Surface–Aerosol Stability and Pathogenicity of Diverse Middle East Respiratory Syndrome Coronavirus Strains, 2012–2018

Neeltje van Doremalen; Michael Letko; Robert J. Fischer; Trenton Bushmaker; Jonathan Schulz; Claude K. Yinda; Stephanie N. Seifert; Nam Joong Kim; Maged G. Hemida; Ghazi Kayali; Wan Beom Park; Ranawaka A.P.M. Perera; Azaibi Tamin; Natalie J. Thornburg; Suxiang Tong; Krista Queen; Maria D. van Kerkhove; Young Ki Choi; Myoung-don Oh; Abdullah M. Assiri; Malik Peiris; Susan I. Gerber; Vincent J. Munster


Emerging Infectious Diseases. 2021;27(12):3052-3062. 

In This Article


Stability of MERS-CoV Strains in Aerosols or as Fomites Compared With SARS-CoV-2

We selected 8 MERS-CoV strains and 1 SARS-CoV-2 strain (SARS-CoV-2/WA1-2020) to be used in this study (Table; Figure 1). Five MERS-CoV strains were isolated from human cases and 3 strains were isolated from dromedary camels. Strains were isolated during 2012–2018 and originated from the Middle East,[5] Africa[2] or South Korea[1] (Table). All originally obtained viruses were passaged once in Vero E6 cells, and virus stocks were deep sequenced (Table). We used MERS-CoV sequences to construct a phylogenetic maximum-likelihood tree, which showed a wide distribution of MERS-CoV strains selected. Thus, our panel represents a broad sample of known genetic variation within currently circulating MERS-CoV strains.

Figure 1.

Phylogenetic tree of 446 full Middle East respiratory syndrome coronavirus (MERS-CoV) genomes showing distribution of human-derived (red) and camel-derived (blue) isolates. The tree was constructed with PhyML ( and rooted at the midpoint. Strain EMC/12 was obtained from Erasmus Medical Center (Rotterdam, the Netherlands); U/14, KSA/15, and KSA/18 from the Centers for Disease Control and Prevention (Atlanta, GA, USA); SK/15 from Chungbuk National University (Cheongju, South Korea); and C/KSA/13, C/E/13, and C/BF/15 from Hong Kong University (Hong Kong, China). Scale bar indicates nucleotide substitutions per site. KSA, Kingdom of Saudi Arabia.

We first investigated the stability of MERS-CoV as fomites on polypropylene, stainless-steel, copper, and silver surfaces, which we selected because they represent commonly encountered surfaces in hospital environments or have virocidal properties.[23] For comparison with a pandemic human coronavirus, we also included SARS-CoV-2. Back-titrations of all virus strains showed comparable starting virus titers. Stability of MERS-CoV on polypropylene and stainless-steel surfaces, maintained at 21°C–22°C and a relative humidity of 45%–55% under standard laboratory light conditions, was similar to that reported for MERS-CoV and SARS-CoV-2 stability on surfaces.[21,26] We found major differences in decay rates when comparing EMC/12 to SK/15, KSA/18, C/KSA/13, and C/BF/15 on polypropylene. These differences were not found for the other surfaces (Figure 2; Appendix Figure, Infectious virus titers were low for all strains on copper and silver surfaces at 24 hours. We analyzed data by using linear regression for the first 24 hours for each surface and each virus. Decay, averaged between all virus strains, was higher for copper (−0.11576 log10 TCID50/h) and silver (−0.08744 log10 TCID50/h) surfaces than for polypropylene (−0.0529 log10 TCID50/h) and stainless-steel (−0.0469 log10 TCID50/h) surfaces.

Figure 2.

Stability of MERS-CoV strains on surfaces and in aerosols compared with those for SARS-CoV-2. Simple linear regression of virus was used for different surfaces and in aerosols. For surface stability, 50 μL of MERS-CoV or SARS-CoV-2 was spread on a surface, either polypropylene, stainless steel, copper, or silver; 1 mL of Dulbecco's modified Eagle medium was added at times 0, 1, 24, 48, or 72 hours, and samples were titrated. For aerosol stability, MERS-CoV– or SARS-CoV-2–containing aerosols were sprayed into a Goldberg drum; samples were taken at times 0, 30, 60, 120, and 180 min and then titrated. Linear regression was calculated per virus and indicated as lines. Dotted lines indicate limits of detection. Strain sources are listed in the legend for Figure 1. MERS-CoV, Middle East respiratory syndrome coronavirus; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2; TCID50, median tissue culture infectious dose.

We aerosolized all MERS-CoV strains in a Goldberg drum at 21°C and a relative humidity of 60%–70% in the dark. We then tested samples at 0, 30, 60, 120 and 180 min after aerosolization, titrated them, and compared results with those for SARS-CoV-2. We detected no major differences in linear regression of loss of infectious virus in aerosols between strains. For all MERS-CoV strains, infectious virus could still be detected at 180 min after aerosolization (Figure 2, panel B).

In Vitro Replication of MERS-CoV Strains

To investigate any in vitro growth differences, we grew strains in 2 cell systems, Vero E6 cells and HAE cultures, in comparison to the reference strain EMC/12. At 48 hpi, C/KSA/13 and KSA/15 showed higher titers than EMC/12 in Vero E6 cells. At 72 hpi, C/KSA/13 and C/BF/15 showed lower titers than EMC/12 in HAE cultures. We observed no other major differences in either culture type. Although differences were not always statistically significant, all camel-derived viruses had reduced replication kinetics compared with those for EMC/12 in HAE cells at 24–72 hpi (Figure 3).

Figure 3.

Middle East respiratory syndrome coronavirus replication in Vero E6 cells (A) and human airway epithelium (B). Replication is shown as geometric means; error bars indicate SDs. Vero E6 cells were infected with a multiplicity of infection of 0.01, and human airway epithelium were infected with a multiplicity of infection of 0.1. Samples of supernatants were obtained at 8, 24, 48 and 72 hours postinoculation and titrated. Statistically significant differences compared with those for the prototypical strain, EMC/12, were calculated by using ordinary 1-way analysis of variance, followed by a Bonferroni multiple comparisons test. Dotted lines indicate limits of detection. Strain sources are listed in the legend for Figure 1. TCID50, median tissue culture infectious dose. *p<0.05; **p<0.01.

Disease Progression for MERS-CoV Strains in hDPP4 Transgenic Mice

MERS-CoV enters cells expressing the receptor human dipeptidyl peptidase IV (hDPP4). Our laboratory developed hDPP4 transgenic mice to test MERS-CoV vaccine efficacy.[20] We intranasally inoculated 10 mice/group with 103 TCID50 MERS-CoV/mouse. Mice started to lose weight on days 2–5 postchallenge; weight continued to decrease for all groups, except for mice inoculated with C/BF/15, in which only 1 mouse continued to lose weight (Figure 4, panel A). For all groups, including C/BF/15, weight loss was also associated with other signs: ruffled coat, increased breathing rate, reluctance to move, and hunched posture. Only animals in the groups inoculated with SK/15 (1/6) and the group inoculated with C/BF/15 (5/6) survived. Average time to death was similar for all groups, excluding C/BF/15: EMC/12, 7.33 days; U/14, 6.5 days; KSA/15. 7 days; SK/15. 7.6 days; KSA/18. 7.67 days; C/KSA/13, 7.5 days; and C/E/13, 8 days (Figure 4, panel B).

Figure 4.

In vivo replication of different Middle East respiratory syndrome coronavirus (MERSCoV) strains. hDPP4 mice were inoculated intranasally with 103 TCID50 of MERS-CoV. Four mice were euthanized on day 3, and the remaining 6 mice were monitored for survival. A) Relative weight loss of hDPP4 mice. B) Survival of hDPP4 mice. C) Oropharyngeal shedding of MERS-CoV as measured by using an UpE quantitative reverse transcription PCR. D) Amount of shedding per experiment per mouse calculated by using area under the curve (AUC) analysis of viral load in oropharyngeal swab specimens. Results are displayed per mouse per virus strain. E) Viral load in lung tissue obtained from mice euthanized at day 3. F) Viral mRNA load in lung tissue obtained from mice euthanized at day 3. G) Percentage of positive pixels quantified from lung tissues stained for MERS-CoV antigen. Colors in panels D–F match those for strains in panels A–-C; strain sources are listed in the legend for Figure 1. Statistical significance was compared by using 1-way analysis of variance, followed by a Bonferroni multiple comparisons test. Dotted lines indicate limits of detection. TCID50, median tissue culture infectious dose. *p<0.05.

We measured viral RNA in oral swab specimens obtained during days 1–7 postchallenge and found no major differences in the amount of shedding between different groups (Figure 4, panels C, D). Viral genome RNA was lower in lung tissue collected on day 3 from mice inoculated with SK/15, C/E/15, and C/BF/15. Subgenomic RNA was lower to a major degree only in lung tissue of mice inoculated with C/BF/15 (Figure 4, panels E, F).

We observed no differences in pathology between different groups. Animals rarely showed pulmonary pathology at day 3. However, animals that had lesions showed only a minimal and random lymphocytic infiltrate. Immunohistochemical analysis showed that MERS-CoV antigen was expressed rarely or randomly in type I and II pneumocytes and not located in areas of inflammation. Morphometric analysis of pulmonary tissue that had immunoreactivity showed no major differences between groups (Figure 4, panel G).