MERS coronaviruses from camels in Africa exhibit region-dependent genetic diversity

Abstract

Middle East respiratory syndrome coronavirus (MERS-CoV) causes a zoonotic respiratory disease of global public health concern, and dromedary camels are the only proven source of zoonotic infection. Although MERS-CoV infection is ubiquitous in dromedaries across Africa as well as in the Arabian Peninsula, zoonotic disease appears confined to the Arabian Peninsula. MERS-CoVs from Africa have hitherto been poorly studied. We genetically and phenotypically characterized MERS-CoV from dromedaries sampled in Morocco, Burkina Faso, Nigeria, and Ethiopia. Viruses from Africa (clade C) are phylogenetically distinct from contemporary viruses from the Arabian Peninsula (clades A and B) but remain antigenically similar in microneutralization tests. Viruses from West (Nigeria, Burkina Faso) and North (Morocco) Africa form a subclade, C1, that shares clade-defining genetic signatures including deletions in the accessory gene ORF4b Compared with human and camel MERS-CoV from Saudi Arabia, virus isolates from Burkina Faso (BF785) and Nigeria (Nig1657) had lower virus replication competence in Calu-3 cells and in ex vivo cultures of human bronchus and lung. BF785 replicated to lower titer in lungs of human DPP4-transduced mice. A reverse genetics-derived recombinant MERS-CoV (EMC) lacking ORF4b elicited higher type I and III IFN responses than the isogenic EMC virus in Calu-3 cells. However, ORF4b deletions may not be the major determinant of the reduced replication competence of BF785 and Nig1657. Genetic and phenotypic differences in West African viruses may be relevant to zoonotic potential. There is an urgent need for studies of MERS-CoV at the animal-human interface.

Keywords: Africa; MERS; coronavirus; evolution; zoonosis.

Fig. 1.

Phylogenetic analysis of MERS-CoV full and partial genomes using the maximum-likelihood method (PhyML). Posterior probabilities of nodes with values higher than 0.8 are shown. The tree is rooted against a MERS-CoV–related bat coronavirus Neoromicia/PML-PHE1/RSA/2011 (GenBank accession no. KC869678). The long branch leading to Neoromicia/PML-PHE1/RSA/2011 has been removed to allow greater resolution of the viruses of interest. Fourteen virus genomes from African camels from this study are highlighted in red, and abbreviations for the viruses are given in square brackets. Clade C and subclade C1 MERS-CoV from African camels are denoted. Five deletion events (i–v) for viruses in clade C1 deduced from the deletion patterns affecting ORF3 and ORF4b (Fig. S3) are indicated at the respective branches.

Fig. 2.

Comparison of virus replication kinetics in Calu-3 cells, human DPP4-transduced mice, and ex vivo cultures of the human bronchus and lung. (A) Calu-3 cells were infected at an MOI of 0.01, and virus titers in culture supernatants were determined by the 50% tissue culture infective dose per milliliter (TCID₅₀) assay. The horizontal dotted line denotes the limit of detection in the TCID₅₀ assay. Statistical significance was determined by two-way ANOVA with Bonferroni’s correction: *P < 0.05; **P < 0.01; ***P < 0.001 compared with EMC; ###, P < 0.001 compared with AH13; +, P < 0.05; +++, P < 0.001 compared with BF785. Each experiment was repeated once, and one representative experiment is shown. (B) Ad5-hDPP4–transduced C57BL/6 mice (7 wk old, female) were infected intranasally with 1 × 10⁴ pfu of MERS-CoV strains EMC, AH13, or BF785. Virus titers in the lungs were measured. The horizontal dotted line denotes the limit of detection. Titers are expressed as pfu per gram of tissue. n = 3 mice per group at each time point. Statistical significance was determined by two-way ANOVA with Bonferroni’s comparisons: ***P < 0.001 compared with EMC; ###, P < 0.001 compared with AH13. (C and D) Comparison of virus replication and tropism of MERS-CoV EMC, BF785, and Nig1657 in ex vivo cultures of human bronchus (C) and lung (D). Virus titers in culture supernatants were determined by TCID₅₀ assay. Data are the means and SEM of three individual tissue donors. The horizontal dotted line denotes the limit of detection in the TCID₅₀ assay. Statistical significance was determined by two-way ANOVA with Bonferroni’s comparisons: *P < 0.05; **P < 0.01 compared with EMC.

Fig. 3.

Comparison of innate immune responses in Calu-3 cells infected with different MERS-CoVs: EMC, AH13, BF785, or Nig1657. Calu-3 cells were infected with MERS-CoV at an MOI of 2. Influenza A/HK/483/1997 (H5N1) was used as a positive control for cytokine induction, and mock-infected cells served as negative controls. Cell lysates were collected at 6, 24, and 48 h postinfection, and mRNA was extracted. The RNA expression of viral gene (UpE) (A), mRNA expression of IFNs (IFN-β, and IFN-λ1) (B and C), IFN-inducible genes IP-10 (D), ISG15 (E), and MX1 (F), and a proinflammatory cytokine TNF-α (G) is shown as the mean and SEM of three biological replicates. The detection limit of each gene was 20 copies. (H) The cell-culture supernatants were tested for IP-10 protein using the BD Cytometric bead-assay (detection limit 7.54 pg/mL). The experiment was carried out with three biological replicates. Statistical significance was determined by two-way ANOVA with Bonferroni’s correction and is indicated as *P < 0.05; **P < 0.01; ***P < 0.001.

Fig. 4.

Comparison of virus replication kinetics and innate immune responses in Calu-3 cells infected with reverse-genetically engineered MERS-CoV, rgEMC-WT, and MERS-CoV with ORF4b deletion, rgEMCΔ4b. (A) Calu-3 cells were infected at an MOI of 0.01. Virus titers in culture supernatants were determined by the TCID₅₀ assay. Means and SEM of three biological replicates are shown. The horizontal dotted line denotes the limit of detection in the TCID₅₀ assay. (B and C) Calu-3 cells were infected at an MOI of 2 with MERS-CoVs rgEMC-WT and rgEMCΔ4b; mock-infected cells served as negative controls. RNA expression of viral gene UpE (B) and mRNA expression of IFN-β and IFN-λ1 (C) is shown as the mean and the SEM of three biological replicates. The detection limit of each gene was 20 copies. Each experiment was repeated once, and a representative experiment is shown. Statistical significance was determined by two-way ANOVA with Bonferroni’s comparisons: *P < 0.05; **P < 0.01; ***P < 0.001.