Dipeptidyl peptidase 4 is a functional receptor for the emerging human coronavirus-EMC

Abstract

Most human coronaviruses cause mild upper respiratory tract disease but may be associated with more severe pulmonary disease in immunocompromised individuals. However, SARS coronavirus caused severe lower respiratory disease with nearly 10% mortality and evidence of systemic spread. Recently, another coronavirus (human coronavirus-Erasmus Medical Center (hCoV-EMC)) was identified in patients with severe and sometimes lethal lower respiratory tract infection. Viral genome analysis revealed close relatedness to coronaviruses found in bats. Here we identify dipeptidyl peptidase 4 (DPP4; also known as CD26) as a functional receptor for hCoV-EMC. DPP4 specifically co-purified with the receptor-binding S1 domain of the hCoV-EMC spike protein from lysates of susceptible Huh-7 cells. Antibodies directed against DPP4 inhibited hCoV-EMC infection of primary human bronchial epithelial cells and Huh-7 cells. Expression of human and bat (Pipistrellus pipistrellus) DPP4 in non-susceptible COS-7 cells enabled infection by hCoV-EMC. The use of the evolutionarily conserved DPP4 protein from different species as a functional receptor provides clues about the host range potential of hCoV-EMC. In addition, it will contribute critically to our understanding of the pathogenesis and epidemiology of this emerging human coronavirus, and may facilitate the development of intervention strategies.

Figure 1. Binding of hCoV-EMC S1 to cells is correlated with infection of hCoV-EMC.

a–d, Shown in the left panels are the FACS analyses of hCoV-EMC S1–Fc binding (red line) to Vero (a), COS-7 (b), Huh-7 (c) and bat cells (d). A feline CoV S1–Fc protein (blue line) and mock-incubated cells (grey shading) were used as controls. In the middle panels, hCoV-EMC-infected cells are visualized using an antiserum that recognizes the non-structural protein NSP4. In the right panels, hCoV-EMC RNA levels in supernatants of the infected cells at 0, 20 and 40 h after infection were quantified using a TaqMan assay and expressed as genome equivalents (GE; half-maximal tissue-culture infectious dose (TCID₅₀) per ml). Error bars indicate s.e.m. PowerPoint slide

Figure 2. hCoV-EMC S1 binding to DPP4.

a, Huh-7 cell lysates were incubated with hCoV-EMC and SARS-CoV S1–Fc proteins and affinity-isolated proteins were subjected to protein electrophoresis under non-reducing conditions. The arrowhead indicates the position of the ∼110-kDa DPP4 protein specifically isolated using the hCoV-EMC S1–Fc protein. b, hCoV-EMC and SARS-CoV S1–Fc proteins were mock-incubated or incubated with soluble DPP4 (sDPP4) or soluble ACE2 (sACE2) followed by protein A sepharose affinity isolation and subjected to protein electrophoresis under non-reducing conditions. PowerPoint slide

Figure 3. DPP4 is present on hCoV-EMC-susceptible cell lines and human bronchiolar epithelial cells.

a, COS-7 cells transfected with plasmids encoding human DPP4 (hDPP4), bat DPP4 (bDPP4) or a control plasmid (pcDNA) were tested for S1 binding and staining with a polyclonal antiserum against DPP4. b, Similarly, COS-7, Huh-7, Vero and bat cells were tested for reactivity with the same antiserum against DPP4 (blue lines) or with a control normal goat serum (grey peak). c, d, DPP4 expression (red) was also found in primary human bronchiolar epithelial cell cultures (c) and human bronchiolar tissue (d) and appeared to be localized to the apical surfaces of non-ciliated cells that do not express β-tubulin IV (green). e, Double-stranded viral RNA (cyan) was detected in hCoV-EMC-infected primary human bronchiolar epithelial cell cultures and appeared to be localized to non-ciliated cells that express DPP4 (red). Stainings were performed using antibodies directed against β-tubulin IV (ciliated cells; green), DPP4 (red), dsRNA (hCoV-EMC; cyan), and DAPI (cell nucleus; blue). All scale bars are 10 μm. PowerPoint slide

Figure 4. DPP4 is essential for virus infection.

a, Inhibition of hCoV-EMC infection of Huh-7 cells by antibodies to DPP4. Supernatants collected at 2 h (open bars) and 20 h (filled bars) were tested for the presence of hCoV-EMC RNA using a TaqMan assay. Results representative of three different experiments are shown as ΔCt values (one-way ANOVA test, *P < 0.05; n = 3 per group), normal goat, normal goat serum. b, Infection of human primary bronchiolar epithelial cells is blocked by the DPP4 antibodies in a dose-dependent manner and samples were analysed at 2 h (open bars) and 20 h (filled bars) after infection (one-way ANOVA test, *P < 0.05; n = 3 per group). c, COS-7 cells transfected with plasmids encoding human DPP4 (hDPP4), bat DPP4 (bDPP4) or a control plasmid (pcDNA) were inoculated with hCoV-EMC at a multiplicity of infection of 1 and left for 1 h. Cells were washed twice and stained at 8 h after infection (original magnification, ×200) or supernatant collected at 2 h (open bars), 20 h (black bars) and 40 h (blue bars) was tested for the presence of hCoV-EMC RNA using a TaqMan assay (d). Results representative of three different experiments are expressed as GE (TCID₅₀ ml⁻¹) values (one-way ANOVA test, *P < 0.05; n = 4 per group). All error bars represent s.e.m. PowerPoint slide