Spread, circulation, and evolution of the Middle East respiratory syndrome coronavirus

Abstract

The Middle East respiratory syndrome coronavirus (MERS-CoV) was first documented in the Kingdom of Saudi Arabia (KSA) in 2012 and, to date, has been identified in 180 cases with 43% mortality. In this study, we have determined the MERS-CoV evolutionary rate, documented genetic variants of the virus and their distribution throughout the Arabian peninsula, and identified the genome positions under positive selection, important features for monitoring adaptation of MERS-CoV to human transmission and for identifying the source of infections. Respiratory samples from confirmed KSA MERS cases from May to September 2013 were subjected to whole-genome deep sequencing, and 32 complete or partial sequences (20 were ≥ 99% complete, 7 were 50 to 94% complete, and 5 were 27 to 50% complete) were obtained, bringing the total available MERS-CoV genomic sequences to 65. An evolutionary rate of 1.12 × 10(-3) substitutions per site per year (95% credible interval [95% CI], 8.76 × 10(-4); 1.37 × 10(-3)) was estimated, bringing the time to most recent common ancestor to March 2012 (95% CI, December 2011; June 2012). Only one MERS-CoV codon, spike 1020, located in a domain required for cell entry, is under strong positive selection. Four KSA MERS-CoV phylogenetic clades were found, with 3 clades apparently no longer contributing to current cases. The size of the population infected with MERS-CoV showed a gradual increase to June 2013, followed by a decline, possibly due to increased surveillance and infection control measures combined with a basic reproduction number (R0) for the virus that is less than 1.

Importance: MERS-CoV adaptation toward higher rates of sustained human-to-human transmission appears not to have occurred yet. While MERS-CoV transmission currently appears weak, careful monitoring of changes in MERS-CoV genomes and of the MERS epidemic should be maintained. The observation of phylogenetically related MERS-CoV in geographically diverse locations must be taken into account in efforts to identify the animal source and transmission of the virus.

FIG 1

Bayesian-inferred phylogeny of the 32 new MERS-CoV sequences combined with the 33 previously available genomes (EMC/2012 [JX869059] , Jordan_N3 [KC776174], Munich_AbuDhabi_2013 [KF192507], England-Qatar_2012 [KC667074], Al-Hasa_1_2013 [KF186567], Al-Hasa_2_2013 [KF186566], Al-Hasa_3_2013 [KF186565], Al-Hasa_4_2013 [KF186564, plus all previously published MERS-CoV sequences 17], England2-HPA [http://www.hpa.org.uk/Topics/InfectiousDiseases/InfectionsAZ/MERSCoV/respPartialgeneticsequenceofnovelcoronavirus/], France_UAE_2013 [KF745068], Qatar_3_2013 [KF961221], and Qatar_4_2013 [KF961222]). All new genome sequences from this study are labeled in red. Clades are marked with vertical bars on the right and (with the exception of clade A and the Al-Hasa clade) named by the initial genome in the clade. The scale bar indicates the genetic distance, in substitutions per site, from the arbitrary midpoint root. Bayesian posterior probabilities for each clade are listed above the relevant node.

FIG 2

Time-resolved phylogenetic tree of all concatenated coding regions of the 42 phylogenetically distinct MERS-CoV genomes (see Materials and Methods for further details). A discrete traits model implemented in BEAST version 1.7.5 (36) was used to determine the most probable geographical location for each branch; a change in branch color indicates a geographical location change during its evolutionary history. Posterior probabilities for the inferred geographical locations are indicated at the nodes, an asterisk at a node indicates a posterior probability of >0.9 for that clade, and time is indicated on the x axis.

FIG 3

Bayesian skyline plot (BSP) showing the changes in effective population size of MERS-CoV across time. The dashed black line indicates the median population size estimated from the BMCMC used in the inference of the time-resolved phylogeny (see Fig. 2 and Materials and Methods). The gray shading indicates the 95% highest posterior density of the estimated population size.

FIG 4

Distribution of MERS-CoV clades in time and space. (A) All available MERS-CoV genomes were stratified by phylogenetic clade (see Fig. 1) and plotted by virus sample date. The length of each clade was determined as the difference in days between the first and last observed sample of that virus and yielded the following values: Al-Hasa (21 April 2013 to 22 June 2013; 62 days), Riyadh_3 (5 February 2013 to 2 July 2013; 147 days), Buraidah_1 viruses (3 May 2013 to 5 August 2013; 84 days), and Hafr-Al-Batin_1 (4 June 2013 to 01 October 2013; 119 days). (B) All available MERS-CoV genomes were stratified by phylogenetic clade (see Fig. 1) and plotted by the case location. Cities are indicated by small black circles, and sequenced viruses by larger circles colored according to phylogenetic clade.

FIG 5

Substitutions in MERS-CoV spike proteins. All available KSA MERS-CoV spike ORFs were translated, the proteins aligned, and amino acid differences from the reconstructed ancestral clade B protein determined; changes observed in more than one genome are marked by vertical colored bars, with the new amino acid residue coded as shown at the bottom. Gray bars indicate a gap in sequence coverage. Functional domains of the spike (S) protein are marked and include the N-terminal domain, the receptor binding domain, the fusion domain (Fusion), heptad repeats 1 and 2 (HR1 and HR2) (20, 42), the transmembrane (TM) domain, and the cytoplasmic (Endo) domain (43).