescent lipid probe, according to the manufacturer’s instructions. Briefly, DCs were washed in RPMI 1640 and resuspended in Diluent C (Sigma-Aldrich) at 2 × 106 cells/ml. Cells were added to an equal volume of 4 μM PKH26 in Diluent C and incubated for 5 min at room temperature. An equal volume of FCS was added for 1 min. Cells were then resuspended in complete medium and washed four times at 400 × g for 5 min each time.
CFSE-labeled DCs were infected with HSV for 18 h. DCs were cocultured with uninfected PKH26-labeled DCs for an additional 4–24 h to allow uptake of apoptotic cells. In some experiments HSV-infected DCs were positively selected for apoptotic cells using immunomagnetic sorting with annexin V microbeads (Miltenyi Biotec). At the end of the culture period, cells were washed and analyzed for double-positive cells (CFSE+PKH26+) using flow cytometry to indicate the uptake of HSV-infected DCs by bystander DCs.
Cross-presentation of HSV Ag
HSV-2-infected DCs from an HLA-A2− donor were cocultured with uninfected HLA-A2.1+ DCs at a ratio of 1:1 for 24 h to allow for uptake. Cells were then extensively washed and cocultured with an HLA-A2.1-restricted CD8+ T cell clone specific for HSV-2 UL47289–298 peptide (25) in a 24-h IFN-γ ELISPOT assay at a DC:T cell ratio of 1:10. HLA-A2 presentation was blocked by incubating HLA-A2+ DCs with anti-HLA-A2 mAb CR11-351 (1/100 dilution) for 30 min before coculture with CD8+ T cells. As a positive control, DCs were loaded with 1 μM UL47289–298 peptide.
IFN-γ ELISPOT assay
The IFN-γ ELISPOT assay was performed in nitrocellulose-lined, 96-well microtiter plates (Millipore). Plates were coated overnight at 4°C with IFN-γ mAb (1D1K; Mabtech) at 5 μg/ml in 50 μl/well. Plates were washed six times with PBS and blocked with complete medium for 1 h at room temperature. After addition of 4 × 103 DCs and 4 × 104 CD8+ T cells/well and incubation for 24 h at 37°C, plates were washed six times with PBS/0.05% Tween 20 (Sigma-Aldrich). Biotinylated IFN-γ mAb (7-B6–1; Mabtech) diluted to 1 μg/ml in PBS/0.05% Tween 20/1% BSA was then added at 100 μl/well. After 2 h at room temperature and six washes with PBS/0.05% Tween 20, 100 μl of streptavidin-alkaline phosphatase (Bio-Rad) diluted at 1/1000 in PBS was added to each well for 45 min. Plates were then washed three times with PBS/0.05% Tween 20, followed by three washes with PBS. After a final wash, 100 μl of 5-bromo-4-chloro-3-indolyl-phosphate/NBT plus substrate (Bio-Rad)/well was added for 20 min at room temperature. The reaction was stopped under running water. The plate was allowed to dry, and spots were counted using a stereomicroscope. For analysis, spot counts from wells without DC (background) were subtracted from all wells.
Two-tailed Student’s t test was used for comparison of means. A value of p < 0.05 was considered statistically significant.
HSV induces apoptosis in human monocyte-derived DCs
We have previously shown that DC viability decreases significantly following HSV-1 infection (16). Concurrently, our experiments on murine bone marrow-derived DCs revealed lower viabilities in HSV-2 than in HSV-1 infected DC cultures (17). However, there are well-described differences in immunobiology of HSV infection between mice and humans. We, therefore, examined whether this observation could be replicated in human system and whether the differences in viabilities were due to different levels of apoptosis induction. Immature DCs were infected with HSV-1 at an MOI of 5 PFU/cell and analyzed for apoptosis and necrosis at 6, 12, or 24 h p.i. For this purpose we used annexin V, a protein that binds with high affinity to phosphatidylserine residues that become exposed on the surface of apoptotic cells. Cells react to annexin V before the plasma membrane loses its ability to exclude a vital dye such as PI. Thus, by staining cells with a combination of fluorescently labeled annexin V and PI, it is possible to differentiate viable cells (annexin V−PI−), early apoptotic cells (annexin V+PI−), and late apoptotic cells (annexin V+PI+). Our results indicated that HSV-1 infection induced apoptosis of DCs that could be observed as early as 6 h p.i. (Fig. 1,A). The percentage of apoptotic cells increased over time, and by 24 h p.i., on the average 33.3% of DCs were annexin V+PI− (four experiments) compared with 4.0% of uninfected cells (p < 0.05; Table I). Infection of DCs with the UV-HSV-1 had no effect on cell viability, suggesting that virus binding, entry, and uncoating were not responsible for the induction of apoptosis (Fig. 1,A and Table I).
|.||Mean (%) .||Range (%) .||SD (%) .||p Valueb .||p Valuec .|
|.||Mean (%) .||Range (%) .||SD (%) .||p Valueb .||p Valuec .|
Cells that bound annexin V in the absence of PI staining were considered apoptotic. The results are a composite of four experiments from different donors.
Mock vs infected cells, by Student’s t test.
HSV-1-infected vs HSV-2-infected cells, by Student’s t test.
An HSV-type-dependent ability to induce apoptosis was also seen in humans, but with different kinetics. On the average, 33.3% of HSV-1-infected and 51.1% of HSV-2-infected DCs were annexin V+PI− at 24 h p.i. (p < 0.05; Fig. 1, A and B, and Table I). The remaining cells consisted of late apoptotic or necrotic cells (annexin V+PI+) and viable cells (annexin V−PI−). Different strains of HSV-1 and HSV-2 were also tested with comparable results (results not shown). Our results indicate that HSV-2 is also more cytopathic to human monocyte-derived DCs and that the level of apoptosis was higher for HSV-2-infected than for HSV-1-infected cells.
Previous studies using specific and general caspase inhibitors indicated that HSV-1-induced apoptosis is caspase dependent (21, 26). Therefore, to confirm the induction of apoptosis by HSV, we used two assays to determine caspase activation. Treatment of cells with a cell-permeable, fluorescently labeled, pan-caspase inhibitor, fluorescein-z-VD-fmk, showed that over a period of 24 h, DCs from HSV-1-infected cultures had significantly higher levels of activated caspases than uninfected DCs or DCs treated with UV-HSV-1 (Fig. 1,C). These observations were confirmed by a colorimetric assay of caspase-3 activation from cell lysates of infected DCs compared with uninfected DC controls (Fig. 1 D). In this assay, HSV-2-infected cells had a higher level of caspase-3 activation, confirming our annexin V staining data. Thus, HSV infection of DCs induces caspase-3 activation.
Apoptosis induces characteristic, well-described changes in cell ultrastructure, most of which occur at late irreversible stages of apoptosis, but they can be distinguished from necrosis. Therefore, we next analyzed the ultrastructural morphology of HSV-infected DCs compared with uninfected controls. Uninfected DCs showed normal ultrastructure with heterogeneous chromatin and intact nuclear and plasma membranes and organelles (Fig. 2,B). In contrast, DCs from HSV-1-infected cultures at 12 h p.i. showed typical features of apoptosis, namely, cell shrinkage, chromatin condensation, nuclear fragmentation, membrane blebbing, and apoptotic body formation, with relatively preserved organelle structure (Fig. 2, A and D). In some apoptotic DCs, HSV-1 nucleocapsids were detected in the nucleus (Fig. 2,D, inset) and occasionally in the cytoplasm. There was no evidence of necrosis, which results in general swelling of cells and disruption of organelles with minimal nuclear changes. These morphological changes were very similar to the observed morphology in positive control cultures of human DCs treated with 1 M d-sorbitol to induce apoptosis (Fig. 2 C). Thus, DCs from HSV-1-infected cultures display typical apoptotic cell morphologies.
Both early and late HSV proteins are responsible for induction of apoptosis in monocyte-derived DCs
To determine the class and kinetics of viral proteins responsible for the observed apoptosis, we treated monocyte-derived DCs with 400 μg/ml PAA, an inhibitor of viral DNA polymerase and, thus, of late viral protein expression. We observed that apoptosis of HSV-1- and HSV-2-infected DCs was reduced, but not abolished, in PAA-treated cultures (Fig. 3), suggesting that both early and late HSV proteins play roles in apoptotic cell death in HSV-infected DCs. The levels of apoptosis due to late proteins were more apparent in HSV-2-infected DCs than in those infected with HSV-1.
HSV-infected apoptotic DCs are phagocytosed by uninfected DCs
Immature DCs can acquire microbial and tumor-derived Ags through an efficient phagocytosis of apoptotic cell fragments (27, 28, 29). Therefore, we investigated whether HSV-infected DCs could be phagocytosed by uninfected DCs. CFSE-labeled DCs (green) were infected with HSV-2 for 18 h and cocultured with uninfected PKH26-labeled DCs (red) for an additional 24 h at 37°C at a ratio of 1:1. Using flow cytometry, we detected double-positive cells, suggesting uptake of apoptotic cells (Fig. 4,A). Because HSV-infected DC cultures contained a mix of viable and apoptotic cells, we wanted to determine which population was being internalized by uninfected DCs. Using microbeads conjugated to annexin V, we separated viable (annexin V−) from early and late apoptotic (annexin V+) cells. Each population from HSV-infected cultures was then incubated with uninfected PKH26-labeled DCs at 37°C as described above and analyzed by flow cytometry. As a negative control, the cells were cocultured at 4°C. Results suggested that both apoptotic and viable cells were taken up by uninfected DCs, although the majority of phagocytosed cells were apoptotic DCs (Fig. 4,B). To confirm that our flow cytometry data reflected internalization of apoptotic cells by uninfected DCs and not simply cell-to-cell contact, cells were analyzed by confocal microscopy. As shown in Fig. 4,C, PKH26-labeled DCs contained cell fragments from CFSE-stained, HSV-infected DCs, thereby excluding cell-cell adherence as the explanation for the double-positive cells seen on FACS dot plots (Fig. 4, A and B). Therefore, uninfected DCs are able to capture and internalize cellular fragments from apoptotic, HSV-infected DCs.
Uninfected DCs that were pulsed with apoptotic HSV-infected DCs stimulate HSV-specific CD8+ T cells
It has previously been shown that DCs can acquire Ags from apoptotic cells for presentation to CD8+ T cells (27, 30, 31). We therefore investigated whether HSV Ags from infected DCs could be cross-presented by uninfected DCs. For this purpose, we infected DCs from a HLA-A2.1− donor with HSV-2 and cocultured the cells with uninfected HLA-A2.1+ DCs as described in Materials and Methods. Fig. 5,A shows that an HLA-A2.1-restricted CD8+ T cell clone, specific for a HSV-2 UL47289–298 peptide, recognized the epitope on the surface of HLA-A2.1+ DCs. The recognition was dependent on the HLA haplotype of uninfected DCs and was inhibited by anti-HLA-A2 Ab, thus excluding the possibility of direct viral presentation by HSV-infected DCs (Fig. 5,A). Because HSV-infected human DCs can produce low levels of infectious virus (16), it was possible that DCs that have taken up infected cells became infected, resulting in direct presentation of viral Ag to HSV-specific CD8+ T cells. To exclude this possibility, initial infection of DCs and coculture with uninfected DCs were conducted in the presence of 400 μg/ml PAA to prevent HSV replication. We observed that uninfected DCs cross-presented HSV Ag even in the absence of productive infection (Fig. 5,B). Moreover, cross-presentation was as effective as direct presentation of UV-HSV-2 added at a concentration equivalent to an MOI of 5 (Fig. 5,B). To exclude direct infection as a possible source of Ag, uninfected DCs were incubated with the supernatant from HSV-infected DCs that had been passed through a 0.22-μm pore size filter, thereby allowing only virus particles and soluble matter to pass through. In this case, there was no specific response, suggesting that cell-to-cell contact was required for Ag uptake (Fig. 5 B). Together, these results indicate that cross-presentation of HSV Ag requires uptake of infected cells and that the levels of HSV or viral Ag exported from infected DCs are insufficient for cross-presentation by DCs to CD8+ T cells.
DCs cross-present HSV Ag from apoptotic DCs
It has recently been shown that DCs are capable of cross-presenting viral Ag from both apoptotic and live cells (32). To determine whether cross-presentation of HSV Ag was due to apoptotic or viable DC uptake, we separated the two populations using annexin V-conjugated beads, as described previously, and cocultured them with uninfected HLA-A2.1+ DCs. Fig. 6 shows that cross-presentation of HSV Ag from apoptotic DCs was much more effective than that of HSV Ag from viable DCs.
The results of the most recent clinical trials of HSV subunit vaccine (gD2 + monophosphoryl lipid A) against genital herpes disease, which provided partial protection in HSV-1- and HSV-2-seronegative women (33), have stimulated efforts to better define the key immunogenic proteins and mechanisms responsible for immune activation. DCs play a crucial role in the initiation of cellular immunity in viral infections, especially after initial epithelial invasion. Experiments in our laboratories and others have shown that HSV has profound effects on the function of acutely infected human and murine DCs, including down-regulation of costimulatory and adhesion molecules, inhibition of cytokine secretion, and maturation, rendering these cells poor T cell stimulators (15, 16, 17, 34). In the present study we confirmed our observation of induced apoptosis in mice and extended this finding to demonstrate the role of both HSV early and late proteins in HSV-induced DC apoptosis. Furthermore, uninfected (bystander) DCs can phagocytose apoptotic DCs and cross-present the contained HSV Ags to virus-specific CD8+ T cells. This represents the first evidence of cooperation between infected and uninfected DCs in herpes infection.
Programmed cell death in DCs was demonstrated by annexin V staining of PI-negative cells, activation of caspases (by intracellular staining and in cell lysates), and electron microscopy. As expected from the activation of caspase-3, which is known as the death caspase, all the irreversible ultrastructural features of apoptosis appeared with subsequent cell death. These included early changes, such as loss of the nuclear membrane, chromatin condensation, and vacuolation, and late changes, such as nuclear fragmentation, which were similar to those observed in the positive control cultures consisting of sorbitol-treated DCs. Similar to that in mice, infection of DCs with HSV-2 resulted in a higher proportion of apoptotic cells than in HSV-1-infected DCs (17). In this study we show that HSV-induced apoptosis of DCs can occur in the absence of viral DNA synthesis, suggesting that the induction phase takes place at the earlier stages of infection. However, late proteins also appear to play a role in apoptosis, especially in HSV-2-infected DCs. These results support the previous findings of Koyama and Adachi (35) that initiation of apoptosis in HSV-1-infected HEp-2 cells is an early event. In our system, initiation of apoptosis does not occur at viral binding or entry, as demonstrated by the inability of UV-HSV to initiate apoptosis of DCs, which is in agreement with the study by Aubert et al. (21), which used human epithelial cells.
Monocyte-derived DCs appear to be one of the few human cell types in which HSV-1 is not able to block virus-induced apoptosis; the others are activated T lymphocytes from cord blood and peripheral blood (36, 37, 38). HSV-1 encodes several pro- and antiapoptotic genes. Those implicated in the prevention of apoptosis include HSV-1 IE protein ICP27, the early US3 kinase, the late US5 glycoprotein (gJ), HSV-2 US3, and ICP10 protein kinase (21, 39, 40, 41, 42, 43, 44, 45). Viral genes responsible for initial induction of apoptosis are ill defined, but US1.5, a product of the immediate-early α22 gene, has proapoptotic capacity (46).
Induction of apoptosis in DCs has significant implications and consequences for the immune response to primary and recurrent HSV infection in vivo. Classically, apoptosis is an innate cellular response to viral infection, which, when it occurs early, can prevent completion of viral replication and spread. HSV has evolved mechanisms to inhibit apoptosis in most human cells, although these mechanisms do not seem to be operative in human monocyte-derived DCs or in murine bone marrow-derived DCs (17). Apoptosis of DCs would then be deleterious to the host, because the number of functional APCs that can activate virus-specific CTLs is reduced unless an HSV-specific T cell response can be induced with cross-presentation of HSV Ags by functional bystander DCs. Cross-presentation has been demonstrated in vitro for DCs that have taken up vaccinia virus- and vaccinia-HIV recombinant-infected monocytes or DCs (31, 47), CMV-infected fibroblasts (28, 30), melanoma Ags in recombinant canarypox virus (48), and EBV-infected B cells (49).
In vivo studies in mice hint at cooperation between DCs found at the site of HSV infection and those initially outside the danger zone (18, 19). Together, these findings suggest that cooperation between DCs subsets may be an important factor in the generation of CD4+ or CD8+ T cell responses to HSV. Our results represent the first evidence that such cooperation is likely. We show that uninfected, PKH26-labeled DCs could phagocytose HSV-infected DCs stained with CFSE and, using an HLA-A2.1+-restricted CD8+ T cell clone specific for an HSV-2 tegument peptide UL47289–298, that uninfected DCs can cross-present HSV Ag from infected DCs. HSV Ag presentation could not be induced by the direct infection of bystander DCs from virus released into the extracellular fluid by initially infected DCs, possibly due to the low productivity of infection, as previously reported (16). In contrast, UV-HSV-2 added to HLA-A2.1+ DCs was sufficient for stimulation of CD8+ T cells. Additionally, cross-presentation occurred in the absence of virus replication, which was blocked by PAA, and to a higher level, confirming that HSV Ag presentation was not due to direct infection of bystander (uninfected) DCs. This result also confirmed that the input viral Ag is sufficient for recognition by this CD8+ clone, as previously shown (25). Ag presentation was effectively blocked by anti-HLA-A2.1 mAb, excluding direct viral presentation by initially infected HLA-A2.1− DCs. Furthermore, there was far greater cross-presentation after uptake of apoptotic than viable infected DCs.
In skin or mucosa, herpes simplex is an epidermal infection and therefore is likely to first infect Langerhans cells, especially because human Langerhans cells are permissive to infection (M. Ruckholdt, L. Bosnjak, S. Turville, and A. Cunningham, unpublished observation). Moreover, Langerhans cell numbers in mouse skin have been shown to be inversely correlated with the severity of cutaneous infection (20). Direct infection of Langerhans cells in human skin explants is also likely to result in apoptosis. This might suggest that only bystander uninfected Langerhans cells are likely to transport HSV protein Ags to draining lymph nodes as has been suggested by Belz et al. (50). However, our findings also suggest a different mechanism. Apoptotic Langerhans cells in primary or recurrent herpetic lesions may be phagocytosed by underlying dermal or submucosal DCs. This phenomenon of migration of underlying DCs into inflammatory lesions has been observed in other conditions, such as contact dermatitis (51). Maturation and migration of such DCs to lymph node would lead to stimulation of specific CD4+ or CD8+ T cells. These would home to the lesions (52) and clear HSV-infected keratinocytes after reinduction of MHC class I on their surface by IFN-γ secreted by CD4+ T lymphocytes. Such hypotheses are consistent with the recent findings of Zhao et al. (19) and Allan et al. (18), who observed that the APCs in lymph nodes differ from those in Langerhans cells, and yet viral infection in their models was restricted to the epidermis. Earlier work by Mueller et al. (53) demonstrated that HSV DNA was not detectable by PCR in the skin-draining lymph nodes, suggesting that the transfer of HSV Ag from initially infected cells to APCs occurs proximal to the lymph nodes, consistent with the above hypothesis. Thus, this DC-DC cross-presentation provides a mechanism for early induction of HSV-specific CD4+ and CD8+ T lymphocytes in lymph nodes. Additional studies using murine models and human epidermal and dermal DC emigrants from ex vivo explants will be required to characterize these events in more detail.
We thank Drs. Lisa Sedger and Zorka Mikloska for their helpful comments and suggestions, and Drs. David Miles and Mark Willcox for the use of their caspase-3 assay kits.
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.
This work was supported by National Health and Medical Research Council of Australia Grant 253684 (to A.L.C. and C.A.J.) and National Institutes of Health Grant AI50132 (to D.M.K). L.B. was supported by the Australian Postgraduate Award and the Millennium Foundation Top-Up Grant.
Abbreviations used in this paper: DC, dendritic cell; MOI, multiplicity of infection; PAA, phosphonoacetic acid; PI, propidium iodide; p.i., postinfection.