|Home | About | Journals | Submit | Contact Us | Français|
The epidemic of the severe acute respiratory syndrome (SARS) has swept through the globe with more than 8000 reported probable cases. In Hong Kong, the hardest hit areas included our teaching hospital and the Amoy Gardens apartment complex. A novel coronavirus, SARS-coronavirus (SARS-CoV), with a single-stranded plus sense RNA genome, was promptly implicated as the causative agent and subsequently fulfilled Koch's postulates. To aid the understanding of SARS-CoV, groups of investigators rapidly sequenced viral isolates around the world. We were the third group in the world to release the complete SARS-CoV genome sequence (isolate CUHK-W1) on the world-wide web. With other isolates from patients of distinct epidemiological backgrounds, we additionally sequenced four complete (CUHK-Su10, CUHK-AG01, CUHK-AG02, CUHK-AG03) and two partial SARS-CoV genomes. The reviewed data obtained from representative patients from the hospital and community outbreaks has documented the evolution of the virus in this epidemic. Their sequence variations also revealed a remarkable epidemiological correlation. We demonstrate that sequence variations in the SARS-CoV genome can be applied as a powerful molecular tool in tracing the route of transmission, when used adjunctively with standard epidemiology.
SARS is a recently recognized, infectious and potentially life-threatening condition.1–3 The World Health Organisation issued a global alert on the condition on 12th March 2003. The epidemic of SARS has swept through the globe, affecting both health care workers and the general community. According to the data compiled by the World Health Organisation in September 2003, the number of reported probable cases with onset of illness from 1st November 2002 to 31st July 2003 amounted to 8098 with 774 deaths.4
A novel coronavirus, SARS-CoV, with a single-stranded plus sense RNA genome, was promptly implicated as the causative agent.5–7 Macaques exposed to the SARS-CoV subsequently developed respiratory symptoms and pathology similar to SARS patients, thus fulfilling Koch’s postulates.8
To aid the understanding of this epidemic, groups of investigators around the world rapidly isolated SARS-CoV from representative patients in major outbreaks and sequenced the SARS-CoV genomes.
In Hong Kong, as of 8th September 2003, the local case number amounted to 1775,9 with two major epidemiological outbreaks. The first took place at the Prince of Wales Hospital where 156 SARS cases were hospitalised between 11th March and 25th March 2003.2 The second major outbreak occurred in the community, where up to 15th April 2003, 321 residents of the Amoy Gardens, a densely populated housing complex comprised of multiple 33 storey apartment blocks with eight units per floor, were infected.10 Since these two outbreaks were characterised by the rapid spread of infection among individuals and associated with more than a quarter of all cases in Hong Kong, we determined and published the genomic sequences from representative patients involved in these two outbreaks.11,12 The protocols for RNA extraction, first strand cDNA synthesis and our DNA sequencing strategy have been described in detail previously.12
The index patient, J, at our hospital was a 26-year-old ethnic Chinese man who was admitted to the Prince of Wales Hospital on 4th March 2003.2,13 J was epidemiologically linked to the first reported local case of SARS, a nephrologist from Zhongshan, Guangdong Province, China, who visited Hong Kong in late February 2003.3 Four days prior to symptom onset, J visited the hotel at the same floor where the nephrologist stayed. Patient K was J’s mother who lived with J and previously enjoyed good health. She started to have fever, chills and rigors on 6th March 2003. Seven days later, she was admitted to the Prince of Wales Hospital in a critical condition, with a temperature of 39.7°C and an oxygen saturation of 77% in room air. Chest X-ray showed bilateral patchy consolidation. She required 100% oxygen to maintain a saturation of 94%. She was transferred to the Intensive Care Unit and died one month later. A nasopharyngeal aspirate was obtained from her on admission and SARS-CoV was cultured from this sample yielding viral isolate CHUK-Su10.
To investigate the possibility that different strains of SARS-CoV might exist, we sequenced the entire genome of another viral isolate obtained from a patient, W. Prior to the onset of symptoms, Patient W had not contacted J or visited the hotel where the Zhongshan nephrologist stayed. He had a history of travelling to Shenzhen, Guangdong Province, China, five days before the onset of his symptoms. He was in Thailand when his symptoms occurred. He complained of fever, chills, myalgia and repeated vomiting. His chest X-ray showed haziness in the right hilar region and left mid zones. On admission, he had a temperature of 38.5°C, with a lymphocyte count of 3.9 x 109/L (normal range: 4 - 10.8 x 109/L). SARS-CoV was cultured from his nasopharyngeal aspirate and this sample yielded isolate CUHK-W1.
In the Amoy Gardens outbreak, the infection spread rapidly with 41% of cases being from Block E. In particular, the cases concentrated in residents of the adjacent units, 7 and 8. These two units shared a common ventilation lightwell and a vertically linked faulty sewage system that were implicated as the environmental factors contributing to the outbreak.10 In addition to the suspected index patient AG01,10 four other residents (patients AG02-AG05) who lived in units 7 and 8 were randomly chosen for our investigation. Their clinical details have been detailed elsewhere.12 All of them presented with serological and virological evidence of SARS-CoV infection. The various samples collected from these patients with the highest viral loads were used for genomic sequencing (serum from AG01; nasopharyngeal aspirates from AG02, AG04 and AG05; stools from AG03).
We obtained 5 complete genomic sequences, for CUHK-Su10, CUHK-W1, CUHK-AG01, CUHK-AG02 and CUHK-AG03 (Genbank accession numbers AY282752, AY278554, AY345986, AY345987, and AY345988). Except for several sequence variations, they are consistent with the Tor2 and CDC-Urbani (Genbank accession numbers AY274119 and AY278741) SARS-CoV sequences published earlier.14,15 The predicted open reading frames (ORFs) are shown in the Figure. They are arranged in the order of genes encoding the orf 1ab polyprotein (including the polymerase protein), spike glycoprotein, envelope protein, membrane protein and nucleocapsid protein, as described for other coronaviruses.16 Phylogenetic analysis of the predicted viral proteins indicates that SARS-CoV does not closely resemble any of the three previously known groups of coronaviruses, thus it has been suggested that SARS-CoV should be classified into a new fourth group within the genus Coronavirus.14
Many sequence variations in the SARS-CoV genome exist between isolates, but some are simply artefacts due to viral culture adaptation, whilst others might be attributed to sequencing errors. To screen for mutations among the genomes of different SARS-CoV isolates, only those sequence variations that appear in at least two isolates either obtained by us or deposited by others at GenBank were considered and tabulated in the Table.
Compared to CUHK-Su10, which is associated with our hospital outbreak, Tor2 only differs by 1 nucleotide substitution that results in amino acid substitution in the membrane protein. In addition to this, CDC-Urbani further differs from CUHK-Su10 by another nucleotide substitution that results in no amino acid substitution in the orf1ab polyprotein. The close resemblance between these three SARS-CoV genomes was in line with epidemiological investigations that a cluster of cases were traceable to the nephrologist from Guangdong Province, China, who checked-in to a hotel in Hong Kong on 21st February 2003.3 Visitors to this hotel, including patient J, subsequently spread SARS to Hong Kong, Canada and Vietnam, where CUHK-Su10, Tor2 and CDC-Urbani were isolated.13
Similarly, CUHK-AG01 reveals a nearly identical sequence to CUHK-Su10, except with two mutations, T3852C and C11493T (Table), which results in no amino acid changes. This provides molecular evidence that the virus infecting-patient AG01 and the virus strain in the hospital infections were epidemiologically linked.12
According to our criteria for mutations, CUHK-AG02 and CUHK-AG03 revealed no further mutations other than those that appeared in CUHK-AG01. To further investigate the prevalence and importance of T3852C and C11493T, we sequenced these nucleotide positions from additional isolates, CUHK-AG04 and CUHK-AG05. Each of these showed identical sequence to CUHK-AG01, CUHK-AG02 and CUHK-AG03. Hence, these two mutations were present in all Amoy Garden isolates that we investigated. This is consistent with the fact that patients AG01-AG05 have visited or lived in the adjacent units of the same block. This again demonstrates the usefulness of SARS-CoV genome sequencing in supplementing epidemiology.
We discovered another strain of SARS-CoV among the patients in our hospital. CUHK-W1, which had been isolated from patient W who was not epidemiologically linked to the nephrologist from Zhongshan. CUHK-W1 is different from CUHK-Su10 at eight nucleotide positions (Table). Among them, seven involved amino acid substitution in the orf1ab polyprotein (including polymerase), spike protein, membrane protein and a putative predicted protein orf10. Remarkably, among these eight sequence variations that distinguish CUHK-W1 from CUHK-Su10, six have been found to be identical to GD01, BJ01, BJ02, BJ03 (Genbank accession numbers AY278489, AY278488, AY278487 and AY278490), which were isolated from mainland China.
Taking a closer look at CUHK-W1, we observed that among the 12 sequence variations that distinguish GD01 (isolated in Guangdong province, Southern China), from CUHK-Su10, seven were identical between GD01 and CUHK-W1. Thus CUHK-W1 is more closely related to GD01 than to CUHK-Su10/Tor2/CDC-Urbani. These results are especially relevant in view of the travel history of patient W to Shenzhen, which is another city of the Guangdong province.
Comparing GD01, CUHK-W1 and CUHK-Su10 together, we noted that CUHK-W1 exhibits features from two SARS-CoV strains, seven from GD01 and five from CUHK-Su10. Taking chronological and geographical information into account, we postulate that CUHK-W1 may represent an intermediary strain of SARS-CoV that bridges the earlier strain, GD01, to the later strains, including CUHK-Su10, that are linked to the Hong Kong hotel visited by the nephrologist from Zhongshan.
Our reviewed data show that SARS-CoV has been undergoing gradual evolution since it acquired the ability to spread amongst human populations. Whether these accumulating mutations may have implications on the possible future re-emergence of SARS remains to be seen.
We have also illustrated that SARS-CoV genome sequencing can supplement epidemiological investigation, since sequence variations showed a remarkable epidemiological correlation. Sequence variations could help to decipher the route of transmission across cities or within communities, particularly where basic epidemiological data is absent or as a tool used adjunctively with standard epidemiology. Preventive measures could then be reinforced to prevent future outbreaks.
The next step of SARS-CoV genomics would be to investigate the possible effects of these sequence variations on the infectivity, tissue tropism and pathogenicity of the virus. In the longer term, it is hoped that a detailed understanding of the biology of this virus will allow us to develop new therapeutic and vaccination approaches against SARS.
Department of Chemical Pathology
Y.M. Dennis Lo, Stephen S.C. Chim, K.C. Allen Chan, Rossa W.K. Chiu, Enders K.O. Ng, Yu Kwan Tong
Department of Biochemistry
S.K.W. Tsui, C.C. Au, K.W. Au, Anna H. Chan, C.W. Chan, Cecy Y.C. Kou, H.M. Lam, W.Y. Lam, S.K. Lau, Y.L. Lau, Y.M. Lau, S.L. Law, T.W. Law, Mandy L.Y. Li, C.H. Tse, Andy W.K. Wan, C.H. Wong, W.H. Yiu, C.Y. Lee, M.M.Y. Waye, K.P. Fung
Department of Microbiology
John S. Tam, Paul K.S. Chan, Carrie Au-Yeung, Jo L.K. Cheung, Ida Chu
Department of Paediatrics
Emily C.W. Hung
We wish to dedicate this review to the patients we have described in this paper and to the frontline healthcare workers who have cared for these SARS patients. This work is partially supported by the Hong Kong Research Grants Council Special Grants for SARS Research (CUHK 4508/03M).