ACE2 internalization induced by a SARS-CoV-2 recombinant protein is modulated by angiotensin II type 1 and bradykinin 2 receptors

Abstract

Aims: Angiotensin-converting enzyme 2 (ACE2) is a key regulator of the renin-angiotensin system (RAS) recently identified as the membrane receptor for the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Here we aim to study whether two receptors from RAS, the angiotensin receptor type 1 (AT1R) and the bradykinin 2 receptor (B2R) modulate ACE2 internalization induced by a recombinant receptor binding domain (RBD) of SARS-CoV-2 spike protein. Also, we investigated the impact of ACE2 coexpression on AT1R and B2R functionality.

Materials and methods: To study ACE2 internalization, we assessed the distribution of green fluorescent protein (GFP) signal in HEK293T cells coexpressing GFP-tagged ACE2 and AT1R, or B2R, or AT1R plus B2R in presence of RBD alone or in combination with AT1R or B2R ligands. To estimate ACE2 internalization, we classified GFP signal distribution as plasma membrane uniform GFP (PMU-GFP), plasma membrane clustered GFP (PMC-GFP) or internalized GFP and calculated its relative frequency. Additionally, we investigated the effect of ACE2 coexpression on AT1R and B2R inhibitory action on voltage-gated calcium channels (Ca_V2.2) currents by patch-clamp technique.

Key findings: RBD induced ACE2-GFP internalization in a time-dependent manner. RBD-induced ACE2-GFP internalization was increased by angiotensin II and reduced by telmisartan in cells coexpressing AT1R. RBD-induced ACE2-GFP internalization was strongly inhibited by B2R co-expression. This effect was mildly modified by bradykinin and rescued by angiotensin II in presence of AT1R. ACE2 coexpression impacted on B2R- and AT1R-mediated inhibition of Ca_V2.2 currents.

Significance: Our work contributes to understand the role of RAS modulators in the susceptibility to SARS-CoV-2 infection and severity of COVID-19.

Keywords: ACE2; AT1R; Angiotensin II; B2R; SARS-CoV-2; Telmisartan.

Fig. 1

(A) Example images of the GFP signal patterns identified in cultures of HEK293T cells transfected with ACE2-GFP incubated with vehicle or RBD (10 μg/mL) for 40 min. Objective magnification: 60×. (B) Example images of HEK293T cells transfected with ACE2-GFP incubated with vehicle or RBD 10 μg/mL for 40 min and immunostained against RBD. For nuclei visualization cells were co-stained with Höescht. The asterisk in merged image of RBD-treated cells shows that a non-transfected cell is negative for RBD. Arrowheads point at colocalization of ACE2-GFP and RBD signal. Objective magnification: 63×.

Fig. 2

(A) Representative images of cultures of HEK293T cells transfected with ACE2-GFP incubated with vehicle or RBD (10 μg/mL) for 5, 10 or 40 min. For membrane visualization cells were treated with Cell Mask. Magnification: 60×. Arrowheads indicate the different GFP signal patterns (white: PMU GFP, light gray: PMC GFP and dark gray: internalized GFP). (B) GFP patterns abundance (%) in HEK293T cells transfected with ACE2-GFP upon vehicle or RBD incubation at the specified times. PMU GFP (plasma membrane uniform GFP), PMC GFP (plasma membrane clustered GFP). Association between time of RBD treatment and GFP patterns frequency was evaluated by Chi-square test (χ² = 190.03, P_chi < 0.0001; * indicates different from the expected frequency of the corresponding GFP pattern, P < P_corr = 0.004; Somers' D = 0.57; n = 45–60 total cells per condition).

Fig. 3

(A) GFP patterns abundance (%) in cultures of HEK293T cells transfected with ACE2-GFP alone (ACE2-GFP) or co-transfected with AT1R (ACE2-GFP + AT1R) incubated with vehicle or RBD (10 μg/mL) for 40 min. Association between treatments and GFP patterns frequency in RBD-treated groups was evaluated by Chi-square test (χ² = 1.237, P_chi = 0.54; n = 30–50 cells per condition). (B) GFP patterns abundance (%) in cultures of HEK293T cells co-transfected with ACE2-GFP and AT1R (ACE2-GFP + AT1R) co-incubated with vehicle or RBD and vehicle (+Veh), telmisartan (+Telm, 10 μM) or angiotensin II (+Ang II, 1 μM) for 40 min. Telmisartan and Ang II were added to the culture medium 5 min prior to RDB. Association between treatments and GFP patterns frequency in RBD-treated groups was evaluated by Chi-square test (χ² = 17.89, P_chi < 0.01; * indicates different from the expected frequency of the corresponding GFP pattern, P < P_corr = 0.004; Somers' D = 0.24; n = 35–47 cells per condition).

Fig. 4

GFP patterns abundance (%) in cultures of HEK293T cells transfected with ACE2-GFP alone (ACE2) or co-transfected with B2R (+B2R) incubated with vehicle or RBD (10 μg/mL) for 40 min. In +B2R + BK condition, cells were co-incubated with BK plus vehicle or BK plus RBD (BK, 0.5 μM) for 40 min. BK was added to the culture medium 5 min prior to RBD. Association between treatments and GFP patterns frequency in RBD-treated groups was evaluated by Chi-square test (χ² = 47.29, P_chi < 0.0001; * indicates different from the expected frequency of the corresponding GFP pattern, P < P_corr = 0.006; n = 26–48 cells per condition).

Fig. 5

(A) GFP patterns abundance (%) in cultures of HEK293T cells co-transfected with ACE2-GFP and AT1R (ACE2-GFP + AT1R) or plus B2R (+B2R) incubated with vehicle or RBD (10 μg/mL) for 40 min. Association between treatments and GFP patterns frequency in RBD-treated groups was evaluated by Chi-square test (χ² = 21.36, P_chi < 0.0001; * indicates different from the expected frequency of the corresponding GFP pattern, P < P_corr = 0.008; n = 28–33 cells per condition). (B) GFP patterns abundance (%) in cultures of HEK293T cells co-transfected with ACE2-GFP, AT1R and B2R (ACE2-GFP + AT1R + B2R) co-incubated with vehicle or RBD (10 μg/mL) and vehicle (+Veh), BK (+BK, 0.5 μM) or angiotensin II (+AngII, 1 μM) for 40 min. For co-incubation, B2R or AT1R agonists were added to the culture medium 5 min prior to RBD. Association between treatments and GFP patterns frequency in RBD-treated groups was evaluated by Chi-square test (χ² = 14.56, P_chi < 0.01; * indicates different from the expected frequency of the corresponding GFP pattern, P < P_corr = 0.006; n = 33–40 cells per condition).

Fig. 6

(A) Representative traces (top) of Ca_V2.2 calcium currents (I_CaV2.2) and time courses (middle) of HEK293T cells co-transfected with Ca_V2.2 and B2R (+B2R, n = 7), or Ca_V2.2, B2R and ACE2 (+B2R + ACE2, n = 8), or Ca_V2.2, B2R and AT1R (+B2R + AT1R, n = 6), or Ca_V2.2, B2R, AT1R and ACE2 (+B2R + AT1R + ACE2, n = 7) in control condition (vehicle) and 0.5 μM BK application (+BK); 0.1 ACE2-GFP or GPCR: Ca_V2.2 molar ratio. Black and gray dots correspond to vehicle and BK traces respectively. Bars (bottom) represent the average I_CaV2.2 inhibition by BK application for each condition. One Way ANOVA and Tukey's post-test (P value estimated versus +B2R). (B) Representative traces (top) of Ca_V2.2 calcium current (I_CaV2.2) and time courses (middle) of HEK293T cells co-transfected with Ca_V2.2, and AT1R (+AT1R, n = 7), or Ca_V2.2, AT1R and B2R (+AT1R + B2R, n = 5), or Ca_V2.2, AT1R and ACE2 (+AT1R + ACE2, n = 4), or Ca_V2.2, AT1R, B2R and ACE2 (+B2R + AT1R + ACE2, n = 4) in control condition (vehicle) and 1 μM Ang II application (+Ang II); 0.1 ACE2-GFP or GPCR: Ca_V2.2 molar ratio. Black and gray dots correspond to vehicle and Ang II traces respectively. Bars (bottom) represent the average I_CaV2.2 inhibition by angiotensin II application for each condition. One Way ANOVA and Tukey's post-test (P value estimated versus +AT1R).

Fig. 7

Representative traces (top) of Ca_V2.2 calcium currents (I_CaV2.2) from HEK293T cells co-transfected with (A) Ca_V2.2 and AT1R (+AT1R, n = 11) or Ca_V2.2, AT1R and ACE2-GFP (+AT1R + ACE2, n = 10), (B) Ca_V2.2 and B2R (+B2R, n = 12) or Ca_V2.2, B2R and ACE2-GFP (+B2R + ACE2, n = 13), and (C) Ca_V2.2, B2R and AT1R (+AT1R + B2R, n = 17) or Ca_V2.2, B2R, AT1R and ACE2-GFP (+B2R + AT1R + ACE2, n = 23); 0.1 ACE2-GFP: or GPCR: Ca_V2.2 molar ratio. Bars (bottom) represent the average I_CaV2.2 for each condition. Unpaired Student's t-test.