Human Infection With MERS Coronavirus After Exposure to Infected Camels

Saudi Arabia, 2013

Ziad A. Memish; Matthew Cotten; Benjamin Meyer; Simon J. Watson; Abdullah J. Alsahafi; Abdullah A. Al Rabeeah; Victor Max Corman; Andrea Sieberg; Hatem Q. Makhdoom; Abdullah Assiri; Malaki Al Masri; Souhaib Aldabbagh; Berend-Jan Bosch; Martin Beer; Marcel A. Müller; Paul Kellam; Christian Drosten


Emerging Infectious Diseases. 2014;20(6) 

In This Article

The Study

On November 3, 2013, the Ministry of Health of Saudi Arabia was notified of a suspected case of MERS-CoV infection in a 43-year-old male patient at King Abdulaziz University Hospital in Jeddah. The patient had cared for ill camels in his herd of 9 animals starting in early October, when the patient noted respiratory signs of illness with nasal discharge in several animals; he continued caring for the sick animals until October 27, the day of onset of his own illness. The patient cared for the animals for ≈3 hours per day 3 days per week, applying herbal remedies to the animals' snouts and nostrils. He did not clean the stables or milk the animals, but he routinely consumed raw, unpasteurized camel milk from the herd.

Presence of MERS-CoV RNA in the patient was confirmed at Jeddah Regional Laboratory by using reverse transcription PCR (RT-PCR) targeting the upE and orfA gene fragments.[11,12] Respiratory swab specimens yielded detectable signal after 28 RT-PCR cycles, indicative of an approximate viral load of 350,000 RNA copies per sample. A nearly complete viral genome was obtained (Jeddah_1_2013; GenBank accession no. KJ556336), confirming the presence of a typical MERS-CoV whose closest relatives were in the Riyadh_3 clade, as defined in[2] (phylogeny shown in online Technical Appendix Figure 1,

To identify potential sources of infection, on November 9, the Ministry of Health investigated 5 close household contacts and the animal attendant on a farm owned by the patient. Nasopharyngeal swab samples were taken and tested at Jeddah Regional Laboratory by using RT-PCR. Deep nasal swab specimens were taken on the same day from 3 of the 9 camels at the farm. Testing of all samples by RT-PCR using the upE assay)[11,12] did not detect MERS-CoV RNA in any of the human patients, but 1 of the 3 camels (camel G) tested positive (cycle threshold [Ct] = 33). On November 13, nasal swab samples were obtained from all 9 animals. upE RT-PCR results were positive for camel G (Ct = 38) and a second camel (camel B; Ct = 39).

Samples from November 13 and a small remaining amount of RNA extract from camel G from November 9 were sent to the Sanger Institute in Cambridge, UK, and confirmation of reactivity (Ct ≈ 38) was obtained for pooled samples with the upE assay from camel G but not for camel B. The result for camel G was confirmed at the Institute of Virology in Bonn, Germany, for the same samples by using real-time RT-PCRs targeting the upE and 1A diagnostic target regions.

A sequence of ≈15% of the camel-derived genome was determined from 8 RT-PCR fragments, 2 of them partially overlapping (4,608 nt total).[13] Phylogenetic analyses supported the conclusion that transmission occurred between camel and patient, but no direction was implied (e.g, camel to human vs. human to camel; online Technical Appendix Figure 1). The human- and camel-derived sequences shared a signature of single nucleotide polymorphisms that occurred in no other known MERS-CoV sequences (Figure). Single nucleotide exchanges occurred at nt positions 21945 and 29662; these exchanges might have come up during virus transmission, as described.[10] However, because of the low RNA concentration in the samples, reamplification of the material and investigation of possible PCR-based mutations could not be done.

Figure 1.

Direct comparison of the Middle East respiratory syndrome coronavirus (MERS-CoV) Jeddah_1_2013 genome sequence, Jeddah_ Camel1_2013 fragments (boxes at bottom), and representative genomes of other clade viruses: 2 additional genomes from the Riyadh_3 clade, Riyadh_3_2013 and Taif_1_2013; and representative genomes from the Al-Hasa and Hafr-Al-Batin_1 and Buraidah_1 clades. A map of the MERS-CoV genome with the major open reading frames (ORFs) indicated is shown at the top. Nucleotide differences for other genomes from Jeddah_1_2013 are shown by vertical colored bars: orange, change to A; red, change to T; blue, change to G; violet, change to C. Gaps in all full-genome sequences are indicated in gray. Hashed index lines indicate the boundaries of the fragments. Positions according to the MERS-CoV genome EMC/2012: fragment 1, 9767–10354; fragment 2, 17507–18394; fragment 3, 21089–22046; fragment 4, 23569–24059; fragment 5, 25349–26056; fragment 6, 27276–28095; fragment 7, 29596–29757. Fragments have been deposited in GenBank (accession nos. KJ556337–KJ556340; others are pending).

Technical Appendix Figure 1.

Bayesian-inferred time-resolved maximum-clade credibility tree of 66 human MERS-CoV genomes and 3 camel genomes or genome-fragments. Highlighted in orange is the partial camel sequence generated in this study (Jeddah_Camel1_2013, GenBank accession number pending), and in blue is the MERS-CoV genome from the camel owner (Jeddah_1_2013, GenBank accession number pending). Additional camel MERS-CoV sequences are indicated in dark red. Asterisks indicate nodes with a posterior probability greater than 0.9. The phylogeny was inferred using BEAST version 1.8.0 under an HKY+G substitution model, an uncorrelated lognormal molecular clock, and a GMRF skyride coalescent model. Trees were visualized and annotated using FigTree version 1.4.0 (

A serum sample taken from camel G during the initial investigation on November 9 was tested by recombinant immunofluorescence assay (IFA) as described[6,14] and showed reactivity to MERS-CoV (titer 320). To investigate signs of recent MERS-CoV infection in the group of camels, we obtained blood samples at short intervals from all 9 animals during November 14–December 9, 2013. All samples were tested by ELISA against recombinant MERS-CoV spike antigen domain S1 fused to human Fc fragment, using a formulation as described.[15] Serum samples from all 9 animals showed reactivity to the MERS-CoV antigen; serum samples from control animals showed no reactivity (online Technical Appendix Figure 2, panel A). ELISA signals were constant over time in most animals, but a small, yet visible, change of signal over the observation period was noted for camels B and G. To clarify the reasons for this putative signal increase, the first and last serum samples (obtained on November 14 and December 9) from all animals were re-tested by IFA. As summarized in the Table, camels B and G showed 4-fold increases of titer for the paired first and last serum samples. In serologic tests that rely on 2-fold endpoint titrations, a titer increase >2 dilution steps is considered a significant sign of recent acute infection. For additional confirmation of rises of titer, sequential samples from camels B and G were compared with sequential samples from camels E and I by using IFA with endpoint titration. These data confirmed the increases of titers for camels B and G (online Technical Appendix Figure 2, panel B).

Technical Appendix Figure 2.

Serologic screening of nine dromedary camels from Jeddah over a period of 4 weeks. A, Dromedary camel sera were tested by a MERS-CoV spike protein S1-subunit based ELISA at a dilution of 1:100. Optical density was measured at 450 nm. The cutoff (dashed line) was set to three times the OD450 value of a selection of seronegative camel sera. Results from camel sera previously tested negative [7] are shown for comparison. B, Reciprocal endpoint titers as determined by recombinant immunofluorescence assay for animals with rise in titer (B, G) and animals with constant titer (E, I).

Because bovine CoV occurs in camels, we tested for antibodies against bovine CoV in camels B and G to exclude potential cross-reactions. Using IFA,[6] we found no bovine CoV antibodies in camel G, but camel B showed rising bovine CoV antibody titers (Table). To obtain further differentiation, we performed neutralization assays against MERS-CoV and bovine CoV.[7] Titers in serum samples from November 14 and December 9, respectively, were 160 and 320 for camel B and 160 and 160 for camel G. None of the animals showed serum neutralization against bovine CoV.