Apoptosis of human peripheral blood mononuclear cells by wild-type measles virus infection is induced by interaction of hemagglutinin protein and cellular receptor, SLAM via caspase-dependent pathway

Authors


Correspondence
Shigeru Suga, Department of Pediatrics, Mie National Hospital, 357 Ozatokubota-cho, Tsu, Mie 514-0125, Japan.
Tel: +81 59 232 2531; fax: +81 59 232 5994; email: sugas@mie-m.hosp.go.jp

ABSTRACT

Although MV infection causes lymphopenia and degradation of cell-mediated immunity, the mechanisms are poorly known. MV interacts with cellular receptors which mediate virus binding and uptake and are on the surface of PBMC. In this study, apoptosis of MV-infected PBMC in vitro was analyzed. Both PBMC treated with UV-inactivated viruses and those infected with live MV underwent apoptosis. Apoptosis of wild-type MV-infected PBMC was blocked by anti-SLAM and anti-MV hemagglutinin antibodies, respectively. Furthermore, addition of soluble MV hemagglutinin recombinant protein induced apoptosis in PBMC. These data suggest that induction of apoptosis in MV-infected PBMC is triggered by interaction between hemagglutinin protein of MV and receptor, without other viral components. To further determine the mechanisms of apoptosis, caspase activity was analyzed by Western blotting. Wild-type virus Yonekawa strain-induced apoptosis was blocked by pretreatment with pan-caspase inhibitor (Z-VAD-fmk). Intriguingly, the laboratory-adapted Nagahata strain-induced apoptosis was not blocked by Z-VAD-fmk, indicating that there may be different apoptosis pathways which depend on the viral receptors, SLAM and CD46. Both extrinsic and intrinsic apoptotic pathways, including activation of caspase-3, -8 and -9, are involved in Yonekawa strain-induced apoptosis. Taken together, the findings of this study could open up a new avenue for understanding the molecular mechanisms of MV-induced PBMC apoptosis and immunosuppression.

List of Abbreviations: 
Ab

antibody

anti-PARP

anti-poly (ADP-ribose) polymerase

DC

dendritic cells

FITC

fluorescein isothiocyanate

F protein

fusion protein

H protein

hemagglutinin protein

IL

interleukin

mAb

monoclonal antibody

MOI

multiplicity of infection

MV

measles virus

MVH

measles virus hemagglutinin

NH

Nagahata strain

PARP

poly (ADP-ribose) polymerase

PBMC

peripheral blood mononuclear cells

PE

phycoerythrin

PHA

phytohemagglutinin

p.i.

postinfection

PI

propidium iodide

SAP

SLAM-associated protein

SLAM

signaling lymphocyte activation molecule

TCID50

50% tissue culture infectious dose

UV

ultraviolet

UV-MV

UV-inactivated MV

YK

Yonekawa strain

MV causes a contagious, acute, infectious disease characterized by high fever, cough, maculopapular rash and conjunctivitis (1). Although effective live vaccines are available, measles is still responsible for 4% of deaths in children younger than 5 years of age worldwide (2). MV uses the respiratory tract as its entry route and spreads to lymphoid organs, eventually resulting in viremia. After an incubation period of 10 to 14 days, MV causes a profound immunosuppression that permits opportunistic infections, which are a major cause of the high morbidity and mortality rates. MV, which belongs to the genus Morbillivirus in the family Paramyxoviridae, is an enveloped virus with a non-segmented single-stranded negative-sense RNA genome. The envelope of MV contains two glycoproteins, the hemagglutinin (H) protein which plays a major role in interactions with cellular receptors, and the fusion (F) protein which is involved in envelope-cell membrane fusion. CD46 has been identified as a cellular receptor for vaccine and laboratory-adapted strains of MV (3, 4). Another receptor, SLAM (also known as CD150), is utilized by all strains (5). Wild-type strains use SLAM but not CD46.

Despite numerous studies both in vitro and in vivo, the mechanisms underlying the severe immunosuppression and lymphopenia in patients with MV infection are not well understood. Several abnormalities of the immune system have been described, suggesting that multifactorial mechanisms account for this immunosuppression. It has been reported that inhibition of lymphocyte proliferation by MV infection is associated with cell cycle arrest in the G0/G1 phase (6–8) and with disruption of Akt kinase activation (9). MV alters cytokine production, increases in IL-4 and IL-10 and downregulation of IL-12 following MV infection (10–12).

MV-infected cells, including Vero cells, promonocytic cells and PBMC, have been found to induce apoptotic cell death (13–17). In previous studies, apoptosis has been found in PBMC of MV-infected patients (18–20). Furthermore, it has been demonstrated that MV-infected DC produce proliferative unresponsiveness of lymphocytes when co-cultured with T cells, the lymphocytes undergoing Fas-mediated apoptosis and impairment of DC maturation through CD40 signaling (21, 22). These reports raise the possibility that apoptosis in MV-infected cells plays an important role in MV-induced immunosuppression, especially lymphopenia. However, the mechanisms of apoptosis in MV-infected cells are not well understood. In this study, in the case of both UV-inactivated wild-type MV and live MV, we have clearly revealed that apoptosis in PBMC is triggered only by interaction of the H protein of MV and the cellular receptor, SLAM, via a caspase dependent pathway.

MATERIALS AND METHODS

Cells, viruses and reagents

B95a cells, a marmoset B-cell line transformed with Epstein-Barr virus, were grown in RPMI 1640 medium containing 10% FBS (10% FBS RPMI 1640), and Vero cells (derived from African green monkey kidney) were grown in minimal essential medium-10% FBS. Wild-type measles virus, Yonekawa strain (YK), was propagated on B95a cells or Vero/hSLAM cells and a laboratory-adapted type, Nagahata strain (NH), was propagated on Vero cells. These viruses were gifted from Dr. Ihara (Mie National Hospital, Japan). Virus titers were determined by measuring TCID50 on B95a or Vero cells. In some experiments, the virus was exposed to UV light (254 nm) for 30 min from a distance of 10 cm. Viral inactivation was confirmed by the absence of cytopathic effect after inoculation of Vero cell or B95a cell monolayers with the UV-irradiated virus solution.

The mAb specific for SLAM, also called CD150, (clone IPO-3) was purchased from Kamiya Biomedical (Seattle, WA, USA) and the mAb specific for CD46 (clone E4.3) from BD Pharmingen (San Diego, CA, USA). The monoclonal antibody recognizing H protein (clone CV1) was purchased from Chemicon International (Temecula, CA, USA). An anti-MV F monoclonal antibody (A504) was a gift from Dr. J. Schneider-Schaulies (University of Würzburg, Würzburg, Germany). Phycoerythrin (PE)-conjugated anti-CD3 mAb was purchased from Becton Dickinson (San Jose, CA, USA), purified human IgG from Sigma (St Louis, MO, USA), goat anti-mouse IgG Ab from MBL (Nagoya, Japan), the mAb to active caspase-3 from Genlantis (San Diego, CA, USA), antibodies for caspase-8 and -9 from MBL and anti-PARP(Asp214) antibody (clone 19F4) from Cell Signaling Technology (Boston, MA, USA). The broad-spectrum caspase inhibitor (Z-VAD-fmk) was purchased from MBL. MV hemagglutinin mosaic recombinant protein, containing the large hemagglutinin immunodominant regions 1–30, 115–150 and 379–410 amino acids, was purchased from GenWay Biotech (San Diego, CA, USA).

Isolation of PBMC and viral infection

PBMC were isolated from healthy adults using Ficoll-Hypaque (Histopaque 1077; Sigma, St Louis, MO) gradient. Purified PBMC were counted and activated with 5 μg/ml of PHA in 10% FBS RPMI 1640 for 24 hr at 37°C. For experiments using SLAM negative fraction, adherent monocytes were depleted from PBMC by their attachment to tissue culture dishes, and non-adherent cells were collected and used for MV infection without PHA stimulation. Activated PBMC were washed three times with PBS and incubated for 1 hr at 37°C with each virus strain at an MOI of 0.05 to 1. Cells were washed with RPMI 1640 three times to free them of unattached virus and cultured in 10% FBS RPMI 1640 containing 5 μg/ml of PHA for various periods at 37°C.

Evaluation of apoptosis

Detection and quantitation of hypodiploid DNA content of apoptotic cells was measured by flow cytometry using FACScan (Becton Dickinson). Mock- and MV-infected cells were harvested at different time points p.i., then washed three times with PBS. Cells were incubated in 0.5 ml of stain solution containing 0.2% triton-X, 10 μg/ml of PI and 100 μg/ml of ribonuclease A for 30 min at room temperature in the dark and the sub-G1 fraction was quantified by flow cytometry using CellQuest software (Becton Dickinson). In experiments in which apoptosis was blocked, either cells were preincubated with 10 μg/ml of anti-SLAM mAb or 10 μg/ml of anti-CD46 mAb for 1 hr at 37°C, followed by MV infection; or MV was preincubated with a 1:50 dilution of anti-H protein mAb (anti-MVH mAb) or anti-F protein mAb (anti-MVF mAb) for 1 hr at 37°C before being added to the cell cultures.

For quantitative analysis of apoptosis, TUNEL assay based on 3′ end labeling of DNA fragments using FITC-conjugated nucleotides was performed using the FlowTACS FITC kit (Trevigen, Gaithersberg, MD, USA) following the manufacturer's instructions. Apoptotic cells were measured by flow cytometry using CellQuest software. Annexin V staining was also performed using an Apoptosis kit (MBL) following the manufacturer's instructions. Briefly, approximately 1 × 105 cells were collected by centrifugation. The cells were resuspended in 85 μl of binding buffer and incubated with 10 μl of Annexin V–FITC and 5 μl of PI for 15 min at room temperature in the dark. For surface marker analyses, PE-conjugated anti-CD3 mAb was added instead of PI solution. Finally, cells were suspended in 400 μl of binding buffer and then analyzed by flow cytometry using CellQuest software.

Virus-cell binding assay and binding inhibition assay

PBMC (2 × 105 cells) in 25 μl of PBS were incubated with viruses at an MOI of 1 for 1 hr at 4°C, washed with PBS and resuspended in 50 μl of PBS. Cells were incubated with 1:100 dilution of mAb recognizing H protein for 45 min at 4°C. The cells were washed three times with PBS and incubated with FITC-conjugated goat anti-mouse IgG for 30 min at 4°C.

In the case of the virus binding inhibition assay, cells were either pretreated with given concentrations of anti-SLAM mAb or control antibody for 1 hr at 4°C, followed by incubation with MV for 1 hr at 4°C. The cells were then washed and incubated with 100 μg/ml of purified human IgG for 1 hr at 4°C. Bound virus was detected with FITC-conjugated goat anti-human IgG Ab and measured by FACscan using CellQuest software. In addition, the virus solution was pretreated with 1:25 dilution of anti-MVH mAb or control antibody for 1 hr at 4°C. Then the pretreated virus solution was incubated with cells for 1 hr at 4°C and bound virus detected as described above.

Western blot analysis

Western blot analysis was performed on approximately 5 to 10 × 106 cells. PBMC were harvested and washed with PBS and then suspended in lysis buffer (20 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid, pH 7.9 10 mM KCl, 1 mM EDTA, 10% glycerol, 1 mM Na3VO4, 1 mM dithiothreitol, 0.2% NP-40), supplemented with the protease inhibitors phenylmethylsulfonyl fluoride (2 mM) and leupeptin (1 μg/ml) for 15 min at 4°C. The cell lysates were centrifuged at 13 000 rpm for 10 min at 4°C. Equal amounts of proteins for each sample were subjected to 8–12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to membrane by using iBlot PVDF Transfer stacks (Invitrogen, Carlsbad, CA, USA). The membrane was blocked in PBS containing 5% nonfat dry milk for 1 hr at room temperature, and was then incubated in the presence of each primary antibody overnight at 4°C. The membrane was washed and treated with a 1:1000 dilution of horseradish peroxidase-conjugated secondary antibody for 1 hr at room temperature. After extensive washing, the signals were detected using Western Lightening Chemiluminescence Reagent Plus (PerkinElmer Life Science, Boston, MA, USA). Imaging was performed with LAS-3000mini (Fujifilm, Tokyo, Japan).

RESULTS

Treatment with wild-type MV and UV-MV induces apoptosis in PBMC

MV-infected PBMC were examined for the hypodiploid DNA content typical of apoptosis, as described previously (17). The time course of cell cycle analysis in uninfected and wild-type MV YK-infected PBMC is shown in Figure 1a. A small percentage (3.3 to 6.8%) of mock-infected cells displayed sub-G1 DNA content at 48 to 96 hr p.i. In contrast, a significantly higher percentage (10.4 to 35.3%) of MV-infected cells had shifted to sub-G1 phase by the same time points p.i. Apoptosis in PBMC was also monitored by TUNEL assay, which detects DNA fragmentation and thus measures a downstream step in apoptosis (Fig. 1b). The percentages of TUNEL positive cells in MV infection were higher (15.1 to 39.8%) than those in mock treatment (12.4 to 17.4%) at 48 to 96 hr p.i. These data support the proposition that wild-type MV induce apoptosis in PBMC. It has been reported that MV productively replicates in PHA-stimulated PBMC (23). To determine whether viral replication contributes to apoptosis induction we also used UV-MV, because it is known that, while UV-MV can still bind to the receptors of permissive cells, it does not replicate in them. Interestingly, the percentages of TUNEL positive cells after treatment with UV-MV significantly increased (15.1% to 33.7%) at 48 to 96 hr p.i. (Fig. 1b). Further, we evaluated whether apoptosis induction was affected by the amount of UV-MV (Fig. 1c). Cell cycle analysis clearly showed the dose response of UV-MV for induction of apoptosis in PBMC. When PBMC were cultured in the presence of increasing amounts of UV-MV for 96 hr, a dose-dependent increase in percentages of cells in sub-G1 phase was found. Therefore, these results suggest that wild-type MV induce apoptosis in PBMC by binding to cell surface receptors rather than by replication.

Figure 1.

Detection of apoptosis in wild-type MV- or UV-MV-treated peripheral blood mononuclear cells (PBMC). (a) Cell cycle analysis of MV- and mock-infected PBMC. The cells were infected with MV (wild-type, YK) or control solution at MOI of 0.1, collected at different times p.i. and labeled with PI solution. The percentages of sub-G1 cells were measured and calculated by flow cytometry. (b) Detection of PBMC apoptosis by TUNEL assay. The cells were infected with MV (YK), UV-MV (YK) or control solution at MOI of 0.1, and collected at different times p.i. (c) Cell cycle analysis of UV-MV-treated PBMC. PBMC were incubated with UV-MV at different MOI and collected at 96 hr p.i. The cells were labeled with PI solution and sub-G1 fractions were measured and calculated by flow cytometry. Results in (a), (b) and (c) are all expressed as mean percentages of the variable in question ± standard deviation (SD) calculated from three independent experiments.

Wild-type MV-induced PBMC apoptosis is mediated through the interaction of hemagglutinin protein and virus receptor, SLAM

However, there still remained a possibility that soluble factors in MV stock may have contributed to MV-induced apoptosis under our experimental conditions. To clarify whether MV stock prepared in B95a cells contains soluble factors, including cytokines and Epstein-Barr virus, which may contribute to apoptosis, blocking experiments were performed using specific antibodies against wild-type MV receptor, SLAM or H protein of MV. PBMC were preincubated with 10 μg/ml of anti-SLAM mAb for 1 hr at 37°C, followed by wild-type UV-MV (YK) treatment. Cell cycle analysis in the absence of blocking antibody showed sub-G1 cells at 96 hr p.i. were 7.3% and 22.2% in mock- and UV-MV-treated PBMC, respectively (Fig. 2a). When PBMC were pretreated with anti-SLAM mAb, sub-G1 cells were decreased to 8.7% in UV-MV-treated PBMC. Pretreatment with control antibody had no effect on sub-G1 cells treated with UV-MV (data not shown). Blocking with anti-MVH mAb was performed as follows: UV-MV and anti-MVH Ab were first incubated for 1 hr, and then the complex was added to the PBMC culture. The cells were collected at 96 hr after treatment for cell cycle analysis. As shown in Figure 2b, sub-G1 fractions were 8.4% and 25.5% for mock- and UV-MV-treated PBMC, whereas those of UV-MV-treated PBMC were reduced to 13.4% by pretreatment with anti-MVH mAb. These data clearly demonstrate blocking effects of anti-SLAM and anti-MVH mAb on UV-MV-induced apoptosis in PBMC. To determine whether these blocking effects of anti-SLAM and anti-MVH mAb result from interference with binding of MV to PBMC, we performed virus-cell binding assay and binding inhibition assay. As shown in Figure 3, anti-SLAM and anti-MVH mAb clearly blocked virus-cell binding. These results indicate these mAbs abrogate MV binding to cells, resulting in inhibitory effects on UV-MV-induced PBMC apoptosis. In addition, wild-type MV prepared in Vero/hSLAM cells induced apoptosis of PBMC to a similar extent as did MV prepared in B95a cells (data not shown). Taken together, it is unlikely that soluble factors in MV stock are involved in induction of apoptosis. However, it has been described previously that a soluble factor produced by MV-infected T cells can inhibit the proliferation of uninfected T and B cells (24, 25). We examined whether UV-MV treatment might make PBMC secrete soluble factors which induce apoptosis. Supernatants of one day primary cultures of UV-MV-treated PBMC were harvested. Secondary cultures were performed by culturing freshly isolated PBMC with supernatants. After 3days of culture, apoptosis of PBMC was not observed by FACS analysis (data not shown).

Figure 2.

Inhibition of apoptosis in wild type UV-MV-treated PBMC by pretreatment with anti-SLAM or anti-MVH Ab. (a) PBMC were treated with anti-SLAM Ab (IPO-3; 10 μg/ml), 1 hr prior to treatment with UV-MV (YK). The cells were collected at 96 hr after treatment for cell cycle analysis. (b) UV-MV (YK) was pretreated with anti-MVH Ab (×50) for 1 hr before addition to cell cultures. The cells were collected at 96 hr after treatment for cell cycle analysis. Results are expressed at the mean percentages of sub-G1 cells ± SD calculated from three independent experiments.

Figure 3.

Inhibition of wild-type MV binding to human PBMC by anti-SLAM or anti-MVH Ab. (a) PBMC were preincubated for 1 hr in the presence or absence of anti-SLAM Ab (IPO-3; 50 μg/ml) at 4°C. The cells were then infected with wild-type MV (YK) or mock at an MOI of 1 and bound was virus detected by flow cytometry. (b) MV (YK) was treated with or without anti-MVH Ab (×25) for 1 hr at 4°C. The pretreated MV was added to PBMC at an MOI of 1 and bound virus was detected by flow cytometry. FL1-H, fluorescence 1-histogram.

Following UV-MV binding to SLAM on cells, F protein interacts with the cell surface and the virus particle can then be incorporated into cells. Subsequently, viral proteins such as nucleoprotein or other proteins and/or viral RNA that are present in the particle may act to cause the cells to undergo apoptosis. To exclude this possibility, soluble recombinant H protein was used to induce apoptosis of PBMC. Interestingly, FACS analysis using Annexin V staining clearly showed that H protein-treated PBMC underwent apoptosis, suggesting the apoptosis could be induced by H protein binding to cells alone (Fig. 4). Taken together, these results strongly confirm that apoptosis induction by UV-MV treatment is mediated through the interaction of H protein of MV with the virus receptor, SLAM.

Figure 4.

Apoptosis in soluble rMVH -treated PBMC. PBMC were treated with mock or rMVH (10 μg/ml) and collected at 24 hr after treatment. The percentage of AnnexinV+/PI- cells were measured and calculated by flow cytometry. Data shown are representative of three experiments. rMVH, recombinant MV hemagglutinin protein.

Wild-type MV induces apoptosis in SLAM positive lymphocytes, but not in SLAM negative ones

To further confirm involvement of SLAM-H protein interaction in PBMC apoptosis, we analyzed wild-type MV- or UV-MV-induced apoptosis using SLAM-positive or -negative cell fractions. It is well known SLAM molecules are expressed on monocytes and activated T and B cells. Monocyte-depleted PBMC were prepared and cultured with or without PHA. FACS analysis clearly showed that PHA stimulation induced SLAM expression in monocyte-depleted PBMC (Fig. 5a). Using these cells, wild type MV- or UV-MV-induced apoptosis was examined by Annexin V staining combined with lymphocyte surface marker analysis. As shown in Figure 5b, wild-type UV-MV clearly induced apoptosis in both CD3 positive and negative populations of the SLAM positive fraction. However, apoptosis induced by wild-type MV or UV-MV infection was not detected in the SLAM negative fraction (Fig. 5c and data not shown). These results indicate that H-protein of wild-type MV induces apoptosis in both activated T and B cells through its interaction with SLAM.

Figure 5.

Wild-type MV-induced PBMC apoptosis in SLAM positive and negative fractions. (a) Monocyte-depleted PBMC were incubated in the presence (broken and bold lines) or absence (solid line) of PHA for 24 hr. The cells were then stained with anti-SLAM (solid and bold lines) or control (broken line) mAb, and the degree of expression of SLAM analyzed by flow cytometry. (b) Annexin V staining combined with a marker analysis for SLAM positive cells after 24 hr treatment with UV-MV (YK) or control solution. The percentages of cells present in each fraction are indicated. (c) Annexin V/PI staining for SLAM negative cells after 24 hr treatment with MV (YK) or control solution. The percentages of cells present in each fraction are indicated.

Caspase activation is involved in PBMC apoptosis induced by wild-type, but not by laboratory-adapted, MV

To further investigate the mechanisms of MV-induced apoptosis in PBMC, we analyzed caspase activation. First, we examined the effect of the broad-spectrum caspase inhibitor Z-VAD-fmk on MV-induced apoptosis. Cells were pretreated with increasing concentrations of Z-VAD-fmk for 1 hr at 4°C and were then infected with MV. After incubation for 72 hr at 37°C, the cells were harvested and cell cycle analysis performed using flow cytometry. As shown in Figure 6a, apoptosis was inhibited by the caspase inhibitor in a dose dependent manner, and completely suppressed at a concentration of 500 μM. Next, we performed the same experiment using UV-MV for induction of apoptosis. As shown in Figure 6b, in the absence of the pan-caspase inhibitor the mean percentages of sub-G1 cells were 9.8% and 26.9% for mock- and UV-MV-treated PBMC, respectively. In contrast, the mean percentage of sub-G1 cells decreased to 10.8% for UV-MV-treated PBMC in the presence of the pan-caspase inhibitor. These data suggest that caspase activation is involved in PBMC apoptosis induced by H protein-SLAM interaction. To further define caspase activation in MV-infected PBMC, we examined the activation of caspase-3, which is the main effector caspase, by Western blotting. As shown Figure 7a, the active form of caspase-3 was detected at 48 hr p.i. and had increased by 72 hr p.i. We also examined the activation of caspase-3 in the presence of the pan-caspase inhibitor. Figure 7b shows that caspase-3 activation in MV (YK)-infected PBMC was clearly inhibited by Z-VAD-fmk. The results suggest that the induction of apoptosis by wild-type MV depends on caspase-3 activation.

Figure 6.

Involvement of caspases in wild-type MV- and UV-MV-induced apoptosis of PBMC. (a) Dose-dependent effect of pan-caspase inhibitor on MV-induced apoptosis. Cells were pretreated with increasing concentrations of Z-VAD-fmk before MV infection and apoptotic cells were analyzed at 72 hr p.i. by flow cytometry. (b) Effect of pan-caspase inhibitor on UV-MV-induced apoptosis. Cells were preincubated with Z-VAD-fmk (500 μM) for 1 hr 4°C. The cells were treated with UV-MV (YK) at an MOI of 0.25 and apoptotic cells were analyzed at 72 hr p.i. by flow cytometry. Results are expressed as the mean percentages of sub-G1 cells ± SD calculated from three independent experiments.

Figure 7.

Caspase-3 activation in PBMC apoptosis induced by wild-type MV. (a) PBMC were infected with MV (YK) at an MOI of 0.2 for 48 hr or 72 hr and were harvested for detection of cleavage of procaspase-3 by Western blot. (b) Cells were pretreated with Z-VAD-fmk (500 μM) for 1 hr at 4°C. The cells were then infected with MV (YK) at an MOI of 0.2 and harvested at 72 hr p.i. for detection of cleavage of procaspase-3 by Western blot.

We next asked whether the laboratory-adapted type, NH, also induces PBMC apoptosis. Cell cycle analysis and Annexin V staining showed apoptosis of PBMC with MV (NH) infection as well as UV-MV (NH) (Fig. 8a and b). Intriguingly, the pan-caspase inhibitor failed to block apoptosis of PBMC induced by MV (NH) or UV-MV (NH) (Fig. 8a and b), although the amount of active form of caspase-3 was decreased to a similar degree as occurred with the mock-infected sample (Fig. 8c), suggesting that laboratory-adapted type induced-PBMC apoptosis is independent of caspase activation pathways. Laboratory-adapted strains of MV use either SLAM or CD46 as a receptor. We therefore asked whether interaction of H protein and CD46 was involved in apoptosis of PBMC induced by UV-MV (NH). Blocking experiments with anti-SLAM or anti-CD46 mAb were performed. Although anti-SLAM mAb only partially reduced the percentage of apoptotic cells, anti-CD46 mAb completely blocked apoptosis of PBMC induced by UV-MV (NH), suggesting that UV-MV (NH) mainly utilize CD46 as a cellular receptor (Fig. 8d). Taken together, these data indicate that the interaction of H protein and CD46 can trigger PBMC apoptosis via caspase-independent pathway(s).

Figure 8.

Effects of pancaspase inhibitor or blocking antibodies on PBMC apoptosis and caspase-3 activation induced by laboratory-adapted strain MV (NH). (a) Cell cycle analysis of MV- and mock-treated PBMC. PBMC were pretreated with Z-VAD-fmk (1 mM) for 1 hr at 4°C. The cells were treated with MV (NH) or control solution at MOI of 0.25, collected at 72 hr p.i. and labeled with PI solution. The percentages of sub-G1 cells were measured and calculated by flow cytometry. (b) Annexin V/PI staining of MV- and mock-treated PBMC. PBMC were pretreated with 1 mM of Z-VAD-fmk for 1 hr at 4°C. The cells were treated with MV (NH) or control solution at MOI of 0.25, collected at 24 hr p.i. and stained with Annexin V-FITC/PI solution. The percentages of Annexin V+/PI- cells were measured and calculated by flow cytometry. (c) Cells were pretreated with 500 μM of Z-VAD-fmk for 1 hr at 4°C. The cells were infected with MV (NH) at an MOI of 0.2 and harvested at 72 hr p.i. for detection of cleavage of procaspase-3 by Western blot. (d) PBMC were treated with anti-SLAM mAb (IPO-3; 10 μg/ml) or anti-CD46 mAb (E4.3; 10 μg/ml), 1 hr prior to treatment with UV-MV (NH) at an MOI of 0.25. The cells were collected at 24 hr after treatment for AnnexinV analysis. Results in (a), (b) and (d) are expressed as the mean percentages of the variables in question ± SD calculated from three independent experiments.

Activation of caspase-8 and -9 in wild-type MV-infected PBMC

Since caspase-3 activation was involved in wild-type MV-induced PBMC apoptosis, we further investigated the upstream signals of the caspase activation pathway. First, activation of caspases-8 and -9 was analyzed by Western blotting (Fig. 9). Compared with mock-infected cells, increased amounts of cleaved caspase-8 and -9 were observed in MV-infected cells at 96 hr p.i. These results suggest that both caspase-8 and caspase-9 activation pathways are involved in wild-type MV-induced apoptosis. Finally, we examined cleavage of PARP, which is a well-known substrate for caspase-3 in apoptotic events. As shown in Figure 9, a greater amount of cleaved PARP was detected in MV-infected than in the mock-infected cells, reflecting caspase-3 activation.

Figure 9.

Activation of caspase-8 and -9 in MV-infected PBMC. Cells were infected with MV (YK) at an MOI of 0.2 and harvested at 96 hr p.i. for detection of cleavage of procaspase-8, -9, and PARP by Western blot.

DISCUSSION

The cell death mechanisms in MV infection are still unclear in spite of the fact that lymphopenia during natural measles has been clinically described (19, 26). Several hypotheses have been proposed to explain MV-induced cell death. It has been reported that apoptotic cells can be detected in PBMC of MV-infected patients (19) and MV-infected PBMC have shown apoptotic cell death in vitro (17). These studies have suggested the possibility that apoptosis is one of major mechanisms for lymphopenia in MV-infected patients. In this study, we investigated the molecular mechanisms of PBMC apoptosis in MV infection in vitro. Cell cycle analysis and TUNEL assay clearly showed that wild-type MV induces PBMC apoptosis. Interestingly, apoptosis was observed in UV-MV treated cells as well, suggesting that MV may induce apoptosis purely by binding to the cells. MV has two envelope glycoproteins, H and F protein. It has been shown that down regulation of cellular receptors, such as CD46 and SLAM, is caused by expression of the MV H protein alone (27, 28). In another study, it has been reported that cell-surface contact of MV glycoproteins can induce inhibition of cell proliferation in vitro (29). These reports suggest the possibility that interaction of H protein and cellular receptor may play an important role in regulating cellular functions such as proliferation and apoptosis. To determine whether a glycoprotein-cellular receptor interaction is involved in inducing apoptosis of cells, we performed blocking assay of UV-MV-induced apoptosis by using anti-SLAM, anti-MVH or anti-MVF mAb. Pretreatment with anti-SLAM and anti-MVH mAb clearly inhibited UV-MV-induced PBMC apoptosis. On the contrary, a blocking experiment using anti-MVF mAb showed no effect on apoptosis (data not shown). Furthermore, soluble MV H protein treatment could induce PBMC apoptosis. These results indicate that apoptosis of PBMC can be induced only by virus binding to cells via interaction of H protein with the cellular receptor SLAM.

MV-infected lymphocytes were revealed to be only a small proportion of the apoptotic cells in measles patients (18–20), proposing the possibility of a bystander effect. Several hypotheses have been proposed to explain this bystander effect. MV-infected monocytes induce apoptosis by interaction with uninfected T cells via cell surface molecules (17). A soluble factor that is produced by MV-infected T cells, but not HeLa or Vero cells, inhibits proliferation of uninfected T and B cells (24, 25), suggesting the possibility of involvement of unidentified soluble factor(s) in apoptosis induction. MV nucleoprotein, which is released into the extracellular compartment, interacts with its receptor, inducing apoptosis of uninfected cells (30). However, the present data exclude the possibility that soluble factors or viral components other than H protein are involved in apoptosis induction. Together, we propose the following explanation for bystander effect of MV infection: uninfected PBMC could undergo apoptosis purely through interaction of cell surface SLAM with H protein of MV particles, or H protein expressed on the surface of MV-infected neighboring cells. To confirm this hypothesis, further experiments using H protein expressing cell lines and PBMC are underway.

Several studies on a number of viruses have reported that apoptosis of MV-infected cells is also induced through a caspase-dependent pathway (31–34). One investigation revealed that strong activation of caspase-3 and -6 occurs in MV-infected PBMC (17). In another study, caspase-3 activation was demonstrated in an MV-infected human melanoma cell line (30). Our results also indicate that wild-type MV-induced apoptosis in PBMC is dependent on caspase activation. Western blotting analysis revealed caspase-3 activation in MV-infected cells. Furthermore, induction of apoptosis was suppressed by treatment with a broad-spectrum caspase inhibitor (Z-VAD-fmk). Caspase-3 activation was inhibited in the presence of this caspase inhibitor.

Caspases are critical mediators of classical apoptosis; upstream initiator caspases (caspases-8 and -9) and downstream effector caspases (such as caspase-3, -6 or -7). Caspases-8 and -9 are the first activated in response to apoptotic stimuli and are responsible for processing and activation of effector caspases. There are two caspase-dependent pathways (extrinsic and intrinsic), for inducing apoptosis. Caspase-8 is involved in the extrinsic pathway and caspase-9 in the intrinsic one. Caspase-3 is an effector caspase common to both pathways. To determine the pathway by which wild-type MV (YK) infection induces PBMC apoptosis, activation of initiator and effector caspases was analyzed. Our data showed wild-type MV infection induces activation of both caspases-8 and -9. Activation of effector caspase-3 was also detected, followed by cleavage of the death substrate PARP. These results suggest that both the extrinsic and intrinsic pathways are involved in wild-type MV-induced PBMC apoptosis.

Interestingly, when laboratory-adapted type (NH) MV was used for apoptosis induction, different phenomena were observed. Although inhibition of caspase-3 activation was demonstrated in the presence of Z-VAD-fmk, induction of apoptosis was not suppressed. These results suggest that the laboratory-adapted type can induce apoptosis without caspase activation. It has been reported that there are many pathways by which apoptosis is induced (35). MV-infected DCs apoptosis is mediated via the Fas pathway and tumor necrosis factor-related apoptosis-inducing ligand system (19, 21, 36). Although there is not yet enough data to explain the different mechanisms by which laboratory-adapted type or wild-type induce apoptosis, we speculate that apoptosis induction by the laboratory-adapted type may be mediated by caspase independent pathways, involving apoptosis-inducing factor and endonuclease G. Also, there is a possibility that intracellular signaling of CD46 may be involved in apoptosis induction because the laboratory-adapted type can utilize both CD46 and SLAM as cellular receptors. Our data imply that interaction of CD46 with MV H protein can trigger apoptosis of PBMC, but more work is required to dissect the downstream signal transduction.

The SLAM family (SLAM, 2B4, NTB-A, Ly9, CD48) may play an important role in regulating both innate and adaptive immune responses. The functions of SLAM (CD150) signaling are IL-4 secretion by T cells, and IL-12 and tumor necrosis factor-α production by macrophages (37–39). The intracellular signaling pathway is mediated by SAP, following recruitment of the protein tyrosine kinase Fyn and facilitation of downstream signaling (39, 40). However, the contribution of the intracellular signaling pathway through the cell surface receptor SLAM in response to MV infection is still unclear. Recent studies have shown that loss of SAP expression in normal T cells results in defective activation-induced cell death and reduced amounts of activated caspase-3 and -9 (41, 42). Together with our data, there is a possibility that MV H protein binding to SLAM may affect intracellular SLAM-SAP signaling and caspase-dependent apoptotic pathways.

In conclusion, our results demonstrate that wild-type measles virus, Yonekawa strain, can induce apoptosis of PBMC in vitro. Moreover the apoptosis is mediated only by interaction between the viral envelope H protein and cellar receptor SLAM via a caspase-dependent pathway. The apoptosis mechanism seems differ according to receptor selection. Our study could open up a new avenue for understanding the molecular mechanism of MV-induced PBMC apoptosis and immunosuppression.

ACKNOWLEDGEMENT

We thank Dr. J. Schneider-Schaulies for providing the anti-MV F mAb.

Ancillary