Endothelial β1 Integrins are Necessary for Microvascular Function and Glucose Uptake

Abstract

Microvascular insulin delivery to myocytes is rate limiting for the onset of insulin-stimulated muscle glucose uptake. The structural integrity of capillaries of the microvasculature is regulated, in part, by a family of transmembrane adhesion receptors known as integrins, which are composed of an α and β subunit. The integrin β1 (itgβ1) subunit is highly expressed in endothelial cells (EC). EC itgβ1 is necessary for the formation of capillary networks during embryonic during development and its knockdown in adult mice blunts the reactive hyperemia that manifests during ischemia reperfusion. In this study we investigated the contribution of skeletal muscle EC itgβ1 in microcirculatory function and glucose uptake. We hypothesized that loss of EC itgβ1 would impair microvascular hemodynamics and glucose uptake during insulin stimulation, creating 'delivery'-mediated insulin resistance. An itgβ1 knockdown mouse model was developed to avoid lethality of embryonic gene knockout and the deteriorating health resulting from early post-natal inducible gene deletion. We found that mice with (itgβ1^fl/flSCLcre) and without (itgβ1^fl/fl) inducible stem cell leukemia cre recombinase (SLCcre) expression at 10 days post cre induction have comparable exercise tolerance and pulmonary and cardiac functions. We quantified microcirculatory hemodynamics using intravital microscopy and the ability of mice to respond to the high metabolic demands of insulin-stimulated muscle using a hyperinsulinemic-euglycemia clamp. We show that itgβ1^fl/flSCLcre mice compared to itgβ1^fl/fl littermates have, i) deficits in capillary flow rate, flow heterogeneity, and capillary density; ii) impaired insulin-stimulated glucose uptake despite sufficient transcapillary insulin efflux; and iii) reduced insulin-stimulated glucose uptake due to perfusion-limited glucose delivery. Thus, EC itgβ1 is necessary for microcirculatory function and to meet the metabolic challenge of insulin stimulation.

Figure 1 –. Deletion of EC itgβ1 in adult mice is lethal.

A. Schematic of the treatment regime used to downregulate Itgβ1 in endothelial cells (EC). itgβ1^fl/fl mice were crossed with mice expressing the stem cell leukemia (SCL) promoter driven tamoxifen-inducible CreER transgene. At 10 weeks old, itgβ1^fl/fl and itgβ1^fl/flSCLcre male mice were administered tamoxifen (TMX) for 5 consecutive days. B, C. Body weight and body composition as a percent of body weight was measured at 30 days post-TMX treatment. Data in panel C are presented as pie graphs with means and standard error for lean, fat, and water % included within each respective slice. D. Incremental exercise test procedure. E. Mice were subjected to an exercise stress test until they were no longer able to match the running speed of the treadmill for >5 seconds. This was defined as the time to exhaustion. F. Gross image of a single left lung lobe indicating hemorrhage in an itgβ1^fl/flSCLcre mouse. H&E micrographs of representative lung sections 30 days post-TMX showing widespread pulmonary hemorrhage in itgβ1^fl/flSCLcre mice. Cartoons in panels A and D were generated using Bioredner.com. Values are mean ± SE. n=3–8 mice/group. Panel C: Two tail student t test were used to determined statistical significance. ***p<0.001, ****p<0.0001.

Figure 2 –. Downregulation of EC-itgβ1 decreases capillary density in skeletal muscle.

A. Representative immunofluorescence images of gastrocnemius muscle at 10X, with 63X merged inset (far right panel) stained with ITGB1, CD31, and DAPI. B. Immunofluorescence microscopy shows that itgβ1 is decreased in the skeletal muscle of itgβ1^fl/flSCLcre mice 10 days after the last dose of TMX. C. CD31 is decreased in the skeletal muscle of itgβ1^fl/flSCLcre mice. D. Co-localization of CD31 and itgβ1 shows an overall decrease in the gastrocnemius muscle of itgβ1^fl/flSCLcre mice. E. Representative images of individual capillaries (Lectin-LEA) and anti-ITGB1 co-expression in itgβ1^fl/flSCLcre and itgβ1^fl/fl mice. Arrows indicate strong ITGB1 signal colocalized with Lectin positive capillaries. Panel E representative images were taken using the Zeiss LSM710 confocal microscope. Two tail student t test were run to compare groups. Three to four mice per genotype were used to generate sections and images. The average fluorescent intensity across multiple fields of view (FOV) were conducted to generate a single mean, resulting in a biological n=3–4/group. Data are mean ± SE. *p<0.05; **p<0.01.

Figure 3 –. EC itgβ1 downregulation does not affect pulmonary or cardiac function at 10 days post-TMX.

A. In vivo respiratory gas analysis was performed 10 days post TMX. No differences in pulmonary function were found between groups. B. Echocardiography was performed 10-days post TMX. A representative image showing the left ventricle (LV) and right ventricle (RV) of the myocardium. Real-time video images were analyzed to quantify C. quantify pulmonary artery acceleration, right ventricular stroke volume, right ventricular wall thickness, and right ventricular internal dimension. These parameters were similar between genotypes. Rrs, resistance of the total respiratory system; Ers, elastance of the total respiratory system; Crs, compliance of the total respiratory system; Rn, airway resistance; G, tissue dampening; H, tissue elastance; FEV0.5, forced expiratory volume in 0.05 sec; FEV0.1, forced expiratory volume in 0.1 sec; FEV0.2, forced expiratory volume in 0.2 sec. 10x image scale bar = 200 μm. 60x image scale bar = 50 μm. Two tail student t test was used to compare groups. Data are mean ± SE. n=4–9 mice/group.

Figure 4 –. Downregulation EC itgβ1 does not affect body weight or exercise tolerance at 10 days post-TMX.

At 10 weeks-of-age itgβ1^fl/flSCLcre and itgβ1^fl/fl mice were administered tamoxifen (TMX) for 5 days. Ten days after the last TMX treatment, A. body weight and B. body composition were measured C. Mice were subjected to an exercise stress test to evaluate whole-body physical function. The treadmill speed increased every 3 minutes until mice could no longer maintain running speed, indicating exercise exhaustion. D. Summary of health validation between 10 days and 30 days post-TMX. At 10 days post-TMX itgβ1^fl/flSCLcre mice have normal exercise tolerance, cardiac and pulmonary function. By 30 days post-TMX, these mice are severely exercise intolerant and moribund. Panel D was generated via Biorender.com Two tail student t test were run to compare groups. n=7–13 mice/group. Data are mean ± SE.

Figure 5 –. EC itgβ1 regulates microvascular hemodynamics without affecting capillary insulin efflux.

Intravital microscopy (IVM) was performed in anesthetized mice 10 days post-TMX. A. A decrease in flow velocity through skeletal muscle capillary beds in itgβ1^fl/flSCLcre mice was observed. B. The proportion of perfused vessels was reduced in itgβ1^fl/flSCLcre mice than itgβ1^fl/fl mice. C. Within one field of view (FOV), there was significantly more heterogeneity in flow velocity when comparing all vessels in the itgβ1^fl/flSCLcre mice than the itgβ1^fl/fl mice. D. Within one FOV, there was more heterogeneity in hematocrit distribution in the itgβ1^fl/flSCLcre than itgβ1^fl/fl mice. E. Insulin efflux was determined via intravital microscopy following an INS-647 bolus. T = 0.25 minutes indicates the beginning of imaging, which occurs ~15 seconds after injection of INS-647. Interstitial space is defined as the region emanating 1–3 μm from the capillary wall. Ratio of plasma to interstitial INS-647 as a function of time following INS-647 injection, normalized to the ratio at t = 0 minutes. Decay constant of the plasma/interstitial INS-647 ratio, a measure of transendothelial insulin transport kinetics. F. The fractional removal rate of capillary insulin. Two tail student t test was used to compare groups. n=6–9/group. Data are mean ± SE. *p<0.05; **p<0.01.

Figure 6 –. Downregulation EC itgβ1 impairs whole-body insulin action.

Hyperinsulinemic-euglycemic clamps were performed 10 days after the last TMX dose in conscious unstress mice. A. Arterial glucose was administered and monitored during the glucose infusion and insulin clamp. B. Glucose infusion rate was monitored every 10 minutes for each group. C. The rate of endogenous glucose appearance was measured for each group. D. Glucose flux was measured between groups over 100 minutes. E. Insulin was measured at baseline and during the clamp period. F. The glucose metabolic index (Rg’) was measured in gastrocnemius muscle, vastus lateralis muscle, epididymal (eWAT), and inguinal white adipose tissue (iWAT). G. Ratio of tissue water [³H]-glucose over plasma [³H]-glucose at the end of the clamp period. Panel A&B: Two-way ANOVA with repeated measures were run with Tukey adjustment. Panel C-E: Two-way ANOVA with group and condition as factors were run with Tukey adjustment. Panel F&G: Two tail student t test were run to compare groups. Data are mean ± SE n=9–12/group. *p<0.05; **p<0.01.

Figure 7 –. Myocellular insulin signaling and ex vivo glucose uptake is not dependent on EC itgβ1.

A. Lysates were prepared from gastrocnemius muscle collected from 10-day post-TMX mice and analyzed by Western blot for levels of activated and total AKT, GSK3β, and ERK under basal and insulin stimulated (clamp) conditions. B-D. Bands were quantified and values are expressed as activated/total proteins as fold changes relative to the basal itgβ1^fl/fl condition E. Ex vivo 2DG uptake in isolated soleus muscle in the absence or presence of insulin (100 nM). Two-way ANOVA with repeated measures were run with Tukey adjustment. Data are mean ± SE. Panels A-D n=4–5/condition. Panel E, bilateral soleus muscle was isolated from n=3–5 mice per genotype. Soleus from one leg was stimulated with insulin and the contralateral muscle with vehicle. *p<0.05; **p<0.01