|Home | About | Journals | Submit | Contact Us | Français|
Herpes simplex virus type 1 (HSV-1) is one of the most frequent and successful human pathogens. It targets immature dendritic cells (iDCs) to interfere with the antiviral immune response. The mechanisms underlying apoptosis of HSV-1-infected iDCs are not fully understood. Previously, we have shown that HSV-1-induced apoptosis of iDCs is associated with downregulation of the cellular FLICE-inhibitory protein (c-FLIP), a potent inhibitor of caspase-8-mediated apoptosis. In this study, we prove that HSV-1 induces degradation of c-FLIP in a proteasome-independent manner. In addition, by using c-FLIP-specific small interfering RNA (siRNA) we show for the first time that downregulation of c-FLIP expression is sufficient to drive uninfected iDCs into apoptosis, underlining the importance of this molecule for iDC survival. Surprisingly, we also observed virus-induced c-FLIP downregulation in epithelial cells and many other cell types that do not undergo apoptosis after HSV-1 infection. Microarray analyses revealed that HSV-1-encoded latency-associated transcript (LAT) sequences, which can substitute for c-FLIP as an inhibitor of caspase-8-mediated apoptosis, are much less abundant in iDCs as compared to epithelial cells. Finally, iDCs infected with an HSV-1 LAT knockout mutant showed increased apoptosis when compared to iDCs infected with the corresponding wild-type HSV-1. Taken together, our results demonstrate that apoptosis of HSV-1-infected iDCs requires both c-FLIP downregulation and diminished expression of viral LAT.
Herpes simplex virus type 1 (HSV-1) is an important human pathogen and causes mucocutaneous lesions in the orolabial region (48). Less often, HSV infection can result in life-threatening encephalitis in otherwise healthy individuals or disseminated infection in immunocompromised persons. Neonates infected during delivery can also develop severe disseminated infection. After primary infection of epithelial cells, HSV-1 establishes a state of latency in ganglia and persists for the lifetime of the host (59). In response to certain stimuli, the virus reactivates from latency without neuronal death. Thereafter, newly assembled daughter virions migrate down the axon toward epithelial tissue, causing recurrent infections of mucocutaneous regions and virus shedding. On the molecular level, HSV-1 consists of more than 80 genes that are expressed sequentially in a strongly regulated cascade (23, 24).
Apoptosis of host cells represents an important defense mechanism against viral invasion by preventing viral replication and dissemination. The extrinsic pathway of apoptosis induction is triggered by ligation of death receptors (36) or by injection of granzymes (63). Intrinsic triggers of apoptosis such as DNA damage, oxidative stress, deprivation of growth factors, and viral infection disrupt the integrity of the mitochondrial membrane, resulting in release of cytochrome c into the cytoplasm (46, 68). Transduction of the apoptosis-inducing signal within the cell occurs via initiator caspases: caspase-8 in the extrinsic pathway and caspase-9 in the intrinsic pathway. Both activate downstream effector caspases that cleave cellular substrates, leading to the characteristic morphological and biochemical features of apoptosis. The apoptosis signaling network is negatively regulated by caspase inhibitor molecules such as the cytokine response modifier A (CrmA) and the cellular FLICE-inhibitory protein (c-FLIP) (3, 6). Both splice variants of the cellular FLICE-inhibitory protein, the long form (c-FLIPL) and the short form (c-FLIPS), represent important inhibitors of death receptor-induced apoptosis (38). They act as catalytically inactive caspase-8 homologues, which compete with caspase-8 at the level of the death-inducing signaling complex (DISC) to prevent its activation.
Viruses, including HSV-1, have evolved numerous strategies to counteract apoptosis (3, 6, 45). Apoptosis is induced in response to HSV-1 at several signaling checkpoints, and the virus in turn has evolved multiple mechanisms that block apoptosis to prevent premature cell death (20). It has been shown that HSV-1 renders infected cells resistant to apoptosis induced by cytotoxic T lymphocytes (31). Moreover, studies in rabbits (52) and mice (1, 10) have shown that latently infected neurons are resistant to apoptosis, although only noncoding latency-associated transcript (LAT) and no viral proteins could be detected.
A major immune evasion strategy of HSV-1 is to attack immune cells, including T cells (21, 27, 28, 57), B cells (21), and macrophages (18). Most importantly, like many other herpesviruses (8), HSV-1 also targets human immature dendritic cells (iDCs) (9, 39, 41, 44, 55, 60) and induces apoptosis (9, 44, 55). As these cells play a central role in the induction of efficient immune responses (54, 64), HSV-1-induced apoptosis of iDCs represents a mechanism of immune evasion giving the virus the opportunity to replicate efficiently in epithelial cells and establish latency in neurons.
We have previously shown that HSV-1-induced apoptosis of iDCs is associated with decreased expression of the antiapoptotic cellular c-FLIP protein (44). In this study, we further analyze the mechanism of c-FLIP downregulation in iDCs and investigate whether c-FLIP downregulation is sufficient to drive uninfected iDCs into apoptosis. Furthermore, we prove that many different cell types other than iDCs, including epithelial cells, also decrease expression of c-FLIP protein in response to the virus, although they do not undergo apoptosis after infection. Finally, we address the question why iDCs but not epithelial cells, the main factory for production of HSV daughter virions, are prone to apoptosis after HSV infection.
Monocytes were obtained from buffy coats of healthy donors (DRK, Berlin, Germany) by density gradient centrifugation and subsequent isolation with Monocyte isolation kit II (Miltenyi Biotec GmbH, Bergisch Gladbach, Germany). Monocytes were then differentiated into iDCs in RPMI 1640 with 10% heat-inactivated fetal calf serum (FCS), 10 mM HEPES, 4 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, 200 U/ml interleukin-4 (IL-4) (Immunotools, Friesoythe, Germany), and 500 U/ml granulocyte-macrophage colony-stimulating factor (GM-CSF) (Immunotools, Friesoythe, Germany) for 6 days. Human umbilical vein endothelial cells (HUVECs) were prepared by the method of Jaffe et al. (29). Primary keratinocytes and fibroblasts were isolated and cultured as described previously (34). A549, HeLa, and HaCaT (provided by N. Fusenig, DKFZ, Heidelberg, Germany) cells were maintained in Dulbecco's modified Eagle's medium (DMEM) with 10% FCS, 4 mM l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin. L428 and L1236 cells were cultured in RPMI 1640 with 10% heat-inactivated FCS, 10 mM HEPES, 4 mM l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin.
HSV-1 strains KOS, F, and 17 as well as the LAT knockout mutant, 17N/H (7), were propagated on Vero E6 cells. The ICP27 deletion mutant of HSV-1 was propagated on Vero 2-2 cells, which are stably transfected to complement ICP27. Cell culture supernatant of infected Vero E6 or Vero 2-2 cells was collected and centrifuged at 2,000 × g for 10 min. Titers in the cleared supernatant were determined by measuring the 50% tissue culture infective dose (TCID50). For infection experiments, suspension cells were resuspended in and adherent cells were overlaid with a minimal volume of culture medium and the appropriate volume of HSV-1-containing supernatant was added to achieve the desired multiplicity of infection (MOI). After incubation for 1 h at 37°C to allow virus adsorption, the virus-containing medium was removed and cells were washed two times with medium. A sufficient volume of culture medium was added, and cells were maintained under normal culture conditions. As controls, infection was performed with heat-inactivated (56°C, 30 min) or UV-inactivated (10 min) HSV-1 or cells were left uninfected but otherwise treated in exactly the same way as infected cells.
Flow cytometry was performed on a FACScalibur (Becton Dickinson, Heidelberg, Germany). For detection of apoptosis cells were stained with Annexin V-fluorescein isothiocyanate (FITC) (Santa Cruz Biotechnology, Santa Cruz, CA) and propidium iodide (PI) and directly analyzed by flow cytometry. Antibody against HSV-1 glycoprotein D (gD) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA), and as a secondary antibody, we used FITC-conjugated goat anti-mouse F(ab′)2 (Immunotools, Friesoythe, Germany).
Cells were washed with cold phosphate-buffered saline (PBS) and lysed for 1 h on ice in NP-40 buffer (50 mM Tris-HCl, 10 mM EDTA, 80 mM KCl, 1% NP-40 [pH 7.5] with one tablet of protease inhibitor cocktail from Roche Diagnostics GmbH, Mannheim, Germany). After centrifugation at 15,000 × g for 30 min, the supernatant containing soluble cytoplasmatic proteins was mixed with loading buffer (0.5 M Tris-HCl, 10% SDS, 10% glycerol, 0.05% bromphenol blue, 5% β-mercaptoethanol), incubated at 95°C for 10 min, and separated by 12 to 15% SDS-PAGE. In order to achieve equal loading of lanes, the protein content of lysates was determined using the bicinchoninic acid (BCA) protein assay kit (Pierce, Rockford, IL). After transfer to polyvinylidene difluoride (PVDF) membranes (Millipore Corporation, Bedford, MA), the blots were blocked with 5% bovine serum albumin (BSA) or 5% milk powder in Tris-buffered saline with 0.1% Tween and incubated with anti-c-FLIP (clone NF-6) mouse anti-human monoclonal antibody. After washing, blots were incubated with peroxidase-conjugated goat anti-mouse F(ab′)2 (Jackson ImmunoResearch, West Grove, PA), developed by enhanced chemiluminescence using Super Signal West Dura extended duration substrate (Perbio, Bonn, Germany), and detected with a charge-coupled device (CCD) camera (KODAK Image Station 4000MM digital imaging system; Carestream Molecular Imaging, New Haven, CT). After being stripped off with 0.4 M glycine-0.2% SDS (pH 2.2), blots were blocked again and stained with anti-β-actin mouse anti-human monoclonal antibody (clone AC-15; Abcam, Cambridge, United Kingdom) to check for equal loading. Densitometric analysis of c-FLIP and β-actin bands was performed by using ChemiImager 4000 software (Alpha Innotech Corporation, San Leandro, CA).
A549 cells were transiently transfected with different c-FLIP-expressing plasmids using FuGENE 6 (Mannheim, Germany) according to the manufacturer's instructions.
Control nonsilencing small interfering RNA (siRNA) (target sequence, TTTATGTGTGCCCGTGTGGAA) and FLIP siRNA (target sequence, TTGTGCCGGGATGTTGCTATA) were purchased from MWG (Ebersberg, Germany). Monocytes were isolated from buffy coats of healthy donors (DRK, Berlin, Germany) with CD14 MicroBeads (Miltenyi Biotec GmbH, Bergisch Gladbach, Germany) and transfected with 1 μmol siRNA per 2 × 107 cells using an Amaxa Nucleofector (Amaxa, Cologne, Germany) according to the manufacturer's protocol (human monocyte Nucleofector kit; Nucleofector program Y-001). The transfected cells were then cultured in RPMI 1640 supplemented with 10% FCS, 4.5 mM glutamine, 100 U penicillin, and 100 μg/ml streptomycin together with 200 U/ml IL-4 and 500 U/ml GM-CSF for 5 days in order to get iDCs. For the apoptosis assay, 1 × 105 iDCs were stimulated with 0.125 μg/ml agonistic anti-CD95 monoclonal antibody (clone CH-11) for 12 h or were left untreated.
At different time points postinfection (p.i.), cells were harvested and total RNA was extracted with TRI reagent (Sigma-Aldrich, Deisendorf, Germany). Fluorescence-labeled cDNA was prepared from purified poly(A)-containing RNA by random hexamer-primed polymerization using Superscript II reverse transcriptase (Gibco-BRL, Invitrogen, Carlsbad, CA) as described in detail elsewhere (67). Generation of HSV-1 microarrays by oligonucleotide selection, synthesis, and deposition on the chip has been described previously (65, 69). Hybridization of microarrays with labeled cDNA and scanning was carried out as reported recently (66, 67). Briefly, microarrays were hybridized with cDNA probes for 18 h and rinsed for scanning with a proprietary HiLight dual-color kit (Genicon Sciences, San Diego, CA) at 52°C in an MAUI hybrid mixer assembly (BioMicro Systems, Salt Lake City, UT). After hybridization, the slides were washed, blocked, and bound by the gold and silver RLS method (two-color nucleic acid microarray resonance light-scattering-based method; Genicon Sciences, San Diego, CA). Microarrays were scanned with a GSD-501 HiLight reader (designed for RLS; Genicon Sciences, San Diego, CA). First, for each gene or coterminal transcript family, a median value was calculated from the net (minus background) hybridization values of the three replicate probe spots on the chip. These median hybridization values were then expressed as a percentage of the total viral signal of the given hybridization (median hybridization value of probe multiplied by 100, divided by the sum of all hybridization values) to obtain a relative abundance for the respective detected transcript. Microarray analysis of HSV-1 gene expression in each cell type was performed three times with RNA from three independent infection experiments. Relative abundances of each transcript from the different infections were pooled to calculate final median values. Relative abundances in iDCs and HaCaT or HeLa cells were compared by Student's two-tailed t test, assuming unequal variance and with the null hypothesis being that the true values in the different cell types are identical.
Cells were lysed with MagNA Pure lysis buffer (Roche, Mannheim, Germany), and mRNA was isolated with a MagNA Pure-LC device using standard protocols. RNA was reverse transcribed with avian myeloblastosis virus (AMV) reverse transcriptase (RT) and oligo(dT) primer (First Strand cDNA synthesis kit; Roche, Mannheim, Germany). For amplification of target sequences, LightCycler primer sets (Search-LC, Heidelberg, Germany) were used with LightCycler FastStart DNA Sybr green I kit (Roche, Mannheim, Germany). RNA input was normalized by the average expression of the housekeeping genes encoding β-actin and cyclophilin B. By plotting a known input concentration of a plasmid to the PCR cycle number at which the detected fluorescence intensity reached a fixed value, a virtual standard curve was generated. This standard curve was used to calculate transcript copy numbers. The presented relative copy numbers are mean averages of data of two independent analyses for each sample and parameter.
A549 cells were transfected with plasmids expressing c-FLIPL (F23G/F114G), which does not aggregate when overexpressed in cells (26) or c-FLIPS-green fluorescent protein (GFP) and lysed after 48 h in Triton buffer (50 mM HEPES, 5 mM EDTA, 150 mM NaCl, 1% Triton X-100 [pH 7.5]). After 1 h on ice and sedimentation of debris by centrifugation, the lysates were divided into four aliquots. One aliquot was frozen at −20°C until SDS-PAGE. Another aliquot was incubated at 37°C for 2 h and then loaded on SDS-PAGE. The two remaining aliquots were each mixed with an equal volume of lysate from HSV-1-infected (strain KOS; MOI = 10, lysed at 18 h p.i.) but untransfected A549 cells and also incubated at 37°C before loading on SDS-PAGE. To one of the aliquots mixed with lysates from HSV-1-infected A549 cells, the proteasome inhibitor MG132 (Sigma-Aldrich, Deisendorf, Germany) was added (50 μM) before incubation at 37°C. SDS-PAGE and immunodetection of c-FLIP were performed as described above.
We have previously reported that HSV-1 infected iDCs downregulate expression of the antiapoptotic c-FLIPL protein and undergo apoptosis at relatively late time points during infection (44). This was not the consequence of general protein degradation in HSV-1-infected cells since expression of other apoptosis-regulating proteins remained unaltered (44). We now investigated the underlying mechanism in more detail. The c-FLIPL level started to decline at 10 to 12 h p.i. (Fig. (Fig.1A).1A). Moreover, HSV-1-induced apoptosis in iDCs significantly increased between 14 and 20 h p.i. (Fig. 1B and C). Later during HSV-1 infection, there was an increase in Annexin V-positive cells that were also propidium iodide (PI) positive (Fig. (Fig.1C).1C). It is likely that those double-positive cells represent for the most part infected iDCs in the late stage of apoptosis. The reduction in the amount of c-FLIPL could have been due to decreased nuclear export of c-FLIPL mRNA. Therefore, we tested if the transport of c-FLIPL mRNA from the nucleus into the cytoplasm was blocked by HSV-1. The relative numbers of c-FLIPL transcripts in the nucleus and cytoplasm of HSV-1-infected iDCs were determined separately by quantitative reverse transcription (RT)-PCR (Fig. (Fig.2A).2A). Indeed, the abundance of c-FLIPL mRNA in the nucleus was upregulated during HSV-1 infection or infection with UV-inactivated HSV-1, suggesting a block in nuclear export of c-FLIPL transcripts. However, in both cases the number of c-FLIPL transcripts in the cytoplasm of HSV-1 infected iDCs did not decrease compared to uninfected cells or cells infected with heat-inactivated virus. This finding excludes the possibility that the observed reduction in c-FLIPL protein was due to a mechanism operating on the mRNA level.
We next determined whether c-FLIP degradation in HSV-1-infected iDCs is mediated by the proteasome as turnover of c-FLIP in uninfected cells is proteasome dependent (56). However, treatment of HSV-1-infected iDCs with the proteasome inhibitor MG132 could not block c-FLIP downregulation (Fig. (Fig.2B).2B). Next we explored whether c-FLIP protein downregulation was due to degradation by a viral or a virus-induced cellular protease (Fig. (Fig.2C).2C). For this purpose, lysates derived from A549 cells transfected with plasmids expressing c-FLIPL or c-FLIPS were mixed with lysates derived from HSV-1-infected but untransfected A549 cells. In the absence of viral proteins, the amount of c-FLIP protein was not decreased. In contrast, after adding lysates from HSV-1-infected cells, both c-FLIPS and c-FLIPL were strongly degraded in the presence or absence of a proteasome inhibitor. Taken together, these data suggest that either a viral or a virus-induced cellular protease degrades c-FLIP independently of proteasomal function.
Reduction of c-FLIPL in iDCs is associated with apoptosis induction (44, 70). This does not necessarily reflect a causative relationship between both events. Moreover, HSV-1 modulates the expression of many cellular proteins. Therefore, we addressed the question whether reduction of c-FLIPL by siRNA is sufficient for induction of apoptosis in iDCs. Treatment with c-FLIPL-specific but not control siRNA strongly decreased the amount of c-FLIPL in iDCs (Fig. (Fig.3A).3A). As a consequence, the viability of iDCs and their resistance to death receptor-mediated apoptosis declined (Fig. (Fig.3B).3B). After transfection with control siRNA, 85% of iDCs were neither apoptotic nor necrotic, whereas this percentage dropped to 20% after transfection with c-FLIPL-specific siRNA. Simultaneously, an increase in apoptotic (Annexin V positive) cells from 2% to 16% was detected. Stimulation of the extrinsic apoptosis pathway by addition of agonistic anti-CD95 antibodies led to a further reduction in the number of viable cells in iDCs treated with c-FLIPL-specific siRNA. Moreover, a large number of iDCs that had been transfected with c-FLIPL-specific siRNA and left untreated or treated with agonistic anti-CD95 antibodies were Annexin V and PI positive (61% and 89%, respectively). Those double-positive cells most likely represent iDCs in the late stage of apoptosis. Collectively, these data suggest that c-FLIPL downregulation is sufficient to induce apoptosis in some iDCs and strongly enhances susceptibility to death receptor signaling in others.
We tested many different cell types other than iDCs for c-FLIPL downregulation and apoptosis induction after HSV-1 infection to prove whether the observations made with iDCs can be generalized. For this purpose, different cell types were infected with different MOIs (ranging from 1.5 to 5) to achieve optimal infection rates. Human umbilical vein endothelial cells (HUVECs), primary fibroblasts, primary keratinocytes, and the Hodgkin/Reed-Sternberg (HRS) cell lines L428 and L1236 expressed c-FLIP and could be efficiently infected with HSV-1, as determined by staining for viral glycoprotein D (gD). (The numbers of infected cells after subtraction of values obtained with uninfected cells or cells infected with heat-inactivated virus are as follows: iDCs, 86%; HUVECs, 89%; fibroblasts, 91%; keratinocytes, 40%; L428, 83%; and L1236, 67%.) Intriguingly, the amount of c-FLIPL was reduced after infection in all cell types tested compared to uninfected cells or cells treated with heat- or UV-inactivated virus (Fig. (Fig.4A).4A). In some of these cell types (iDCs, HUVECs, primary keratinocytes, and HRS cell lines), c-FLIPS, another c-FLIP isoform, was detectable and was also downregulated after HSV-1 infection (data not shown). As shown in Fig. Fig.4B,4B, the percentage of apoptotic (Annexin V positive) iDCs increased from 5% to 34% after infection with HSV-1. In contrast, in the other cell types, no change in the percentage of apoptotic cells occurred and the percentage of PI-positive (necrotic) cells strongly increased. Taken together, these data suggest that c-FLIP downregulation occurs in all c-FLIP-expressing cells during HSV-1 infection but only iDCs become apoptotic. This implies that besides c-FLIP downregulation there are other viral or cellular parameters which are important for the decision whether HSV-1-infected cells undergo apoptosis.
For efficient replication, viruses, including HSV-1, have to prevent premature cell death. In the case of HSV-1, epithelial cells are the main producers of viral progeny. In contrast, iDCs contribute little to the overall production of free HSV-1 virions. Thus, the different outcome of HSV-1-induced c-FLIP downregulation in iDCs as compared to other cell types could have been due to cell-type-dependent viral counterbalance of apoptosis. To test this hypothesis, iDCs and epithelial cells (HeLa cells and HaCaT cells) were infected with HSV-1. After 18 h, the numbers of infected (Fig. (Fig.5A5A and Table Table1)1) and apoptotic (Fig. (Fig.5B5B and Table Table1)1) cells were determined. Most of the epithelial cells could be infected, whereas the infectability of iDCs showed some donor-dependent variation. In contrast to iDCs, HeLa or HaCaT cells were relatively resistant to apoptosis induction after HSV-1 infection. Even at a very high MOI of 15, the HSV-1-infected epithelial cells did not undergo substantial apoptosis (data not shown). This suggests that HSV-1 implements an efficient antiapoptotic program in these cells. Therefore, epithelial cells were infected with an ICP27 deletion mutant of HSV-1 (ΔICP27), which is unable to prevent induction of apoptosis after infection (4, 62). Figure Figure5C5C shows that 35 to 45% of epithelial cells were driven into apoptosis 18 h after infection with ΔICP27. In conclusion, the comparatively low level of apoptosis in HSV-1-infected epithelial cells is due to viral counterbalance of apoptosis but not due to an inherent apoptosis resistance.
Having established that viral antiapoptotic genes are responsible for the low extent of apoptosis in HSV-1-infected epithelial cells as compared to iDCs, we comparatively studied viral gene expression in these cell types. For this purpose, total RNA was extracted at 3, 6, 10, and 18 h after HSV-1 infection and subjected to microarray analyses which include detection of LAT. The latter have been shown to block apoptosis initiated by caspase-8 (11, 22, 33). Strikingly, at 10 and 18 h p.i., specific probes annealing to HSV-1-encoded LAT (LATI, RH6, LATX, LAT3, and LAT5C) (Fig. (Fig.6)6) revealed a significantly lower abundance of antiapoptotic LAT in iDCs compared to HaCaT or HeLa cells on the microarray. Figure Figure77 shows that at 10 h p.i., the relative abundance of LAT in iDCs compared to HaCaT cells was between 1.4-fold (LAT3) and 7.9-fold (LAT5C) lower (corresponding to log10 iDC/HaCaT ratios of −0.15 for LAT3 and −0.90 for LAT5C, respectively). In comparison to HeLa cells, the relative abundance of LAT in iDCs at this time point was between 2.3-fold (LAT3) and 6-fold (LAT5C) lower (corresponding to log10 iDC/HeLa ratios of −0.36 for LAT3 and −0.78 for LAT5C, respectively).
The expression kinetics of LAT in iDCs and HaCaT and HeLa cells as detected by our microarray analyses is shown in Fig. Fig.8.8. The abundance of ICP4, an essential regulatory protein required for expression of most E and L genes, was found to be similar in iDCs and epithelial cells. There was a significantly higher expression of unique short region 3/4 (US3/4) in iDCs compared to HeLa cells at 6 h p.i. However, this was not necessarily due to increased expression of US3, a known antiapoptotic HSV-1 gene, since we could not discriminate between transcripts belonging to a coterminal transcript family, e.g., US3 and US4. In epithelial cells, the abundance of LAT increased strongly between 6 and 10 h p.i. and either further increased (LATI and RH6) or remained at the same level (LATX and LAT3) until 18 h p.i. In contrast, in iDCs only low numbers of LAT could be detected between 6 and 18 h p.i., when cells became apoptotic. Besides LAT, we did not find significant differences between iDCs and epithelial cells with regard to other antiapoptotic HSV-1 genes, including unique long region 54 (UL54), US1, US5, US6, and US11 (data not shown). In conclusion, in the critical time period of HSV-1-induced c-FLIP downregulation, the relative abundance of LAT was much lower in iDCs as compared to epithelial cells.
In order to evaluate the functional relevance of LAT in iDCs, we compared the extent of apoptosis in iDCs infected with 17N/H, a LAT deletion mutant of HSV-1 (Fig. (Fig.6)6) and the respective wild-type virus (strain 17). Intriguingly, we observed significantly (P ≤ 0.05; directional Wilcoxon signed-rank test) more apoptosis in 17N/H-infected iDCs compared to iDCs infected with the respective wild-type strain (Fig. (Fig.9A).9A). Increased apoptosis was not due to enhanced infection efficiencies as the numbers of gD-positive iDCs after infection with wild-type strain and LAT deletion mutant virus were not different (Fig. (Fig.9B).9B). As expected, in iDCs infected with 17N/H no signal with LAT-specific probes LAT-A and LAT-B was obtained by quantitative RT-PCR (Fig. (Fig.9C).9C). The signal detected with probe LAT-C in iDCs infected with 17N/H (Fig. (Fig.9C)9C) is likely to originate from aberrant LAT sequences downstream of the deletion in 17N/H (Fig. (Fig.6),6), which covers a region not associated with the antiapoptotic function of LAT (1). There was the possibility that the higher apoptosis-inducing capacity of the LAT deletion mutant virus was due to altered expression of viral lytic genes in iDCs, thus inducing more apoptosis. We found a slight but statistically not significant increase in copy numbers of UL10 transcripts and transcripts originating from the antiapoptotic viral genes US3, US6, US1, and US11 after infection with 17N/H compared to iDCs infected with the wild-type strain (Fig. (Fig.9D).9D). Taken together, our experiments demonstrate a higher susceptibility of iDCs to HSV-1-induced apoptosis when the antiapoptotic function of LAT is lacking, despite slightly increased or unaltered expression of the other antiapoptotic viral genes. These data support the conclusion that in principle LAT can contribute to protection of iDCs from apoptosis. However, iDCs express much less LAT than other cell types such as epithelial cells. Thus, HSV-1 does not efficiently counterbalance apoptosis induction in iDCs, which play a crucial role in the antiviral immune response but are irrelevant for the overall production of viral progeny.
In this study, we observed that HSV-1-induced apoptosis in iDCs results from cell-type-specific imbalance in proapoptotic and antiapoptotic viral functions.
Several studies have reported that HSV-1 drives iDCs into apoptosis after infection (9, 39, 41, 44, 55, 60). Previously, we described how HSV-1-infected iDCs downregulate expression of c-FLIPL (44), which blocks the caspase-8-dependent apoptosis pathway (38), but the underlying mechanism was not clarified. We can exclude that c-FLIP protein synthesis was shut off by HSV-1-encoded ICP27, which is known to inhibit pre-mRNA splicing and therefore nuclear export of mRNA (61). In this study, increased numbers of c-FLIPL transcripts were detected in the cytoplasm of HSV-1-infected iDCs compared to uninfected cells or cells infected with heat-inactivated HSV-1. The enhanced abundance of c-FLIPL transcripts can be explained by HSV-1-induced activation of NF-κB (2, 47) which increases transcription of the c-FLIP gene (37, 40). Thus, the mechanism of HSV-1-induced c-FLIP downregulation is not operating on the mRNA but on the protein level. Indeed, we found that both c-FLIPL and c-FLIPS are degraded independently of proteasomal function by either a viral protease or a virus-induced cellular protease, excluding inhibition of translation by RNA interference as an underlying mechanism. In contrast, proteasome-dependent degradation of c-FLIP has been observed in other situations, including activation of p53 (19), treatment with peroxisome proliferator-activated receptor γ ligand (35), and adenovirus infection (51).
Subsequently, we investigated the functional relevance of c-FLIP downregulation for the reduced survival of iDCs after HSV-1 infection. This question is difficult to address in the context of HSV-1 infection as the virus alters the expression of many cellular proteins. Knocking down c-FLIPL expression in the absence of viral infection resulted in apoptosis of iDCs and enhanced their susceptibility to death receptor signaling. In line with this finding, other investigators described how downregulation of c-FLIPL expression in iDCs after treatment with bisindolylmaleimide promotes CD95-mediated apoptosis (70). It is possible that, in iDCs, c-FLIP prevents ligand-independent death signaling through the caspase-8-mediated apoptosis pathway, as shown recently for breast cancer cells (14). In accordance with this view, death ligands, including CD95, TRAIL receptor, and tumor necrosis factor (TNF) receptor, contributed weakly and only in some donors to apoptosis of HSV-1-infected iDCs (44). Taken together, these findings indicate that HSV-1-induced c-FLIP degradation is indeed functionally relevant for iDC apoptosis.
Besides iDCs, all other cell types susceptible to HSV-1 infection downregulated c-FLIP, including HUVECs, primary fibroblasts, primary keratinocytes, and the Hodgkin/Reed-Sternberg (HRS) cell lines L428 and L1236. However, these cells were resistant to apoptosis induction by HSV-1. The latter finding is in accordance with previous reports from several other groups analyzing HSV-1-infected cells other than iDCs (4, 5, 20). In another study, CD95-mediated apoptosis of human epithelial cells was blocked after HSV-1 infection (42), suggesting that HSV-1 can actively interfere with the caspase-8-mediated apoptosis pathway in this cell type. Similarly, we found that epithelial cells reduce c-FLIP expression in response to HSV-1 infection without being forced into apoptosis. Taken together, these results suggest that HSV-1-encoded antiapoptotic molecules can counterbalance the loss of c-FLIP in epithelial cells but not in iDCs.
To test the hypothesis that expression of HSV-1-encoded antiapoptotic genes depends on the cell type, comparative microarray analyses of infected iDCs and epithelial cells were performed. We did not find significant differences in expression of most of the known antiapoptotic viral genes, including the α4, UL54, US1, US5, US6, and US11 genes. It is unclear whether US3, a virus-encoded kinase with known antiapoptotic activity, is differentially expressed, since the detected signal originates from a coterminal transcript family that also includes US4 transcripts. However, it is likely that these viral genes are not involved in compensation of c-FLIP downregulation as they mostly block apoptosis induced by the intrinsic pathway (45). Most strikingly, at 10 and 18 h p.i. HSV-1-encoded LAT was only weakly expressed in iDCs compared to epithelial cells. This finding is in accordance with a previous report showing that the LAT promoter activity varies greatly according to cell type and that LAT is highly expressed in epithelium (30). How LAT expression is downregulated in iDCs during lytic HSV infection remains to be analyzed.
The antiapoptotic function of LAT is crucial for HSV-1 reactivation from latency (1, 10, 43, 52). Moreover, LAT has also been demonstrated to inhibit apoptosis in transfected cells (1, 11, 12, 25, 33, 49). Importantly, apoptosis initiated by caspase-8 can be blocked by LAT products (11, 22, 33) and cell lines producing high numbers of LAT do not activate the executioner caspase-3 in response to apoptosis stimuli (12, 50). Recently, it has been demonstrated that c-FLIP can substitute for LAT function during reactivation of HSV-1 (32). This finding indicates that LAT is able to block the caspase-8 pathway. Thus, it is conceivable that HSV-1-encoded LAT can compensate for c-FLIP downregulation in cells productively infected with HSV-1. A role for LAT in counterbalancing c-FLIP downregulation is supported by further evidence. Significantly more apoptotic iDCs were found after infection with an HSV-1 LAT null mutant (7) than after infection with the corresponding parental strain, which can provide at least some LAT.
So far, it is unclear how LAT exerts its c-FLIP-like function. LAT comprises a group of RNA molecules that are transcribed from within the long repeats that flank the unique long (UL) segment of the viral genome (58). Splicing of the polyadenylated 8.3- to 8.5 kb primary LAT generates a stable 2-kb LAT intron which is neither capped nor polyadenylated (17). This primary LAT, detected by probes LATX and LAT3, showed a significantly lower abundance in iDCs as compared to epithelial cells, suggesting that the stable intron is also less abundant in iDCs. Additional processing of the 2-kb LAT intron results in a 1.5-kb product which is sufficient for blocking apoptosis (25, 33, 49). As LAT most likely does not code for a protein (16), non-protein-coding RNAs (e.g., microRNAs) could mediate its antiapoptotic function (13, 53).
Our results demonstrate that HSV-1 uses its antiapoptotic machinery in a cell-type-dependent manner that precisely serves its needs. In epithelial cells, which represent the major site of viral replication, LAT is sufficiently expressed to counterbalance c-FLIP downregulation in the late phase of the viral replication cycle and prevent premature apoptosis. In this way, HSV-1 ensures that enough viral progeny are produced to transmit the virus to other individuals. In contrast, iDCs are not only dispensable for overall production of HSV-1 virions but also play a crucial role in orchestrating the antiviral immune response. By drastically reducing LAT expression in iDCs, HSV-1 may allow c-FLIP downregulation to become functional, thereby driving iDCs into apoptosis. In this way, the virus could severely impair presentation of viral antigen by virus-infected iDCs very late in the infection cycle, whereas the immediate-early protein ICP47 (71), which blocks transport of peptides into the endoplasmic reticulum (ER), downregulates antigen presentation in the early phase. It is also tempting to speculate that HSV-1 avoids early induction of apoptosis in iDCs because early in infection it uses this cell type as a vehicle to enhance viral spread in the organism. In conclusion, the cell-type-specific and timely counterbalance of apoptosis-inducing responses may allow HSV-1 to efficiently spread and replicate and at the same time subvert the antiviral immune response.
We thank C. Johnen and K. Bräutigam (Department of Surgery, Charité Universitätsmedizin Berlin, Berlin, Germany) for providing primary keratinocytes; N. E. Fusenig (Deutsches Krebsforschungszentrum, Heidelberg, Germany) for providing HaCaT cells; N. W. Fraser (Department of Microbiology, University of Pennsylvania School of Medicine, Philadelphia, PA) for permission to work with the 17N/H LAT mutant virus; T. Kaiser (Deutsches Rheuma-Forschungszentrum, Berlin, Germany) for assistance with flow cytometry; P. Krammer and H. Walczak (Deutsches Krebsforschungszentrum, Heidelberg, Germany) for providing monoclonal antibody clone NF6; and M. Naito (Institute of Molecular and Cellular Biosciences, University of Tokyo, Tokyo, Japan) for making c-FLIP mutant plasmids available to us.
This study was made possible by a grant from the Charité-Universitätsmedizin Berlin (to A.K. and G.S.).
Published ahead of print on 11 November 2009.