Critical role of endothelial regulator of G protein signaling 5 in postnatal angiogenesis

Abstract

Aims

Regulator of G protein signaling 5 (RGS5) modulates G-protein coupled receptor (GPCR) signaling and is markedly upregulated in angiogenesis. However, its function in endothelial cells and its role in mediating postnatal angiogenesis remain unclear. The purpose of this study is to define the role of endothelial RGS5 in regulating VEGF signaling and angiogenesis.

Methods and results

In a hindlimb ischemia model, we found impaired postnatal angiogenesis, reduced perfusion recovery, and increased rates of auto-amputation in mice with either global or endothelial-specific deletion of Rgs5. Rgs5 deficiency in ECs led to reduced VEGFR2 phosphorylation at Tyr¹¹⁷⁵ and its downstream signaling pathways. This was accompanied by increased activity of the tyrosine phosphatase, SHP-1, which dephosphorylates and inactivates VEGFR2. In vitro, Rgs5-deficient ECs exhibited diminished migration, increased apoptosis, and reduced viability in response to VEGF stimulation. These findings indicate that RGS5 is essential for maintaining VEGF-mediated angiogenic signaling through inhibition of SHP-1 activity.

Conclusion

Endothelial RGS5 is a critical regulator of VEGF signaling and postnatal angiogenesis. By attenuating SHP-1–mediated dephosphorylation of VEGFR2, RGS5 preserves VEGF signaling and angiogenesis. Targeting the RGS5–SHP-1 axis may be a useful therapeutic strategy for improving angiogenesis in response to ischemia.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10456-026-10079-z.

Introduction

G protein-coupled receptors (GPCRs) constitute the largest family of transmembrane receptors, mediating a wide array of physiological processes [1, 2], including cardiovascular homeostasis and vascular remodeling [3, 4]. These receptors transduce extracellular cues into intracellular responses via heterotrimeric G proteins, whose signaling is tightly regulated by the Regulator of G protein Signaling (RGS) family of GTPases. By enhancing the intrinsic GTPase activity of Gα subunits, RGS proteins serve as key modulators of GPCR signal duration and intensity [5, 6]. RGS5, a member of the R4 subfamily, primarily modulates Gα_q and Gα_I signaling [7] and is notably expressed in pericytes, endothelial cells, and vascular smooth muscle cells [8]. Its expression is markedly induced following vascular injury and during both developmental and pathological angiogenesis, including tumorigenesis [9–12]. Indeed, RGS5 expression in vascular smooth muscle cells is upregulated during arteriogenesis, where it promotes collateral vessel remodeling by shifting G-protein signaling from Ca²⁺-dependent contractile pathways toward RhoA-mediated remodeling responses [10]. Loss of RGS5 impairs smooth muscle cell proliferation and collateral growth, resulting in defective flow recovery after arterial occlusion [10]. While these findings define the importance of RGS5 in mural cell biology, its specific function in endothelial cells, particularly in VEGF-dependent angiogenic signaling, remains unknown.

Angiogenesis is orchestrated by a coordinated balance of pro- and anti-angiogenic factors, with vascular endothelial growth factor (VEGF) serving as a central mediator of endothelial cell proliferation, migration, and new vessel formation [13]. Dysregulated VEGF signaling contributes to the formation of immature, leaky vasculature, a hallmark of pathological angiogenesis. Among the endogenous negative regulators of VEGF signaling is the Src homology region 2 domain-containing phosphatase-1 (SHP-1), which attenuates receptor tyrosine kinase activity, including VEGFR2, via dephosphorylation of key signaling intermediates. Importantly, SHP-1 has been reported to directly dephosphorylate VEGFR2 at Tyr¹¹⁷⁵, a critical site regulating VEGF-induced angiogenic signaling [14, 15]. SHP-1 has been implicated in maintaining vascular quiescence and modulating endothelial activation thresholds.

Given the importance of VEGF in angiogenesis and the emerging roles of both RGS5 and SHP-1 in endothelial cell biology, we hypothesized that RGS5 may influence VEGF-mediated angiogenesis, potentially via SHP-1-dependent mechanisms. The aim of this study, therefore, was to elucidate the mechanisms by which endothelial RGS5 regulates VEGF signaling during postnatal angiogenesis, with a particular focus on its interaction with SHP1.

Methods

Animals

We used 12–20 weeks old, age-matched global (Rgs5^–/–) and endothelial cell-specific Rgs5-deficient male mice (EC-Rgs5^–/–) with corresponding control littermates. All mice were congenic strains on C57Bl/6J background. To generate these models, we first created a conditional Rgs5^flox/flox strain by inserting loxP sites flanking exons 3 to 5, which encode the RGS5-binding domain on chromosome 1. For global knockout mice (Rgs5^–/–), Rgs5^flox/flox were crossed with Pgk-Cre mice (Fig. 1a).

Fig. 1.

Loss of Rgs5 Results in Impaired Angiogenesis After Hind Limb Ischemia (HLI) a Mouse breeding strategy, b Representative laser Doppler images of global Rgs5 deficient (Rgs5^−/−) and control (Pgk-Cre) mice hind limbs at indicated times after HLI (left panel) with corresponding quantifications of blood flow and functional recovery (right panel). A color scale for the perfusion range is shown, with blue indicating impeded perfusion. (n = 10). c Representative images of different outcomes of Rgs5^–/– and Pgk-Cre (arrows indicate ischemic foot, distal hindlimb, and digits gangrene) (left panel) and exclusion rate after HLI (right panel) (n = 10). d Capillary density assessed by using anti-CD31 immunostaining (left panel, Bar = 50 μm), and quantification of mean capillary density (right panel) of ischemic gastrocnemius muscles recovered 28 days after HLI and at baseline (n = 10 − 5). e Images of Rgs5^–/– and Pgk-Cre mice aortic rings (left panel) and quantification of length and number of sprouting vessels (right panels) treated with 10 ng/ml VEGF for 10 days (n = 10). f Representative images of Matrigel plugs recovered after two weeks, their relative hemoglobin content, and CD31 immunostaining (n = 5, Bar = 50 μm). Data are presented as mean. b) One-way ANOVA with Holm correction for multiple testing, d, e, and f) Unpaired t-test. ^nsP > 0.05, **P < 0.01, ****P < 0.0001

For endothelial-specific deletion, we generate endothelial-specific inducible RGS5 knockout mice (EC-Rgs5^flox/flox) by crossing Rgs5^flox/flox mice with a tamoxifen-inducible Cre recombinase line under the control of the vascular endothelial cadherin promoter (Cdh5-CreERT2) [16, 17]. To induce endothelial-specific Rgs5 deletion, EC-Rgs5^flox/flox mice were administered tamoxifen intraperitoneally at a dose of 20 mg/kg/day for five consecutive days (Sigma-Aldrich, St. Louis, MO) (Fig. 2a). Experiments were conducted at least five days after the final tamoxifen injection. Control groups included Pgk-Cre, wild-type (WT), Rgs5^flox/flox, Cdh-CreERT2, EC-Rgs5^flox/flox, and Rgs5^flox/flox mice treated with tamoxifen intraperitoneally at a dose of 20 mg/kg/day for five days (Rgs5^flox/flox+TMX).

Fig. 2.

Endothelial-specific Rgs5 deficiency (EC-Rgs5^–/–) Reduces Angiogenesis After Hind Limb Ischemia (HLI). a Mouse breeding strategy, b Representative laser Doppler images of endothelial-deficient Rgs5 mice (EC-Rgs5^–/–) and control groups ((WT-wild type, Rgs5^flox/flox-conditional Rgs5 mice, EC-Rgs5^flox/flox-conditional Rgs5 mice with inducible Cre recombinase under the control of the vascular endothelial cadherin promoter (Cdh5-CreERT2) without tamoxifen treatment, Rgs5^flox/flox-conditional Rgs5 mice with tamoxifen treatment Rgs5^flox/flox+TMX) (left panel)). Evaluation of blood flow, functional recovery, and exclusion rate after HLI (right panel) (n = 7). c Representative images of capillaries assessed by using anti-CD31 immunostaining and quantification of mean capillary density of ischemic gastrocnemius muscles recovered 28 days after HLI and at baseline (n = 5, Bar = 50 μm). d Images of EC-Rgs5^–/– and EC-Rgs5^flox/flox mice aortic rings (left panel) and quantification of length and number of sprouting vessels (right panels) treated with 10 ng/ml VEGF for 10 days (n = 5). (e) Representative images of Matrigel plugs recovered after two weeks, their relative hemoglobin content, and CD31 immunostaining (n = 5, Bar = 50 μm). Data are presented as median for b, and means for the rest. Data are presented as median for b and means for the rest. b) One-way ANOVA with Holm correction for multiple testing, c) and d), e) Unpaired t-test, ^nsP>0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Female mice undergo estrous cycles, which cause fluctuations in hormone levels (like estrogen and progesterone) every 4–5 days. These hormonal variations can affect vascular physiology and experimental outcomes, leading to increased variability in data [18, 19]. By using only male mice, we aimed to reduce this variability and simplify the interpretation of results. Animals were housed under a 12:12-hour light/dark cycle (light: 124 lx white light from 6 am to 6 pm; dark: <2 lx red light from 6 pm to 6 am) with free access to food (Harlan2918) and water.

Study approval

The animal protocols used in this study comply with NIH Guidelines for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Chicago and the University of Arizona.

Hindlimb ischemia

Animals were initially anesthetized in an induction chamber with 1–3% isoflurane in 100% oxygen at a flow rate of 2 L/min until they were unresponsive to pinch-toe stimuli. Animals were then placed in the supine position on the operating table, connected to continuous isoflurane flow and a heating pad to maintain the internal temperature at 37 °C with a rectal probe connected to the heating pad. Limb hair was removed and disinfected with iodine and alcohol swabs. Skin dissection was made approximately 1 cm long, from the knee towards the medial thigh. The femoral artery was carefully dissected from the vein and nerve and occluded with a double knot of 7 − 0 suture underneath the proximal end of the femoral artery. The second double suture knot was placed close to the knee, underneath the distal end of the femoral artery, just proximal to the popliteal artery. The femoral artery was carefully transected and removed between knots. The incision was closed with continued 5 − 0 Vicryl sutures. Limb perfusion was assessed with Laser Doppler Imager (LDI, Moor Instruments, Wilmington, DE.) before surgery, immediately post-surgery, and at days 7, 14, 21, and 28 post-ischemia. Color-coded images were analyzed with software (Laser Doppler Perfusion measure, V308, Moor Instruments, Wilmington, DE.). Limb perfusion was presented as the ratio of the perfusion of the ischemic limb relative to the contralateral limb. The procedure was successfully performed, and complete ischemia was confirmed as an 80% perfusion decrease [20]. At 28 days post-reperfusion, animals were euthanized by using 100% carbon dioxide (CO₂) inhalation delivered at a gradual displacement rate of 30–70% of the chamber volume per minute, added to the existing air in the chamber. CO₂ flow was maintained for at least 1 min after respiratory arrest, followed by cervical dislocation to ensure death, in accordance with the AVMA Guidelines for the Euthanasia of Animals (2020).

Functional assessment of hindlimb function

Hindlimb functional assessment was evaluated with serial measurements of spontaneous movement pre-surgery, immediately after surgery, and thereafter, every 7 days for 28 days. Animals were scored as follows: 4 (normal function), 3 (plantar flexion but no flexing toes), 2 (no dragging and no plantar flexion), and 1 (dragging foot). Animals with signs of gangrene and autoamputation were excluded from the experiment [21].

Immunofluorescent staining

Capillary density was examined 28 days after hindlimb ischemia (HLI). Isolated gastrocnemius muscles were fixed in 4% paraformaldehyde and embedded in Tissue-Tek (OCT compound, Sakura, Horgen, Switzerland). Capillary density was observed on 5-micron cryosections permeabilized with ice-cold acetone for 5 min and then blocked with 1% normal goat serum for 1 h. The cryosections were immunostained with CD31 antibody at 4 °C overnight, followed by 1 h incubation with goat anti-rabbit Alexa Fluor 488-labeled secondary antibodies. Staining was observed with a fluorescent microscope (Olympus Corporation BX41, Tokyo, Japan). Capillary density was expressed as the number of CD31-positive vessels per microscopic field.

Matrigel plug assay

The Matrigel plug assay was used to evaluate the effect of RGS5 deficiency on angiogenesis in vivo. Matrigel matrix (0.4 ml, Corning Life Science, Tewksbury, MA.) containing 80 units/ml heparin (Sigma Aldrich, St. Louis, MO) and 60 ng/ml VEGF-A (Peprotech, Rocky Hill, NJ) was subcutaneously injected into the flanks of mice. After two weeks, mice were sacrificed, and the Matrigel plugs were removed. The relative hemoglobin content in the plugs was measured with Drapkin’s reagent kit (Sigma-Aldrich, St. Louis, MO) according to the manufacturer’s instructions. Histological analysis of Matrigel plugs was evaluated by CD31 immunofluorescence staining.

Aortic ring assay

Mouse aortas were excised and cleaned of the surrounding tissue in a cold, sterile saline solution. Aortas were cut into approximately 1 mm in length rings. Individual rings were embedded in the Matrigel matrix (Corning Life Science, Tewksbury, MA.) on individual wells of a 48-well plate, with and without the addition of 10 ng/ml of VEGF-A (Peprotech, Rocky Hill, NJ). Solidified Matrigel with aortic rings was covered with growth factor-deficient culture medium and incubated at 37 °C. Neovessel formation arising from aortic rings was evaluated by measuring the number of sprouts and their length under the standard phase contrast microscope for 10 days [22].

Cell culture

Primary human umbilical vein endothelial cells (HUVEC) were purchased from EMD Millipore (Burlington, MA.) and cultured at 37 °C with 5% CO₂ per manufacturer’s instructions. Each experiment represents an independent biological replicate performed with distinct cell batches at early passages (P3–P5). RGS5 was knocked down by lentiviral-RGS5 shRNA particles (Sigma-Aldrich, St. Louis, MO). For control, HUVECs were transduced with non-targeting shRNA particles. Successful transduction was confirmed by Western blot and immunofluorescence analysis.

Tube formation assay

Endothelial tube formation was evaluated using 24-well plate bottoms that were covered with 300 µl of Matrigel matrix (Corning Life Science, Tewksbury, MA.) and incubated at 37 °C for 30 min. HUVECs were plated at a density of 1.2 × 10^⁵ cells/well onto Matrigel in an endothelial medium recommended by the manufacturer. The plate was incubated overnight at 37 °C and then imaged at 20x magnification. Cell-to-cell branch points were counted using ImageJ software (NIH, Bethesda, MD.).

Transwell migration assay

Endothelial cell migration was assessed using modified 24-well Boyden chambers with a polycarbonate membrane with 8 μm pores (EMD Millipore, Burlington, MA.).

HUVECs were starved in basal medium (serum/growth factors free) at 37 °C overnight before being seeded at a density of 1 × 10^⁵ cells/well, on the top layer of a Trans-well membrane. The lower chamber was filled with 0.5 ml of medium with 20 ng/ml VEGF as a chemoattractant. After 24 h of incubation, the remaining cells on the upper part of the membrane were scraped, and the remaining cells on the bottom surface were stained with 0.1% crystal violet. Stained cells were imaged under the microscope (Olympus Corporation BX41, Tokyo, Japan) and counted using ImageJ software (NIH, Bethesda, MD.).

Proliferation assay

Endothelial cell proliferation was assessed using the Click-iT™ Plus EdU Cell Proliferation Kit with Alexa Fluor™ 488 dye according to the manufacturer’s instructions (Thermo-Fisher, Waltham, MA). Briefly, endothelial cells were cultured and treated as indicated. Cells were incubated with 10 µM EdU for 2 h at 37 °C, fixed with 4% paraformaldehyde, and permeabilized with 0.5% Triton X-100 in PBS. Incorporated EdU was detected using the Click-iT reaction cocktail containing Alexa Fluor™ 488. Nuclei were counterstained with DAPI. Fluorescent images were acquired using a fluorescence microscope (Olympus Corporation BX41, Tokyo, Japan). Alexa Fluor™ 488 fluorescence intensity was quantified using a Varioskan™ LUX multimode microplate reader (Thermo Fisher Scientific, Waltham, MA).

Wound healing assay

HUVECs were grown in a 24-well plate until confluency and starved for 3 h. The media was changed with fresh basal endothelial media (with or without 100ng/ml VEGF), and a scratch (wound) was made across the wells. Images were captured after 0, 4, 8, and 12 h of incubation at 37 °C, 5% CO₂. The area of clearing was analyzed and quantified with ImageJ software (NIH, Bethesda, MD.).

Apoptosis (TUNEL) and viability (MTT) assay

To measure apoptosis, we used an In-situ Cell Death Detection Kit with fluorescein (Tunnel Assay, Sigma-Aldrich, St. Louis, MO) according to the manufacturer’s instructions. For cell viability, we used MTT Assay (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, a tetrazole assay) (Abcam, Cambridge, MA.) according to the manufacturer. The assay relies on the reduction of MTT by the mitochondrial dehydrogenases to the purple-colored formazan crystals analyzed spectrophotometrically (OD = 590 nm) after DMSO dissolution. All viability and apoptosis measurements were performed in HUVECs at passage 5, without serum deprivation, chemical stress, or any external apoptotic stimulus.

Western blot

HUVECs were lysed in a buffer (Cell Signaling Technology, Danvers, MA) containing PMSF (1 mM), sonicated, and centrifuged at 12.000 rpm for 5 min. Supernatants were collected, and protein concentration was measured with Bradford Assay (Bio-Rad Laboratories, Hercules, CA). Protein samples were separated on SDS-PAGE and transferred to the nitrocellulose membrane. Membranes were blocked with 5% BSA and immunoblotted with primary antibodies at 4 °C overnight. As a secondary antibody, we used an anti-mouse or anti-rabbit secondary antibody (Supplemental materials), and detection was accomplished using the enhanced chemiluminescence kit (ECL Substrate; Bio-Rad Laboratories, Hercules, CA). Images were obtained and quantified using the ChemiDoc™ MP System (Bio-Rad Laboratories, Hercules, CA).

PTP assay

To determine tyrosine phosphatase activity, we used PTP Assay Kit 1 (Sigma-Millipore, St. Louis, MO) according to the manufacturer’s recommendations. This assay was used to detect phosphatase activity, confirmed by malachite green and pNPP hydrolysis assay. To visualize phosphatase activity, we used the ELF-97 phosphatase substrate that produces a bright yellow-green fluorescence precipitate upon hydrolysis by phosphatases (Thermo-Fisher Scientific, Waltham, MA.) in 4% formaldehyde fixed cells and gastrocnemius muscle cryosections.

Real-time quantitative polymerase chain reaction (qPCR)

Total RNA was isolated using an RNA purification kit according to the manufacturer’s instructions (Thermo-Fisher, Waltham, MA). RNA (0.5 µg) was reverse transcribed into cDNA, and qPCR was performed using the StepOnePlus™ Real-Time PCR System (Thermo-Fisher Scientific, Waltham, MA) with mouse and human primer sets (Supplemental materials). PCR settings after initial denaturation (95 °C for 30 s) were as follows (40 cycles): denaturation (95 °C for 15 s), annealing/extension, and plate read (60 °C for 60 s) (Applied Biosystems, StepOnePlus, Real-Time PCR System, Foster City, CA.). The expression level was normalized with GAPDH and quantified using the 2^{− DDCt} method [23].

Statistics

Statistical analyses were performed with RStudio 2024.12.1, and plots were made using GraphPad Prism 8 (San Diego, CA). The results with normal distribution are expressed as mean ± SEM, and data without normal distribution are presented as median ± interquartile range (IQR). The data were tested for normality using the Shapiro-Wilk test. To analyze the statistical significance between two groups with normal data distribution, we used the Welch t-test, and for non-normal distribution, we used the Mann-Whitney U test. For multiple comparisons of normally distributed data, we used 1- or 2-way ANOVA and the Kruskal-Wallis test for non-normally distributed data sets. P-values were adjusted using Holm’s method for multiple comparisons. All tests were 2-tailed, and P < 0.05 was considered statistically significant.

Results

Loss of RGS5 leads to defective angiogenesis following hind limb ischemia

RGS5 is an important hypoxia-inducible protein [24] that may play a critical role in the initiation and progression of postnatal angiogenesis. To determine whether RGS5 is necessary for postnatal angiogenesis, we employed the hindlimb ischemia (HLI) model. Before femoral artery ligation, the left-to-right hindlimb blood perfusion ratio was ~ 1.0 in all groups. Following femoral artery ligation, unilateral blood perfusion was decreased to < 20% of the baseline value in all groups (n = 10). Laser Doppler imaging (LDI) measurements following femoral artery ligation showed complete blood flow recovery within 28 days after surgery in controls. However, compared to controls, the blood perfusion recovery in Rgs5^–/– mice was substantially lower on postoperative day 14 (30.4 ± 2.9 vs. 64.3 ± 3.5%, n = 9–10, P < 0.0001), day 21 (43.6 ± 2.6 vs. 79.7 ± 2.3%, n = 7–10, P < 0.0001), and day 28 (49.6 ± 2.2 vs. 95.7 ± 1.1%, n = 7–10, P < 0.0001). These findings were associated with greater motor deficits in Rgs5-deficient mice at day 14, (1.8 vs. 2.9, n = 9–10, P < 0.0001), day 21, (2.0 vs. 4.0, n = 7–10, P < 0.0001), day 28, (2.4 vs. 4.0, n = 7–10, P < 0.0001) (Fig. 1b).

Following femoral artery ligation, 30% of Rgs5-deficient mice developed distal hindlimb gangrene (10% on day 7 and 20% and day 14, n = 10) (Fig. 1c). Compared to control littermates, capillary densities in the gastrocnemius muscle of Rgs5-deficient mice were substantially lower at 4 weeks after femoral artery ligation (126 ± 7.1 vs. 217 ± 6.8, n = 10, P < 0.0001) (Fig. 1d). Importantly, baseline capillary density did not differ between the groups (n = 5, P > 0.05, Fig. 1d).

Similarly, aortic rings from Rgs5^–/– mice exhibited decreased length and number of sprouting vessels compared with that of control littermates (0.19 ± 0.02 vs. 0.56 ± 0.06 mm, n = 10, P < 0.01, and 12.1 ± 1.6 vs. 56.3 ± 1.9, respectively, n = 10, P < 0.0001) (Fig. 1e). Similarly, the Matrigel plug assay showed a decreased new blood vessel formation in Matrigel plugs (0.8 ± 0.1 vs. 4.6 ± 0.3, n = 5, P < 0.0001) (Fig. 1f).

Pro-angiogenic effect of endothelial RGS5 after hindlimb ischemia

Similar to global Rgs5-deficiency, endothelial-specific Rgs5 deficient mice (EC-Rgs5^–/–) also show substantial decrease in blood flow restoration (EC-Rgs5^–/– 58.6 ± 3.5 vs. EC-Rgs5^flox/flox 98.9 ± 0.8, n = 6–7, day 28, P < 0.0001) and functional recovery (EC-Rgs5^–/– 2.5 vs. EC-Rgs5^flox/flox 3.7, n = 6–7, day 28, P < 0.0001) following femoral artery ligation (Fig. 2b). Furthermore, 14.3% of EC-Rgs5^–/– mice developed distal hindlimb gangrene. (Fig. 2b). Tamoxifen administration in Rgs5^flox/flox mice lacking Cre recombinase (Rgs5^flox/flox +TMX) did not affect postnatal angiogenesis compared with the other control groups (n = 7, P > 0.05, Fig. 2b). Capillary density in the gastrocnemius muscles of EC-Rgs5^–/– mice was substantially reduced 4 weeks after femoral artery ligation compared to control littermates (145.8 ± 12 vs. 229.6 ± 9.2, n = 5, P < 0.001) (Fig. 2c). Baseline measurements showed no difference in capillary density between the groups (n = 5, P > 0.05, Fig. 2c). The aortic ring assay obtained similar results. The length and the number of vascular branches were significantly decreased in EC-Rgs5^–/– mice compared to those of control littermates. (length 0.28 ± 0.04 vs. 0.5 ± 0.05 mm, n = 5, P < 0.01, and count 110.2 ± 11.8 vs. 191.6 ± 14.3, n = 5, P < 0.01) (Fig. 2d). Similarly, the Matrigel plug assay showed a decreased new blood vessel formation in Matrigel plugs (2.2 ± 0.3 vs. 6.1 ± 0.2, n = 5, P < 0.0001) (Fig. 2e).

RGS5 deficiency decreases endothelial angiogenic capacity and VEGFR2 activity

To determine the potential mechanism by which endothelial RGS5 regulates postnatal angiogenesis, we used human umbilical vein endothelial cells (HUVECs) at passage 2 and knocked down RGS5 expression. The RGS5-deficient cells were generated by lentiviral transduction of RGS5 shRNA. Control cells were transduced with non-mammalian shRNA (Mission™ shRNA, Sigma-Aldrich, St. Louis, MO). To visualize and measure transduction efficiency, we used the TurboGFP™ constructs, followed by a Western blot analysis of RGS5 expression (Fig. 3a). Endothelial RGS5 deficiency leads to decreased migration (511.8 ± 44.4 vs. 1279 ± 90, n = 5, P < 0.001, Fig. 3b) and the ability to form tubes after VEGF stimulation (39.4 ± 1.5 vs. 64.3 ± 2.7, n = 7, P < 0.0001, Fig. 3c). Indeed, RGS5-deficient endothelial cells were not able to fill the gap in a wound healing assay even after 12 h (RGS5-KD 58.1 ± 4% vs. Control 100%, n = 5, P < 0.0001, and RGS5-KD VEGF 82.9 ± 2.6% vs. Control VEGF 100%, n = 5, P < 0.01 Fig. 3d).

Fig. 3.

Loss of RGS5 Reduces Endothelial Angiogenic Capacity and VEGFR2 Activity. a TurboGFP™ (left panel) and Western blot (right panel) evaluation of transduction and RGS5-KD efficiency. b Trans-well migration, c Matrigel tube formation, and d wound healing assays of RGS5-KD and control endothelial cells with and without VEGF stimulation (10 ng/ml) (n = 5–7). e Apoptosis TUNEL assay of RGS5-KD and control endothelial cells (n = 5). f Proliferation assay (n = 5) and g Western blot analysis of VEGFR2 activation in RGS5-KD and control endothelial cells (n = 5) Bar = 50 μm. Data are presented as means. b), c), e), f and g) Unpaired t-test, d) One-way ANOVA with Holm correction for multiple testing. *** P < 0.001 ****P < 0.0001

To determine whether the lack of VEGF signaling has an important anti-apoptotic effect in RGS5-deficient endothelial cells, we analyzed whether the RGS5 deficiency affects apoptosis [25]. TUNEL assay showed a significant increase in the number of apoptotic cells in RGS5-deficient cells compared to that of control cells (152.4 ± 10.3 vs. 40.4 ± 3.7, n = 5, P < 0.001, Fig. 3e). Similar results were obtained with the MTT assay, which assesses viability based on metabolic activity. RGS5-deficient endothelial cells were less metabolically active compared to their respective controls (Str-RGS5-KD 0.39 ± 0.02 vs. Str-Control 0.62 ± 0.02, Str + VEGF- RGS5-KD 0.34 ± 0.03 vs. Str + VEGF- Control 0.56 ± 0.02, n = 5, P < 0.01, Supplement Fig. 1a). HUVECs at passage 5 exhibit minimal baseline apoptosis under standard culture conditions; therefore, the increased apoptotic rate observed in Rgs5-deficient cells reflects an intrinsic defect rather than passage-dependent senescence. Furthermore, endothelial RGS5 deficiency significantly decreased cell proliferation compared to control cells (2.3 ± 0.2 vs. 0.4 ± 0.1, n = 5, P < 0.0001, Fig. 3f).

To determine VEGF-VEGFR2 signaling in RGS5-deficient endothelial cells, we analyzed the expression and activation of VEGFR2. There were no differences in the total expression of VEGFR2 or other important mediators of postnatal angiogenesis in RGS5-deficient endothelial cells compared to controls (n = 5, P > 0.05, Supplement Figure b). However, interestingly, western blot analysis showed decreased phosphorylation of VEGFR2 tyrosine¹¹⁷⁵ and activation in RGS5-KD cells following VEGF stimulation (0.41 ± 0.03 vs. 1.36 ± 0.06, n = 5, P < 0.0001, Fig. 3g), whereas phosphorylation of other VEGFR2 tyrosine residues (Tyr¹⁰⁵⁹, Tyr⁹⁵¹, and Tyr⁹⁹⁶) were not significantly affected (n = 5, P > 0.05, Supplement Fig. 1c). Altogether, these results indicate that RGS5 deficiency decreases endothelial angiogenic capacity, their response to VEGF stimulation, and VEGFR2 activation.

Loss of Endothelial RGS5 Inhibits PI3K/AKT and ERK1/2 signaling pathways

VEGFR2 downstream signaling includes the activation of phosphatidyl inositol 3-kinase (PI3K)/AKT and extracellular signal-regulated kinase 1/2 (ERK1/2) signaling pathways that regulate a variety of cellular functions, including proliferation, migration, growth, viability, and angiogenesis [26]. Given the established link between VEGF, AKT, and ERK1/2 signaling, we tested whether RGS5 deficiency could affect AKT and ERK1/2 activation. Western blot analysis showed decreased activation and phosphorylation of AKT (Ser⁴⁷³) in RGS5-deficient cells when compared to controls (0.5 ± 0.07 vs. 1.06 ± 0.03, n = 5, P < 0.0001, Fig. 4a). However, the expression of AKT was not affected by RGS5 deficiency (n = 5, P > 0.05, Fig. 4a). Similarly, RGS5 deficiency significantly reduced ERK1/2 phosphorylation (Thr²⁰²/Tyr²⁰⁴) without affecting total ERK1/2 expression (Control 1.06 ± 0.03 vs. RGS5-KD 0.6 ± 0.03, and Control+VEGF 1.2 ± 0.02 vs. RGS5-KD+VEGF 0.6 ± 0.04, n = 5, P < 0.0001, Fig. 4b).

Fig. 4.

Deficiency of RGS5 Reduces AKT and ERK1/2 Phosphorylation and Increases Phosphatase Activity in Endothelial Cells. a and b Western blot evaluation of AKT and ERK1/2 activation after VEGF stimulation (10 ng/ml) in RGS5-KD and control cells (n = 5). c Phosphatase assay of RGS5-KD and control endothelial cells (n = 6). Immunofluorescence evaluation of phosphatase activity in d in vitro cultured RGS5-KD endothelial cells and e gastrocnemius muscle of endothelial-deficient RGS5 mice and their controls (n = 5, Bar = 50 μm). f Western blot evaluation of VEGFR2 activation with and without VEGF and sodium stibogluconate (10 µg/ml SSG) treatment in RGS5-KD and control cells (n = 5). g Western blot analysis of SHP-1 activity after VEGF stimulation in RGS5-KD and control endothelial cells (n = 5). Data are presented as means (median, as appropriate for c and g). a) One-way ANOVA with Holm correction for multiple testing, b, c), d), e), f), and g) Two-way ANOVA with Holm correction for multiple testing. *P < 0.05, **P < 0.01, ****P < 0.0001

RGS5 deficiency increases SHP-1 phosphatase activity and mediates VEGF signaling

The phosphorylation state of VEGFR2 reflects a balance between protein tyrosine kinases (PTKs) and protein tyrosine phosphatases (PTPs). Accordingly, PTP assay and ELF-97 staining were used to analyze the potential effect of RGS5 deficiency on phosphatase activity. Phosphatase activity was first assessed in cultured endothelial cells (in vitro, Fig. 4c and d) and subsequently confirmed in ischemic muscle tissue sections (in vivo, Figure. 4e). When compared to controls, RGS5-KD endothelial cells showed increased phosphatase activity (942.3 [IQR, 836.2–1057] vs. 577.1 [IQR, 563.8-613.9], n = 6, P < 0.0001, Fig. 4c). Furthermore, ELF-97 staining as a measure of PTP activity was also increased in RGS5 deficient endothelial cells in vitro (Control 74.8 ± 2.4 vs. RGS5-KD 84.2 ± 3, P < 0.05, and Control+VEGF 123.8 ± 4.3 vs. RGS5-KD+VEGF 164.4 ± 4.43, n = 5, P < 0.0001 Fig. 4d) and in the gastrocnemius muscles of EC-Rgs5^–/– mice after femoral artery ligation in vivo compared to their respective controls (EC-Rgs5^flox/flox 79.1 ± 2.6 vs. EC-Rgs5^–/– 111.5 ± 4.6, n = 6, P < 0.0001, Fig. 4e). These results suggest that PTP activity in endothelial cells is intrinsically inhibited by RGS5.

To determine whether inhibition of PTP could restore VEGF-stimulated VEGFR2 phosphorylation in RGS5-deficient endothelial cells, we treated cells with the PTP inhibitor sodium stibogluconate (10 µg/ml, SSG, Cayman Chemical, Ann Arbor, MI.). Sodium stibogluconate treatment almost completely rescued phosphorylation of tyrosine¹¹⁷⁵-VEGFR2 in RGS5-KD endothelial cells (83 ± 3.5 vs. 96.6 ± 3.8, n = 5, P < 0.05, Fig. 4f). At the low dose (10 µg/ml) used in our study, SSG treatment blocks SHP-1 activity while closely related SHP-2 remains fully functional [27]. Therefore, we decided to analyze the expression and activity of SHP-1 in RGS5-deficient endothelial cells. Although the expression of SHP-1 did not change, phosphorylation of its tyrosine⁵³⁶ was substantially higher in RGS5-KD cells after VEGF treatment (1.05 [IQR, 1.03–1.1] vs. 0.75 [IQR, 0.72–0.76], n = 5, P < 0.01, Fig. 4g). The time curve for VEGF treatment has shown a decrease in VEGFR2^Tyr1175 phosphorylation and an increase in SHP-1 activity in RGS5-KD cells at every time point for 45 min. Maximum differences were observed between 5 and 10 min. (n = 5, P < 0.0001, Supplement Fig. 1d).

To determine whether SHP-1 deletion in RGS5-deficient cells could rescue VEGFR2-tyrosine¹¹⁷⁵ phosphorylation, we generated double RGS5-SHP-1 knockdown endothelial cells with their respective controls (Fig. 5a). Double knockdown was sufficient to completely rescue VEGFR2 phosphorylation after VEGF treatment in RGS5-deficient cells. (RGS5-SHP-1-KD 2.83 ± 0.2 vs. RGS5-KD+Control (SHP-1) 1.17 ± 0.15, n = 5, P < 0.001, Fig. 5b). Furthermore, double knockdown completely rescued endothelial cell’s ability to migrate (RGS5-SHP-1-KD 98.2 ± 1.1% vs. RGS5-KD+Control (SHP-1) 64.4 ± 2.6% at 12 h, n = 5, P < 0.0001, Fig. 5c) and to form new vascular tubes (RGS5-SHP-1-KD 64 ± 3.5 vs. RGS5-KD+Control (SHP-1) 19.6 ± 2.5, n = 5, P < 0.0001, Fig. 5d).

Fig. 5.

Loss of SHP-1 restores VEGFR2 Activity in RGS5 Deficient Endothelial Cells. a Western blot analysis of RGS5 and SHP-1 knockdown and b their effect on VEGFR2^Tyr−1175 phosphorylation. c Wound healing and d tube formation assays of double RGS5-SHP-1 knockdown endothelial cells and respective controls (Control-unmanipulated cells, Control (RGS5-KD)- cells transduced with non-targeting shRNA control for RGS5-KD, RGS5-KD- knockdown of RGS5, Control (SHP-1-KD)- cells transduced with non-targeting shRNA control for SHP-1-KD, SHP-1-KD- knockdown of SHP-1, RGS5-KD+Control (SHP-1-KD)- RGS5 knockdown and non-targeting shRNA control for SHP-1-KD). Bar = 50 μm. Data are presented as mean (median, as appropriate for c). b), and d) One-way ANOVA with Holm correction for multiple testing (n = 5 per group, **P < 0.01, ***P < 0.001, ****P < 0.0001)

Normalization of Rho/ROCK activity does not rescue the angiogenic deficit in Rgs5-deficient endothelial cells

Our previously published work demonstrated that RGS5-deficient brain microvascular endothelial cells exhibit reduced eNOS expression and activity, a defect driven by increased Rho/ROCK signaling [17]. To evaluate whether this mechanism contributes to the impaired angiogenic phenotype observed in Rgs5-deficient endothelial cells in the present study, we first determined the concentration of the pharmacological ROCK inhibitor Y-27f632 required to normalize Rho/ROCK activity. ROCK activity was assessed by measuring the phosphorylation of myosin-binding subunit (MBS) at Thr⁸⁵³ (p-MBS), a well-established downstream target of ROCK. ROCK activity was quantified as the ratio of phosphorylated MBS to total MBS (p-MBS/MBS) by Western blot analysis. We found that 20 µM Y-27632 was sufficient to reduce ROCK activity to levels comparable to control endothelial cells (Supplement Fig. 2a).

Next, we assessed whether ROCK inhibition could rescue the angiogenic defects in Rgs5-deficient endothelial cells. Matrigel tube formation (Control + VEGF 46.54 ± 3.8 vs. RGS5-KD+VEGF + Y 22.1 ± 3.6, n = 5, P < 0.001, Supplement Fig. 2b) and Wound healing assays (4 h- Control + VEGF 36.6 ± 5.4 vs. RGS5-KD+VEGF + Y 28.8 ± 4.9, n = 5, P < 0.05, 8 h- Control + VEGF 98.6 ± 3.1 vs. RGS5-KD+VEGF + Y 53.6 ± 5.9, n = 5, P < 0.01, 12 h- Control + VEGF 100 vs. RGS5-KD+VEGF + Y 79 ± 9, n = 5, P < 0.01, Supplement Fig. 2c) revealed that while normalization of Rho/ROCK activity partially improved endothelial behavior, it did not restore angiogenic capacity to control levels after VEGF stimulation, indicating that additional RGS5-dependent mechanisms underlie the angiogenic impairment.

Discussion

We have shown that RGS5 is required to suppress SHP-1 phosphatase activity, thereby preserving VEGFR2 phosphorylation and downstream pro-angiogenic signaling. Although RGS5 subcellular localization in endothelial cells may be dynamically regulated by its activation state and GPCR-dependent pathways, the mechanisms regulating its localization were not directly examined in the present study. It is possible that during VEGF stimulation or endothelial proliferation, increased Gαi/Gαq signaling and elevated intracellular Ca²⁺ levels promote cytosolic retention of RGS5, thereby facilitating its regulatory effects on SHP-1 and VEGFR2 signaling. However, these proposed mechanisms remain speculative and require further investigation. Importantly, deletion of Rgs5, either globally or specifically in endothelial cells, substantially decreased postnatal angiogenesis, resulting in delayed functional recovery, gangrene, and autoamputation in mice following femoral artery ligation. Collectively, these findings establish RGS5 as a critical regulator of endothelial function and neovascularization in ischemic tissues.

Angiogenesis, a critical process in normal embryonic development and wound repair, also plays a central role in various pathological conditions, including cancer, metastasis, eye diseases, rheumatoid arthritis, and ischemic and inflammatory diseases [28]. The pro-angiogenic factor, VEGF, is a major regulator of angiogenesis. Hypoxia is the most important stimulator of VEGF expression, resulting from hypoxia-inducible factor 1 (HIF-1) binding to the VEGF promoter [29]. Similarly, RGS5 is also identified as an HIF-1 inducible gene, involved in angiogenesis and blood vessel stabilization [24]. While previous studies suggested that RGS5 counteracts the pro-angiogenic effects of VEGF with an unknown mechanism [24], our study has shown that Rgs5 deficiency diminishes VEGF signaling and postnatal angiogenesis, likely by increasing SHP-1 activity. Increased SHP-1 activity directly dephosphorylates tyrosine¹¹⁷⁵ on VEGFR2, impacting its downstream signaling pathways, including the critical AKT and ERK1/2 pathways [30, 31]. Furthermore, our study showed that RGS5 deficiency leads to reduced endothelial migration, branching, and tube formation, accompanied by increased endothelial cell apoptosis, partly due to AKT and ERK1/2 signaling inhibition [32]. While some groups reported that RGS5 overexpression increases apoptosis [24], our findings are consistent with majority of studies showing that RGS5 expression in blood vessels correlates positively with new blood vessel growth and tumor progression [10, 12, 33]. Furthermore, a recent study showed that TGFβ in the tumor-residing pericytes leads to the elevation of pSmad2-RGS5 binding and their nuclear trafficking. Nuclear trafficking of pSmad2–RGS5 complex therefore preserves PI3K–AKT signaling and protects pericytes from apoptosis. These findings highlight that RGS5’s biological role is determined by whether it resides in the cytosol (pro-apoptotic) or the nucleus (anti-apoptotic) [33]. Further experiments are needed to determine the role of nuclear RGS5 in postnatal angiogenesis.

The AKT and ERK1/2 signaling pathway not only influences endothelial apoptosis, migration, branching, and tube formation, but also can activate eNOS, leading to increased nitric oxide (NO) production [34]. Nitric oxide is a fundamental regulator of vascular homeostasis, blood pressure, vascular remodeling, and angiogenesis. Inhibition of the PI3K/AKT pathway or mutation of the AKT phosphorylation site at Serine¹¹⁷⁷ on eNOS prevents VEGF-stimulated NO production [30]. This finding agrees with our prior study that the loss of Rgs5 in mice leads to reduced activation of eNOS and hypertension [17]. Interestingly, a previous study suggested that global Rgs5 deletion may cause hypotension [35]. This is in contrast to our earlier study showing that Rgs5-deficient mouse lines exhibit mild hypertension [17, 36]. Inhibition of VEGF-A signaling is known to induce hypertension, primarily through reduced nitric oxide bioavailability and microvascular rarefaction [37, 38]. The angiogenic defects observed in our Rgs5-deficient mice, therefore, could also explain hypertension in these mice, likely arising from increased SHP-1 activity and impaired VEGFR2 downstream signaling.

We found that the increased SHP-1 activity observed in RGS5-deficient endothelial cells potentially inhibits angiogenesis through the dephosphorylation of VEGFR2^Tyr1175 [39]. To restore VEGFR2 phosphorylation in these cells, SHP-1 knockdown was generated in RGS5-deficient endothelial cells. Silencing SHP-1 proved sufficient to rescue VEGFR2^Tyr1175 phosphorylation and restore the angiogenic potential of RGS5-deficient cells. Indeed, emerging evidence suggests that SHP-1 is an important regulator of Ang-1-Tie2 signaling [40]. While our findings suggest the importance of endothelial RGS5 in angiogenesis, further investigations are needed to assess the impact of Rgs5 deletion on hematopoietic cells, pericytes, and smooth muscle cells in postnatal angiogenesis. Furthermore, the potential roles of other protein tyrosine phosphatases (PTPs), including SHP-2, PTP-ε, VE-PTP, and CD148, merit consideration in the RGS5 regulation of postnatal angiogenesis.

Consistent with our previous findings demonstrating increased Rho/ROCK signaling and impaired eNOS activity in RGS5-deficient endothelial cells [17], we investigated whether enhanced ROCK activation contributes to the impaired angiogenic phenotype observed in the present study. Pharmacological inhibition of ROCK with Y-27632 effectively normalized ROCK activity in RGS5-deficient endothelial cells, as demonstrated by reduced phosphorylation of the ROCK downstream target MBS. However, despite partial improvement in endothelial migration and tube formation, ROCK inhibition was insufficient to fully restore angiogenic responses to control levels following VEGF stimulation. These findings suggest that increased Rho/ROCK signaling contributes to endothelial dysfunction in RGS5-deficient cells but does not fully account for the impaired angiogenic phenotype.

In summary, our study identifies RGS5 as a critical regulator of VEGF signaling through its modulation of SHP-1 phosphatase activity. Loss of RGS5 results in enhanced SHP-1 activity, increased VEGFR2 dephosphorylation, and impairment of postnatal angiogenesis. These findings establish RGS5 as an important mediator of VEGF signaling. However, the molecular mechanism by which RGS5 regulates SHP-1 activity remains unknown. Further experiments are needed to determine how RGS5 modulates SHP-1 phosphatase activity, including whether this regulation involves direct interaction, alterations in SHP-1 conformation, post-translational modifications, changes in subcellular localization, or intermediary signaling partners. Elucidating this mechanism could lead to the development of therapeutic strategies that enhance angiogenesis and improve functional recovery in ischemic cardiovascular diseases.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

NS, HCT, and DP performed experiments. ME performed statistical analysis. NS and JKL supervised the project and designed experiments. NS and JKL interpreted the data and wrote the manuscript. All authors contributed intellectually to the project and reviewed the manuscript.

Funding

National Institutes of Health (NS070001, HL052233, and HL136962; JKL).

American Heart Association (17POST33650097; NS).

Data availability

The authors declare that all data supporting the findings of this study are available and will be available from the corresponding author upon request.

Declarations

Conflict of interest

The authors have declared that there are no conflicts of interest.