Reliable typing of MERS-CoV variants with a small genome fragment.

BACKGROUND: Middle East Respiratory Syndrome coronavirus (MERS-CoV) is an emerging pathogen that causes lower respiratory tract infection in humans. Camels are the likely animal source for zoonotic infection, although exact transmission modes remain to be determined. Human-to-human transmission occurs sporadically. The wide geographic distribution of MERS-CoV among dromedary camels and ongoing transmissions to humans provides concern for the evolution of a MERS-CoV variant with efficient human-to-human transmission capabilities. Phylogenetic analysis of MERS-CoV has occurred by analysis of full-length genomes or multiple concatenated genome fragments, which is time-consuming, costly and limited to high viral load samples.
OBJECTIVE: To develop a simple, reliable MERS-CoV variant typing assay to facilitate monitoring of MERS-CoV diversity in animals and humans. STUDY
DESIGN: Phylogenetic analysis of presently known full-length MERS-CoV genomes was performed to identify genomic regions with sufficient phylogenetic content to allow reliable MERS-CoV variant typing. RT-PCR assays targeting these regions were designed and optimized.
RESULTS: A reverse-transcription PCR assay for MERS-CoV targeting a 615 bp spike fragment provides a phylogenetic clustering of MERS-CoV variants comparable to that of full-length genomes. The detection limit corresponds to a cycle treshold value of ∼ 35 with standard upE real time PCR assays on RNA isolated from MERS-CoV EMC. Nasal swabs from RT-PCR positive camels (Ct values 12.9-32.2) yielded reliable sequence information in 14 samples.
CONCLUSIONS: We developed a simple, reliable MERS-CoV variant typing assay which is crucial in monitoring MERS-CoV circulation in real time with relatively little investment on location.

Background

Middle East Respiratory Syndrome coronavirus (MERS-CoV; Family Coronaviridae) may cause severe lower respiratory tract infection in humans [1], [2]. Camels are considered the likely animal source for zoonotic infection; sporadic human-to-human transmission does occur, but is considered to be inefficient based on currently available data [3], [4], [5], [6], [7], [8], [9], [10]. Until recently, new MERS-CoV infections in humans were reported at a steady low rate reaching ∼200 confirmed human cases early 2014. In March and April 2014, however, a surge of MERS-CoV infections occurred mainly in hospitals around Jeddah, Kingdom of Saudi Arabia (KSA) and United Arab Emirates (UAE). This increased case load was in part attributed to an increase in community cases, but mostly to transmission within hospitals, with no evidence for evolution of a MERS-CoV variant with more efficient human-to-human transmission capabilities (WHO;http://www.who.int/csr/disease/coronavirus_infections/MERS_CoV_Update_09_May_2014.pdf?ua = 1). The ongoing occurrence of new cases, however, and the finding that MERS-CoV is endemic in dromedary camels in a wide geographic region [3], [10], [11], stresses the need for surveillance of strain diversity, to help unravel the epidemiology of this newly identified pathogen, and to provide a reference for studies into MERS-CoV evolution.

All currently sequenced human and camel MERS-CoV genomes share >99% nucleotide identity across the ∼30 kb genome. Phylogenetic analysis has occurred mainly by analysing full-length genomes or multiple concatenated genome fragments, to provide reliable phylogenetic information [5], [12], [13], [14]. However, full length genome sequencing capacity is not widely available, and requires relatively high viral load, leading to limited success when trying to sequence animal or human samples [5].

Objectives

An accurate typing of MERS-CoV variants, preferably with a simple assay encompassing a short region of the MERS-CoV genome, is crucial in monitoring the MERS-CoV outbreak in real time with relatively little investments on location. In this study, we describe the development of a MERS-CoV variant typing assay, which can be used in monitoring MERS-CoV circulation, especially when more information on virus type is required rapidly from a large number of viruses from animals/humans.

Study design

In silico analysis

Full or near full-length MERS-CoV genomes encompassing nucleotides 215–29770 (numbering corresponding to MERS-CoV EMC genome JX869059) were aligned with MAFFT version 7 (http://mafft.cbrc.jp/alignment/server/) (Table 1 ). To remove the redundancy from the dataset, FastGroupII analysis (http://fastgroup.sdsu.edu/fg_tools.htm) was performed grouping all currently available viral genomes based on nucleotide composition, resulting in 15 groups (Table 1). One viral genome from each group was taken as representative in subsequent analyses. A summary of nucleotide positions that vary across the genomes was created using BioEdit v7.2.0 [15] and the number of nucleotide positions with variations was plotted over 1000 nucleotide windows. PhyML trees were generated using Seaview 4 software with the approximate likelihood ratio test based on a Shimodaira–Hasegawa-like procedure which used general time reversible as substitution model. Nearest neighbor interchange, subtree pruning, and regrafting-based tree search algorithms were used to estimate tree topologies, as described previously [5].

Table 1

Viral genomes used in this study.

Group name	Sequences	Accession number
RIYADH_9_2013	RIYADH_9_2013	KJ156869
	RIYADH_2_2012	KF600652
AL-HASA_15_2013	AL-HASA_19_2013	KF600632
	BURAIDAH_1_2013	KF600630
	AL-HASA_2_2013	KF186566
	AL-HASA_21_2013	KF600634
	AL-HASA_25_2013	KJ156866
	AL-HASA_17_2013	KF600647
	AL-HASA_15_2013	KF600645
	AL-HASA_16_2013	KF600644
	AL-HASA_12_2013	KF600627
	AL-HASA_3_2013	KF186565
	AL-HASA_4_2013	KF186564
	AL-HASA_18_2013	KF600651
	AL-HASA_1_2013	KF186567
	AL-HASA_15_2013	KF600645
FRA/UAE_2012	KFU_HKU1_2013	KJ650297
	KFU_HKU19DAM2013	KJ650296
	KFU_HKU13_2013	KJ650295
	FRA/UAE_2012	KF745068
HAFR-AL-BATIN_6_2013	HAFR-AL-BATIN_6_2013	KJ156874
	HAFR-AL-BATIN_2_2013	KJ156910
	HAFR-AL-BATIN_1_2013	KF600628
RIYADH_5_2013	RIYADH_5_2013	KJ156944
	RIYADH_4_2013	KJ156952
	TAIF_1_2013	KJ156949
	JEDDAH_1_2013	KJ556336
	WADI-AD-DAWASIR_1_2013	KJ156881
QATAR_3_2013	QATAR_3_2013	KF961221
	QATAR_4_2013	KF961222
RIYADH_1_2012	RIYADH_1_2012	KF600612
	BISHA_1_2012	KF600620
RIYADH_3_2013	RIYADH_3_2013	KF600613
Egypt_Camel_NRCE-HKU205_2014	Egypt_Camel_NRCE-HKU205_2014	KJ477102
ENGLAND/QATAR_2012	ENGLAND/QATAR_2012	KC667074
EMC_2012	EMC_2012	JX869059
RIYADH_14_2013	RIYADH_14_2013	KJ156934
Munich/AbuDhabi_2013	Munich/AbuDhabi_2013	KF192507
Qatar_Camel_2_2014	Qatar_Camel_2_2014	KJ650098
JORDAN/N3_2012	JORDAN/N3_2012	KC776174
	Not in variant analysis
	KSA_CAMEL_363_2013	KJ713298
	KSA_CAMEL_376_2013	KJ713299
	KSA_Camel_378_2013	KJ713296
	KSA_CAMEL_503_2013	KJ713297
	KSA_CAMEL_505_2013	KJ713295
	Jeddah_Camel_1_2013	KF917527
	Jeddah_Human_1_2013	KF958702
	Florida/USA_2_Saudi Arabia_2014	KJ829365
	Indiana/USA-1_Saudi Arabia_2014	KJ813439

MERS-CoV typing RT-PCR

Total nucleic acids were isolated using an automated MagNAPure 96 extraction with the total nucleic acid isolation kit (Roche, Mannheim, Germany) as described previously [13]. The partial S2 domain of MERS-CoV spike (corresponding to nucleotides 23781–24395 of MERS-CoV EMC genome JX869059) was amplified with the OneStep RT-PCR Kit (Qiagen) using 1x QIAGEN OneStep RT-PCR Buffer, 400 μM of each dNTP, 2 μl QIAGEN OneStep RT-PCR Enzyme Mix and 0.6 μM of primers VS804 (5′-TCAGGTTGATCAACTTAATAGT-3′) and VS805 (5′-TTGAGTAATGCCAACACCGTT-3′) in a volume of 25 μl; 30 min 50 °C, 15 min 95 °C, 40 cycles of 0.5 min 94 °C, 0.5 min 50 °C, 1 min 72 °C, and a final extension of 10 min 72 °C. A nested PCR was performed using 1x PCR Buffer, 2 mM MgCl2, 200 μM of each dNTP, 20 μM primers VS804 and VS805 and 2.5 units HotStarTaq DNA polymerase (Qiagen) in a volume of 50 μl; 15 min 95 °C, 40 cycles of 0.5 min 94 °C, 0.5 min 50 °C, 1 min 72 °C, and a final extension of 10 min 72 °C. Amplicons were sequenced directly on both strands with the BigDye Terminator version 3.1 cycle sequencing kit on an ABI PRISM 3100 genetic analyser (Applied Biosystems).

Samples

The detection limit of the assay was determined on RNA isolated from 10x dilutions of cell culture derived MERS-CoV EMC_2012 (JX869059), as described above. Virus stocks were prepared as described previously [16]. RNA was isolated from serial 10-fold dilutions of MERS-CoV EMC_2012 [13]. Serial 10-fold dilutions of this RNA were amplified in parallel with the MERS-CoV variant typing assay described above and the upE and N gene real time PCR assays [17], [18]. The sensitivity of the MERS-CoV variant typing assay was expressed as cycle threshold value based on the upE real time PCR assay.

On May 13 and 15, 2014, the first two MERS-CoV infected patients in the Netherlands who became infected upon travel to Saudi Arabia were reported to WHO; throat swabs from these patients were available [19].

In February and April 2014, nasal swabs were taken from dromedary camels of different age and sex from a slaughterhouse in Doha, Qatar [13], which were available for this study.

Results

To identify genome regions for reliable phylogenetic analysis comparable to that of full-length genome, (near) full-length MERS-CoV genomes (Table 1) were aligned. Viral genomes that were 99.9% identical were grouped and one viral genome from each group was taken as representative in the analysis (Table 1) and a PhyML tree was generated (Fig. 1 A). Four major MERS-CoV clusters can be discerned, represented by MERS-CoV Qatar_3_2014, England/Qatar_2012, EMC_2012, and Egypt_Camel_NRCE-HKU205_2014 (KF961221, KC667074, JX869059, KJ477102), which have been identified previously in analyses based on full-length genomes or concatenated genome fragments [5], [12], [13], [14]. The number of single nucleotide polymorphisms (SNPs) across the 15 representative MERS-CoV genomes was plotted over 1000 nt windows (Fig. 1B). Four genome fragments, two located in ORF1a, one in S and one in ORF4b, showed a relatively high number of SNPs (Fig. 1B) and for this reason already had been used as concatenated genome fragment for phylogenetic analysis [12]. However, phylogenetic trees created for these four fragments separately did not accurately reflect the phylogenetic positions of the currently known full-length MERS-CoV genomes (data not shown).

Fig. 1

Characterization of MERS-CoV variation. (A) PhyML tree of full-length genome sequences of 15 distinct MERS coronavirus variants based on analysis of nucleotide sequence diversity across the genome. Four major MERS-CoV clusters are indicated (A, B1, B2, C). (B) The number of nucleotide positions with variations over 15 representative MERS-CoV genomes (Supplementary Table 1) was plotted over 1000 nucleotide windows. A schematic diagram of the MERS-CoV genome is depicted below the graph, and four fragments with the highest number of single nucleotide polymorphisms (SNPs) are indicated with black boxes. The dotted line indicates the average number of SNPs per 1000 nt window added with 1x standard deviation. In red, nucleotide positions displaying considerable phylogenetic information regarding four identified MERS-CoV variant clusters and the red box indicates the 615 bp fragment of the here-described MERS-CoV variant typing assay. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

In a subsequent analysis all identified SNPs were inspected visually and regions containing SNPs with phylogenetic information regarding the previously identified four clusters of viruses were identified (Fig. 1A). The S2 domain of the spike protein contains a number of these mutations (Fig. 1B). A fragment of 615 bp, containing three of these SNPs, provided a phylogenetic tree similar to the one obtained upon full-genome analysis regarding the previously identified four MERS-CoV clusters (Fig. 2 ), whereas other MERS-CoV genome regions did not provide similar results. As observed for the full-length genomes, human, and camel MERS-CoV genomes shared >99% nucleotide identity across the 615 bp S2 domain fragment. MERS-CoV genomes that were released recently and not taken along in the variation analysis were typed using the 615 bp S2 domain and clustered similar to their phylogenetic positions as upon full genome analysis (Fig. 2 and Table 1), thereby validating the assay.

Fig. 2

Phylogenetic analysis of human and camel MERS-CoVs using the MERS-CoV variant typing assay. A PhyML tree was generated from a spike S2 domain genome fragment corresponding to nt 23781–24395 of MERS-CoV EMC genome (accession number JX869059) for known MERS-CoV genomes (Table 1) and the first two Dutch MERS-CoV patients (boxed). In red are newly identified camel sequences from a slaughterhouse in Qatar. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

A sequencing RT-PCR targeting the identified genome fragment was developed and optimized using RNA isolated from cell culture-derived MERS-CoV. In limiting dilution experiments, the MERS-CoV variant typing assay amplified the 615 bp fragment down to a cycle treshold value of ∼35 as determined by diagnostic upE real time PCR assays [17], [18]. The MERS-CoV variant typing assay was used to type RT-PCR positive nose swabs from the first two human Dutch MERS-CoV cases in 2014. The results showed grouping consistent with previous findings based on long sequence fragments [19] (Fig. 2). In addition, the MERS-CoV variant typing assay was performed on camel samples from a slaughterhouse in Qatar [13] and sequences for 14 MERS-CoV positive animals with cycle threshold values ranging from 12.9 to 32.2 as determined by UpE real time RT-PCR [17], [18] were obtained (Fig. 2). Five different camel MERS-CoV variants in clusters B1 and B2 were detected, without the need for full genome sequencing.

Discussion

Phylogenetic analysis of representative MERS-CoV full-length genomes indicated that four regions in the MERS-CoV genome exist with a substantially higher nucleotide variation across genomes. However, phylogenetic analysis of these genome regions separately did not provide reliable phylogenetic information, in contrast to an analysis of the concatenated fragments [5], [12], [19]. Subsequent analyses revealed a region in the open reading frame that encodes the spike protein with a number of positions in which nucleotide variation occurs between MERS-CoV variants with a strong phylogenetic signal regarding previously identified clusters of viruses based on full-length MERS-CoV genomes. The here-described MERS-CoV variant typing assay based on a 615 bp spike fragment provides a crude indication of the MERS-CoV variant under study on the basis of four identified clusters of MERS-CoV variants, exemplified by MERS-CoV Qatar_3_2014, England/Qatar_2012, EMC_2012, and Egypt_Camel_NRCE-HKU205_2014. In silico, the assay accurately typed all currently known MERS-CoVs for which full-length genomes are available, including the recently released Florida/USA_2_KSA_2014 and Indiana/USA-1_KSA_2014 strains. This was confirmed by typing of samples available in our laboratory from the first two Dutch MERS-CoV cases that phylogenetically grouped with Indiana/USA-1_KSA_2014 and other viruses in cluster I [19] and from camels from a slaughterhouse in Doha, Qatar [13]. The observed detection limit of a cycle treshold value of ∼35 allows variant typing in clinical samples obtained from humans and animals with relatively low viral loads. It enables inclusion of samples in phylogenetic analysis that would not have been included when only full length genomes would have been accepted.

This MERS-CoV variant typing assay, targeting a part of the MERS-CoV spike gene, is a relatively simple RT-PCR sequencing assay that could be performed more widely as initial screening assay in laboratories with basic sequencing capacity. It provides accurate crude MERS-CoV type information, applicable in monitoring viral variants in real time. New variants identified through this initial screening could then be sent to a reference laboratory for further characterization. The continued occurrence of transmission between humans in health care and family settings is an ongoing concern as stated by the World Health Organization (http://www.who.int/csr/disease/coronavirus_infections/MERS_CoV_Update_27_March_2014.pdf?ua=1), although the outbreaks appear to be self‐limiting or extinguishable with rigorous implementation of appropriate infection control guidelines at present. However, as the primary route of transmission to humans is uninterrupted, human-to-human transmissions will continue to occur. The data obtained from the MERS-CoV variant typing assay would aid in informing the most effective international preparedness and response, allowing ad hoc risk assessment and implementation of containment strategies if necessary.

Conflict of interest

Saskia L. Smits is part time employed by Viroclinics Biosciences BV. This does not alter our adherence to all the policies on sharing data and materials.

Bart Haagmans has a patent on MERS-CoV field.

Funding

This work was funded by ZonMW TOP project 91213058.

Ethical approval

All procedures were performed in compliance with relevant laws and institutional guidelines and in accordance with the Declaration of Helsinki.

Authors’ contributions

All authors contributed to gathering and analysis of the information. Saskia Smits, Bart Haagmans, and Marion Koopmans drafted and revised the manuscript based on all authors contributions.