We used a previously developed direct-contact model in which the transmission of a MERS-CoV clade B isolate (Qatar15/2015) was assessed in llamas (13,14). In brief, we kept a group of 5 llamas inside an experimental enclosure (Appendix Figure 1, panel A) to study the transmission capabilities of a MERS-CoV clade C isolate (MERS-CoV/Egypt2013) that was obtained from an infected dromedary (15). We inoculated 2 llamas with MERS-CoV and placed them in direct contact with 3 sentinels at 2 days postinoculation (dpi) (Appendix). We monitored clinical signs and body temperature and collected nasal swab specimens for virologic studies. In addition, we retrieved experimental data from MERS-CoV Qatar15/2015-inoculated and in-contact llamas (13,14) and performed comparative analyses.

We specifically selected the animals used in the transmission studies to be 6-to-10-month-old juveniles of similar geographic origin, sex, and health status background. All animal experimentation and MERS-CoV handling were conducted at the Biosafety Level 3 facilities of the Biocontainment Unit of IRTA CReSA (Barcelona, Spain). Animal handling and experimental procedures were approved by the Ethical and Animal Welfare Committee of IRTA and by the Ethical Commission of Animal Experimentation of the Autonomous Government of Catalonia (approval nos. FUE-2017–00561265 and CEA-OH/10942/1).

Figure 1

Figure 1. MERS-CoV RNA and infectious virus shedding in llamas experimentally infected with MERS-CoV Egypt/2013 (red) or Qatar15/2015 (blue) strains. A–C) Viral RNA and infectious MERS-CoV shedding of inoculated animals. Genomic (A)...

Rectal temperatures of all animals remained at basal levels (37°C–40°C), and no animals displayed clinical signs during the study. We detected no gross or microscopic lesions in the upper or lower respiratory tracts of any studied llama, independent of their experimental group. Animals inoculated with a high dose of either MERS-CoV Egypt/2013 (clade C) or Qatar15/2015 (clade B) had similar levels of genomic and subgenomic viral RNA in nasal swab specimens for 2 weeks (Figure 1, panels A, B). They also shed high titers of infectious virus during the first week after inoculation in a similar biphasic pattern (Figure 1, panel C), indicating that the doses used to inoculate the animals caused productive infection with both strains. Although viral shedding was comparable between both experimental groups, higher infectious titers were detected in Egypt/2013-inoculated llamas. The infection was characterized by a first peak of shedding at 2 dpi and a subsequent reduction in MERS-CoV viral loads, followed by a secondary peak before viral clearance.

Figure 2

Figure 2. Mixed model analyzing the transmission competence of each MERS-CoV strain over time in investigation of llamas experimentally infected with MERS-CoV. Boxplots show daily virus shedding of sentinel llamas infected with...

The African MERS-CoV isolate was transmitted to 2 of 3 in-contact animals in this study, as determined by quantitative reverse transcription PCR (Figure 1, panels D, E), but infectious virus shedding in contact animals largely remained below threshold levels (Figure 1, panel F). Infectious MERS-CoV Egypt/2013 could only be isolated sporadically and at titers close to the limit of detection. In contrast, the Arabian MERS-CoV Qatar15/2015 isolate was transmitted to all direct-contact llamas, leading to productive infection (Figure 1, panels D–F). Of note, genomic and subgenomic MERS-CoV Egypt/2013 RNA was detected at lower levels and cleared faster in direct-contact llamas than in sentinels infected with the MERS-CoV Qatar15/2015 strain (Figure 1, panels D, E). In the remaining sentinel, a productive infection did not develop, but the animal was naturally exposed to MERS-CoV Egypt/2013, as indicated by traces of genomic RNA in NS at 3–7, 10, and 12 dpi (cycle quantitation values >37) and development of serum neutralizing antibodies (nAbs) to MERS-CoV (Appendix Figure 2). Subgenomic RNA analyses indicated no evidence for either viral replication or shedding in this llama throughout the study. Statistical analyses in sentinel animals showed a significant reduction in MERS-CoV Egypt/2013 replication and shedding period compared with those observed in llamas exposed to MERS-CoV Qatar15/2015 strain (Figure 2). Regardless of the MERS-CoV strain investigated, nAbs were detected starting at 2 weeks after infection in all inoculated animals and in-contact sentinels (Appendix Figure 2). We did not find statistical differences in serum nAb levels among experimental groups.

Altogether, our data demonstrated transmission of both MERS-CoV strains, which resulted in decreased viral replication and shedding capabilities of the Egypt/2013 strain compared with the Qatar15/2015 strain in sentinel llamas infected by contact. Therefore, the Egypt/2013 strain might have a lower potential of transmission than the Qatar15/2015 strain.

The results of our experimental investigation might explain why MERS-CoV clade C strains have not been established in the Arabian Peninsula after being introduced through imported camels and competing with enzootic clade B viruses. However, further studies are needed to determine whether this potentially reduced transmissibility is a common feature of the diverse MERS-CoV lineages found in dromedaries in Africa. Specific amino acid substitutions in the spike protein or in other genomic regions of African clade C viruses might be determinant of the low replication phenotype observed in the in-contact animals in our study, as has been previously observed in human cells (12). However, viral or host factors that play a key role in conferring replication and transmission competence remain to be explored in camelid reservoirs. High MERS-CoV genome stability was previously described in llamas infected with the Qatar15/2015 strain (13). Thus, eventual mutations arising from animals infected with the Egypt/2013 strain were not expected, and no sequencing was performed in those infected animals. Nonetheless, our study provides in vivo experimental data demonstrating reduced MERS-CoV fitness of 1 African clade C isolate to in-contact camelids compared with an Arabian clade B isolate. In addition, reduced MERS-CoV shedding from infected camelids might limit spillover to humans. Introducing MERS-CoV clade B strains to Africa through infected camelids must be avoided, because these strains might outcompete African MERS-CoV clade C strains and pose a greater zoonotic and pandemic threat in Africa.