Reduced cofilin activity as a mechanism contributing to endothelial cell stiffening in type 2 diabetes

Abstract

An emerging instigator of endothelial dysfunction in type 2 diabetes (T2D) is stiffening of the cell. Previous reports suggest that polymerization of filamentous actin (F-actin) is a potential mediator of endothelial stiffening. Actin polymerization is limited by active cofilin, an F-actin-severing protein that can be oxidized, leading to its inactivation and loss of severing capability. Yet, whether these mechanisms are implicated in endothelial stiffening in T2D remains unknown. Herein, we report that endothelial cells exposed to plasma from male and female subjects with T2D, and the aortic endothelium of diabetic male mice (db/db), exhibit evidence of increased oxidative stress, F-actin, and stiffness. Furthermore, we show reactive oxygen species, including H₂O_2, are increased in the endothelium of mesenteric arteries isolated from db/db male mice, and that exposure of endothelial cells to H₂O₂ induces F-actin formation. We also demonstrate, in vitro, that cofilin-1 can be oxidized by H₂O₂, leading to reduced F-actin severing activity. Finally, we provide evidence that genetic silencing or pharmacological inhibition of LIM kinase 1, an enzyme that phosphorylates and thus inactivates cofilin, reduces F-actin and cell stiffness. In aggregate, this work supports inactivation of cofilin as a potential novel mechanism underlying endothelial stiffening in T2D.

NEW & NOTEWORTHY

Cell stiffening is an emerging contributor to endothelial dysfunction, a classic feature of type 2 diabetes (T2D). However, the mechanisms underlying endothelial stiffening remain largely unknown. This work provides evidence that oxidative stress-induced inactivation of cofilin, a key F-actin severing protein, may be implicated in increasing endothelial F-actin and cell stiffness in T2D.

INTRODUCTION

Type 2 diabetes (T2D) is increasingly recognized as a cardiovascular-centric disease characterized by oxidative stress, endothelial dysfunction, and arterial stiffening, features that are likely interconnected. A prevailing view is that T2D is associated with increased oxidative stress, endothelial dysfunction, and reduced nitric oxide bioavailability, key contributors to arterial stiffening. However, evidence from preclinical models of obesity and T2D demonstrates that endothelial cells also become stiff early in the disease course, coinciding with indices of impaired endothelium-dependent dilation (e.g., blunted flow-mediated dilation, FMD) (1–4). Because it has been reported that stiffer endothelial cells release less nitric oxide (5, 6), it is conceivable that stiffening of the endothelium is an early pathogenic event that contributes to endothelial dysfunction and vascular disease in T2D. Yet, the mechanisms underlying endothelial stiffening remain largely unknown.

The primary determinant of mammalian cell stiffness is the cytoskeleton (7). This ubiquitous structure comprises a highly dynamic network of actin filaments, microtubules, and intermediate filaments. These structures are sensitive to biochemical and biomechanical forces and are capable of remodeling and stiffening in response to several stimuli (8). Specifically, endothelial cell stiffening is associated with increased polymerization of filamentous-actin (F-actin) (9). The balance between polymerization and depolymerization of F-actin is critical to determining its cellular organization and content. This is regulated by several actin-binding proteins, including cofilin (10, 11). Active cofilin is capable of severing and depolymerizing F-actin filaments (11). In contrast, inactivation of cofilin leads to increased polymerized F-actin and cell stiffening (12, 13). We have previously reported that inhibition of Lim Kinase (LIMK), a protein that phosphorylates and inactivates cofilin, reduces F-actin and stiffness in vascular smooth muscle cells (13). However, whether inactivation of cofilin causes endothelial cell stiffening is yet to be determined.

The activity of cofilin is coordinated by several post-translational modifications, including oxidation (12). Indeed, H₂O₂-induced oxidation of cofilin has been demonstrated to impair its F-actin depolymerizing activity and lead to increased F-actin in non-endothelial cells (12). As such, we hypothesized that oxidative stress-induced inactivation of cofilin leads to increased F-actin and contributes to endothelial stiffening in T2D. Furthermore, we determined whether inhibition of LIMK could reduce endothelial cell F-actin and stiffness.

MATERIALS AND METHODS

Ethics and approvals

All human study procedures were conducted in accordance with the Declaration of Helsinki and approved by the University of Missouri Institutional Review Board (IRB, No. 2028142, 2008181, and 2012106). Written informed consent was obtained from all participants prior to study participation. All animal study procedures received prior approval by the University of Missouri Animal Care and Use Committee and were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Human participants

Seventy-two participants with T2D (39 females and 33 males, age=55±1 yrs, body mass index=35.43±0.8) and 43 aged-matched healthy participants (27 females and 16 males, age=51±2 yrs, body mass index=23.3±0.3) who underwent measurements of brachial artery FMD and carotid-to-femoral pulse wave velocity (cf PWV) were included in this retrospective analysis. Brachial artery FMD and cf PWV were assessed after an overnight fast, as previously described (14, 15) and according to published guidelines (16, 17). All participants were free of a history of recent cardiovascular events (<12 mo), renal or hepatic diseases, autoimmune diseases, active cancer, current tobacco use, excessive alcohol consumption (>7 drinks/wk for females; and >14 drinks/wk for males), pregnancy, and uncontrolled hypertension (>180 mmHg systolic, >110 mmHg diastolic). It should be noted that data from a subset of participants included in this retrospective analysis were also included in previous publications examining different research questions (15, 18–20).

Animal studies

C57BL/6J (wild-type) and db/db (a mouse model of T2D) male mice on the same background (12 to 14 wk-old) were obtained from Jackson Laboratory. Given the lack of a significant T2D by sex interaction on vascular outcomes in our human cohort, we proceeded with ex vivo experiments using mice from only one sex. Mice with ad libitium access to water were fed standard chow (5053-PicoLab Rodent Diet 20, LabDiet) and kept under 12h:12h dark/light cycles. Mice were euthanized via inhalation of 2% isoflurane (AKORN Animal Health) with room air (250mL/min), followed by pneumothorax and exsanguination. Arteries were excised for microscopy and functional testing. Experimenters were blinded when possible.

Endothelial cell culture experiments

Human umbilical vein endothelial cells from pooled donors (HUVECs; Lonza) were used in all cell culture experiments and cultured under standard conditions in VascuLife EnGS culture media (2% FBS). Cells were starved for 24h in culture media (0.5% FBS) prior to treatment. To assess the impact of the circulating milieu of T2D on endothelial cells, plasma samples from a subset of T2D (n=10; “T2D plasma”) and healthy age-matched control subjects (n=10; “healthy plasma”) were pooled and diluted (30% of final volume) in cell culture media (0.5% FBS) before exposure to cells for 24h. To test the effects of oxidative stress, cells were treated with vehicle (ddH₂O) versus H₂O₂ (50μM) for 1.5h. To assess the effects of LIMK inhibition, cells were treated with vehicle (<0.01% DMSO) versus LIMKi3 (1μM) for 18h (13). To test the effects of F-actin induction and stabilization, cells were treated with vehicle (~0.01% DMSO) versus jasplakinolide (200nM) for 1h. To assess the effects of reduced cofilin activity, cells were transfected with DNA plasmids encoding non-activatable mutant cofilin (pEGFP-N1 human cofilin S3E) or wild-type cofilin (pEGFP-N1 human cofilin WT) for 6h using Lipofectamine 3000. pEGFP-N1 human cofilin S3E and pEGFP-N1 human cofilin WT were gifts from James Bamburg. To assess the role of LIMK1, cells were transfected with an siRNA targeting LIMK1 (20nM) or a scramble control (20nM) using Lipofectamine^™ RNAiMAX transfection reagent complexed in Opti-MEM^™ medium. Although HUVECs proliferate easily, are reliable, easy to modify genetically, and more readily available than other cell types, we acknowledge that future studies should extend these observations to endothelial cells from other vascular beds. All current and future reagents, stains, and antibodies are detailed in the supplementary materials which can be accessed here: 10.6084/m9.figshare.27328101

Assessment of FMD in isolated arteries

To determine the effects of pharmacologically induced endothelial stiffening on endothelial function, FMD was assessed in isolated mesenteric arteries, as previously described (21). Briefly, arteries were cannulated, pressurized at 70mmHg, and exposed intraluminally to jasplakinolide (200nM; 1h) before preconstriction with 10⁻⁶ M phenylephrine and exposure to incremental increases in intraluminal flow (0.0–3.3 ml/h). Arteries were exposed to cumulative concentrations of sodium nitroprusside (SNP; 10⁻⁹–10⁻⁴ M) to evaluate endothelium-independent vasodilation.

Determination of endothelial cell stiffness

Endothelial cell stiffness in aortic explants and cultured cells was assessed following nanoindentation at room temperature using a silicon nitride cantilever (Bruker, MLCT, Billerica, MA) with a NanoWizard IV Atomic Force Microscope (JPK Bruker, Berlin, Germany) mounted on a Leica DMi8 automated microscope (Leica Microsystems, Inc., Morrisville, NC). The endothelial surface of en-face aortic explant preparations and cultured human endothelial cells fixed in 4% paraformaldehyde (PFA) for 15mins, were indented and calculation of Young’s modulus of Elasticity from the generated force curves was performed, as previously described (2, 13). Fixed cells have been reported to maintain relative changes in stiffness (22).

Immunofluorescent assessment of F-actin and reactive oxygen species (ROS)

Immunofluorescent images of cultured endothelial cells were obtained from cells seeded on 15-well-ibidi plates (Ibidi, #81506, Fitchburg, WI). Cells were fixed in 4% PFA and permeabilized with 0.1% Triton X-100 for 15mins. Cultured endothelial cells were incubated for 1h in 1.5nM 40,6-diamidino-2-phenylindole (DAPI, 1:1000), 200nM Alexa Fluor 568 phalloidin (1:400) to stain for nuclei, and F-actin, respectively. Widefield epifluorescent images were acquired with a Leica DMi8 automated microscope, with THUNDER imaging technology (Leica Microsystems, Inc., Morrisville, NC) using a 63X/1.4 NA oil or 40X/0.8 NA air objective. Positive fluorescent areas or fluorescence intensity were quantified using ImageJ (NIH, Bethesda, MA) or Imaris software (Bitplane, Inc., Concord, MA), respectively, and normalized by cell number. Isolated mesenteric endothelial tubes were prepared as previously described (23). Endothelial tubes were incubated with the fluorogenic probes dihydroethidium (DHE, 10μM) and dichlorofluorescein (DCF, 15μM) to visualize production of ROS, using a MV PLAPO 2XC/0.5 NA objective coupled to a XR/Mega10 camera on an Olympus MVX10 microscope. Fluorescence intensity was quantified at 5min intervals and fluorescence accumulation rate was determined by linear regression, as previously described (24).

Determination of aortic ring oxidative stress with Immunohistochemistry

Aortic rings (2mm) isolated from wild-type and db/db mice were fixed in 4% paraformaldehyde and paraffin-embedded before transversely sectioned (5μm slices) and incubated with a 3-nitrotyrosine antibody (1:200). Aortic sections were counter-stained with hematoxylin to visualize the nuclei. Transmitted light images of the sections were acquired with an Olympus light microscope (Olympus, BX43) using a 40X/0.65 NA air objective. The obtained images were color-deconvoluted using the H-DAB vector on ImageJ and R-channel images specific to 3-nitrotyrosine were used for quantification. Using ImageJ, the endothelium was segmented from the medial layer to quantify the endothelium-specific intensity of 3-nitrotyrosine.

Determination of protein expression via western blotting

Protein expression was assessed in cultured endothelial cell lysates prepared in RIPA buffer, EDTA (5mM), and HALT^™ phosphatase inhibitor cocktail. Proteins within samples were separated in Criterion Tris-Glycine eXtended-PAGE precast-gels (Bio-Rad) and transferred onto polyvinylidene difluoride membranes. Specific proteins were probed using the following primary antibodies: cofilin (1:000), phosphorylated cofilin Ser³ (1:500), and LIMK1 (1:1000). Secondary antibody was Goat Anti-Rabbit (1:5000). Blots were imaged using a ChemiDoc XRS+ (Bio-Rad) and protein band intensities were quantified via densitometry using Image Lab Software (v.6.1, Bio-Rad). All specific protein bands were normalized to total protein determined by UV-activated stain-free gels as a loading control.

F-actin depolymerization assay

F-actin depolymerization was measured using a commercially available Actin Polymerization Biochem Kit^™, according to manufacturer instructions. Briefly, 200μL of pyrene-labeled F-actin (0.2mg/mL) per well was added to a 96-well plate with 200μL of buffer control. Baseline fluorescence was measured using a plate reader (BioTek Synergy H1, Agilent, Santa Clara, CA). Twenty μL of human recombinant cofilin-1 (1mg/mL), previously incubated in either nanopure water, 50μM of H₂O₂ (cofilin oxidation confirmed by mass spectroscopy; data not shown), 1mM of dithiothreitol (DTT), with gentle agitation at 37°C for 0.5h, or subjected to heat denaturation (10mins, 100 °C; data not shown), was added to wells and fluorescence decay (indicative of F-actin depolymerization) data were reported every 5min for 1.5h.

Proteomics

Proteomics analysis of T2D plasma-treated cultured endothelial cells were performed at the University of Missouri Charles W. Gehrke Proteomics Center. Cell samples were prepared and analyzed as previously described (25). Briefly, cell lysates were precipitated in acetone prior to peptide analysis using a Bruker timsTOF pro2 connected to an Evosep LC system with DIA acquisition method using 44min LC gradient.

Statistical analyses

Statistical analyses were performed using GraphPad Prism (version 10, Prism Software, La Jolla, CA). Shapiro-Wilk test was used to determine normality. The Robust regression and Outlier removal (ROUT) test was used to detect and remove outliers based on a false discovery rate with Q=5%. Unpaired Student’s t-test and Mann-Whitney were used to compare independent samples, as appropriate. Two-way analysis of variance (ANOVA) with repeated-measurements was followed with Bonferroni’s post-hoc test accordingly to assess group/condition by time/concentration effects. Data are presented as Mean±SEM. Significance was accepted at P≤0.05. Data supplements and detailed statistical results can be accessed here: 10.6084/m9.figshare.27328101

RESULTS

T2D is associated with arterial stiffening, endothelial cell stiffening, and endothelial dysfunction

We show that brachial artery FMD is decreased while cf PWV is increased in a cohort of females and males with T2D, relative to healthy counterparts (Figure 1A). The effect of T2D on these two vascular outcomes was independent of sex (T2D by sex interaction P>0.05). Thus, data from both sexes were combined (Figure 1A). Corroborating our human data, we show that FMD is decreased while stiffness (E_inc) is increased in mesenteric arteries isolated from db/db versus wild-type male mice (Figure 1A). Furthermore, we report that aortic acetylcholine-induced relaxation is reduced while endothelial cell stiffness, as assessed by atomic force microscopy, is increased in db/db versus wild-type male mice (Figure 1A). There were no significant differences in SNP endothelium-independent vasodilation between db/db and wild-type mice (data not shown). These observations support the notion that, in T2D, vascular stiffening is coupled with endothelial dysfunction.

Figure 1. T2D is associated with impaired FMD and endothelial stiffening, whereas pharmacologically induced endothelial F-actin impairs FMD in isolated arteries.

A: Human brachial artery FMD and cf PWV (females, n=66; males, n=49); FMD (n=6–7) and incremental modulus of elasticity (E_inc) of isolated mesenteric arteries (n=6–7), and area-under-the-curve (AUC) (n=7–9) of endothelium-dependent relaxation to acetylcholine and endothelial stiffness in isolated aortic rings from wild-type and db/db mice (11–13 weeks of age). B: F-actin content (yellow), representative images, and C: quantification of cortical endothelial cell stiffness in cells exposed to T2D or healthy plasma (n=14–15 and 7, respectively). D: Proteomic analysis of endothelial cells exposed to T2D plasma or healthy plasma. Significant differences (Q-value 0.05, 1.2-fold change) in proteins associated with the cytoskeleton are visualized. E: F-actin content (yellow) (n=7–8), with representative images; scale bar = 30μm. F: Quantification of endothelial cell stiffness in cells exposed to jasplakinolide (jaspk; 200nM, 1h) or vehicle (n=6). G: Mesenteric arteries isolated from C57BL/6J male mice exposed intraluminally to either jaspk (200nM, 1h) or vehicle and exposed to increasing intraluminal flow rates to induce FMD (n=9), or vasodilatory responses to SNP(n=9). Inset data for FMD and SNP are expressed as AUC (n=9). Data are presented as mean±SEM. Two-way ANOVA with repeated measures was used to assess FMD and SNP-induced vasodilation (G). Two-tailed paired (insets in G) and unpaired Student’s t-tests were used in all other analyses. *P≤0.05 vs. respective controls and main effect of condition.

Exposure to T2D plasma enhances endothelial F-actin and cell stiffness in cultured endothelial cells

We demonstrate that F-actin and cell stiffness increase in cultured human endothelial cells treated with plasma collected from subjects with T2D compared to those treated with plasma from healthy subjects (Figure 1B,C). In further support of the association between T2D, F-actin, and endothelial stiffness, proteomic analyses reveal changes in several actin cytoskeletal proteins in cultured endothelial cells exposed to T2D plasma versus healthy plasma (Figure 1D).

Endothelial cell stiffening impairs FMD in isolated mesenteric arteries

We show that pharmacological manipulation of the actin cytoskeleton alters endothelial cell stiffness and FMD. Specifically, jasplakinolide, which induces actin polymerization and stabilization, enhances F-actin and increases cell stiffness in cultured endothelial cells (Figure 1E,F). Furthermore, in isolated arteries, intraluminal exposure to jasplakinolide impairs FMD, without significantly affecting endothelial-independent vasodilation to SNP (Figure 1G).

Endothelial ROS and markers of oxidative stress are enhanced in T2D

We provide evidence that production of ROS, including H₂O₂, is elevated in mesenteric endothelial tubes from db/db compared to wild-type mice (Figure 2A). Moreover, aortic endothelial 3-nitrotyrosine, a marker of oxidative stress, is enhanced in aortas from db/db mice (Figure 2B). Of note, when we segment the aortic endothelium, 3-nitrotyrosine remains elevated in db/db compared to wild-type mice (Figure 2B). In further support of enhanced endothelial oxidative stress in T2D, endothelial cells exposed to plasma from humans with T2D demonstrate increased 3-nitrotyrosine compared to healthy plasma (Figure 2C).

Figure 2. T2D is associated with increased endothelial cell production of ROS and markers of oxidative stress, while endothelial cells exposed to H₂O₂ demonstrate enhanced F-actin, likely mediated by inhibition of cofilin depolymerizing activity.

A: Fluorescent intensity over time (red, reported every 5min) and normalized fluorescent rate (dF/dt; insets) of DHE and DCF in mesenteric endothelial tubes isolated from wild-type and db/db mice (11–13 weeks of age; n=7), representative images; scale bar = 50μm. Fluorescent representative images were acquired with a Nikon E800 microscope using a 40X/0.8 NA water objective coupled to a DS-Qi2 camera with Elements software (version 4.51). B: IHC staining of the aortic wall and segmented endothelium intensity of 3-nitrotyrosine from wild-type and db/db mice (11–13 weeks of age; n=20–21 and a subset of n=8–10), with representative images; scale bar = 50μm. C: Fluorescent intensity (red) of 3-nitrotyrosine in endothelial cells exposed to T2D or healthy plasma (n=20–23), with representative images; scale bar = 50μm. D: F-actin (yellow) content in endothelial cells exposed to H₂O₂ (50μM; 1.5h) (n=19–20), with representative images; scale bar = 30μm. E: Fluorescent intensity of isolated polymerized F-actin, over time (reported every 5min), following incubation with reduced (DTT; 1mM, 0.5h pre-treatment) and oxidized cofilin-1 (H₂O₂; 50μM, 0.5h pre-treatment) (n=9). #P≤0.05 vs. reduced cofilin F: Fluorescent intensity of F-actin (yellow) in endothelial cells transfected with either active wild-type pEGFP-N1 human cofilin WT (GFP, cyan) or constitutively inactive mutant pEGFP-N1 human cofilin S3E (GFP, cyan) representative images (n=10–11). Two-way ANOVA with repeated measures was used to assess the change in fluorescence of DHE and DCF in isolated mesenteric endothelial tubes (A) and polymerized F-actin in mesenteric arteries (E). Data are presented as mean±SEM. One-way ANOVA was used to assess differences in fluorescent intensity of polymerized F-actin (F). Two-tailed unpaired Student’s t-tests were used in all other analyses. *P≤0.05 vs. *P≤0.05 vs. respective controls and main effect of condition.

H₂O₂ induces F-actin and impairs cofilin depolymerizing activity

Cultured endothelial cells exposed to H₂O₂ (50μM, 1.5h), a model of oxidative stress, demonstrate increased F-actin (Figure 2D). Furthermore, H₂O₂ exposure in vitro causes oxidation of human recombinant cofilin-1, impairing its F-actin depolymerizing activity (Figure 2E). In support of these findings, we demonstrate that overexpression of a constitutively inactive mutant of cofilin enhances F-actin in cultured endothelial cells compared to overexpression of wild-type cofilin (Figure 2F).

LIMK inhibition or silencing of LIMK1 reduces cultured endothelial cell F-actin and aortic endothelial stiffness in db/db mice

We demonstrate that LIMK inhibition attenuates phosphorylation of cofilin in endothelial cells (Figure 3A), decreases F-actin content (Figure 3B), and shows a tendency to reduce cell stiffness (P=0.066; Figure 3C). Dissecting the role of LIMK isoforms, we show that siRNA-mediated silencing of LIMK1 alone reduces cofilin phosphorylation and F-actin (Figure 3D,E). Moreover, we demonstrate that LIMK inhibition also abrogates phosphorylation of cofilin and lowers F-actin in endothelial cells co-incubated with H₂O₂ (Figure 3F,G), and LIMK inhibition, ex vivo, ameliorates endothelial cell stiffness in aortic explants from db/db mice (Figure 3H).

Figure 3. Targeting LIMK1 prevents inactivation of cofilin (phosphorylation), reduces F-actin, and lowers endothelial stiffness in aortas isolated from db/db mice.

A: Expression of phosphorylated cofilin Ser³ relative to total cofilin, with representative blots (n=6), B: fluorescent intensity of F-actin (n=15), with representative images; scale bar=30 μm, and C: quantification of cortical cell stiffness in endothelial cells following treatment with an inhibitor of LIMK (LIMKi3; 1μM, 18h) or vehicle (DMSO) (n=6). D: Expression of LIMK1 relative to total protein (n=11–12), and phosphorylated cofilin Ser³ relative to total protein (n=12), with representative blots, and E: fluorescent intensity of F-actin, with representative images, in endothelial cells transfected with an siRNA against LIMK1 (LIMK1 siRNA) or a scramble control (Scramble siRNA) (n=10–13). F: Expression of phosphorylated cofilin Ser³ relative to total cofilin (n=11–12), with representative blots, and G: fluorescent intensity of F-actin (n=7–8), with representative images, in endothelial cells co-incubated with H₂O₂ (50μM,1.5h) and LIMKi3 (1μM, 1.5h) or vehicle (DMSO); scale bar = 50μm. H: Quantification of en-face endothelial cell stiffness in paired analysis from aortic rings isolated from db/db mice and treated ex vivo with LIMKi3 (1μM, 18h) or vehicle (DMSO) (n=10). Data are presented as mean±SEM. Two-tailed unpaired (A-G) and paired (H) Student’s t-tests were used in all other analyses. *P≤0.05 vs. respective controls.

DISCUSSION

Endothelial stiffening is becoming widely accepted as an early event and contributor to endothelial dysfunction and overt vascular disease. Indeed, endothelial stiffening has been shown to occur early in the pathogenesis of endothelial dysfunction and prior to stiffening of the subendothelial layer (3, 4). However, the molecular mechanisms underlying endothelial stiffening remain largely unknown. Here, we provide supporting evidence that oxidation of cofilin, which reduces its F-actin severing and depolymerizing activity, may be implicated in increasing endothelial F-actin and cell stiffness in T2D.

Specifically, we report that endothelial cells exposed to T2D plasma display oxidative stress and increased F-actin accompanied by elevated stiffness compared to cells exposed to plasma from healthy subjects. Identifying the molecules in the diabetic plasma responsible for the induction of this endothelial stiffening phenotype, while important, is beyond the scope of this short report. As expected, glucose levels were higher in plasma samples from T2D, compared to healthy controls, and these data were included in a previous publication (18). High glucose is a potential culprit based on previous work demonstrating that exposure of endothelial cells to high glucose elicits oxidative stress and increased stiffness (26, 27). However, other pro-oxidant factors, including lipids and inflammatory cytokines, are likely also implicated (28). Of note, we show that pharmacological induction and stabilization of F-actin, with jasplakinolide, increases endothelial cell stiffness and impairs FMD in isolated arteries. This finding, along with prior work indicating that enhanced endothelial F-actin interferes with eNOS activity (5, 29), supports the notion that F-actin-dependent cell stiffening may contribute to endothelial dysfunction.

Given the integral role of the actin cytoskeleton in governing cell stiffness, we interrogated the role of cofilin activity in regulating endothelial cell stiffening. We demonstrate that endothelial oxidative stress, a feature of T2D confirmed in our mouse model, impairs cofilin’s actin depolymerizing activity and enhances F-actin in endothelial cells. These findings agree with other studies in which H₂O₂ treatment inactivates cofilin-1 and causes an increase in F-actin and stiffness in non-endothelial cells (12). Strengthening the notion that inactivation of cofilin causes increased F-actin in endothelial cells, we show that overexpression of an inactive form of cofilin increases F-actin compared to overexpression of wild-type cofilin. While phosphorylation of cofilin is also a mechanism that leads to inactivation of cofilin, when we measured levels of phosphorylated cofilin in aortic sections from wild-type and db/db mice, and in endothelial cells exposed to plasma from T2D and healthy counterparts, we observed no differences between conditions (Figures S1-S3). These findings suggest that cofilin phosphorylation may not contribute to endothelial stiffening in these contexts, lending further support toward an alternative mechanism leading to cofilin inactivation (i.e., oxidation of cofilin).

In further support of the role of cofilin inactivation in mediating endothelial cell stiffening, our data show that inhibition of LIMK lowers phosphorylated cofilin (inactive form), reducing F-actin and stiffness. These data mirror our previous findings in vascular smooth muscle cells (13). While we acknowledge that LIMK inactivates and reduces cofilin activity in a different manner to oxidation of cofilin, it is still significant that LIMK inhibition, and thus reduced inactivation of cofilin, lowered endothelial stiffness in cells exposed to H₂O₂ and diabetic arteries. Future studies should further interrogate the interactions between these pathways. For example, we cannot rule out the possibility that oxidative stress also acts through LIMK to induce F-actin. Moreover, follow-up experiments are needed to strengthen the notion that oxidative stress leads to increased F-actin and endothelial stiffening. These include, but are not limited to, exposing cells to H₂O₂-scavenging compounds to increase antioxidant status, exposing cells to other ROS beyond H₂O₂ (e.g., potassium superoxide), or by inducing endogenous oxidative stress by genetic or pharmacological inhibition of antioxidant enzymes (e.g., superoxide dismutase). Evidence also points to the existence of two or more distinct pools of actin which differentially interact with signaling pathways regulating F-actin (25). One makes up the cortical membrane and is composed primarily of γ-actin which is important for endothelial cell motility, whereas β-actin is concentrated near the basal membrane forming stress fibers and connections to focal adhesions, important for cell adhesion and intercellular-contacts (30). This may explain the existence of several pathways capable of modulating cofilin activity.

Taken together, this work supports the idea that reduced cofilin activity, as a result of oxidation, may drive F-actin formation in the endothelium, leading to cell stiffening and contributing to endothelial dysfunction in T2D (Figure 4). However, more research is required to elucidate the role of different actin isoforms and F-actin regulation in mediating endothelial stiffening and dysfunction in T2D, as well as determine whether in vivo targeting of cofilin (e.g., LIMK inhibition) can serve as a strategy to dampen endothelial stiffening and improve cardiovascular outcomes in T2D.

Figure 4. Schematic representation of proposed mechanism.

Oxidative stress impairs the actin-severing and depolymerizing activity of cofilin in endothelial cells, leading to cell stiffening (top section). Alongside, targeting cofilin using a LIMK inhibitor (LIMKi3) lowers phosphorylated cofilin (inactive form), reducing F-actin and stiffness, thus representing a promising strategy to ameliorate endothelial dysfunction in T2D (bottom section). Created in BioRender. Power, G. (2024) BioRender.com/s64d918

Supplementary Material

All supplemental data can be accessed here: 405 10.6084/m9.figshare.27328101

DATA AVAILABILITY

The data underlying this article will be shared upon reasonable request to the corresponding authors.