SARS-CoV: The Pathogen |
|
| The Pathogen The Pathogen Coronavirus - Human Coronavirus - Animal Coronaviruses Viral Replication Virus Structure - Replicase - Nucleocapsid - Spike Protein Viral Entry - Membrane Fusion - ACE2 as Receptor Home Epidemiology Disease Pathology Diagnosis & Treatment The Pathogen Immune Response Vaccine Development Global Health Response References Disclaimer |
The etiological agent of SARS is a novel coronavirus, and the virus is thus referred to as the SARS-coronavirus, abbreviated SARS-CoV. The name coronavirus derives from the halo/corona appearance of viral particles when viewed under a microscope, which is a result of the protruding spike proteins that coat the surface of the virus. The enveloped viral particle is 60-130 nm in diameter and contains a single-stranded positive RNA strand. The SARS-CoV genome is 29,272 nucleotides in length with 41% being G/C residues. Large genome size is characteristic of coronaviruses, which actually exhibit the largest genomes of all RNA viruses. There are 11 open reading frames (ORFs) in SARS-CoV, and genome organization of the major structural proteins is typifies that of coronaviruses, with short untranslated regions at both termini: [5'- replicase (rep), spike (S), envelope (E), membrane (M), nucleocapsid (N)-3'] Two distinct SARS-CoV genotypes were identified during the most recent (and only) outbreak. One genotype originated from Hotel M in Hong Kong, while the other originated from cases in Guangdong, Hong Kong, and Beijing. Due to the error prone process of viral replication in SARS-CoV, there is the real possibility that strains and clades (as in HIV) will emerge in future outbreaks.
Coronaviruses are large, enveloped viruses ranging from 60 to 220 nm in diameter. When viewed under a microscope, the "corona" that surrounds the virus is due to a coating of protruding surface spike proteins. Coronaviruses are positive-stranded RNA viruses and have the largest genomes of all RNA viruses. Due to discontinuous transcription and polymerase "jumping" during viral transcription and replication, there is a high frequency of RNA recombination. Phylogenically, coronaviruses belong to the order Nidovirales , family Coronaviridae , and genus Coronavirus . Before the identification of SARS-CoV, known coronaviruses were divided into three groups. With the sequencing of the complete genome, SARS-CoV has been assigned a fourth, novel, monophyletic group within the genus:
Coronaviruses typically have narrow host ranges. However, it appears that SARS-CoV may be relatively less stringent in this aspect and has the ability to infect different cell cultures, such as African green monkey ( Ceropithecus aethiops ) kidney cells (Vero cells), fetal rhesus kidney cells (FrhK-4), and human colorectal adenocarcinoma cells (Caco-2). Coronavirus: Human Coronavirus Prior to the advent of SARS, there were two other strains of HCoV known to cause illness. HCoV 229E and OC43 cause respiratory illness that ranges from the common cold to severe and sometimes fatal pneumonia. These illnesses are not generally severe in healthy individuals but present major health complications in immunocompromised populations. On sequencing of the SARS genome, it was found that it was only mildly related to previous human strains and was much closer to several animal strains from Chinese civet cats. In March 2004, a fourth strain, HCoV-NL63, was identified in a seven month old. This unique strain does not appear to be a recombination of earlier identified viruses and can be best identified by a unique N-terminus fragment. (88) Coronavirus: Animal Coronaviruses Coronaviruses tend to be highly host-specific. They do not create infections in very genetically different organisms. For instance, infectious bronchitis virus (IBV), a virus causing bronchitis in chickens, does not infect cows or humans. Because coronaviruses are so specific, it's difficult to assign a good animal model for research on a virus that infecrts humans like SARS. However, because so little work has been done on SARS and because it is such a new pathogen, it is useful to see what is known about vaccines to known coronaviruses. In understanding what has been accomplished in the realm of animal coronavirus vaccines, we can quickly rule out some vaccine options and give weight to others, an important step in the development of vaccines. The most frequently researched animal coronaviruses are in livestock and poultry bcause of the economic problem that sick animals cause farmers and the incentive to produce a good vaccine against the pathogens that do the most damage. In all of the animal models that have been subject to reasearch, attenuated live vaccines tend to be the most efficacious in current use. This works well in the poultry world because infectious bronchitis virus (IBV) is relatively stable in its genotype. Keep in mind, however, that this vaccine strategy is probably not a good one for SARS because of the high rates of mutation that the virus is subject to. Reversion to the virulent genotype would be much more possible in SARS than IBV (the genotype of IBV has changed 2% since 1940), so few people would be willing to introduce the whole, live, albeit attenuated virus, into their systems. However, new vaccine design methods that incorporate recombinant DNA in their design are changing the way researchers are attacking the coronavirus problem. In this field, most vaccines focus on the elution and sequencing of spike protein motifs. Some vaccines incorporate nucleocapsid and membrane peptides as well, but the majority of research is inspecting the use of spike protein peptides in eliciting a virus-nullifying response. Some novel work has been done with porcine transmissable gastroenteritis virus (TGEV). Other recombinant vectors are being investigated for use against bovine coronavirus (BcoV). Other animal models of coronavirus infection exist in cats and mice. Poultry are frequently affected by infectious bronchitis virus (IBV). Some work has been done with the creation of live attenuated virus vaccines and these tend to work fairly well. As long as antibody is present in the system when antigenic challenge occurs, infection is prevented. In all animal models studied, the best antibody response was elicited by live attenuated vaccines. Studies were done in which chickens were vaccinated in ovo and these were highly successful in attaining protection against IBV. Many mature vaccinated birds still shed virus, but their organs were free of infiltration (a sign of systemic, more serious infection). Because the spike protein of this virus is so highly conserved, it is appropriate that an antibody response would be sufficient to contain it. It is also highly unlikely to affect research in a SARS vaccine because the methods used to obtain the highest degree of protection are unfeasible and the vaccine model itself does not lend itself well to the highly mutable SARS. A vaccine has been deeloped for IBV that incorporates a recombinant plasmid in a fowl poxvirus vector. The plasmid is loaded with spkie protein and delivered to the poultry via fowl poxvirus. This method of immunization is less effective in chickens than the traditional attenuated live vaccine. This discrepancy in the viability of a plasmid vaccine says nothing about this method's utility in SARS research as, again, the viruses at the root of each infection are very different. A recombinant adenovirus vaccine against bovine coronavirus has been developed that shows good potential for protectivity in the mucosa where the virus attacks. The target of the vaccine was an antibody response against the hemagglutinin-esterase which is a membrane protein of the virus. In rats, this approach provided good hemagglutinin-specific IgA production when a vaccine was delivered intraduoudenally with an oral booster. It is these antibodies that would have the most influence in nullifying a challenge by environmental BcoV. For its time, this vaccine was novel in its use of a plasmid vector administered mucosally. There have also been vaccines developed that incorporate an inactivated BcoV strain. The viruses were prepared for delivery by cultivation in a porcine testicular cell line. These vaccines provided solid protection against the pathogen. Pigs are affected by porcine transmissable gastroenteritis virus (TEGV). Interesting research has been done on vaccine designs for TGEV. These studies have tended to move away from the classic live attenuated virus vaccine and into models that use sequenced peptides in plasmid or vector structures. One interesting design was to load Salmonella with plasmids coding for the most antigenic spike protein sequences. Mice were used as test subjects for the efficacy of this approach. The bacteria, once modified with the engineered plasmid, now display antigenic proteins of TGEV. Upon introduction into the murine host, the Salmonella were phagocytosed by macrophages, which subsequently displayed the pertinent peptides on both MHC classes 1 and 2. This stimulated both arms of the adaptive immune system, eliciting both an antibody-mediated response and a CTL response. Other work done on TGEV used a naked DNA approach. Again, a plasmid was engineered that contained genes coding for antigenic spike proteins on TGEV. This time, the plasmid alone was injected intramuscularly into the host. Once injected, the plasmid was absorbed into host cells and transcription and translation of the spike peptide genes that it was carrying. The spike protein peptides were then presented on MHC class 1 and a Th1-type cell activation ensued. This method elicits a relatively low antibody response when compared to live attenuated vaccines for TGEV but provides excellent in vivo protection against the pathogen. Cats are affected by feline coronavirus (FCoV) which is the cause of feline infectious perionitis (FIP). Research has been focused on the use of live attenuated virus as a vaccination method. There have been tests on delivery methods for the vaccine but, by and large, the only vaccine model being used is a live attenuated vaccine. The sole protein that has shown to elicit antibody immune response in the case of FCoV is the spike protein. Work has been done to determine which sections of the spike protein on FCoV stimulate the most neutralizing antibodies. Fairly conserved peptides were discovered, leading to the conclusion that a DNA plasmid vaccine may be possible. Some of the most antigenic portions of the spike protein on FCoV correspond with sequences on porcine TGEV. This indicates that some of the measures useful in research on TGEV might also be applicable towards FCoV vaccine design. Mice are inflicted with a hepatitis infection caused by a murine coronavirus (A59). One vaccine developed to combat this virus incorporated spike protein-specific antibodies bound to influenza hemagglutinin. The SARS-CoV genome is contained in a positive single-stranded RNA (+RNA). A complementary negative RNA (-RNA) strand is synthesized from the +RNA, and mRNA transcripts are subsequently synthesized from the complementary -RNA strands. The mRNA transcripts are then translated into gene products for new virion assembly: +RNA (genome) >> -RNA (complementary) >> mRNA (transcript) >> translation It should be noted that ribosomes can only transcribe proteins from single-stranded RNA and are unable to recognize double-stranded RNA. As such, it has been suggested that complementary RNA strands could be used to inhibit translation by annealing to mRNA transcripts, forming double-stranded RNA inaccessible to ribosomes. The mRNA transcripts are described as a "nested" set of mRNA's. That is, there is overlap between different genes in the +RNA genome. It is unclear how the mRNA's are transcribed, as there are two lines of evidence in opposition to each other:
Interestingly, it has been found that N protein, which is involved in replication and transcription of viral RNA, distributes in the cytoplasm and not in the nucleolus. This suggests that SARS-CoV interacts with cytoplasmic proteins and not nucleolar proteins. This lack of nuclear dependency means that both nucleated and anucleated cells could potentially be infected. Because this mode of operation differs from other known coronaviruses, it has been suggested that the unusual localization behavior of SARS N protein may have contributed to the severity of the intial outbreak. Finally, it should be noted that SARS-CoV replication is error prone and that mutations have already been detected, although the outbreak lasted for a relatively short period of time. Such mutations could be the result of immune pressure and could lead to the development of clades and strains, as in HIV. It is therefore important to develop vaccines now that will protect across regional strains before further changes in the genome can take place and re-emerge. In fact, when considering the large genome size of SARS-CoV, the amount of mutation possible could match or even exceed that found in HIV should the virus emerge in future outbreaks.
The rep gene makes up 2/3 of the SARS-CoV genome. The two ORFs (ORF1a and ORF1b) are predicted to encode two proteins that undergo proteolytic processing: RNA polymerase and 3 chymotrypsin-like protease (3C-like protease or 3CLpro) (117). SARS-3CLpro is a 35.8 kDa cysteine protease that engages in the proteolytic cleavage of precursor polyproteins into functional proteins required for viral replication. As such, this protease is an appealing target for the development of anti-SARS agents, given its key role in the processing of viral polyproteins in the replication cycle. It has in fact been suggested that inhibitors of protease activity in other viruses could be useful starting points for the development of anti-SARS agents targeting 3CLpro, including:
Virus Structure: Nucleocapsid (N Protein) The N protein is 422 amino acids in length with a mass of 47 kDa. It shares only 20-30% sequence homology with the N proteins of other known coronaviruses. In other coronaviruses, N protein plays a role in
Such roles have thus far not been reported in SARS-Cov. However, it has been reported by He et al. (51) that N-N self-interaction may be important for initiating ribonucleoprotein (RNP) formation and subsequent encapsidation during the virion assembly process. In this study, a Ser/Arg-rich motif (SSRSSSRSRGNSR) was identified as being crucial for N protein oligomerization, as detected by sucrose gradient fractionation experiments. Deletion of the Ser/Arg-rich sequence resulted in the loss of N-N self-interaction. Using fluorescence tracking techniques, while expression of wild-type N protein showed localization around the nucleus, expression of the sequence-deleted N protein showed scattered localization, indicating that the Ser/Arg-rich motif may play a role in sub-cellular localization of N protein, which may be important for virus assembly. N protein also appears to be immunogenic, as humoral responses against N protein can be detected in SARS-infected patients. This of course has implications of developing vaccines vased on N protein. Virus Structure: Spike Protein (S Protein) The S protein is seen under the microscope as a layer of protrusions on the viral surface, giving coronaviruses their characteristic halo/corona appearance. The SARS-CoV S protein is 1255 amino acids in length and consists of an extracellular domain (residues 1-1199), a transmembrane domain (residues 1200-1215), and an intracellular domain (1216-1255). In addition, there is a short signal peptide (residues 1-13). Coronavirus S proteins consist of two domains S1 and S2, which may be cleaved into separate subunits. The S1 domain is the knob region of the spike and is involved in binding to host cell receptors. The S2 domain is the stem region in which coiled-coil and transmembrane regions take part in host cell entry and cell-cell fusion. Though S1 and S2 domains are present in SARS-CoV, the protein does not appear to have a cleavage site. As in other coronaviruses, S2 is more conserved than S1 and contains heptad repeat regions HR1 and HR2, similar to HIV gp41 (see Viral Entry). S protein also appears to be immunogenic, as humoral responses against S protein have been detected in SARS-infected patients. Additionally, there is new evidence that S protein could be a critical antigen, since viral vaccines based on S protein have proved successful in preventing SARS-CoV infection. Studies have shown that the spike (S) protein plays an important role in viral invasion of host cells. It is believed that the S1 domain of the S protein mediates cellular receptor recognition ( specificity ), while the S2 domain mediates fusion ( virulence ). This organization parallels HIV where two non-valently associated glycoprotein subunits perform the same functions: gp120 functions in target-cell recognition through interaction with CD4 and a coreceptor, while gp41 promotes fusion. Furthermore, one study by Zhang and Yap (140) found that SARS-CoV S2 and HIV gp41 share similar helix structures on residues 879-942. Such similar structural motifs between SARS-CoV S2 and HIV gp41 provide further evidence for a similar membrane fusion mechanism between the two viruses. As such, it has been suggested that inhibitors targeting HIV viral fusion might be useful starting points for the development of anti-SARS agents. In fact, a study by Liu et al. (81) applied the hypothesis that HIV and SARS-CoV infect via similar fusogenic mechanisms toward the design of compounds targetting the SARS-CoV fusogenic mechanism. In HIV, gp41 contains two heptad repeat regions HR1 and HR2. Binding of gp120 to CD4 and coreceptor leads to a change in conformation in gp41 mediated by HR1 and HR2 interaction, resulting in the formation of a fusogenic hexameric core structure and subsequent fusion with the host cell. C-peptide can bind gp41 HR1 and thus block fusion by abrogating the interaction between HR1 and HR2. The S2 domain of S protein also contains HR1 and HR2 sequences, and it was hypothesized that fusion was mediated by an HR1-HR2 interaction similar to that in HIV gp41. Six peptides corresponding to SARS HR1 and HR2 were synthesized and used as probes to identify possible antiviral peptides. CP-1, derived from the HR2 region, was found to inhibit SARS-CoV infection in the micromolar range. CP-1 could thus be used as lead in the design of future compounds for the treatment or prophylaxis of SARS-CoV infection. The mechanism of SARS-CoV entry is a fusogenic one. After binding to the host cell, the virus uncoats at the plasma membrane in a membrane fusion event, releasing its RNA genome into into the host cell cytoplasm. Interestingly, the virus is purported to leave a "footprint" on the cell membrane, since glycoproteins on the viral membrane envelope integrated into the host cell membrane during fusion. It has been suggested that these proteins could be targeted, and T-cells could be induced to recognize and clear infected cells. Several studies have provided evidence that angiotensin-converting enzyme 2 (ACE2) is the functional receptor for SARS-CoV. It is proposed that the fusogenic spike protein binds to ACE2 on host cells and thence mediates viral entry into the host. Wang et al. (128) describe the process in which cDNA of SARS-CoV sensitive Vero cells was transfected into HeLa cells, which are not sensitive to SARS-CoV infection. This new transfected HeLa cell clone, called F5, became highly susceptible to SARS-CoV infection, indicating that it now expressed a receptor for SARS-CoV. Next, a second HeLa cell line expressing SARS-CoV S protein was established, called S-HeLa. Not surprisingly, co-culture of S-HeLa with F5 lead to syncytia formation, indicating receptor engagement and fusion between S-HeLa and F5 cells. Finally, when susceptible F5 cells were preincubated with anti-ACE2 antibodies, infection was abrogated, and it was concluded that ACE2 is the functional receptor for SARS-CoV. Additionally, when ACE2 cDNA cloned from F5 cells was used to transfect non-infectable NIH/3T3 cells, the transfected NIH/3T3 line exhibited infe ctivity similar to F5, further demonstrating that ACE2 is the functional reeceptor for SARS-CoV. Homology modeling has been used to build a 3D model of ACE2 (100). In this model, the top of ACE2 is characterized by a deep channel containing the catalytic site. The surface of the channel and the surrounding ridges are highly negatively charged, with key residues D136, E150, N154, D157, D292, D295, and D299. Complementary to this, the hypothesized S protein receptor binding domain (RBD) is characterized by positive charges, particularly in the electronegative loop containing K439, R441, R444, H445, and K447. In addition to complementarity of charges, hydrophobic patches on ACE2 (Phe, Trp, and Tyr residues) are also hypothesized to contribute to receptor-ligand binding. In summary, the ACE2 model provides structural evidence suggesting the feasibility of a receptor-ligand interaction with SARS-CoV S protein:
|
