Glucose transporter 4 (GLUT4) distribution on metabolic healthy obese (MHO) vs metabolic unhealthy obese (MUO)

Background. Adipose tissue is a key factor during metabolic dysregulation, whereas adipose tissue dysfunction is promoted if enlarged adipocytes, impaired adipocyte differentiation, in�ammation, remodeling and/or �brosis. As evidenced by in vitro studies and animal models, adipocyte hypertrophy alone is su�cient to impair glucose uptake, probably explained by changes of cell distribution of Glucose Transporter 4 (GLUT4) and tra�cking, re�ecting a new potential regulatory pathway. Patients with obesity can be subclassi�ed into cardiometabolic risk phenotypes, with signi�cant difference in adipose tissue characteristics and glucose uptake. Therefore, GLUT4 distribution in adipocytes from the so known: 1) metabolic healthy obese (MHO) and 2) metabolic unhealthy obese (MUO), may help to understand cardiometabolic risk differences between them. Objectives. To analyze GLUT4 distribution in adipocytes from MHO vs MUO phenotypes. Subjects and Methods. Cross-sectional observational study. Patients programmed for bariatric surgery were included. Patients were classi�ed as MHO or MUO according to NCEP-ATPIII criteria. Samples from subcutaneous (SAT) and visceral adipose tissue (VAT) were obtained during bariatric surgery. GLUT4 expression and distribution in adipocytes were evaluated by immuno�uorescence and image analysis. Results. GLUT4 molecules showed a lower number and more spared distribution in adipocytes from MUO phenotype. Likewise, GLUT4 distribution was related to insulin resistance in both metabolic phenotypes, with differences between VAT and SAT. Conclusion. GLUT4 expression pattern in adipocytes was related with the type of adipose tissue, the cardiometabolic risk and insulin resistance, suggesting potential and speci�c regulatory role.


Introduction
Obesity is a multifactorial disease involving social, economic, genetic, microbiome and individual behavior differences; along with higher risk to develop diabetes mellitus, cardiometabolic risk, obstructive sleep apnea-hypopnea syndrome, non-alcoholic fatty liver disease and certain types of cancer.It is the fth leading cause of death, accounting for 3.4 million deaths each year [1][2][3].The population with obesity can be divided according to the cumulative number of dysmetabolic components, being "metabolically healthy obese (MHO)", those characterized by a mild increase in cardiometabolic risk, but not signi cantly different from normal-weight population; and "metabolically unhealthy obese (MUO)", characterized by a signi cantly increased cardiometabolic risk [3][4][5].
In the other hand, insulin resistance is associated with clusters of metabolic and cardiovascular disorders.Quali-and/or quantitative lipids disorders promote dysregulation pathways in adipose tissue, liver, heart, muscle and pancreas; leading to an impaired glucose metabolism.It is well recognized that adipose tissue is a highly active metabolic and endocrine organ that plays a key role in insulin resistance and metabolic syndrome.Glucose Transporter 4 (GLUT4) is an insulin-dependent transporter located in adipocytes; whereas abnormalities in GLUT4 tra cking and intracellular translocation may promote insulin resistance [5][6][7][8][9].Several murine models have described that adipocyte´s morphology is able to predict a low degree of chronic in ammation and insulin resistance; suggesting that adipocyte size plays a key role during insulin resistance development [10][11][12] and also hypothesize subsequent effects on GLUT4 tra cking and distribution, as well as actin structures impairment [13][14][15][16].Our study aimed exploring whether GLUT4 number and/or distribution in different sized adipocytes may impact cardiometabolic phenotype and insulin resistance.

Results
Ten patients with morbid obesity, mean BMI 48 kg/m 2, mean aged 46 years old and no particular sex prevalence constituted the study population.
Then, patients were divided according to their metabolic risk phenotype (MHO n = 5 and MUO n = 5).MUO group showed higher diastolic blood pressure (Table 1) and a biochemical pro le of increased plasma fasting glucose, HOMA-IR and decreased HDLc (Table 2).The analysis of the pattern of distribution of GLUT-4 expressed in adipocyte membranes, evidenced a lower amount of GLUT-4 molecules and more spared distributed in MUO phenotype compared with MHO phenotype (Fig. 1).
Moreover, the amount of GLUT-4 and the distance between them (inter-GLUT distance) were negatively and positively related with HOMA-IR, respectively; whereas no signi cant difference was observed between metabolic risk phenotypes (Fig. 2).

Discussion
The main nding of the present study was that the amount and distribution of GLUT4 were related with insulin resistance, and further differentiation between cardiometabolic risk pro les.
Despite the small sample size of the study population, main demographic characteristics were homogenously distributed, whereas cardiometabolic risk were different between MHO and MUO, there for they are comparable to other studies.
The relation between GLUT4 distribution in adipocyte`s membrane with insulin resistance is consistent with initial observations in murine model, where enlarged lipid droplets and adipocyte hypertrophy were able to impair insulin sensitivity rather than in ammation alone [12]; likewise, adipocyte's size has been positively related with cardiometabolic risk factors and endothelial dysfunction [16][17][18].Possible explaining mechanisms include: 1) adipocyte size affects cortical actin structures and GLUT4 tra cking vesicles, therefore in uencing processes regulating insulin sensitivity; 2) higher sized adipocytes may promote to keep adipocytes away from vasculature structures, then increasing hypoxia mediated changes, and secretion of pro-in ammatory adipokines [19,20].
In addition, GLUT4 distribution was further affected by the adipose tissue source.GLUT4 was less dense in adipocytes from VAT, which was associated with insulin resistance.However, this effect was observed in the MHO phenotype, but not in the MUO phenotype, suggesting that this lack of effect may be related to the dysfunction of the adipose tissue observed in MUO phenotype [4,21].
Our ndings propose new pathophysiological mechanism linking adipocyte size with insulin resistance, as well as potential diagnostic and therapeutic targets according to cardio-metabolic risk pro les.Some limitations of the present study were related to the small sample size and consequent potential for bias; as well as the interpretation limitation from a cross-sectional study design.
In conclusion, GLUT4 molecules were lower and more spared in adipocytes from MUO phenotype, which was related with insulin resistance markers; with an additional potential role of the source of the adipose tissue.These ndings suggest a key role of GLUT4 in the pathophysiology of adipocyte-derived cardiometabolic risk.

Design and study population
Cross-sectional study evaluating the role of adipocyte's expression of GLUT-4 in insulin resistance and cardiometabolic risk factors.We enrolled patients older than 18 years old, candidate for bariatric surgery (Body Mass Index [BMI] higher than 40 kg/m 2 or BMI higher than 35 kg/m 2 experiencing obesity-related health conditions, such as t2DM, hypertension or obstructive sleep apnea/hypopnea), attended between January 2016 and December 2017 at National Medical Center "20 de Noviembre", ISSSTE at Mexico City.Patients did not received weight-reducing therapy during 6 months previous to the enrollment and they were excluded in case of second bariatric surgery, the existence of in ammatory diseases, severe renal and/or hepatic disease, active malignancy, pregnancy or evidence of history of cardiovascular disease, considered if self-reported or diagnostic evidence of ischemic heart disease, coronary artery disease, myocardial structural abnormalities, cardiac interventions or being under treatment for any of such conditions.

Clinical-Demographic Data
Most clinical-demographic data were obtained from a direct interview performed by experienced physicians.BMI was calculated as weight/height 2 [23].Waist circumference was measured halfway between the lowest rib margin and the iliac crest at the end of a normal expiration.Blood pressure was obtained while the patient was in a seated position and was considered as the mean of three readings obtained 5 min apart using an aneroid sphygmomanometer (Welch Allyn Inc.; Skaneateles Falls, NY, USA).

Adipose Analyses
During the surgery, samples of 2cm 3 of abdominal Subcutaneous Adipose Tissue (SAT) and Visceral Adipose Tissue (VAT) from the omentum, were obtained, and preserved.
For GLUT4 analysis, immuno uorescence assay was performed.Tissues were stained with primary antibody (GLUT4 Monoclonal Antibody (1F8

Statistical analysis
Clinical and biochemical quantitative data were described as mean and standard deviation, whereas qualitative data were resumed as n (%).For inferential analysis, two-tailed, U-Mann Whitney or T-test were applied according to normality of data distribution.Finally, Pearson correlation coe cient was used to evaluate the relationship between GLUT4 distribution pattern and insulin resistance markers.All statistics were performed using GraphPad Prism software (v.7).Statistical signi cance was considered as p < 0.05.InterGLUT4 spare distance according to adipose tissue source and cardiometabolic phenotype.
according to ethical guidelines of the 1975 Declaration of Helsinki and approved by the Local Committees of Research, Ethics in Research and Biosafety of the National Medical Center '20 de Noviembre' ISSSTE, Mexico City.All participants provided written informed consent.

Figure 2 Distribution
Figure 2
GLUT4 spare distance and GLUT4 density of uorescence) were performed with Image pro-plus software.
). Thermo Fisher Scienti c, cat No. MA1-83191) conjugated with antibody linked to a uorophore (Goat anti-Mouse IgG (H + L) Secondary Antibody, FITC.Thermo Fisher Scienti c Cat. No. 62-6511); and further added with mounting medium containing DAPI for nuclear counterstaining.Images were captured with uorescence confocal microscope, and image analyses (inter-