SARS-CoV-2 Entry: At the Crossroads of CD147 and ACE2

Abstract

In late 2019, the betacoronavirus SARS-CoV-2 was identified as the viral agent responsible for the coronavirus disease 2019 (COVID-19) pandemic. Coronaviruses Spike proteins are responsible for their ability to interact with host membrane receptors and different proteins have been identified as SARS-CoV-2 interactors, among which Angiotensin-converting enzyme 2 (ACE2), and Basigin2/EMMPRIN/CD147 (CD147). CD147 plays an important role in human immunodeficiency virus type 1, hepatitis C virus, hepatitis B virus, Kaposi's sarcoma-associated herpesvirus, and severe acute respiratory syndrome coronavirus infections. In particular, SARS-CoV recognizes the CD147 receptor expressed on the surface of host cells by its nucleocapsid protein binding to cyclophilin A (CyPA), a ligand for CD147. However, the involvement of CD147 in SARS-CoV-2 infection is still debated. Interference with both the function (blocking antibody) and the expression (knock down) of CD147 showed that this receptor partakes in SARS-CoV-2 infection and provided additional clues on the underlying mechanism: CD147 binding to CyPA does not play a role; CD147 regulates ACE2 levels and both receptors are affected by virus infection. Altogether, these findings suggest that CD147 is involved in SARS-CoV-2 tropism and represents a possible therapeutic target to challenge COVID-19.

Keywords: ACE2; CD147; COVID-19; EMMPRIN; SARS-CoV-2; basigin; entry; infection.

Figure 1

Inhibition of CyPA binding to CD147 does not play a role in SARS-CoV-2 entry. CaLu3 cells were treated with 1 µg/mL MEM-M6/6 CD147 blocking Ab (α-CD147 Ab) or equivalent mouse IgG (as isotype control) or plain medium (as control, ctr) for one hour before, and during SARS-CoV-2 infection. Samples were collected 48 hpi. (A) N1 RNA levels were analyzed by ddPCR and their relative abundance in cells is shown. N1 RNA levels were normalized to total RNA. N1 levels in CD147-blocking antibody treated samples were normalized to the levels in the corresponding IgG isotype-treated control. Mean ± s.e.m out of three biological replicates is shown. n.s. not significative, Student’s t test. (B,C) Lysates were analyzed by WB. (B) One experiment is shown as representative of three. Immunolabelling for CD147 showed that Ab treatment did not affect the levels of expression of the receptor either in the absence or in the presence of SARS-CoV-2. (C) Spike protein levels were normalized to GAPDH (loading control). Spike levels in CD147-blocking antibody treated samples were normalized to the levels in the corresponding IgG isotype-treated control. Mean ± s.e.m. out of three biological replicates is shown. n.s. not significant, Student’s t-test.

Figure 2

CD147 silencing reduces ACE2 protein levels and impairs SARS-CoV-2 infection. (A–C) CaLu3 cells were transfected with a non-targeting (NT) or a CD147-specific siRNA. Samples were collected 48 h and 96 h after transfection. (A) One experiment is shown as representative of three. (B,C) Receptor levels were analysed by WB (protein, (B)) and by ddPCR (RNA, (C)) analysis. (B) Receptors’ protein levels were normalized to the loading control (GAPDH). Receptors levels in CD147 siRNA-treated samples were normalized to the levels in the corresponding NT siRNA-treated controls. Mean ± s.e.m. out of three biological replicates is shown. **, p = 0.005 for CD147 and p = 0.01 for ACE2, Student’s t-test. (C) Receptors RNA levels were normalized to total RNA. Receptors levels in CD147 siRNA-treated samples were normalized to the levels in the corresponding NT siRNA-treated controls. Mean ± s.e.m. out of three biological replicates is shown. ***, p = 0.001, n.s. not significative, Student’s t-test. (D,E) CaLu3 cells were transfected with a non-targeting (NT) or a CD147-specific siRNA before SARS-CoV-2 infection. Samples were collected 48 hpi. (D) N1 RNA levels were analyzed by ddPCR, and their relative abundance in both cells and supernatants is shown. N1 RNA levels were normalized to total RNA. N1 levels in CD147 siRNA-treated samples were normalized to the levels in the corresponding NT siRNA-treated controls. Mean ± s.e.m. out of three biological replicates is shown. **, p < 0.01; *, p < 0.05, Student’s t-test. (E) Lysates were analyzed by WB. One experiment is shown as representative of three. ACE2 labeling was detected after Spike protein immunostaining on the same membrane. Arrow indicates ACE2 specific band. GAPDH was used as the loading control.

Figure 3

CD147 and ACE2 expression is affected by SARS-CoV-2 infection. (A,B) CaLu3 cells were infected with SARS-CoV-2 at 0.05 MOI and samples collected 48 hpi. Receptors levels were analyzed by ddPCR (RNA, (A)) and WB (protein, (B)). (A) Receptors RNA levels were normalized to total RNA. Receptors levels in infected samples (+SARS-CoV-2) were normalized to the levels in the corresponding uninfected controls (-SARS-CoV-2). Mean ± s.e.m. out of three biological replicates is shown. *, p = 0.02; ***, p = 0.001, Student’s t-test. (B) Receptor protein levels were normalized to the loading control. Receptor levels in infected samples (+SARS-CoV-2) were normalized to the levels in the corresponding uninfected controls (-SARS-CoV-2). Mean ± s.e.m. out of 10 biological replicates is shown. ***, p < 0.0001, Student’s t-test.