Latrophilin-2 mediates fluid shear stress mechanotransduction at endothelial junctions

Abstract

Endothelial cell responses to fluid shear stress from blood flow are crucial for vascular development, function, and disease. A complex of PECAM-1, VE-cadherin, VEGF receptors (VEGFRs), and Plexin D1 located at cell-cell junctions mediates many of these events. However, available evidence suggests that another mechanosensor upstream of PECAM-1 initiates signaling. Hypothesizing that GPCR and Gα proteins may serve this role, we performed siRNA screening of Gα subunits and found that Gαi2 and Gαq/11 are required for activation of the junctional complex. We then developed a new activation assay, which showed that these G proteins are activated by flow. We next mapped the Gα residues required for activation and developed an affinity purification method that used this information to identify latrophilin-2 (Lphn2/ADGRL2) as the upstream GPCR. Latrophilin-2 is required for all PECAM-1 downstream events tested. In both mice and zebrafish, latrophilin-2 is required for flow-dependent angiogenesis and artery remodeling. Furthermore, endothelial-specific knockout demonstrates that latrophilin plays a role in flow-dependent artery remodeling. Human genetic data reveal a correlation between the latrophilin-2-encoding Adgrl2 gene and cardiovascular disease. Together, these results define a pathway that connects latrophilin-dependent G protein activation to subsequent endothelial signaling, vascular physiology, and disease.

Keywords: Fluid Shear Stress; G Protein-coupled Receptor; Latrophilin; PECAM-1; Vascular Development.

Figure 1. Gα proteins specific for endothelial flow responses.

(A) Flow-induced endothelial alignment after Gα knockdown. HUVECs were subjected to fluid shear stress (FSS) at a rate of 12 dynes/cm² for 16 h, and nuclear orientation quantified as histograms showing the percentage of cells within each 10° of the flow directions from 0° to 90° (see “Methods”) ****P < 0.0001; one-way ANOVA with Tukey’s multiple comparisons test. (B) Src family kinase activation, quantified in (C). n = 5 for control, Gi knockdown, Gq11 knockdown and n = 3 for simultaneous knockdown of Gi and Gq11. Values are means ± SEM. ****P < 0.0001, ***P < 0.001; one-way ANOVA with Tukey multiple comparison test. (D) Rescue of Gq/11 and Gi knockdown by re-expression of siRNA-resistant versions of the indicated proteins. ****P < 0.0001; one-way ANOVA with Tukey’s multiple comparison test. (E) Amino acid sequences of Gαi1, Gαi2, and Gαi3 at the mutation sites of Gαi1 gain-of-function mutant. (F) Rescue of Gq/11 and Gi knockdown with indicated Gα proteins. Each point corresponds to one measurement averaged from >500 cells. N = 4. ****P < 0.0001; one-way ANOVA with Tukey’s multiple comparisons test. (G) GINIP pulldown assay for activation of Gαi2 by FSS. N = 3. Results quantified in (H). **P = 0.0185, Student’s t test. (I) GRK2N pulldown assay for activation of Gq. N = 4, quantified in (J). *P < 0.05, Student’s t test. (K) GINIP pulldown assay for FSS-induced activation of wild-type and Q306K Gαi1. N = 4, quantified in (L). *P = 0.0304, ***P = 0.0017; Student’s t test. (M) Gi2 activation after PECAM-1 knockdown. HUVECs expressing GluGlu-tagged Gi2 were transfected with scrambled siRNA or PECAM-1 siRNA, exposed to FSS, and Gi2 activation assayed as described above, quantified in (N). Values are means ± SEM, normalized to input Gα protein levels. ***P < 0.001; Student’s t test, N = 4.

Figure 2. Identification of the upstream flow-responsive GPCR.

(A) Strategy for identification of GPCRs that bind the gain-of-function Gi1Q306K mutant but not the wild-type Gi1. (B) Co-immunoprecipitation of Gi1Q306K(ins4A) with endogenous LPHN2 with and without FSS for 2 min. N = 3, quantified in (C). *P = 0.0374, Student’s t test. (D) Co-immunoprecipitation of LPHN2-GFP with the indicated Gα proteins containing internal GluGlu epitope tags. (E) Gi2 pulldown assay after LPHN2 knockdown. N = 4. Results quantified in (F). ***P < 0.001, Student’s t test.

Figure 3. Latrophilins regulate endothelial flow responses.

(A) HUVECs with or without latrophilin-2 knockdown were subjected to FSS for 16 h, then fixed and stained with Hoechst (nuclei), phalloidin (F-actin) and an antibody against VE-Cadherin. Scale bar: 100 µm. (B) Alignment of HUVECs (each bar = 10° increments) after knockdown of latrophilin isoforms was quantified as in Fig. 1 from >2000 cells/experiment, N = 3. ****P < 0.0001; one-way ANOVA with Tukey’s multiple comparisons test. (C) Localization of LPHN2-mClover3. Scale bar: 100 µm. N = 6. (D) Gi1(Q306K) pulldown from ECs ± FSS for 5 min or for 24 h, probed for PECAM-1. N = 3, quantified in (E). (F) Gi1(Q306K) pulldown from ECs ± FSS for 5 min or for 24 h probed for VE-cadherin. N = 3, quantified in (G). (H) Activation of Src family kinases and Akt by FSS in HUVECs depleted for latrophilin isoforms. N = 3–4, quantified in (I, J). *P = 0.0184, **P < 0.001; one-way ANOVA with Tukey’s multiple comparisons test. (K) Alignment of LPHN2-depleted HUVECs rescued by re-expression of the indicated latrophilin isoforms. ****P < 0.0001; one-way ANOVA with Tukey’s multiple comparisons test. Quantification of >2000 cells for each condition, N = 3. (L) LPHN2 knockdown HUVECs were rescued by re-expression of the indicated LPHN2 mutants. ****P < 0.0001, one-way ANOVA with Tukey’s multiple comparisons test. Quantification of >500 cells for each condition.

Figure 4. Latrophilins in flow-mediated endothelial cell morphology in vivo.

(A) Representative image of HUVECs transfected with scrambled siRNA or LPHN2 siRNA. Outlines of ECs from VE-cadherin staining are shown to next to each image. Scale bar: 100 µm. (B) Quantification of junctional linearization index from (A). n = 30 for each condition. **P = 0.0019, Statistics used two-way ANOVA with Sidak’s multiple comparison test. (C) Dorsal aortas from 48 hpf Tg(kdrl:ras-mCherry) WT or LPHN mutant zebrafish embryos stained for ZO-1. EC were outlined as in Methods. LPHN2 mutant indicates adgrl2a+adgrl2b.1 sgRNAs. (D, E) Quantification of cell size and eccentricity in embryos with latrophilin-2 CRISPR sgRNAs ± silent heart morpholinos (MO). N = 6 per condition. ****P < 0.0001, *P = 0.0267, **P = 0.0015, ***P = 0.0008; one-way ANOVA with Tukey’s multiple comparisons test. (F) ECs in the thoracic aorta in 18-week-old mice at 10 weeks after tamoxifen injections, stained for β-catenin (left) and outlined (right). Data are representative of 6 mice for each condition. Scale bar: 100 µm. (G, H) Quantification of cell eccentricity and junction linearity in mouse aorta. N = 4 per condition. **P = 0.0013; ****P < 0.0001; Student’s t test. (I) Aortas from mice 30 min after Evans blue dye injection, thresholded images on the right. The fractional blue area quantified in (J). n = 3 for each condition. *P = 0.0419; Student’s t test. (K) In vitro endothelial permeability assay using FITC–streptavidin and biotin-conjugated fibronectin. Scale bar: 100 µm. (L) Permeability was quantified by measuring FITC–streptavidin area per image field from (K). n = 5, *P = 0.0457, one-way ANOVA.

Figure 5. Cholesterol depletion activates Latrophilin-dependent signaling.

(A) ECs were treated with FSS or 5 mM methyl-β-cyclodextrin (mβCD) for 1 min then incubated with D4H-mClover3, rinsed and bound mClover3 quantified as described in “Methods”. N = 27 for each condition. ****P < 0.0001; one-way ANOVA with Tukey multiple comparisons test. (B) Phalloidin staining following short acute treatment of methyl-β-cyclodextrin (mβCD). Scale bar: 100 µm. (C) Co-immunoprecipitation of Gi1Q306K(ins4A) with endogenous Latrophilin-2 (LPHN2) after 5 mM MβCD for 1 min. (D) Activation of Src family kinases and VEGFR2 after 5 mM MβCD for 1 min in HUVECs ± LPHN2 knockdown. Results quantified in (E, F), respectively. Values are means ± SEM. N = 5. *P = 0.0305, ***P = 0.0002; one-way ANOVA with Tukey multiple comparison test. (G) Model for junctional endothelial shear stress mechanotransduction.

Figure 6. Latrophilin-2 in flow-induced vascular remodeling.

(A) Representative images of ISVs from 48 hpf embryos with F(0)/mosaic depletions of latrophilin genes or controls. (B) Quantification of diameters of all ISVs in (A). n = 15 per condition. *P = 0.0452, **P = 0.0150; one-way ANOVA with Tukey’s multiple comparisons test. (C) Quantification of diameters of arterial and venous ISVs at 72 hpf of zebrafish. n = 6 per condition. **P = 0.0013; Student’s t test. (D) Ratio of blood flow in the right ischemic vs left control foot by laser doppler, quantified in (E). N = 7–8 per condition. *^,#P = 0.0439, *P = 0.0135, **P = 0.0028, ^##P = 0.0056, ^#P = 0.0163; Student’s t test. (F) Micro-CT of arterial vasculature after surgery. Representative reconstructed micro-CT images from LPHN2 ECKO and WT littermates on day 22 after surgery, quantified in (G). n = 8 per condition. ****P < 0.0001, ***P = 0.0007, *P = 0.0167; two-way ANOVA with Tukey’s multiple comparisons test. (H) Rectus femoris muscle from thighs of 10-week-old mice of LPHN2 ECKO or wild-type littermates, stained for CD31 and SMA to visualize the entire vasculature and arteries. Scale bar: 100 µm. (I) Quantification of capillary density in the entire stitched thigh muscle images. n = 3 for control and n = 6 for Lphn2 ECKO. ****P < 0.0001; Student’s t test. (J) Schematic of mouse treadmill fatigue test. (K) Maximum running during test. n = 4 for control and n = 5 for Lphn2 ECKO. **P = 0.0082; Student’s t test. (L) Running distance for each mouse. n = 4 for control and n = 5 for Lphn2 ECKO. **P = 0.0064; Student’s t test. (M) Oxygen consumption during treadmill fatigue test. *P < 0.0001; two-way ANOVA with Tukey’s multiple comparisons test. Control: N = 5, LPHN2 ECKO: N = 4. (N) Representative 3D reconstructed images at 20° downward angle of ECs plated on top of collagen gels and sprouting under static or flow conditions. ECs were stained with phalloidin; pseudocolors indicate distance from the gel surface as per the color scale on the right. (O) Histograms of cell area vs. distance from gel surface in (N). (P) Mean sprout length. N = 4. ***P = 0.0006; one-way ANOVA with Tukey’s multi-comparison test.