key: cord-333754-copxoyqu
authors: Ma, Hsin-Chieh; Fang, Chiu-Ping; Hsieh, Yi-Ching; Chen, Shih-Chi; Li, Hui-Chun; Lo, Shih-Yen
title: Expression and membrane integration of SARS-CoV M protein
date: 2008-04-09
journal: J Biomed Sci
DOI: 10.1007/s11373-008-9235-1
sha: 
doc_id: 333754
cord_uid: copxoyqu

SARS-CoV M gene fragment was cloned and expressed as a recombinant protein fused with a V5 tag at the C-terminus in Vero E6 cells. In addition to un-glycosylated and glycosylated proteins, one product with smaller size initiated in-frame from the third Met residues probably through ribosomal re-initiation was also detected. Translation initiated in-frame from the third Met is unusual since the sequence around the first Met of SARS-CoV M protein contains the optimal consensus Kozak sequence. The function of this smaller translated product awaits further investigation. Similar to other N-glycosylated proteins, glycosylation of SARS-CoV M protein was occurred co-translationally in the presence of microsomes. The SARS-CoV M protein is predicted as a triple-spanning membrane protein lack of a conventional signal peptide. The second and third trans-membrane regions (a.a. 46–68 and 78–100) are predicted to be the primary type helices, which will be able to penetrate into membrane by themselves, while the first trans-membrane region (a.a. 14–36) is predicted to be the secondary type helix, which is considered to be stabilized by the interaction with other trans-membrane segments. As expected, the second and third trans-membrane regions were able to insert a cytoplasmic protein into the endoplasmic reticulum membrane more efficiently than the first one. These results should be important for the study of SARS-CoV morphogenesis. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1007/s11373-008-9235-1) contains supplementary material, which is available to authorized users.

Severe acute respiratory syndrome (SARS), a new infectious disease typically associated with fever, shortness of breath, cough, and pneumonia, first emerged in southern China in November, 2002. Within months of outbreak, SARS had spread globally, affecting over 8,000 patients in 29 countries with 774 fatalities [1] . The aetiology of SARS is associated with a newly discovered coronavirus, SARSassociated coronavirus (SARS-CoV) [2] . Subsequent studies have indicated that the SARS coronavirus is of animal origin [3] , and its precursor is still present in animal populations within the region. The live-animal markets in southern China may have provided the animal-human interphase that allowed this precursor virus to adopt to human-human transmission. Therefore, even in a situation of no new infections, SARS remains a major health hazard, as new epidemics may arise. Further basic and clinical research is required to control the disease.

By May 9, 2003 , 14 genomes of SARS coronavirus had been sequenced [4] [5] [6] [7] [8] . SARS coronavirus is phylogenetically distinct, and only distantly related to the other coronavirus clades [9, 10] . Coronaviruses are exceptionally large RNA viruses and employ complex regulatory mechanisms to express their genomes [11] . The genome structure, gene expression pattern and protein profiles of SARS-CoV are similar to those of other coronaviruses. Nine SARS-CoV specific mRNAs were synthesized in virus-infected cells [12] . These RNAs were predicted to encode two large replicative polyprotein (pp1a and pp1ab), four structural proteins (spike, membrane, envelope, and nucleocapsid proteins), and other auxiliary proteins. In the cases of other coronaviruses, the four structural proteins (S, M, E, and N) play roles in virion morphogenesis [13] . N binds to viral RNA to form nucleocapsid. Co-expression of M and E proteins together can form virus-like particles [14] . Interactions between the M and E proteins and nucleocapsids result in virus budding through cellular membrane. Through the interaction with M protein, S protein is incorporated into the viral envelope and the mature virions are released from the cells. It has also been demonstrated that virus-like particles of SARS-CoV could be formed by expressing M and E proteins in insect cells [15] . Therefore, the M protein plays a crucial role in coronavirus assembly. Characterization of SARS-CoV M protein will shed light on the studies of the morphogenesis of SARS-CoV. Here, we report the expression and membrane integration of recombinant SARS-CoV M protein in Vero E6 cells.

The construction of the plasmid expressing full-length membrane protein plus V5 and His tag encoded from pcDNA3.1/V5-His A vector sequence (M-V5-His) was described previously [16] .

To mutate amino acid 4 of SARS-CoV M protein from Asn to Asp, PCR primers (5 0 CGGAATTCATGGCAGAC GACGGTACTATTACCG3 0 and 5 0 TGCTCTAGACTGTA CTAGCAAAGCAAT3 0 ) were used to amplify the gene fragment. After PCR, the DNA fragment was digested by restriction enzymes (EcoRI/XbaI) and cloned into pcDNA3.1-V5-His A (linearized by EcoRI/XbaI) expression vector. This expression plasmid encodes full-length SARS-CoV M protein with mutation in amino acid 4 from Asn to Asp.

To mutate amino acid 83 of SARS-CoV M protein from Met to Leu, PCR primers (5 0 CCGGAATTCATGGCAGAC AACGGTACTA3 0 and 5 0 AATACAAGCCAGTGCAATC GCAATC3 0 ) were used to amplify the 5 0 -end of the membrane gene fragment while PCR primers (5 0 GCGA TTGCACTGGCTTGTATTGTAG3 0 and 5 0 TGCTCTAGA CTGTACTAGCAAAGCAAT3 0 ) were used to amplify the 3 0 -end fragment. These two DNA fragments were linked together by PCR using primers (5 0 CCGGAATTCATG GCAGACAACGGTACTA3 0 and 5 0 TGCTCTAGACTGT ACTAGCAAAGCAAT3 0 ). After PCR, the DNA fragment was digested by restriction enzymes (EcoRI/XbaI) and cloned into pcDNA3.1-V5-His A (linearized by EcoRI/ XbaI) expression vector. This expression plasmid encodes full-length SARS-CoV M protein with mutation in amino acid 83 from Met to Leu. Similar approach was used to mutate amino acid 32 of SARS-CoV M protein from Met to Leu.

To clone entire SARS-CoV M gene fragment including 5 0 -untranslated region (starting from nt. 26361 of GI: 30027610), PCR primers (5 0 CCGGAATTCTGGAACTT TAACATT3 0 and 5 0 TGCTCTAGACTGTACTAGCAAAG CAAT3 0 ) were used to amplify the gene fragment. After PCR, the DNA fragment was digested by restriction enzymes (EcoRI/XbaI) and cloned into pcDNA3.1-V5-His A (linearized by EcoRI/XbaI) expression vector. This expression plasmid contains the full-length SARS-CoV M coding sequence and 22 nucleotides of the 5 0 -untranslated region.

To clone the DNA fragment with the first 115 amino acids of HCV core protein into pcDNA3.1-V5-His A, the expression plasmid (M-C-V5-His) encoding a fusion protein with full-length SARS-CoV M protein and the first 115 amino acids of HCV core protein [16] was digested with BamHI to remove the DNA fragment with full-length SARS-CoV M protein and self-ligated.

To clone the DNA fragment with the first 115 amino acids of HCV core protein and the first 50 amino acids of SARS-CoV M protein, the expression plasmid (C-M-V5-His) encoding a fusion protein with the first 115 amino acids of HCV core protein and full-length SARS-CoV M protein [16] was used as template and primers (5 0 CG GAATTCAGGTCTCGTAGACCG3 0 and 5 0 TGCTCTAG AAAGCTTTATTATGTA3 0 ) were used to perform PCR and amplify the gene fragment. After PCR, the DNA fragment was digested by restriction enzymes (EcoRI/ XbaI) and cloned into pcDNA3.1-V5-His A (linearized by EcoRI/XbaI) expression vector. This expression plasmid encodes a fusion protein consisting of the first 115 amino acids of HCV core protein, plus the first 50 amino acids of SARS-CoV M protein, followed by V5 and His tag.

To link the first 115 amino acids of HCV core protein with a.a. 46-68 of SARS-CoV M protein, PCR primers (HCV-1 and 5 0 CGCGGATCCCCTACGCCGGGGGTCT GT3 0 ) were used to amplify the first 115 amino acids of the core gene fragment while PCR primers (5 0 CGCGGATC CTACATAATAAAGCTTGT3 0 and 5 0 TGCTCTAGAAA CAGCAAGCACAAAAC3 0 ) were used to amplify the DNA fragment encoding a.a. 46-68 of SARS-CoV M protein. After PCR reaction, these two DNA fragments were digested by restriction enzymes (EcoRI/BamHI and BamHI/XbaI separately) and cloned into pcDNA3.1-V5-HisA (Invitrogen, USA) expression vector (linearized by EcoRI/XbaI). To link the first 115 amino acids of HCV core protein with a.a. 78-100 of SARS-CoV M protein, a similar approach was performed except using primers (5 0 CGCGGATCCGGGATTGCGATTGCAAT3 0 and 5 0 TG CTCTAGACCTGAAGGAAGCAACGA3 0 ) to amplify the DNA fragment of SARS-CoV M protein a.a. 78-100.

All the expression plasmids were verified by sequencing.

The Vero E6 cells were maintained in RPMI 1640 medium containing 10% fetal calf serum, 1% Glutamine (200 mM, Biological Industries, USA), and 100 lg/ml penicillin/ streptomycin (Gibco BRL, USA). 2.5-2.7 9 10 5 cells were plated in the 35-mm dish. After an overnight incubation, cells were infected with a recombinant vaccinia virus carrying the T7 phage RNA polymerase gene [17] . Two hours after infection, cells were transfected with 0.4 lg plasmid DNA by using Effectene transfection reagent (Qiagen, Germany). 21 h after transfection, recombinant proteins in the cells were analyzed. The mRNAs transcribed from either CMV promoter or T7 promoter in this expression system use the same AUG to initiate translation.

For Western blotting analysis, cells were dissolved in sample preparation buffers after washed by PBS twice. In our previous study [16] , SARS-CoV M protein could not be detected in SDS-PAGE in regular boiling treatment. Therefore, non-heated treatments [18] were used in antigen preparations (sample buffer containing 50 mM Tris-HCl (pH 6.8), 100 mM dithiothreitol, 2% SDS, 0.1 % bromophenol blue, 10% glycerol; no boiling treatment) to detect the expression of SARS-CoV M protein. 4 .5% (acrylamide percentage) gel was used as the stacking gel and 12% gel as the separating gel in this study. When proteins with smaller size were analyzed (e.g. deletion mutants of membrane protein), 15% gel was used as the separating gel. After electrophoresis, the SDS-PAGE gel was transferred to PVDF paper (Pall Corporation, USA). All procedures were then carried out at room temperature following our previous procedures [19] . The proportion of different forms of M protein (in Figs. 1a, b, 2a) was quantified using software ''Quantity one'' (Biorad, USA).

Vero E6 cells (1 9 10 6 ) were infected with recombinant vaccinia virus carrying the T7 phage RNA polymerase gene [17] for 2 h, then transfected with the plasmid expressing 

Full-length SARS-CoV M gene fragment was cloned and expressed as a recombinant protein (221 a.a.) with a C-terminal V5-His tag (29 a.a.) in Vero E6 cells (Fig. 1a) . In addition to the protein with the expected molecular weight (27.5 kDa, marked by thin arrow), several protein products of different sizes were also detected (lane 2 in Fig. 1a ). One such protein with larger size (marked by thick arrow) was the glycosylated M protein since this band disappeared after treatment with either N-glycosidase F (lane 3 in Fig. 1a) or endoglycosidase H (lane 4 in Fig. 1a) . The Asn-Gly-Thr (a.a. 4-6) at the N-terminal of SARS-CoV M protein is a defined consensus sequence (Asn-X-(Ser/Thr)) for N-linked glycosylation [21] . This is verified by the observation that production of the glycosylated M protein was blocked when amino acid residue 4 was changed from Asn (lane 2 in Fig. 1b) to Asp (lane 3 in Fig. 1b) .

One smaller product translated in-frame from the third Met was detected

In addition to the glycosylated and un-glycosylated SARS-CoV M proteins, two smaller protein products (marked by thick line and thin line, respectively) were also detected when M gene was expressed in Vero E6 cells ( Fig. 1a and b) . The size of these two smaller proteins corresponds to the products translated in-frame from the second (amino acid 32) and the third (amino acid 83) Met residues. This hypothesis was supported by the expression of several M deletion mutants (Supplement Fig. 1 ). The protein assumed to be translated in-frame from the third Met disappeared when amino acid 83 was mutated from Met to Leu (lanes 4 and 5 in Fig. 2 ). Instead, one additional protein (marked by double arrow) corresponding to the product initiated in-frame from the fourth Met (amino acid 90) was detected when this mutant M protein was expressed. However, the protein assumed to be translated in-frame from the second Met still exists when amino acid 32 was mutated from Met to Leu (lanes 3 and 5 in Fig. 2) . Thus, this protein should be a degraded product. The protein translated in-frame from the third Met (amino acid 83) could still be detected when the authentic 5 0 -untranslated region of SARS-CoV M gene was included in the expression vector (lane 3 in supplement Fig. 2 ). We have also quantified the proportion of these two different forms of M protein (in Figs. 1a, b and 2 ). Comparing with the unglycosylated form of M protein (as ''100%''), the expression level of the protein with a.a. 32-221 is 40% ± 19% while the expression of the protein with a.a. 83-221 is 20% ± 6%. To rule out the possibility that the read-out products of SARS-CoV M gene were generated by the C-terminus V5-His tag, we carried out experiments with the same tag fused to the C-terminus of a different gene (pyruvate kinase, supplement Fig. 3 ) No other bands are detected below molecular weight 67 kDa (574 a.a. of pyruvate kinase plus 29 a.a. of tag = 603 a.a.), although pyruvate kinase does contain several internal in-frame AUGs (e.g., a.a. 65, a.a.  107, a.a. 112, etc.) . The results indicated that V5-His tag would not facilitate the processing of read-through products.

To determine the N-glycosylation of SARS-CoV M protein is co-translational or post-translational event, SARS-CoV M protein was in vitro translated in the reticulocyte lysate. N-glycosylated product was detected only when microsome was added before but not after of translation (Fig. 3) . Thus, glycosylation of SARS-CoV M protein should occur co-translationally.

The second and third trans-membrane regions could insert a cytoplasmic protein into the endoplasmic reticulum membrane more efficiently than the first one SARS-CoV M protein will go to E.R. membrane though lack of a conventional signal peptide and predominately localizes in the Golgi complex [22] . There are three predicted trans-membrane domains in the SARS-CoV M . 14-36, a.a. 46-68, and a. a. 78-100 (Supplement Fig. 4) . The second and third trans-membrane regions (a.a. 46-68 and 78-100) are predicted to be the primary type helices while the first trans-membrane region (a.a. 14-36) is predicted to be the secondary type helix (Supplement Fig. 4) . The SARS-CoV M protein mutants with deletion of anyone trans-membrane region could still enter into E.R. (data not shown). To determine which trans-membrane domain is responsible for the integration of SARS-CoV M protein into E.R., each of these three trans-membrane domains (Fig. 4) was linked with a cytoplasmic protein (a.a. 1-115 of hepatitis C virus core protein) (Fig. 5 ) and the recombinant fused protein was expressed in Vero E6 cells. To quantitate the average percentage of co-localization from 50 cells for each test, Image J (NIH web) program was used. The average R values of pDsRed-ER + C115, pDs-Red-ER + CM50, pDsRed-ER + CM46-68 and pDsRed-ER + CM78-100 are 0.3379, 0.6632, 0.8624, and 0.8442, respectively. Thus, the second (Fig. 6) and third (Fig. 7) trans-membrane regions could insert a cytoplasmic protein into the endoplasmic reticulum membrane more efficiently than the first one (Fig. 8 ).

In this study, SARS-CoV M gene fragment was cloned and expressed as a recombinant protein fused with a C-terminal V5 tag in Vero E6 cells (Fig. 1) . One protein larger than the expected size was the product of N-glycosylation since its size was reduced to the expected molecular weight when treated with either endo-F or endo-H (Fig. 1a) , and its production was blocked when Asn 4 in the predicted glycosylation site (Asn-X-(Ser/Thr)) was mutated to Asp (Fig. 1b) . In other well studied coronaviruses, the M protein is involved in the assembly and budding of virions together with the E protein [14] . Moreover, the M proteins of other coronaviruses also contain highly conserved glycosylation sequences [23] , and their glycosylation may be related to the interaction between virus and host. Whether N-glycosylation of SARS-CoV M protein is related to the interaction between virus and host awaits further investigation. The N-glycosylation of SARS-CoV M protein should occur in the endoplasmic reticulum rather than Golgi complex since both endo-F and endo-H can remove the glycosylated moiety [24] .

Two other expressed SARS-CoV M protein products with smaller size than the full-length one were also detected in Vero E6 cells (Figs. 1 and 2) . These two smaller proteins seem to be the translated products initiated in-frame from second (amino acid 32) and third (amino acid 83) Met residues. It was verified with a Met 83 but not Met 32 mutated SARS-CoV M protein, which blocked the production of the expected product (Fig. 2) . Translation initiated in-frame from the third Met could be due to ribosomal leaky scanning, re-initiation, and even internal initiation [25, 26] . It is least likely due to internal initiation since there is no conventional IRES (internal ribosomal entry site) in the SARS-CoV M gene sequence. It is less likely through ribosomal leaky scanning since neither the second nor the fourth in-frame AUG was used as the initiation codon (Figs. 1 and 2) . It is probably through ribosomal re-initiation. This explains why neither the second nor the fourth in-frame AUG was used as the initiation codon (too close to the first or third in-frame AUG). And, this also explains why the protein initiated in-frame from the fourth Met was detected when the third in-frame AUG was mutated (Fig. 2) . Production of this smaller product could be due to the replacement of authentic 5 0 -untraslated region (5 0 CUUAUCAUGG3 0 ) of M gene with vector cloning sequence (5 0 GAAUUCAUGG3 0 ). To rule out this possibility, expression plasmid with authentic 5 0 -untraslated region of SARS-CoV M gene was cloned and expressed. The protein product initiated from Met 83 was also detected in this expression construction (Supplement Fig. 2, lane 3) . Translation initiated in-frame from the third Met is unusual since the sequence around the first Met (5 0 CUUAUCAUGG3 0 ) of SARS-CoV M protein is the optimal consensus Kozak sequence (5 0 GCCA/GCCAUG G3 0 ). Moreover, there is one out-of-frame AUG (5 0 GGA-ACAAUGG3 0 ) with consensus Kozak sequence between the first and the second in-frame Met (GI: 30027610). Production of these smaller SARS-CoV M proteins should not be an artifact of this expression system, since another protein (pyruvate kinase) over-expressed using this system Fig. 3 ) did not generate smaller products initiated from downstream in-frame AUGs. The function of this smaller translated product awaits further investigation. Similar to other N-glycosylated proteins [24] , the glycosylation of SARS-CoV M protein is occurred cotranslationally but not post-translationally (Fig. 3) . The M protein of mouse hepatitis virus strain A59 is also a triplespanning membrane protein. Which of these three hydrophobic domains in A59 M protein is the insertion and anchor signal for the E.R.-integration of this protein is not conclusive [27] [28] [29] . The second and third trans-membrane regions (a.a. 46-68 and 78-100) are predicted to be the primary type helices, which will be able to penetrate into membrane by themselves, while the first trans-membrane region (a.a. 14-36) is predicted to be the secondary type helix, which is considered to be stabilized by the interaction with other trans-membrane segments [30] . To determine which transmembrane domain is responsible for the integration of SARS-CoV M protein into E.R., each of these three transmembrane domains was linked with a.a. 1-115 of hepatitis C virus core protein, which is suspected to be a nuclear protein since it contains the nuclear localization signal [31] , and the recombinant fused protein was expressed in Vero E6 cells. Unexpectedly, the protein encoding a.a. 1-115 of hepatitis C virus core protein with V5 tag localizes in the cytoplasm (Fig. 5) but not the nucleus maybe due to the core proteins derived from different isolates. Just like the prediction, the second and third trans-membrane regions were able to insert a cytoplasmic protein into the endoplasmic reticulum membrane more efficiently than the first one (Figs. 6-8) . These results suggest that the second and/or third transmembrane regions are more likely than the first one to target SARS-CoV M protein into E.R.

In summary, when SARS-CoV M gene was expressed in Vero E6 cells, in addition to the full-length un-glycosylated M protein, glycosylated M protein and one smaller product initiated in-fame from the third Met was also detected. Nglycosylation of SARS-CoV M protein is a co-translational event. The second and third trans-membrane regions were able to insert a cytoplasmic protein into the endoplasmic reticulum membrane more efficiently than the first one. These findings should be important for the morphogenesis of SARS-CoV.

The aetiology, origins, and diagnosis of severe acute respiratory syndrome

Epidemiological and genetic analysis of severe acute respiratory syndrome

Infectious diseases. Clues to the animal origins of SARS

Identification of a novel coronavirus in patients with severe acute respiratory syndrome

Newly discovered coronavirus as the primary cause of severe acute respiratory syndrome

Characterization of a novel coronavirus associated with severe acute respiratory syndrome

Characterization and complete genome sequence of a novel coronavirus, coronavirus HKU1, from patients with pneumonia

Virology. The SARS coronavirus: a postgenomic era

SARS virus: the beginning of the unraveling of a new coronavirus

Coronaviridae: the viruses and their replication

Mechanisms and enzymes involved in SARS coronavirus genome expression

Coronavirus envelope glycoprotein assembly complexes

Nucleocapsid-independent assembly of coronavirus-like particles by co-expression of viral envelope protein genes

Assembly of human severe acute respiratory syndrome coronavirus-like particles

Thermal aggregation of SARS-CoV membrane protein

Eukaryotic transient-expression system based on recombinant vaccinia virus that synthesizes bacteriophage T7 RNA polymerase

Study of Dengue virus infection in SCID mice engrafted with human K562 cells

The first hydrophobic domain of the hepatitis C virus E1 protein is important for interaction with the capsid protein

Membrane topology of the hepatitis C virus NS2 protein

The M protein of SARS-CoV: basic structural and immunological properties

Differential maturation and subcellular localization of severe acute respiratory syndrome coronavirus surface proteins S, M and E

Structural requirements for O-glycosylation of the mouse hepatitis virus membrane protein

Initiation of translation in prokaryotes and eukaryotes

Pushing the limits of the scanning mechanism for initiation of translation

Membrane assembly of the triple-spanning coronavirus M protein. Individual transmembrane domains show preferred orientation

Membrane integration and intracellular transport of the coronavirus glycoprotein E1, a class III membrane glycoprotein

Signal recognition particle-dependent insertion of coronavirus E1, an intracellular membrane glycoprotein

SOSUI: classification and secondary structure prediction system for membrane proteins

Nuclear localization signals in the core protein of hepatitis C virus

Acknowledgements This work has been supported by grants from Tzu Chi University (TCMRC 93017 and TCMRC-C95002-01) and from National Science Council of Taiwan (NSC 95-3112-B-320-001) to Dr. Shih-Yen Lo.