Emerging Microbes & Infections 2022, VOL. 11 https://doi.org/10.1080/22221751.2021.2019559 CORONAVIRUSES Enhanced replication fitness of MERS-CoV clade B over clade A strains in camelids explains the dominance of clade B strains in the Arabian Peninsula Nigeer Te a*, Jordi Rodon a *, Mónica Péreza, Joaquim Segalés b,c , Júlia Vergara-Alert a and Albert Bensaid a a IRTA, Centre de Recerca en Sanitat Animal (CReSA, IRTA-UAB), Bellaterra, Cerdanyola del Vallès, Spain; bUAB, CReSA (IRTA-UAB), Bellaterra, Cerdanyola del Vallès, Spain; cDepartament de Sanitat i Anatomia Animals, Facultat de Veterinaria, UAB, Bellaterra, Cerdanyola del Vallès, Spain ABSTRACT Middle East respiratory syndrome coronavirus (MERS-CoV) continues infecting humans and dromedary camels. While MERS-CoV strains from the Middle East region are subdivided into two clades (A and B), all the contemporary epidemic viruses belong to clade B. Thus, MERS-CoV clade B strains may display adaptive advantages over clade A in humans and/or reservoir hosts. To test this hypothesis in vivo, we compared an early epidemic clade A strain (EMC/ 2012) with a clade B strain (Jordan-1/2015) in an alpaca model monitoring virological and immunological parameters. Further, the Jordan-1/2015 strain has a partial amino acid (aa) deletion in the double-stranded (ds) RNA binding motif of the open reading frame ORF4a protein. Animals inoculated with the Jordan-1/2015 variant had higher MERS-CoV replicative capabilities in the respiratory tract and larger nasal viral shedding. In the nasal mucosa, the Jordan-1/2015 strain caused an early IFN response, suggesting a role for ORF4a as a moderate IFN antagonist in vivo. However, both strains elicited maximal transcription of antiviral interferon-stimulated genes (ISGs) at the peak of infection on 2 days post inoculation, correlating with subsequent decreases in tissular viral loads. Genome alignment analysis revealed several clade B-specific amino acid substitutions occurring in the replicase and the S proteins, which could explain a better adaptation of clade B strains in camelid hosts. Differences in replication and shedding reported herein indicate a better fitness and transmission capability of MERS-CoV clade B strains than their clade A counterparts. ARTICLE HISTORY Received 11 August 2021; Revised 7 December 2021; Accepted 13 December 2021 KEYWORDS Alpaca; IFN; Middle East respiratory syndrome coronavirus; MERS-CoV; orf4a; strain Introduction Although limited human-to-human transmission has Middle East respiratory syndrome coronavirus been reported within families, contagion mostly (MERS-CoV) causes a severe respiratory disease in occurred in health-care settings [13]. humans. The Arabian Peninsula continues to be the MERS-CoV is phylogenetically divided into three global epicenter of the disease [1,2]. As of November major clades, temporarily named clade A, B and C 2021, the World Health Organization (WHO) [14–16]. Sequence examination of various MERS- reported 2,578 infections and 888 fatalities (∼34.4%) CoV strains revealed that nowadays only clade B in 27 countries [3]. While most cases have been docu- viruses are circulating in the Arabian Peninsula and mented in The Kingdom of Saudi Arabia, a major tra- have replaced clade A in humans and dromedary vel-associated outbreak in South Korea reminds that camels [16–18]. Therefore, clade B viruses might pos- MERS-CoV is of global concern [4,5]. Due to negative sess an evolutional advantage against clade A isolates, impacts on public health and the economy, MERS- exhibiting enhanced replicative fitness when infecting CoV is listed as one of the WHO Research and Devel- human/dromedary hosts. However, data characteriz- opment Blueprint priority pathogens. ing phenotypic and pathogenic differences between So far, the dromedary camel is the only confirmed these two clades of MERS-CoV are scarce. Recently, source of zoonotic infection, while new world came- Wang et al. [19] showed that a clade B virus (Chi- lids (alpacas and llamas) are also susceptible to naGD01) led to more weight loss, severe lung lesions MERS-CoV under both natural and experimental con- and higher viral titres in the lung of human dipeptidyl ditions [6–11]. The human acquisition is primarily peptidase-4 (h-DPP4) transduced mice than did the due to close contact with infected camels [12]. prototype virus EMC/2012 (clade A). However, in a CONTACT Júlia Vergara-Alert julia.vergara@irta.cat IRTA, Centre de Recerca en Sanitat Animal (CReSA, IRTA-UAB), Campus de la UAB, 08193 Bellaterra, Cerdanyola del Vallès, Spain *Both authors are co-first authors. Supplemental data for this article can be accessed at https://doi.org/10.1080/22221751.2021.2019559 © 2022 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrest- ricted use, distribution, and reproduction in any medium, provided the original work is properly cited. EMERGING MICROBES & INFECTIONS 261 previous study, apparent contradictory results were chemokines along the respiratory tract was monitored reported since clade A and B viruses exhibited compar- for both strains during three consecutive days. Thus, able lung titres and pathology in h-DPP4 transduced together with previous results obtained using the mice [14]. Also, there was no significant difference in Qatar15/2015 isolate [29], a rational explanation is replication competence of these two clade viruses in suggested for the predominance of MERS-CoV clade human bronchial and lung explants [20]. B strains in the Arabian Peninsula. Despite its role in antagonizing interferon (IFN) defense mechanisms [21–23], the accessory protein ORF4b is probably not implicated in virus replication Materials and methods in vivo [14]. Rather, non-structural proteins (nsp), Ethics statement which also interfere with IFN production [24,25], are thought to play a more important role in virus replica- All animal experiments were approved by the Ethical tion than accessory proteins. In that respect, clade and Animal Welfare Committee of IRTA (CEEA- B-specific mutations have been described in ORF1ab IRTA) and by the Ethical Commission of Animal [19]. Also, the role of ORF4a as an IFN antagonist Experimentation of the Autonomous Government has not been assessed in vivo but mainly determined of Catalonia (file N° FUE-2018-00884575 – Project under in vitro conditions [21,26–28]. A recombinant N°10370). The work with infectious MERS-CoV virus lacking ORF4a showed increased expression levels was performed in the biosafety level-3 facilities of type I and III IFNs and slightly decreased replication (BSL-3) of the Biocontainment Unit of IRTA- in human airway epithelium-derived A549 cells [27]. CReSA in Barcelona, Spain. Nevertheless, whether deletions/mutations of these viral factors lead to increased zoonotic potential of Cell culture and virus clade B strains lacks direct proof in reservoir hosts. Recently, we characterized the pathogenesis of a MERS-CoV EMC/2012 (GenBank Accession clade B MERS-CoV strain (Qatar15/2015) in an alpaca JX869059) and Jordan-1/2015 (GenBank Accession model [29]. Upon MERS-CoV infection, few infected KU233364) isolates were propagated in Vero E6 cells nasal epithelial cells could be detected by immunohis- and titrated, as previously described [10]. Viral stocks tochemistry (IHC) as soon as 1-day post inoculation used for experimental infection procedures were pro- (dpi). Nasal tissues were maximally infected on 2 dpi vided by Dr Bart Haagmans from the Erasmus Medi- and infection was extended to the bronchus, where cal Center (Rotterdam, The Netherlands) and were some infected epithelial cells were observed. On the sequenced prior shipment to IRTA-CReSA. No following days, the virus was gradually cleared. Alpa- mutations were found and virus was not propagated cas responded at the peak of MERS-CoV infection before inoculation to alpacas. with early and transient type I and III IFN responses detected exclusively in the nasal mucosa, concomitant Animal study and experimental design with an induction of antiviral ISGs along the whole respiratory tract. Exacerbated inflammation was not Twenty-one 6-8-month-old alpacas (Vicugna pacos) observed in these animals, which displayed attenuated were purchased and randomly assigned a number NF-κB and NLRP3 signalling cascades in conjunction (AP1-AP21). AP19-AP21 were kept as non-infected with increased mRNA levels of the anti-inflammatory controls and euthanized upon arrival. The remaining cytokines IL10 and type III IFNs. animals were acclimated for one week at the animal To assess differences in virulence, replication and BSL-3 facilities. Alpacas, AP1-AP9 (group 1) and pathogenesis explaining field observations that AP10-AP18 (group 2) were housed in two different MERS-CoV clade B strains have replaced clade A in pens and intranasally inoculated with a 10 [7] camelids, we selected an early epidemic clade A iso- TCID50 dose of the MERS-CoV EMC/2012 and Jor- late, EMC/2012 [30], and a clade B strain, Jordan-1/ dan-1/2015 strains respectively, as described pre- 2015 [31], to experimentally infect alpacas. The Jor- viously [10]. Three animals from each group were dan-1/2015 strain harbours a 16 amino acid (aa) del- euthanized per day with an overdose of pentobarbital etion in ORF4a suspected to affect the recognition of followed by exsanguination on 1, 2, and 3 dpi. Clinical dsRNA by host RNA sensors [31]. Therefore, this signs were monitored and nasal swab (NS) samples experimental set up also allowed to gain insights on were collected on the day of euthanasia for MERS- the role of ORF4a in MERS-CoV pathogenesis. Alpa- CoV titration and RNA detection, as previously cas inoculated with Jordan-1/2015 exhibited higher described [10]. Complete necropsies were also per- MERS-CoV loads in the respiratory tract and larger formed. Collected respiratory tissues included: nasal viral shedding, demonstrating the increased replica- turbinates (frontal, medial and caudal), large and tive fitness of clade B strains in a camelid host. In small bronchus, and lung parenchyma (frontal, medial addition, mRNA expression of cytokines and and caudal). Respiratory tissues were immersed in 262 N. TE ET AL. media for viral RNA detection by RT-qPCR; or fixed section, as previously described [29]. Briefly, four con- in formalin or methacarn for immunohistochemistry secutive sections from the same MFPE-block contain- [10] and cytokine quantification respectively [29]. ing nasal specimens were cut. One of the sections was subjected to IHC to localize infected/non-infected cells in the tissues and served as a reference (template) Viral genomic and subgenomic RNA detection for the three subsequent sections which were stained by RT-qPCR with cresyl violet and then subjected to LCM (Sup- NS and tissue samples were processed as previously plementary Figure 1). described [10]. Following euthanasia, each respiratory tract tissue was cut into 0.5 × 0.5 cm pieces (weighing approximately 200–300 mg) per each sample at differ- Total RNA isolation and cDNA synthesis ent dpi and homogenized using a TissueLyser II (Qia- Total RNA from micro-dissected or scraped tissues gen, Germany). Viral RNA from both NS and tissue was extracted, and converted into cDNA synthesis, homogenates was extracted using a NucleoSpin® following previous standard protocols [29]. RNA virus kit (Macherey-Nagel) following the manu- facturer’s instructions. Genomic and subgenomic MERS-CoV RNA was quantified by the UpE [32] Relative quantification of cytokines mRNA and and M mRNA [33] RT-qPCR assays, respectively. viral genomic and subgenomic loads on MFPE Samples with a cycle threshold of less than 40 were tissues considered positive. A microfluidic RT-qPCR assay was used to quantify gene expression from the MFPE samples. The selec- Virus isolation from NS tion of innate immune genes and specific pair of pri- NS samples collected were evaluated for the presence mers to amplify their transcriptional products of infectious virus at various dpi by titration in Vero (Supplementary Table S1) were described in a pre- E6 cells, as previously reported [10]. The amount of vious work [29]. In addition, specific primers for infectious virus in each sample was calculated by the genomic (UpE) and subgenomic (M mRNA) determining the TCID50/mL. viral regions were added to the assay [29]. Data were analysed with the software 4.1.3 (Fluidigm Corporation, South San Francisco, USA) and the Histopathology and immunohistochemistry DAG expression software 1.0.5. 6, as previously Respiratory tissue samples for pathological studies described [29]. The up-or down-regulated were fixed by immersion in 10% neutral-buffered for- expression of each cytokine gene was expressed in malin, embedded in paraffin, sectioned at 2.5 µm for fold change (Fc), all obtained values are available slides and stained with haematoxylin and eosin. A in the Supplementary Data S1. The quantification monoclonal mouse anti-MERS-CoV N protein anti- of MERS-CoV genomic (UpE) and subgenomic (M body (Sino Biological Inc., Beijing, China) was used mRNA) RNA, obtained with the microfluidic to detect the presence of MERS-CoV antigen, as qPCR assay, expressed in Cq values, are also avail- described before [10,34]. able in the Supplementary Data S1. Methacarn-fixed, paraffin-embedded (MFPE) Statistical analysis tissue specimens To use the Student’s t-test, a logarithmic 10 transform- For the assessment of cytokine mRNA profiles, nasal ations were applied on Fc values to approach a log- turbinates, trachea and lung were fixed by immersion normal distribution. Thus, the means of the logarith- in methacarn and paraffin embedded. MFPE-tracheal mic Fc obtained at different dpi for each group of ani- and lung samples were processed for RNA isolation mals could be compared. Significant upregulation or by scraping the whole section from the prepared slides downregulation of genes was considered if they met and referred to as “scraped” tissues. MFPE-nasal the criteria of a relative Fc of ≥2-fold or ≤2-fold specimens were processed for laser capture microdis- respectively with P < 0.05. The Shapiro-Wilk normal- section (LCM) prior to RNA extraction. All these pro- ity test was used to determine the normal distribution cedures were previously established [29]. of RT-qPCR data (Cq values), prior to applying Tukey’s multiple comparisons test to compare Cq values of tissue samples at different dpi. Differences LCM were considered significant at P < 0.05. Correlation LCM was used to micro-dissect IHC positive and coefficients were determined using Spearman’s corre- negative nasal epithelia areas from the same tissue lation test. EMERGING MICROBES & INFECTIONS 263 Alignment of MERS-CoV proteins from clade A Nasal viral shedding of alpacas upon MERS-CoV and B strains challenge The aa sequences from twenty-one MERS-CoV strains NS were obtained from all animals on the day of encompassing seven lineages (including the EMC/ euthanasia (0, 1, 2 and 3 dpi). All MERS-CoV inocu- 2012 and the Jordan-1/2015 strains) were retrieved lated alpacas shed viral RNA at all time points. At from the NCBI database and aligned using the Uni- 2 dpi, Cq values of NS were statistically lower for Prot alignment tool (https://www.uniprot.org/align/). alpacas inoculated with the Jordan-1/2015 strain Details including the accession number of the strains when compared to those challenged with the are provided in Table 1. EMC/2012 strain. NS were subsequently tested for the presence of infectious virus in Vero E6 cells. Consistently, on 2 dpi, alpacas infected with the Jor- Modelling of MERS-CoV PLpro and S proteins dan-1/2015 strain had remarkably higher viral titres The crystal structures of MERS-CoV PLpro and S pro- than animals infected with EMC/2012. At both 2 teins were obtained from the PDB (PDB ID: 4REZ for and 3 dpi, only one out of three animals inoculated PLpro and PDB ID:5W9H for the S protein). Mutation with the EMC/2012 strain shed infectious virus sites were mapped onto structures using PyMOL (Figure 1). Jordan-1/2015 strain, as opposed to the (PyMOL molecular graphics system, version 2.5.0; EMC/2012 strain, replicated better in the alpaca Schrödinger, LLC). model, suggesting an increased transmission poten- tial for clade B strains. Results MERS-CoV Jordan-1/2015 strain possesses an Clinical signs increased epithelial tropism along the respiratory tract compared to the EMC/2012 To compare the pathology induced by strains EMC/ strain 2012 (clade A) and Jordan-1/2015 (clade B) two groups of alpacas, defined as groups 1 and 2 respect- To gain further information on viral replication and ively, were intranasally inoculated with each strain. tissue tropism we performed IHC for MERS-CoV Three animals from each group were sacrificed per detection along alpaca respiratory tracts infected day and during three consecutive days. This timing with both strains. Viral antigen (N protein) was multi- was based on previous experiments performed in alpa- focally located in the nasal turbinates of alpacas inocu- cas with the EMC/2012 [6] and Qatar15/2015 (clade lated with the MERS-CoV EMC/2012 strain (Figure 2 B) [29] strains where maximal viral shedding loads (a)). On 1 and 3 dpi, MERS-CoV positive nasal epi- in tissues were observed at 2 dpi. None of the animals thelial cells were only detected in one out of three ani- showed clinical signs and basal body temperatures mals whereas on 2 dpi all three animals harboured remained normal (below 39.5°C) in response to both viral antigen in their nose. By contrast, the virus was MERS-CoV strains throughout the study. found in the nasal epithelium of all alpacas inoculated Table 1. Information of the different MERS-CoV strains used for amino acid sequence alignment. Name of strain Origin Clade Lineage Accession Number (NCBI) EMC/2012 Human A 1 JX869059 Jordan-N3/2012 Human A 1 KC776174 KSA_CAMEL-376 Dromedary camel B 1 KJ713299 Camel/UAE/D1209/2014 Dromedary camel B 1 KP719933 Abu Dhabi_UAE_18_2014 Human B 1 KP209307 Al-Hasa25/2013 Human B 2 KJ156866 Riyadh_2_2012 Human B 2 KF600652 England-KSA/1/2018 Human B 3 MH822886 Camel/Riyadh/Ry23N/2014 Dromedary camel B 3 KT368825 Makkah_C9355/KSA/Makkah/2014-04-15 Human B 4 KM027261 KSA-CAMEL-505 Dromedary camel B 4 KJ713295 Hu/UAE_032_2014 Human B 4 KY581693 Jeddah/D90/2014 Dromedary camel B 4 KT368844 Qatar15/2015 Human B 5 MK280984 Jordan-1/2015 Human B 5 KU233364 ChinaGD01 Human B 5 KT006149 Riyadh_1_2012 Human B 6 KF600612 Bisha_1_2012 Human B 6 KF600620 Camel/UAE_B8_2015 Dromedary camel B 7 MF598600 D2731.3/14 Dromedary camel B 7 KT751244 Camel/UAE_B65_2015 Dromedary camel B 7 MF598655 264 N. TE ET AL. MERS-CoV Jordan-1/2015 infected cells were unevenly detected within the tracheal and bronchial mucosa on 2 dpi by IHC (Figure 2(b)). Also, viral anti- gen was only scarcely found in bronchiolar epithelial cells of a single animal on 3 dpi (Figure 2(b) and Sup- plementary Table S2). By contrast, tissues from the trachea and the lower respiratory tract (LRT) were all negative by IHC in alpacas inoculated with the MERS-CoV EMC/2012 strain (Figure 2(b) and Sup- Figure 1. Viral loads in nasal swabs (NS) of MERS-CoV EMC/ 2012 and Jordan-1/2015 infected alpacas. Viral RNA (left) plementary Table S2). and infectious MERS-CoV (right) loads from NS samples col- Both MERS-CoV isolates led to similar degrees of lected at different time points. Viral RNA shedding was deter- histological lesions that were mainly limited to the mined by the UpE RT-qPCR assay. Each bar represents the nasal turbinate, being of multifocal distribution and mean Cq value ± SEM of NS from 3 animals euthanized on 0, mild. From 1 to 3 dpi, lesions in the nasal turbinates 1, 2 and 3 dpi, respectively. Statistical significance was deter- were characterized by mild rhinitis, segmental hyper- mined by Tukey’s multiple comparisons test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. Dashed lines depict the plasia of the nasal epithelium, lymphocytic exocytosis detection limit of the assays. Abbreviations: Cq, quantification and epithelial vacuolation. Additionally, small num- cycle; MERS-CoV, Middle East respiratory syndrome corona- bers of lymphocytes with fewer macrophages infil- virus; TCID50, 50% tissue culture infective dose. trated the underlying submucosa (Figure 2(a)). No notable microscopic lesions were observed in the tra- with the Jordan-1/2015 strain throughout the whole chea and bronchus in response to both strains. study (Supplementary Table S2). While a moderate Besides, lung lobes showed very mild multifocal peri- number of pseudostratified columnar epithelial cells vascular and peribronchiolar infiltration by lympho- in the nose contained MERS-CoV Jordan-1/2015 anti- cytes in few animals. Control animals euthanized on gen on 1 dpi, the number of positive epithelial cells day 0 (AP19 to AP21) did not display any significant became remarkably high on 2 dpi; such number stea- histological lesion in collected tissues and IHC dily decreased in the alpacas necropsied on dpi 3 resulted negative in all of them (Supplementary (Figure 2(a)). Table S2). Thus, the Jordan-1/2015 strain showed a Figure 2. Histopathological changes and viral detection in respiratory tracts of alpacas inoculated with two MERS-CoV strains. Respiratory tissues (nasal turbinate, trachea, bronchus and lung) were harvested upon necropsy and immediately fixed in 10% neutral-buffered formalin. (a) Immunohistochemical (IHC) visualization of apical turbinate of alpacas infected with MERS-CoV EMC/2012 (left panel) and Jordan-1/2015 (right panel) euthanized and necropsied on 1 to 3 dpi. H/E staining of the apical nose collected on 2 dpi for both strains was displayed at the bottom. (b) IHC results of trachea, bronchus and lung of alpacas infected with MERS-CoV EMC/2012 (left panel) and Jordan-1/2015 (right panel) necropsied on 2 or 3 dpi. See Supplementary Table S2 for the detailed distribution of MERS-CoV antigen. Abbreviations: H/E, haematoxylin and eosin stain. EMERGING MICROBES & INFECTIONS 265 better replication competence than did EMC/2012 were significantly higher in nasal turbinates of the Jor- along the alpaca respiratory tract, despite similar his- dan-1/2015 inoculated alpacas. In contrast to group 2 topathological alterations provoked by both strains. animals, subgenomic RNA was not detected in the tra- chea, bronchus and lung of alpacas inoculated with the EMC/2012 strain (Supplementary Figure 2(a)). These MERS-CoV RNA loads in alpaca respiratory data corroborated results obtained by IHC and tracts confirmed that the Jordan-1/2015 strain replicated Upon infection with both strains, MERS-CoV geno- better than the EMC/2012 strain in all respiratory tis- mic RNA was detected in non-fixed respiratory tissue sues from alpacas, in particular during the peak of specimens. The highest loads were found in nasal tur- infection (2 dpi). binates of Jordan-1/2015-inoculated animals at 2 and Loads of MERS-CoV genomic UpE and subge- 3 dpi when compared to the EMC/2012 infected indi- nomic M mRNA were also evaluated by a microfluidic viduals. Viral loads were lower in the trachea, RT-qPCR assay on individual MFPE respiratory tissue bronchus and lung for both strains, but animals that sections or micro-dissected mucosa that were used received the Jordan-1/2015, exhibited significantly also for cytokine mRNA quantification. Since the higher rates of infection for the duration of the exper- nasal mucosa is the major driver of innate immune iment (Figure 3(a)). responses in MERS-CoV infected alpacas [29], LCM To ascertain active viral gene translation/replica- was used to isolate MERS-CoV infected/non-infected tion within the whole respiratory tract, the presence nasal epithelial areas based on IHC examination. of MERS-CoV subgenomic RNA in tissue macerates Also, MFPE trachea and lung sections were directly was assessed. Throughout the study, M mRNA loads scrapped from the slide. Due to a lower infectivity of Figure 3. Viral loads in respiratory tissues of MERS-CoV EMC/2012 and Jordan-1/2015 infected alpacas. (a) Genomic viral loads from respiratory tissues of alpacas obtained at 0, 1, 2 and 3 dpi (1 to 3 dpi, n = 3 per day and per strain). Black circles: frontal turbinate, apical trachea and apical lung; blue circles: medial turbinate, medial trachea and medial lung; red circles: caudal turbi- nate, caudal trachea and caudal lung. Green and yellow circles represent values obtained from large and small bronchus, respect- ively. Genomic viral RNA loads were determined by the UpE RT-qPCR assay. Each bar represents the mean Cq value ± SEM of infected tissues from 3 animals euthanized on 0, 1, 2 and 3 dpi, respectively. Statistical significance was determined by Tukey’s multiple comparisons test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. (b) Genomic viral RNA loads in MFPE samples. Micro- dissected MFPE-nasal epithelia were prepared based on an overlapping template section stained by IHC to localize the MERS-CoV N protein as described in Supplementary Figure 1. Nasal samples from AP14 sacrificed on 2 dpi could not be processed due to bad preservation. The trachea and lung were scrapped from MFPE tissue sections. RNA extracted from these samples were converted into cDNA and the MERS-CoV genomic RNA loads were determined by UpE primers in a microfluidic PCR assay (Fluidigm Biomark). Error bars indicate SEM when results were positive in more than one animal. Dashed lines depict the detection limit of the assays. Abbreviations: MPNE, MERS-CoV IHC positive nasal epithelia; MNNE, MERS-CoV IHC negative nasal epithelia; MFPE, methacarn- fixed paraffin embedded-tissues; LCM, laser capture microdissection. Cq, quantification cycle. Dashed lines depict the detection limit of the assays. 266 N. TE ET AL. the EMC/2012 strain, extended nasal epithelial areas (a), Supplementary Figure 3(a, b) and Supplementary labelled by IHC could only be harvested from two ani- Data S1). Additionally, transcription levels of genes mals (AP4 on 2 dpi and AP9 on 3 dpi). Nasal speci- coding for the NLRP3 inflammasome components mens from animal AP14, infected with the Jordan-1/ NLRP3 and PYCARD, downstream signalling 2015 strain, could not be examined due to bad preser- adaptors (NFκBIA and CARD9), enzymes (CASP10, vation. All other nasal epithelia retrieved from animals AZI2, PACT and TBK1) and the IFNLR1 receptor inoculated by the Jordan-1/2015 strain could be pro- fluctuated around basal values independently of the cessed as expected. As shown in Figure 3(b) and Sup- animals, days of sample collection and the strain plementary Figure 2(b), viral genomic and used for infection (Supplementary Figure 3(a) and subgenomic RNA loads found in MERS-CoV IHC Supplementary Data S1). positive nasal epithelial (MPNE) areas were higher Type I and III IFNs were not detected or marginally than those displayed by MERS-CoV IHC negative upregulated in MNNE areas from alpacas inoculated nasal epithelial (MNNE) areas. In addition, MPNE by both strains. On 2 dpi, corresponding to the peak micro-dissected areas of both strains showed compar- of infection, expression of ISGs, RLRs, STAT1, IRF7, able viral RNA loads, while the Jordan-1/2015 strain IL8, CCL2, CCL3 and CASP1 were moderate to highly had a slightly better replication in MNNE areas than induced in MNNE areas from animals infected by that of EMC/2012. Similar to results obtained from tis- both strains while most of the remaining genes were sue homogenates (Figure 3(a)), genomic and subge- unaltered or slightly down-regulated upon infection nomic viral RNA was absent in trachea and lung (Figure 4(b), Supplementary Figure 3(a, c) and Sup- sections of alpacas inoculated with the EMC/2012 plementary Data S1). Again, high variations between strain. In contrast, genomic and subgenomic RNA animals within groups were noticed. loads were observed in trachea and lung section of ani- In a previous study [29], we found that transcrip- mals infected with the Jordan-1/2015 strain (Figure 3 tional levels of IFNs were positively correlated with (b) and Supplementary Figure 2(b)). increased viral loads in the nasal mucosa of alpacas inoculated with the MERS-CoV Qatar15/2015 strain. To confirm this relation for the Jordan-1/2015 and Expression of innate immune genes in the nasal the EMC/2012 strains, mRNA levels of IFNs, epithelia of alpacas infected with the EMC/2012 expressed in Fc, were plotted against Cq values or Jordan-1/2015 strains found in the MERS-CoV UpE microfluidic PCR To compare antiviral and inflammatory pathways assay for both MPNE and MNNE areas. As shown induced by both strains, relative mRNA expression in Figure 4(c), genomic RNA in micro-dissected tis- levels for 37 innate immune response genes [29] sues were strongly or moderately correlated to the were monitored in the upper respiratory tract (URT) inductions of IFNβ (r = 0.72, P = 0.0007) and IFNλ1 and LRT tissues upon infection. All genes were (r = 0.8, P < 0.0001), or to IFNα (r = 0.65, P = 0.0005) detected at basal levels in micro-dissected samples of and IFNλ3 (r = 0.66, P = 0.004), respectively. Thus, control non-infected animals. In MPNE areas, type I our previous finding showing that the expression of and III IFNs were induced as early as 1 dpi in alpacas IFNs depended on levels of viral replication was from group 2. However, MPNE areas could not be col- confirmed and extended to the MERS-CoV EMC/ lected on 1 dpi due to the paucity of MERS-CoV 2012 clade A strain. EMC/2012 infected epithelial cells in group 1 animals. To further explore whether the Jordan-1/2015 On 2 and 3 dpi, IFNs were upregulated in the MPNE strain induces higher IFN responses, IFN relative areas infected by both strains. Of note, the Jordan-1/ quantification values from nasal epithelia of alpacas 2015 strain induced much higher relative levels of infected with Clade B MERS-CoV Qatar15/2015 [29] type I and III IFNs (Figure 4(a) and Supplementary and Jordan-1/2015 strains were normalized against Figure 3(a)). Genes coding for pattern recognition the mean of six non-infected control animals, includ- receptors (PRRs) including RIG1 and MDA5, antiviral ing three animals from a previous study [29]. No sig- ISGs (OAS1, CXCL10, MX1 and ISG15), chemokines nificant variability was observed on IFN levels among (CCL2 and CCL3), the transcriptional factors IRF7 the six control animals, which allowed normalization and the NLRP3 inflammasome component CASP1 procedures to be considered for the data sets obtained were highly upregulated on 2 dpi and levels of with both strains (Supplementary Table S3). Interest- expression decreased on 3 dpi. The same kinetic of ingly, no IFNs were induced until viral loads reached mRNA transcription but with a lower induction was a threshold of Cq = 6.46 (Supplementary Figure 4 observed for the pro-inflammatory cytokines (IL1β, (a)). Thus, MPNE samples with Cq < 6.46 were chosen IL6, IL8, IL15, TNFα and CXCL1) and the anti-inflam- to compare their IFN levels. As indicated in Sup- matory gene IL10 in MPNE areas infected by both plementary Figure 4(b), the Jordan-1/2015 strain strains. Nonetheless, levels of expression of these cyto- induced significantly higher relative levels of IFNλ1 kines could drastically vary between animals (Figure 4 than did the Qatar15/2015 strain despite similar viral EMERGING MICROBES & INFECTIONS 267 Figure 4. Kinetics of innate immune response genes and correlation coefficients between MERS-CoV genomic RNA load and IFN induction in alpaca nasal epithelia. The nasal epithelia of each alpaca (AP1 to AP21, except for AP14) was micro-dissected and MPNE and MNNE areas were selected and isolated for RNA extraction and conversion to cDNA. The Fluidigm Biomark microfluidic assay was used to amplify and quantify transcripts of innate immune genes at different dpi (1 to 3 dpi). HPRT1, GAPDH and UBC genes were used as normalizer genes, and values obtained from the infected animals were compared to those obtained from non- infected alpacas. Average Fc of genes in MERS-CoV positive nasal epithelia (a) and negative nasal epithelia (b) of EMC/2012 (gray bars) and Jordan-1/2015 (white bars) were shown. Data were displayed as means of ± SEM. Statistical significance was determined by Student’s t-test. *P < 0.05; **P < 0.01; ***P < 0.001 (n = 3) compared with non-infected alpacas; #P < 0.05; ##P < 0.01; ###P < 0.001 (n = 3) compared between groups on different dpi. (c) Cq values of MERS-CoV genomic RNA in nasal epithelia were plotted against the relative expression levels of IFNα, IFNβ, IFNλ1 and IFNλ3. Black and red spheres represent samples obtained from EMC/ 2012 and Jordan-1/2015 strains infected groups, respectively. Correlation coefficients were established using Spearman’s corre- lation test. 268 N. TE ET AL. loads. Moreover, same differences were observed Innate immune response genes follow the same between both strains for levels of expression of expression kinetics in the trachea and lung IFNα, β and λ3, although the results were not statisti- upon infection with the EMC/2012 and Jordan- cally significant. Considering that both viral strains 1/2015 strains mainly differ by the ORF4a sequence and that the Expression of innate immune response genes was Qatar15/2015 strain does not induce IFN mRNAs at further assessed along the URT and LRT. Except 1 dpi, this accessory protein may inhibit temporally for IFNλ3, which was not expressed, all other genes earlier IFN transcription but may have only a moder- were detected at basal levels in the trachea and lung ate antagonistic role in IFN responses during the of control non-infected animals. Transcription of all course of infection IFNs was not induced in the trachea and lungs of Figure 5. Kinetics of innate immune response genes in the trachea and lung of alpacas upon infection with MERS-CoV EMC/2012 and Jordan-1/2015 strains. Trachea and lung samples were obtained by scraping MFPE sections from control and infected alpacas (AP1 to AP21). After RNA extraction and conversion to cDNA, a Fluidigm Biomark microfluidic assay was used to amplify and quan- tify transcripts of innate immune genes at different dpi (1 to 3 dpi). HPRT1, GAPDH and UBC genes were used as normalizers, and values obtained from the infected animals were compared to those obtained from noninfected alpacas. Average fold change of highly induced genes in the trachea (a) and the lung (b) of alpacas infected with EMC/2012 (gray bars) and Jordan-1/2015 (white bars) were shown. Data were displayed as means of ±SEM. Statistical significance was determined by Student’s t-test. *P < 0.05; **P < 0.01; ***P < 0.001 (n = 3) compared with non-infected alpacas. EMERGING MICROBES & INFECTIONS 269 270 N. TE ET AL. alpacas upon infection with any of the strains (Sup- repeat (HR) 1 of the S protein (Figure 6). Interestingly, plementary Figure 5(a, b) and Supplementary Data this mutation allows the formation of hydrogen bonds S1). In the trachea, expression of PRRs, ISGs, IRF7 with the residue E357, located between the N terminal and STAT1 were moderately up regulated in both domain (NTD) and the receptor-binding domain groups during the course of infection (Figure 5(a) (RBD), and may affect the fusion mechanism (Figure and Supplementary Figure 5(a)). Compared to non- 7). The remaining aa sequences of clade A strains infected controls, there were few alterations of were highly conserved at nsp2, nsp3 and nsp4 where mRNA transcription in the remaining genes tested all 4 clade B-specific mutations occurred. All in all, during the infection period (Supplementary Figure compared with the EMC/2012 strain, the two clade 5(a) and Supplementary Data S1). B viruses used so far to infect experimentally alpacas In the lung, the EMC/2012 and Jordan-1/2015 (Qatar15/2015 and Jordan-1/2015) share 36 specific strains induced the same pattern of gene transcrip- aa mutations, mostly found in the ORF1ab gene and tion. For both strains, ISGs, RLRs (RIG1 and to a lesser extent, in the ORF3, the membrane protein MDA5), IRF7, IL8, CXCL1, CCL2, CCL3 and (M), the ORF4a, the nucleocapsid (N) and the spike CASP10 were slightly to moderately upregulated (S) gene products. The Jordan-1/2015 harboured a from 1 dpi, reaching a peak on 2 dpi to further unique 16-aa deletion in ORF4a, which is not found decay on 3 dpi (Figure 5(b) and Supplementary in all clade A and B strains analysed. The deleted Figure 5(b)). However, as for the trachea, animals region contains a predicted β-sheet belonging to the from group 2 tended to show higher induction of classical double-stranded RNA binding αβββα fold the above-mentioned cytokines. The rest of the of this protein (Figure 6, boxed in blue). Overall, the genes fluctuated around basal levels independently genomic and the structural protein prediction analysis of the animals and days of sample collection (Sup- revealed that clade B viruses possess several mutations plementary Figure 5(b) and Supplementary Data S1). that could affect the viral fusion with plasmatic mem- branes, viral replication and host antiviral responses such as the IFN signalling cascade. Protein alignment of selected clade A and B MERS-CoV strains and protein structure mapping Discussion To explore whether the increased replication compe- In the Arabian Peninsula, all contemporary zoonotic tence of the Jordan-1/2015 strain could be related to MERS-CoV strains belong to clade B. Hence, the clade B-specific mutations, all protein coding genes evolution of clade B viruses throughout the epidemic of two MERS-CoV clade A and nineteen clade B must be an advantage over early epidemic clade A strains, encompassing lineages 1 to 7 (Table 1), were strains in the reservoir host. RNA viruses are prone randomly selected and compared using the UniProt to undergo natural selection to create advantageous alignment tool. All selected MERS-CoV clade B strains variants due to a combination of high mutation share 5 specific aa mutations in the replicase genes of rates and large population sizes [35,36]. Indeed, the ORF1ab (Figure 6). Of note, in nsp3, a L864I substi- dromedary population is large in the Middle East tution occurred at the palm and fingers subdomains and with a very high seroprevalence of MERS-CoV of PLpro (Figure 6) that cleaves the viral replicase [37]. Here, our study provides a rational explanation polyproteins at three sites releasing viral nsp1, nsp2, on why MERS-CoV B strains substituted clade A and nsp3. Comparative protein structure analysis strains in the field. We compared the viral replication using PyMOL indicated that the L864I mutation is dis- competence, tissue tropism and induced innate host tant from the PLpro active site and thus might not immune responses of the prototype MERS-CoV affect polar contacts with nearby residues (Sup- strain EMC/2012 (clade A) to the Jordan-1/2015 plementary Figure 6). Further, all MERS-CoV clade strain (clade B), all isolated from humans, in a came- B strains show a Q1020R substitution in the heptad lid host model. None of the inoculated alpacas had Figure 6. Schematic diagram of divergent amino acids found in proteins of MERS-CoV clade A and B strains. Full aa sequences of twenty-one MERS-CoV strains were aligned with the UniProt alignment tool. Differences in aa residues affecting each protein are indicated by single-letter aa codes with positions indicated at the top of the diagram. The black and white letters indicate the aa of the prototype strain EMC/2012 and the clade B strains (Qatar15/2015 and Jordan-1/2015), respectively. The conserved aa sequences within clade A and B strains were highlighted in purple and yellow, respectively. The predicted β-sheet belonging to the classical double-stranded RNA binding αβββα fold of ORF4a is blue boxed. Abbreviations: nsp, non-structural protein; NTD, N terminal domain; CTD, C terminal domain; Ubl, ubiquitin-like domain; MD, Macro domain; SUD-M, middle region of SARS-CoV unique domain; PLpro, papain-like protease; NAB, nucleic-acid binding domain; GM2, betacoronavirus-specific marker; TM, transmembrane domain; ZBD, Zinc binding domain; MID, middle domain; SP, signal peptidase; RBD, receptor-binding domain; RBM, receptor-binding motif; FP, fusion peptide; HR, heptad repeat; CP, cytoplasmic domain. EMERGING MICROBES & INFECTIONS 271 clinical signs at any time, and no macroscopic lesions intact ORF4a sequences. In addition, the peak of were observed upon infection with both strains. MERS-CoV replication and maximal IFN transcrip- However, alpacas challenged with the Jordan-1/2015 tion coincided in all three strains on 2 dpi. We strain had higher titres of infectious virus in their confirmed that enhanced IFN expression levels corre- nasal cavity than did the EMC/2012 strain, suggesting lated with higher viral replication in nasal tissues; but, a higher transmission capability of clade B strains. despite similar viral loads found in MPNE areas, the Indeed, an increased replication capacity of the Jor- Jordan-1/2015 strain had only a moderate impact on dan-1/2015 strain was observed in the nasal epithelia IFN mRNA transcription when compared to the of infected alpacas. Moreover, compared to the EMC/ Qatar15/2015 strain. Together, our data suggest that 2012 strain, the Jordan-1/2015 strain exhibited a the IFN antagonistic role of the ORF4a protein is broader tissue tropism, showing productive infection exerted in vivo only during the first hours following in tracheal, bronchial and bronchiolar epithelial cells. infection. However, more experiments are needed to In fact, the clade B Qatar15/2015 strain, which was definitely ascertain the role of ORF4a in vivo. used in a previous study, had comparable replication Despite an enhanced replication competence and competences in the nasal mucosa as observed with early IFN induction in the nasal mucosa, the Jordan- the Jordan-1/2015 strain [29]. Thus, clade B viral 1/2015 strain led to similar moderate pro-inflamma- strains are displaying a better replicative fitness in tory responses as those provoked by the EMC/2012 camelids than the clade A EMC/2012 strain. strain along alpaca respiratory tracts. As observed At the first 24 h of infection, type I or III IFNs were with the Qatar15/2015 strain [29], and contrary to moderate to highly induced in the nasal mucosa of ISGs, IFNs were not induced at any time of the infec- alpacas in response to the Jordan-1/2015 strain but tion in trachea or lungs, supporting the hypothesis not to the EMC/2012 strain (Supplementary Data that IFNs produced in the nasal mucosa can act in S1). Our previous results with the Qatar15/2015 strain an endocrine manner. Additionally, the same histo- showed that only IFNβ started to be transcribed at pathological alterations were noted in nasal turbinates very low levels at 1 dpi [29]. Thus, such an early IFN of alpacas infected by the strains used in the present response could be due to the partial aa deletion in study and were equal to those caused by the the ORF4a protein affecting the Jordan-1/2015 strain Qatar15/2015 strain [29]. Differences between the but does not interfere with the early replication capa- Qatar15/2015 and the Jordan-1/2015 strains resided bility of this strain in the nasal mucosa. Indeed, anti- in higher numbers of IHC MERS-CoV positive epi- viral ISGs, such as MX1, OAS1 or ISG15, were thelial cells in the LRT and larger infiltration of leuko- maximally expressed at 2 dpi coinciding with a dra- cytes in the lungs of alpacas undergoing an infection matic decay of viral loads in nasal tissues at 3 dpi. with the Qatar15/2015 strain. This might be due to Of note, the deletion found in the Jordan-1/2015 an earlier IFN response provoked by the Jordan-1/ ORF4a contains a predicted β-sheet that is suspected 2015 strain. By contrast, the EMC/2012 strains did to affect the binding of dsRNA [31]. By contrast, the not replicate within the alpaca lungs, suggesting that EMC/2012 and the Qatar15/2015 strains harbour the Qatar15/2015 strain might be the most virulent Figure 7. Comparative structural view of the S protein HR1 region in clade A and B MERS-CoV strains. Details of inter- and intra- helical interactions for residue 1020 of the MERS-CoV spike protein (PDB ID: 5W9H) in clade A (a) and B (b) strains. Q1020 (yellow), R1020 (magenta) and aa residues proximal to 1020 (light blue) are shown as sticks; Polar contacts with residue 1020 are shown as red dot lines. The figures were produced using PyMOL. 272 N. TE ET AL. MERS-CoV experimentally tested so far in camelids. effective innate immune response based on early Indeed, as opposed to results obtained in the present recruitment of antiviral mechanisms and a controlled study, upregulation of pro-inflammatory genes such inflammation. Further reverse genetic studies are as NLRP3, TNFa and IL15 was observed in lungs of needed to unravel the consequences of mutations Qatar15/2015 infected animals [29]. By contrast, in differentiating MERS-CoV clades and influencing lungs of h-DPP4 transduced mice infected with the increased replication capability as their interaction clade B strain, ChinaGD01, only IRF3, IRF5 and with host components of innate immune pathways. IL15 were significantly upregulated in relation to the EMC/2012 strain. More severe inflammation was reported in the lung of mice infected with the Chi- Acknowledgements naGD01 strain but no differences in IFN transcription We thank Dr. Bart L. Haagmans from Erasmus Medical levels were observed between the two strains [19]. Center (EMC, Rotterdam) for providing the MERS-CoV Based on the comparison of coding sequences EMC/2012 and Jordan-1/2015 strains; Montse Amenós between all clade A and representative clade B strains, and Joana Ribes from the Centre for research in agricultural it appears that multiple punctual aa mutations are Genomics (CRAG, Cerdanyola del Vallès, Barcelona) for mostly distributed along the replicase gene. One of guidance on the use of LCM and Fluidigm BioMark microfl- uidic assay. We are particularly indebted with staff of the most striking changes differentiating clade A and the BSL-3 biocontainment animal facility at IRTA-CReSA B strains is the L864I substitution at the palm and (Cerdanyola del Vallès, Barcelona). This study was per- fingers subdomain of PLpro of nsp3 but, unless formed as part of the Veterinary Biocontained facility involved in substrate binding, it might not alter the Network (VetBioNet) [EU Grant Agreement INFRA-2016- cleavage of viral polyproteins during replication. Like- 1 N°731014] and the Zoonotic Anticipation and Prepared- wise, the other four clade B-specific mutations affect- ness Initiative (ZAPI project) [Innovative Medicines initiat- ive (IMI) grant 115760 to NT], with contributions from ing nsp2 (A395T), nsp3 (M517I and A1261V) and EFPIA partners, both supported by the European commis- nsp4 (A40 V) may have marginal effect on the viral sion. IRTA is supported by CERCA Programme / General- fitness since these sites do not belong to any of resi- itat de Catalunya. N.T. is a recipient of a China Scholarship dues involving MERS-CoV replication machinery. Council grant (CSC NO. 201608150108). The funders had Remarkably, a key clade B-specific mutation no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. (Q1020R) happened in HR1, essential for the fusion N.T., J.R., J.S., J.V.-A., and A.B. designed research; N.T., mechanism of the S protein allowing passage of the J.R., M.P., J.S., and J.V.-A. performed research; N.T., J.R., virus across the endosomal membrane [38]. In fact, M.P., J.S., J.V.-A., and A.B. analyzed data; and N.T., and Arginine at position 1020 of the S protein was believed A.B. wrote the paper. to be the hallmark of clade B strains [39] and was posi- tively selected during virus evolution [18]. Protein structure analysis revealed that R1020, instead of Disclosure statement Q1020, could interact with E357 by hydrogen bonds No potential conflict of interest was reported by the author(s). providing a protonated residue and a potential pro- tease cleavage site that may affect the S protein mem- brane fusion process. Forni et al. reported that Funding Q1020R replacement reduces the stability of a six- This work was supported by European Commission: [Grant helix bundle formation in the post-fusion confor- Number INFRA-2016-1 N°731014, Innovative Medicines mation [40]. A similar level of destabilization was initiative (IMI) grant 115760]. Jordi Rodon was partially also observed for a nearby mutation (T1015N), funded by VetBioNet (EU Grant Agreement INFRA-2016- which increases MERS-CoV infection efficiency 1 N°731014) and the crowdfunding initiative #Yomecorono. under in vitro condition [40,41]. Indeed, whether the interaction between R1020 and E357 is responsible ORCID for the destabilization process and influences fusion intermediate or pre-fusion state of S protein requires Nigeer Te http://orcid.org/0000-0002-5794-5764 further experimental proof. Our results taken Jordi Rodon http://orcid.org/0000-0002-1032-9091 Joaquim Segalés http://orcid.org/0000-0002-1539-7261 together, with data obtained by others [40,41], point Júlia Vergara-Alert http://orcid.org/0000-0001-7484- out the possible role of the Q1020R substitution for 444X an enhanced replicative capability of MERS-CoV Albert Bensaid http://orcid.org/0000-0001-6493-3756 clade B strains. Overall, we showed experimental evi- dence that early epidemic clade A strains could be out- competed by contemporary viruses due to a higher References replication fitness in camelid populations. Nonethe- [1] Memish ZA, Perlman S, Van Kerkhove MD, et al. less, none of the strains tested so far in camelids Middle East respiratory syndrome. Lancet. caused disease in these species, but rather elicited an 2020;395:1063–1077. EMERGING MICROBES & INFECTIONS 273 [2] Ministry of Health – Kingdom of Saudi Arabia. https:// [19] Wang Y, Sun J, Li X, et al. Increased pathogenicity and www.moh.gov.sa/en/CCC/events/national/Pages/2020. virulence of Middle East respiratory syndrome coro- aspx (accessed 23 Jun2021). navirus clade B in vitro and in vivo. J Virol. 2020;94: [3] WHO. Middle East respiratory syndrome coronavirus e00861-20. (MERS-CoV). https://www.who.int/emergencies/ [20] Chan RWY, Chu DKW, Poon LLM, et al. Tropism mers-cov/en/ (accessed 18 Jun2021). and replication of Middle East respiratory syndrome [4] Kim KH, Tandi TE, Choi JW, et al. Middle East respir- coronavirus from dromedary camels in the human atory syndrome coronavirus (MERS-CoV) outbreak respiratory tract: An in-vitro and ex-vivo study. in South Korea, 2015: epidemiology, characteristics Lancet Respir Med. 2014;2:813–822. and public health implications. J Hosp Infect. [21] Yang Y, Zhang L, Geng H, et al. The structural and 2017;95:207–213. accessory proteins M, ORF 4a, ORF 4b, and ORF 5 [5] Oh M, Park WB, Park S-W, et al. Middle East respir- of Middle East respiratory syndrome coronavirus atory syndrome: what we learned from the 2015 out- (MERS-CoV) are potent interferon antagonists. break in the Republic of Korea. Korean J Intern Protein Cell. 2013;4:951–961. Med. 2018;33:233–246. [22] Matthews KL, Coleman CM, van der Meer Y, et al. [6] Adney DR, Bielefeldt-Ohmann H, Hartwig AE, et al. The ORF4b-encoded accessory proteins of Middle Infection, replication, and transmission of Middle East respiratory syndrome coronavirus and two East respiratory syndrome coronavirus in alpacas. related bat coronaviruses localize to the nucleus and Emerg Infect Dis. 2016;22:1031–1037. inhibit innate immune signalling. J Gen Virol. [7] Crameri G, Durr PA, Klein R, et al. Experimental 2014;95:874–882. infection and response to rechallenge of alpacas with [23] Yang Y, Ye F, Zhu N, et al. Middle East respiratory Middle East respiratory syndrome coronavirus. syndrome coronavirus ORF4b protein inhibits type I Emerg Infect Dis. 2016;22:1071–1074. interferon production through both cytoplasmic and [8] Reusken CBEM, Schilp C, Raj VS, et al. MERS-CoV nuclear targets. Sci Rep. 2015;5:17554. infection of alpaca in a region where MERS-CoV is [24] Mielech AM, Kilianski A, Baez-Santos YM, et al. endemic. Emerg Infect Dis. 2016;22:1129–1131. MERS-CoV papain-like protease has deISGylating [9] David D, Rotenberg D, Khinich E, et al. Middle East and deubiquitinating activities. Virology. 2014;450– respiratory syndrome coronavirus specific antibodies 451:64–70. in naturally exposed Israeli llamas, alpacas and camels. [25] Lokugamage KG, Narayanan K, Nakagawa K, et al. One Heal. 2018;5:65–68. Middle East respiratory syndrome coronavirus nsp1 [10] Vergara-Alert J, van den Brand JMA, Widagdo W, inhibits host gene expression by selectively targeting et al. Livestock susceptibility to infection with mRNAs transcribed in the nucleus while sparing Middle East respiratory syndrome coronavirus. mRNAs of cytoplasmic origin. J Virol. Emerg Infect Dis. 2017;23:232–240. 2015;89:10970–10981. [11] Rodon J, Okba NMA, Te N, et al. Blocking trans- [26] Niemeyer D, Zillinger T, Muth D, et al. Middle East mission of Middle East respiratory syndrome corona- respiratory syndrome coronavirus accessory protein virus (MERS-CoV) in llamas by vaccination with a 4a is a type I interferon antagonist. J Virol. recombinant spike protein. Emerg Microbes Infect. 2013;87:12489–12495. 2019;8:1593–1603. [27] Comar CE, Goldstein SA, Li Y, et al. Antagonism of [12] Azhar EI, El-Kafrawy SA, Farraj SA, et al. Evidence for dsRNA-induced innate immune pathways by NS4a camel-to-Human transmission of MERS coronavirus. and NS4b accessory proteins during MERS corona- N Engl J Med. 2014;370:2499–2505. virus infection. MBio. 2019;10:e00319-19. [13] Killerby ME, Biggs HM, Midgley CM, et al. Middle [28] Siu K-L, Yeung ML, Kok K-H, et al. Middle East res- East respiratory syndrome coronavirus transmission. piratory syndrome coronavirus 4a protein Is a Emerg Infect Dis. 2020;26:191–198. double-stranded RNA-binding protein that suppresses [14] Chu DKW, Hui KPY, Perera RAPM, et al. MERS PACT-induced activation of RIG-I and MDA5 in the coronaviruses from camels in Africa exhibit region- innate antiviral response. J Virol. 2014;88:4866–4876. dependent genetic diversity. Proc Natl Acad Sci. [29] Te N, Rodon J, Ballester M, et al. Type I and III IFNs 2018;115:3144–3149. produced by the nasal epithelia and dimmed inflam- [15] Kiambi S, Corman VM, Sitawa R, et al. Detection of mation are features of alpacas resolving MERS-CoV distinct MERS-coronavirus strains in dromedary infection. PLOS Pathog. 2021;17:e1009229. camels from Kenya, 2017. Emerg Microbes Infect. [30] van Boheemen S, de Graaf M, Lauber C, et al. 2018;7:1–4. Genomic characterization of a newly discovered coro- [16] Sabir JSM, Lam TT-Y, Ahmed MMM, et al. Co-circu- navirus associated with acute respiratory distress syn- lation of three camel coronavirus species and recombi- drome in humans. MBio. 2012;3:e00473-12. nation of MERS-CoVs in Saudi Arabia. Science. [31] Lamers MM, Raj S V., Shafei M, et al. Deletion var- 2016;351:81–84. iants of Middle East respiratory syndrome coronavirus [17] El-Kafrawy SA, Corman VM, Tolah AM, et al. from humans, Jordan, 2015. Emerg Infect Dis. Enzootic patterns of Middle East respiratory syn- 2016;22:716–719. drome coronavirus in imported African and local [32] Corman VM, Eckerle I, Bleicker T, et al. Detection Arabian dromedary camels: a prospective genomic of a novel human coronavirus by real-time reverse- study. Lancet Planet Heal. 2019;3:e521–e528. transcription polymerase chain reaction. [18] Naeem A, Hamed ME, Alghoribi MF, et al. Molecular Eurosurveillance. 2012;17:20285. evolution and structural mapping of N-terminal [33] Coleman CM, Frieman MB. Growth and quantification domain in spike gene of Middle East respiratory syn- of MERS-CoV infection. Curr Protoc Microbiol. drome coronavirus (MERS-CoV). Viruses. 2020;12: 2015;37:1521–1529. doi:10.1002/9780471729259.mc15 502. e02s37. 274 N. TE ET AL. [34] Te N, Vergara-Alert J, Lehmbecker A, et al. Co- [38] Gao J, Lu G, Qi J, et al. Structure of the fusion core and localization of Middle East respiratory syndrome inhibition of fusion by a heptad repeat peptide derived coronavirus (MERS-CoV) and dipeptidyl peptidase- from the S protein of Middle East respiratory syn- 4 in the respiratory tract and lymphoid tissues of drome coronavirus. J Virol. 2013;87:13134–13140. pigs and llamas. Transbound Emerg Dis. 2019;66: [39] Lau SKP, Wong ACP, Lau TCK, et al. Molecular evol- 831–841. ution of MERS coronavirus: dromedaries as a recent [35] Holmes EC. The evolution and emergence of RNA intermediate host or long-time animal reservoir? Int viruses. New York (NY): Oxford University Press; J Mol Sci. 2017;18:2138. 2009. [40] Forni D, Filippi G, Cagliani R, et al. The heptad repeat [36] Simon-Loriere E, Holmes EC. Why do RNA viruses region is a major selection target in MERS-CoV and recombine? Nat. Rev. Microbiol. 2011;9:617–626. related coronaviruses. Sci Rep. 2015;5:14480. [37] Aljasim TA, Almasoud A, Aljami HA, et al. High rate [41] Scobey T, Yount BL, Sims AC, et al. Reverse genetics of circulating MERS-CoV in Dromedary camels at with a full-length infectious cDNA of the Middle slaughterhouses in Riyadh, 2019. Viruses. 2020;12: East respiratory syndrome coronavirus. Proc Natl 1215. Acad Sci USA. 2013;110:16157–16162.