The SARS-Coronavirus Membrane protein induces apoptosis through modulating the Akt survival pathway

Construction of mammalian expression vector

Full length ORF of the SARS-CoV Membrane locus was PCR amplified from viral cDNA template (CUHK-Su10 SARS-CoV isolate; GenBank Accession No. AY282752) and subcloned into pcDNA3.1(+) vector using EcoRI and XbaI enzymes to generate the pcDNA3.1(+)-Membrane construct.

Mammalian cell culture and transient transfection

Human embryonic kidney cell line HEK293T was maintained at 37 °C in Dulbecco’s modified Eagle’s medium (DMEM, Invitrogen) supplemented with 10% heat-inactivated fetal bovine serum (Gibco-BRL), streptomycin (100 g/ml), and penicillin (100 U/ml). Cells were seeded onto 6- or 24-well plates 24 h prior to transfection. Four (6-well plate) or 0.8 (24-well plate) μg of plasmid DNA were used for transient transfection with Lipofectamine 2000 reagent (Invitrogen). Cells were collected for the subsequent analyses 48 h post-transfection.

Semi-quantitative RT-PCR analysis

Total RNA was prepared as previously described [34]. Primers used were GAPDH-F: 5′ ACC ACA GTC CAT GCC ATC AC 3′; GAPDH-R: 5′ TCC ACC ACC CTG TTG CTG TA 3′; M-F: 5′ ATT ACC GTT GAG GAG CTT AAA CAA 3′; and M-R: 5′ CAA TGA CAA GTT CAC TTT CC 3′.

Immunofluorescence staining of HEK293T cells

Cells were seeded onto poly-d-lysine-coated coverslips at a density of 2 × 10⁴  cells/coverslip. After transfection, cells were fixed with 3.7% formaldehyde for 15 min and then permeabilized by 1% Triton X-100 for 5 min. After blocking with 1% goat serum for 30 min, cells were incubated with rabbit anti-SARS-virus-PUPM antibody (C-term) (1:100; Abgent) alone or in combination with mouse anti-native cytochrome c clone 6H2.B2 (1:100; Pharmingen) at 4 °C overnight, and followed by secondary antibody incubation with goat anti-rabbit IgG (H+L)−FITC (1:250; Zymed) alone or in combination with goat anti-mouse IgG (H+L)−TRITC (1:250; Zymed) at room temperature for 1 h. Golgi body was stained by BODIPY^® TR ceramide complexed to BSA (5 μM, Molecular Probes) and cell nuclei were labeled with Hoechst 33342 (trihydrochloride trihydrate) (5 μM, Molecular Probes) at room temperature for 10 min. Cell permeable synthetic caspase inhibitors (Merck) Z-DQMD-fmk (caspase-3 inhibitor V), z-IETD-fmk (caspase-8 inhibitor II) and z-LEHD-fmk (caspase-9 inhibitor I) were dissolved in DMSO. HEK293 cells were treated with caspase inhibitors (50 μM in 1% DMSO) 24 h post-transfection, and were further incubated for another 24 h. Fluorescence images were captured using an Olympus BX51 upright fluorescence microscope or a Leica NT confocal microscope.

Drosophila genetics

Fly strains were grown at 29 °C on standard cornmeal medium supplemented with dry yeast. PDK-1 ¹ and PDK-1 ² lines were kind gifts of Jongkyeong Chung [35]; gmr-GAL4, UAS-DIAP1, UAS-P35, P[GMR-Akt1.Exel]2, Exelixis, and DrosDel lines were obtained from Bloomington Drosophila Stock Center; and the dc3 ^EP2305 and PDK-1 ^EP837 lines were obtained from Szeged Drosophila Stock Centre.

Generation of transgenic fly lines

Full length ORF of the SARS-CoV Membrane gene was PCR amplified from viral cDNA template and subcloned into pUAST vector [36] using EcoRI and XbaI enzymes to generate pUAST-Membrane plasmid. Standard microinjection technique was employed to generate transgenic lines. A total of 10 UAS-Membrane transgenic fly lines were generated. When crossed to gmr-GAL4, all lines showed dominant rough-eye phenotype. A transgenic fly line J3 carrying an UAS-Membrane insert on the third chromosome was used in this study.

Scanning electron microscopy of adult fly eyes

In brief, fly heads were fixed in 2.5% glutaraldehyde (EM grade, Electron Microscopy Sciences) in phosphate buffer (pH 7.4) for 4 h, then post-fixed with 1% osmium tetroxide (Electron Microscopy Sciences), dehydrated to 100% ethanol and critical-point dried with liquid CO₂. Gold-palladium-coated specimens were examined with a Jeol JSM-6301FE microscope operated at 5 kV [37].

Acridine orange and immunofluorescence staining of larval eye discs

Acridine orange staining of third-instar larval eye discs was performed as previously described [38]. Quantification of acridine orange-positive cell was performed by Image-Pro Plus 5.1 (Media Cybernetics). Immunofluorescence staining was performed as previously described [16]. Rabbit anti-SARS-virus-PUPM antibody (C-term) (1:100; Abgent) and goat anti-rabbit IgG (H+L)–FITC secondary antibody (1:250; Zymed) were used. Propidium iodide (10 μg/ml, Molecular Probes) was used to label cell nuclei. Images were captured using an Olympus BX51 upright fluorescence microscope or a Leica NT confocal microscope.

Western blot analysis

For fly protein sample preparation, 16 adult fly heads of appropriate genotypes were homogenized in 75 μl of 6× SDS sample buffer. SDS–PAGE separation and Western blotting were performed as previously described [16]. Primary antibodies used include phospho-Drosophila Akt (Ser505) antibody (1:1000; Cell Signaling), total Akt antibody (1:1000; Cell Signaling) and anti-β-tubulin E7 (1:2000; Developmental Studies Hybridoma Bank, under the auspices of the NICHD and maintained by The University of Iowa, Department of Biological Sciences, Iowa City, IA 52242, USA). Secondary antibodies used were goat anti-rabbit IgG-HRP antibody (1:2000; Cell Signaling) and goat anti-mouse IgG-HRP antibody (1:2000; Zymed).

Statistical analyses

Statistical analyses were performed using Student’s t test. Data were presented as means+SEM. p values < 0.05 were considered statistically significant.

Expression and subcellular localization of the SARS-CoV Membrane protein in HEK293T cells

To study the function of the SARS-CoV Membrane protein, the M ORF of the CUHK-Su10 SARS-CoV isolate was subcloned into a mammalian expression vector pcDNA3.1(+). By RT-PCR, mRNA expression of the M gene was detected 48 h post-transfection (Fig. 1 A). Immunofluorescence was performed to determine the subcellular localization of the M protein. We observed that M protein displayed a punctate localization in the cytoplasm (Fig. 1E) and was partially co-localized with the Golgi body (Fig. 1G, inset).

Induction of apoptosis by SARS-CoV Membrane protein in HEK293T cells

When the cell nuclei of M-transfected HEK293 cells were stained with Hoechst dye, we found that ∼56% of the cell nuclei were condensed as indicated by size reduction (Fig. 2 I and M) whereas only around 10% of the control cells had condensed nuclei (Fig. 2B, E and M). Since nuclear condensation is a hallmark feature of apoptosis, we reasoned that M over-expression induced apoptosis in HEK293T cells. To confirm this, we treated untransfected HEK293T cells with staurosporine, a commonly used apoptosis inducer, and nuclear condensation was also observed (Fig. 2K and M).

Since cytochrome c release from the mitochondria represents one form of apoptotic cell death, we performed cytochrome c immunofluorescence staining to examine whether mitochondrial release of cytochrome c was detected in M-transfected cells. Unlike the untransfected control (Fig. 3 A), we found that cells either transfected with M (Fig. 3C) or treated with staurosporine (Fig. 3E) showed extensive mis-localization of cytochrome c protein which is indicative of mitochondrial release of cytochrome c protein [39]. The appearance of M-transfected (Fig. 3D) and staurosporine-treated cells (Fig. 3F) were also morphologically distinct from the untransfected control (Fig. 3B). We further showed that M-induced nuclear condensation in HEK293T cells could be suppressed by caspase-3, -8 and -9 inhibitors (Fig. 3G).

Over-expression of SARS-CoV Membrane protein in Drosophila

We previously used a transgenic Drosophila model [16] to demonstrate the pro-apoptotic role of the SARS-CoV 3a protein [17]. In this study, we established UAS-M transgenic fly lines, and over-expressed the SARS-CoV M protein in the Drosophila eye using GAL4/UAS transgene expression system [36]. Over-expression of M caused a rough-eye phenotype in adult flies, as shown by the loss of regularity of external eye morphology (Fig. 4 C), while the over-expression of the EGFP control protein (Fig. 4B) did not show any obvious dominant eye malformation phenotype. To determine the subcellular localization of M in Drosophila, we performed immunofluorescence staining in third instar larval eye discs (Fig. 4D and F). Similar to mammalian cells (Fig. 1E), M displayed a punctate cytoplasmic localization in fly cells (Fig. 4F), whereas the EGFP control protein showed homogeneous intracellular localization (Fig. 4E).

Over-expression of SARS-CoV Membrane protein induced apoptosis in Drosophila

Since over-expression of many genes would result in rough-eye phenotype in Drosophila, we therefore performed standard acridine orange (AO) staining [38] in third instar larval imaginal eye discs to examine the pro-apoptotic roles of M in vivo. Acridine orange is a dye that specifically stains apoptotic cells, and we found that the number of AO-positive cells in M-over-expressing eye discs (Fig. 5 G and K) was significantly higher than that of the control (Fig. 5F and K). To further confirm the pro-apoptotic role of M, we co-expressed two anti-apoptotic factors, P35 (a viral caspase inhibitor) and the Drosophila inhibitor of apoptosis 1 (DIAP1) protein, independently with M. As expected, both the rough-eye phenotype (Fig. 5B–D) and elevated number of AO-positive cells (Fig. 5G–I and K) were mostly suppressed back to the control level, respectively.

We next questioned whether M-induced apoptosis would act through the route of cytochrome c in flies as in HEK293T cells (Fig. 3C). It has been reported that over-expression of the apo-form of cytochrome c (without heme group) can block apoptosis [40]. We made use of an EP insert line dc3 ^EP2305 to over-express the endogenous apo-cytochrome c gene dc3 [16] and asked if dc3 over-expression could suppress M-induced apoptosis in Drosophila. Both the rough-eye phenotype (Fig. 5B) and elevated number of AO-positive apoptotic cells in eye discs (Fig. 5G and K) were largely suppressed upon co-expression of M with apo-cytochrome c (Fig. 5 E, J and K).

Expression of phosphoinositide-dependent kinase-1 dominantly suppressed SARS-CoV Membrane-induced apoptosis

We performed a forward genetic modifier screen in Drosophila in order to gain further insights into the mechanistic details of M action in vivo. By analyzing the modifying effects of overlapping genomic deletion lines, from both the Exelixis [41] and DrosDel [42] collections, on the M-induced rough-eye phenotype (Fig. 4C); 30 modifying genomic regions were identified (C.M. Chan, W.M. Chan, C.S. Chan, C.W. Ma, K.W. Ching, L.L. Ho, K.M. Lau, H.Y. Chan, unpublished observations). Detailed characterization of one of these modifying regions (61B1-C1) is described here. The 61B1-C1 genomic region is uncovered by a deletion line Df(3L)ED201 [42]. We further found that one of the EP insert lines in the 61B1-C1 region, PDK-1 ^EP837, dominantly suppressed the M-induced rough-eye phenotype (Fig. 6 B and C). The PDK-1 ^EP837 insert is located at the 5′ upstream region of the phosphoinositide-dependent kinase-1 (PDK-1) gene transcription unit, and had previously been shown to cause over-expression of endogenous PDK-1 [35].

Since PDK-1 over-expression was able to rescue M-induced rough-eye phenotype, we investigated if PDK-1 suppressed the phenotype through inhibiting apoptosis. We showed that over-expression of SARS-CoV M induced apoptosis in vivo as indicated by the increased number of AO-positive cells (Figs. 5G and 6E). When M was co-expressed with PDK-1, the number of AO-positive cells was largely reduced (Fig. 6F and G).

Down-regulation of Akt phosphorylation by SARS-CoV Membrane protein in vivo

Both Akt and PDK-1 kinases play a central role in the cell survival pathway [43]. Since the Akt pathway can modulate apoptosis [44], we therefore investigated whether over-expression of Drosophila Akt1 could suppress M-induced apoptosis. Unlike PDK-1, the M-induced rough-eye phenotype was not suppressed by Akt1 over-expression (data not shown). Since the phosphorylation state of Akt is an essential cell survival indicator, we then argued if M-over-expression would affect Akt1 phosphorylation. Using phospho-specific Akt1 (Ser505) antibodies, we found that over-expression of M down-regulated the phosphorylation states of Akt1 (Fig. 7 A and B) without affecting total Akt1 protein levels (Fig. 7A and C).