Association of physical activity with increased PI3K and Akt mRNA levels in adipose tissues of obese and non-obese adults

Phosphatidylinositol-3-kinase (PI3K)/Akt signaling pathway regulates glucose and lipid metabolism. We examined the association of PI3K and Akt expression in visceral (VAT) and subcutaneous adipose tissue (SAT) with daily physical activity (PA) in non-diabetic obese and non-obese adults. In this cross-sectional study, we included 105 obese (BMI ≥ 30 kg/m2) and 71 non-obese (BMI < 30 kg/m2) subjects (aged/ ≥ 18 years). PA was measured using a valid and reliable International Physical Activity Questionnaire(IPAQ)-long-form, and the metabolic equivalent of task(MET) was calculated. Real-time PCR was performed to analyze the mRNA relative expression. VAT PI3K expression had a lower level in obese compared to non-obese (P = 0.015), while its expression was higher in active individuals than inactive ones (P = 0.029). SAT PI3K expression was increased in active individuals compared to inactive ones (P = 0.031). There was a rise in VAT Akt expression in the actives compared to the inactive participants (P = 0.037) and in non-obese/active compared to non-obese/inactive individuals (P = 0.026). Obese individuals had a decreased expression level of SAT Akt compared to non-obsesses (P = 0.005). VAT PI3K was directly and significantly associated with PA in obsesses (β = 1.457, P = 0.015). Positive association between PI3K and PA suggests beneficial effects of PA for obese individuals that can be partly described by PI3K/Akt pathway acceleration in adipose tissue.

ligand for PI3K; its binding activates PI3K/Akt signaling pathway functions, resulting in increased glucose transporters mainly at the skeletal muscle cells and adipocytes 12 . Reduced expression of the alpha subunit of PI3K in rat adipose tissue, the most basic catalytic subunit of class 1A of this molecule, is associated with increased adipocytes and glucose intolerance 15 . Akt, protein kinase B, which regulates glucose and lipid metabolism, is divided into three isoforms based on serine and threonine amino acid differences 16 . Akt2 is mainly expressed in insulin-sensitive tissues, such as skeletal muscle, adipose tissue, and liver 17 . Akt protein plays a role in inhibiting the ectopic expansion of adipose tissue by inducing adipogenesis through the upregulation of PPARγ. Akt is also interplaying with Mammalian Target of Rapamycin (mTOR) and Sterol Regulatory Element-Binding Proteins (SREBPs) pathways 18,19 .
Considering the role of the PI3K/Akt signaling pathway in underlying mechanisms involved in lipid metabolism regulation, a comprehensive understanding of this pathway, its downstream molecules, and environmental factors affecting the regulation of involved genes expression can be beneficial in finding new ways to prevent and treating obesity.
Previous studies have shown alteration in PI3K/Akt pathway gene expression and protein levels following exercise in different tissues related to metabolic disorders. Nadi et al. showed that exercise increases the PI3K and Akt mRNA levels in nerve-muscle of diabetic rats 20 . Huang et al. and Li et al. detected increased levels of PI3K, Akt, and proteins after exercise training in the myocardial tissue of Wistar rats (compared to the sedentary group) 21,22 . Wang et al. showed that exercise increased the relative expression of PI3K and Akt proteins and activated PI3K/Akt signaling pathway in (hippocampus of) rats' brain 23 . Despite the critical effect of PA on adipose tissue (metabolic adaptation and changes in size and fat content), few studies on the association between PA and PI3K and Akt gene expression in adipose tissue 15,24 . Therefore, in the present study, we investigated the relationship between PI3K and Akt expression with daily PA in adipose tissues of non-diabetic obese and non-obese adults. VAT and SAT were examined due to reported differences in their metabolic effect 6 and their response to PA 25 .

Materials and methods
Participants. In this cross-sectional study, 176 individuals (134 women and 42 men) aged 19-69 years were selected among patients who were admitted to the Mostafa Khomeini and Khatam Al-Anbia hospitals, (Tehran, Iran) for minimal abdominal surgery including appendectomy and hernia repair for both obese and non-obese patients between 2012 and 2015. People with diabetes or cancer and women who were pregnant or breastfeeding were excluded. Demographic information, history of medicine were collected. Before surgery, 5 ml of fasting blood samples were gathered for biochemistry measurements. Subcutaneous and visceral adipose tissues (SAT and VAT) were collected by a single expert surgeon during elective surgery. During the surgery, SAT was harvested from the 5 cm periumbilical area; VAT was collected from the upper left part of omentum (subomental) fat, directly frozen in liquid nitrogen, and stored at − 80 °C. This study was conducted following the Declaration of Helsinki and RIES institutional guidelines. Written informed consent was obtained from all the participants. Anthropometric and laboratory measurements. Weight, height, waist, hip, and neck circumference were measured according to standard protocols described in detail previously 26 . Body mass index (BMI) was calculated as weight (kg) divided by the square of height (m 2 ), and participants were divided according to their BMI status into two groups, including obese with BMI ≥ 30 kg/m 2 and non-obese with BMI < 30 kg/m 2 . After 15 min of rest, systolic and diastolic blood pressures (SBP and DBP) were measured twice in one-minute intervals with a mercury sphygmomanometer, and their mean was recorded 27 .
All serums were gathered after centrifuging the fasting blood samples at 3000×g for 5 mins. An enzymatic colorimetric method with glucose oxidase was used to measure fasting plasma glucose (FPG), and the inter-and intra-assay coefficients of variation (CV) were both 1.0%. Triglyceride (TG) levels were measured by enzymatic calorimetry using glycerol phosphate oxidase, and there were 0.4 and 2.1% CVs between inter-and intra-assays of TG. Total cholesterol (TC) was assayed with the cholesterol esterase and cholesterol oxidize method; interand intra-assay CV of TC were 0.5 and 1.7%, respectively. The measurements of FPG, TG, and TC levels were performed using commercial kits (Pars Azmoon Inc., Tehran, Iran). The enzyme-linked immunosorbent assay (ELISA) measured insulin levels (Mercodia kits, Uppsala, Sweden). The inter-and intra-assay CVs of insulin were 1.7 and 2.3%.
Assessment of physical activity. The validated International Physical Activity Questionnaire (IPAQ)long-form 28 was used to accumulate participants' PA during seven days whose reliability, validity, and reproducibility have been assessed in Iranian society 29 . The IPAQ data was converted into metabolic equivalents (METmin/1 week) for each type of activity by multiplying the number of minutes allocated to each activity by the associated MET score (One MET is equal to energy expenditure during rest and is approximately equal to 3.5 ml O2/kg/min in adults). Therefore, participants were categorized into active (≥ 600 MET-minutes/week) and inactive (< 600 MET-minutes/week) groups 30 . Evaluation of genes expression in adipose lesions. Tissue samples were weighted and incised before RNA extraction. Totally, 30-50 mg visceral and subcutaneous fat tissues were added to 1 mL TRIzol reagent (Invitrogen, U.S.). Total RNA was extracted based on the manufacturer's instructions. Proteins, lipids, carbohydrates, and cell debris were eliminated through the extraction of the aqueous. Then, the Nanodrop spectrophotometer (ND-1000, Thermo Scientific, USA) was used to measure the 260/280 and 260/230 nm absorption www.nature.com/scientificreports/ ratios to evaluate the quality of the extracted RNA. cDNA was synthesized from total RNA according to the instruction of the cDNA synthesis kit (BIOFACT, South Korea). GAPDH (housekeeping gene) was used as an internal control gene in a real-time quantitative PCR instrument (qPCR) Rotor-Gene 6000 (Corbett Life Sciences, Sydney, Australia). The primer sequences of PI3K, Akt and GAPDH were as follows: PI3K (NM_006218.4) Forward: 5′-CAG AAC AAT GCC TCC ACG A-3′, Reverse: 3′-CAC GGA GGC ATT CTA AAG TC-5′ with product size 122 bp; Akt (NM_001382431.1) Forward: 5′-CAC TTT CGG CAA GGT GAT CC-3′, Reverse: 3′-GTC CTT GGC CAC GAT GAC TT-5′ with product size 94 bp, and GAPDH (NM_001357943.2) Forward: 5′-CTG CTC CTC CTG TTC GAC AGT-3′, Reverse: 3′-CCG TTG ACT CCG ACC TTC AC-5′ with product size 100 pb. The annealing temperature for the PI3K and Akt was determined to be 60 and 61 °C, respectively.
The qPCR was performed in 20 μL volumes containing 1 μL of the cDNA, 1 μL forward primers, 1 μL reverse primers, 10 μL 2X SYBR Green Master mix (BIOFACT, South Korea), and 7 μL DEPC water. Then samples were run in duplicate for target genes and GAPDH with the thermal cycling conditions: 10 min at 95 °C, followed by 40 cycles 30 s at 95 °C, 30 s at 60 °C (for PI3K) or 30 s at 61 °C (for Akt) and 30 s at 72 °C. The reference gene normalized the mRNA levels of PI3K and Akt in each sample. The relative quantitation expression was performed by the comparative Ct method according to Livak et al. 31 according to the following formula: Statistical analysis. Statistical analyses were performed using SPSS (version 20) software. The Kolmogorov-Smirnov test was used to evaluate the normal distribution of variables. Normal variables including age, BMI, waist, wrist, hip, and neck circumference, fasting blood glucose, total cholesterol, and PA expressed as mean ± standard deviation (SD), and non-normal variables including SBP, DBP, and TG levels as median (interquartile 25, 75). Insulin was log-transformed to better approximate a normal distribution.
To compare quantitative variables between non-obese and obese groups, Independent-Sample T-test was used for variables with normal distribution, and the Mann-Whitney U test for variables with non-normal distribution. Analysis of variance (ANOVA) and tukey HSD post-hoc tests were carried out to compare the insulin levels between four studied groups. The Chi-Square test was also used to compare qualitative variables such as gender between non-obese and obese groups. Mann-Whitney U test was used to compare the expression of genes in two different tissues and between non-obese and obese groups. Comparative graphs of relative expression between study groups were drawn with GraphPad Prism8 software with Mean ± 1SEM.
Correlation between PI3K and Akt gene expression in VAT and SAT with PA was examined using the Pearson Correlation test. STATA software was used to confirm a linear relationship between gene expression and PA by the Restricted Cubic Splines model using the multivariate regression splines (mvrs) command.
The interaction of PA with BMI and sex on gene expression was tested through linear regression analysis. Due to the significant interaction between PA and BMI, the analysis was performed for two distinct groups (physically active and inactive). Furtheremore, there was no significant interaction between PA and sex in gene expression. The association of PA and gene expression was also examined using linear regression analysis with age and sex adjustments.

Results
Characteristics of the participants. This study evaluated 176 adults (aged 19-65 years). Seventy-one subjects (50 women, 21 men) who had a mean age of 48.3 ± 14.9 years were in the non-obese group, and 105 subjects (84 women, 21 men) who had a mean age of 37.0 ± 10.7 years were in the obese group. Information on anthropometric, biochemical, and PA variables in non-obese and obese individuals is shown in Table 1. Non-obese individuals had a higher mean age than the obese group. The mean BMI, waist, wrist, hip and neck circumference, FBS, total cholesterol, and median of the SBP, DBP, TG, and insulin levels were also higher in obese subjects than in non-obese participants. Although the mean MET of non-obese individuals was higher compared to obese ones, this was not statistically significant (Table 1). Results showed no differences in insulin levels between active and inactive individuals neither in obese nor non-obese groups based on the ANOVA Posthoc (Tukey HSD) test. The significant ANOVA P-value is due to the difference in insulin levels between obese and non-obese subjects ( Table 2).
PI3K gene expression analysis. Based on Mann-Whitney U test results, there was a significant decrease in the VAT PI3K gene expression in obese individuals compared to non-obese individuals (P = 0.015) (Fig. 1A). Moreover, it was upregulated significantly in active individuals compared to inactive ones (P = 0.029) and in non-obese/active subjects compared to non-obese/inactive individuals (P = 0.038) (Fig. 1B,C). However, VAT PI3K expression had no significant difference between obese/active and obese/inactive individuals (P = 0.594) (Fig. 1D). Moreover, The SAT PI3K mRNA level was not different between obese and non-obese individuals (P = 0.091) ( Fig. 2A). Nevertheless, it was increased significantly in active individuals compared to inactive one www.nature.com/scientificreports/ (P = 0.031) and in non-obese/active participants compared to non-obese/inactive group (P = 0.032) (Fig. 2B,C). On the other hand, no significant difference was observed between obese/active and obese/inactive groups (P = 0.115) (Fig. 2D).
Akt gene expression analysis. Gene expression levels were compared between groups using the Mann-Whitney U test. The results showed that VAT Akt mRNA level was marginally different between obese and non-obese groups (P = 0.060) (Fig. 3A). However, it was significantly raised in active group compared to inactive group (P = 0.037) and in non-obese/active compared to non-obese/inactive individuals (P = 0.026) (Fig. 3B,C). The expression of this gene was not significant between the obese/active group and obese/inactive subjects (P = 0.401) (Fig. 3D). Besides, SAT Akt expression was significantly decreased in obese individuals compared to non-obese one (P = 0.005) (Fig. 4A). It was elevated in active compared to inactive group (P = 0.086), in nonobese/active participants compared to non-obese/inactive one (P = 0.057) and in obese/active individuals compared to obese/inactive group (P = 0.192), which were not statistically significant (Fig. 4B-D).

Correlation and linear regression analysis.
The results of the Pearson Correlation test showed a positive correlation between PI3K and Akt expressions with PA in VAT of obese individuals (r = 0.254, P = 0.009 and r = 0.265, P = 0.006, respectively). However, no significant correlation was found between PI3K and Akt expressions with PA in VAT and SAT for non-obese group (Table 3). VAT PI3K expression is directly and significantly associated with PA in obese individuals (β = 1.457, P = 0.015), using linear regression analysis. However, other groups had no significant association between PA with VAT and SAT PI3K and Akt levels. Besides, there was no significant effect of sex and age (Table 4). Table 1. Anthropometric, biochemical, and PA characteristics in non-obese, obese, and total individuals. Qualitative data are reported as percent; Normal and non-normal variables were expressed as mean ± standard deviation (SD) and median (IQ 25-75), respectively. Independent-Sample T-test and Mann-Whitney U test were used to compare quantitative normal and non-normal variables, respectively. The Chi-Square test was used to compare qualitative variables between the two groups.  Table 2. Comparison of the insulin levels between four studied groups. Analysis of variance (ANOVA) and tukey HSD post-hoc tests were carried out to compare the insulin levels between four studied groups.

Discussion
The present study showed that VAT PI3K expression was different between obese and non-obese participants according to their activity and inactivity. This gene's relative expression was lower in the obese group than the non-obese group (regardless of activity status). However, it was higher in active subjects than inactive ones (regardless of BMI status). In non-obese individuals, daily activity (MET) was associated with a higher expression of PI3K. In obese individuals, PA showed a significant positive correlation with PI3K expression. Increased SAT PI3K expression level was also observed in the active group compared to the inactive group and in active obese subjects compared to the inactive obese ones. Regarding the Akt gene, the expression changes of this gene were observed in both VAT and SAT. Higher Akt mRNA level was shown in active individuals compared to inactive subjects (regardless of BMI status) and the non-obese/active subjects compared to non-obese/inactive counterparts in VAT. Moreover, in the obese group, PA had a significant positive correlation with Akt expression. SAT Akt expression was significantly decreased in obese people compared to non-obese people (regardless of activity status).
In 2005, Peres et al. studied the relationship between endurance exercise training and adipocytes' response to insulin in Wistar rats 24 . Male rats ran on a treadmill for one hour a day for seven weeks compared to age-matched sedentary controls. At the end of the study, the final weight of rats in the two groups was not significantly different. The epididymal adipocytes were smaller in the trained group, and plasma glucose and insulin levels were lower than in the control group. Insulin-induced glucose uptake, PI3K-dependent phosphorylation in insulin receptors-1 and 2, and Akt (Ser 473) were also higher in the training group than in the control group. For the first time, the researchers attributed the improved adipocyte insulin responsiveness following exercise to the modulatory effect of the PI3K/Akt signaling pathway.
Our results in this study show a significant decrease in PI3K expression in inactive individuals compared to active ones, regardless of their obesity status. During PA, glucose uptake increases in adipose tissue, and therefore, it is necessary to increase the glucose transporter 4 (GLUT4) at the adipocyte surface 32,33 . GLUT4 is the insulin-responding glucose transporter, a major mediator of glucose removal from the circulation, and a key regulator of whole-body glucose homeostasis. The mechanism of this increase in adipose tissue is unknown,   www.nature.com/scientificreports/ but the consequence of PI3K/Akt pathway activation via insulin is the rise of GLUT4 transportation to the cell surface. Therefore, it seems logical that PA is associated with a higher level of genes expressed in this pathway. However, Ferrannini et al. indicated that insulin-dependent glucose uptake in VAT following exercise was not different between insulin-resistant and non-insulin-resistant individuals. Therefore, according to this result, they attributed the increased glucose uptake in this tissue to the non-insulin-dependent mechanisms 34 . On the other hand, a study in 2019 showed that the treatment of cultured human adipocytes (isolated from subcutaneous abdominal adipose tissue) with obtained serum from subjects after 60 min of cycling exercise elevated the expression of GLUT4. They suggest circulating factors mediate exercise-induced effects on adipose tissue GLUT4 expression 35 . The origin and identity of this serum transmitted potential factors are still unknown. Since PI3K can be activated by non-insulin ligands, including hormones and growth factors 36 , a rise in their serum concentration occurs after exercise, PI3K may play a role in the initiation of the pathway and increase glucose uptake.
One of the important topics investigated in this study was the evaluation of PA effect on gene expression in both SAT and VAT. In addition to anatomical position and mechanical behavior; SAT and VAT have metabolic differences. The adipocytes in the VAT are metabolically more active, more sensitive to lipolysis and more insulin-resistant than SAT adipocytes. VAT is also associated with overstated inflammatory, which explains close association between VAT accumulation and cardiometabolic risk and morbidity 37,38 . Moreover, PI3K/Akt axis expression pattern was not studied in human VAT and SAT separately. We assumed that the response of PI3K/Akt genes expression in VAT might be different from their expression in SAT. Therefore, we collected both adipose tissues. In this study, based on the relative gene expression analysis, the results indicated that PI3K mRNA levels in VAT and the Akt mRNA levels in SAT of obese subjects significantly decreased. Meanwhile, PI3K expression increased in both SAT and VAT of subjects with higher PA. This may indicate that PA has the same effect on PI3K mRNA level in both fat tissues. On the other hand, the relationship between PA and Akt expression was observed only in VAT, which can indicate the greater effect of PA on the Akt expression inVAT. The present study www.nature.com/scientificreports/ revealed that the difference in the expression levels of the studied genes was more prominent in VAT than SAT. Likely, intrinsic depot-specific differences such as higher androgen receptor levels, lower leptin secretion, and mRNA expression, greater glucocorticoid density, and greater IL-6 secretion may play a role in how PA affects them 39 . Quantifying fatty acid and glucose uptake using positron emission tomography (PET) has shown that VAT is metabolically more active than SAT 40,41 . Currently, it is hypothesized that the dysfunction of VAT plays a significant role in obesity pathophysiology; due to its higher level of adrenergic activity, it responds more to PA changes [30][31][32] . The effect of PA on VAT reduction can be partly explained by the IL-6, which is secreted from the contracting muscle and mainly stimulates lipolysis in VAT 42,43 . IL-6 functions as an anti-inflammatory cytokine and is one of the first cytokines to be upregulated after exercise 44,45 . It has been shown that IL-6 receptor blockade impairs this exercise-induced lipolysis 46 .
In 2020, Meng et al. showed that extracted flavonoids from mulberry leaf extract improved hyperglycemia, glycolipid metabolism, adiponectin, and leptin secretion in the insulin-resistant model of 3T3-L1 adipocytes 47 . They revealed this metabolic improvement was associated with increased expression of PI3K, Akt, and IRS1 proteins and thus activation of the PIK3/Akt pathway and high membrane transfer capacity of GLUT4. Wartmannin, a PI3K inhibitor, significantly reduced glucose uptake and secretion of leptin and adiponectin. Therefore, increased expression of PIK3/Akt genes was associated with better glycolipid metabolism, glucose uptake, and adiponectin secretion.
In concordance with these findings, we recognized lower Akt expression in SAT of obese participants than non-obese and VAT of inactive individuals (obese and non-obese) compared to active ones. There was also a positive correlation between the expression of the Akt gene and PA in obese participants. Findings may verify the PIK3/Akt pathway mediatory role in responding to PA.
Adiponectin is one of the essential hormones in adipose tissue, and the reduction of its expression is directly related to the incidence of obesity 48 . This adipokine is recognized as a potential biomarker for the beneficial effects of PA 49    www.nature.com/scientificreports/ positive correlation between PI3K expression and PA in our study suggests that perhaps part of the health benefits of PA for obese individuals is due to an increase in adiponectin secretion via PI3K/Akt pathway acceleration. Based on previous studies and as we expected, increased adipose tissue suppresses the PI3K/Akt signaling pathway. The genes of this pathway are less expressed in obese people than in non-obese individuals. Although mRNA levels alone are not always sufficient to predict protein levels, as a critical indicator of the regulation of gene expression, it shows the development of an adaptive response. In a recent study on children with different degrees of obesity, the levels of PI3K mRNA and Akt mRNA in peripheral blood mononuclear cells were lower in each obesity group than in the control group 51 . In another study on the blood samples of diabetic patients, PI3K and Akt2 mRNA levels were significantly lower in the obese diabetic group compared to other groups 52 .  The numbers below the figures indicated the mean ± SEM and also median (Interquartile ranges 25-75%). The displayed P values obtained from the Mann-Whitney U test, which was used to compare the differences between two groups. www.nature.com/scientificreports/ Insulin is the main activator of PI3K-Akt signaling pathway. Binding to its transmembrane receptor, insulin promotes the signaling cascade, leading to GLUT4 translocation to the cellular membrane and glucose uptake 53 . This process explains lower PI3K and Akt expression in the obese group with higher insulin and FBS levels in serum.
On the other hand, considering the importance of this pathway in increasing glucose uptake and PA's role in energy consumption, it seems that PA has a role in improving obesity status by affecting the genes of the PI3K/ Akt pathway. The association between PA and VAT PI3K gene expression in the obese group has been shown better after adjustment for age and sex. This observation revealed the importance of PA.
According to the literature, this study is the first study investigating the relationship between PA and PI3K/ Akt gene expression in VAT and SAT of human samples. The limitations of this study include: due to limited funding, it was not possible to measure these genes' protein production. It was not possible to examine this pathway's upstream and downstream genes. Collecting more tissue specimens for matching was impossible due to the study on human samples.
The results of this cross-sectional study can be used to design experimental studies; in the clinical trial, studies to investigate the effect of PA on gene expression of this pathway in fatty tissues.

Data availability
The datasets used and analyzed in the present study are available from the corresponding author on reasonable request.  Table 4. The association between PI3K and Akt gene expression and physical activity in VAT and SAT of nonobese and obese groups. The P values obtained from linear regression analysis after sex and age adjustment. PA physical activity, VAT visceral adipose tissue, SAT subcutaneous adipose tissue, CI confidence interval, *P value < 0.001.