Chronic low-dose glucocorticoid treatment increases subcutaneous abdominal fat, but not visceral fat, of male Wistar rats

ABSTRACT

Aim: Most studies developed to investigate the effects of glucocorticoids chronic treatment on white adipose tissue uses high doses of these hormones. This study analyzes some effects of a chronic, continuous and steady infusion of low-dose hydrocortisone and the relationship with lipid accumulation in white adipose depots in rats.Main methods: Nineteen male Wistar rats were divided into control (CON) and cortisol (CORT) groups. Along six weeks CORT group received continuous infusion of 0.6 mg/kg/day of hydrocortisone, while CON group received saline. After euthanasia, subcutaneous and visceral (retroperitoneal and mesenteric) fat pads were excised, weighted and analyzed for: lipogenic enzymes activity; molecular changes of 11- hydroxysteroid dehydrogenase type 1 (11βHSD1) enzyme; enzymes involved in lipid uptake, incorporation, and metabolism and in fatty acids esterification. Besides, morphometric cell analysis was performed.

INTRODUCTION
Glucocorticoids (GC) are steroid hormones synthesized in the cortex of the adrenal glands and its basal secretion is rhythmic and occurs in a circadian fashion under the influence of a neuroendocrine control system – the hypothalamic-pituitary- adrenal axis (HPA). GC can also be locally activated in various tissues through 11β- hydroxysteroid dehydrogenase-1 (11βHSD1) enzyme action, which converts circulating inactive versions of these hormones in their active forms (cortisol in humans and corticosterone in rats) (Seckl & Walker 2001). The increased content or activity of this enzyme increases intracellular GC concentration, amplifying their actions. At the physiological level, GCs are considered catabolic hormones for their action in obtaining substrates (glucose, glycerol, free fatty acids and amino acids) for glucose maintenance and energy mobilization in stressful conditions, in order to supply the increasing energy demand in these situations (Peckett et al. 2011). Although GCs perform several vital functions to the organism (Macfarlane et al. 2009) their primary function is to provide glucose production for maintenance of its blood levels.White adipose tissue (WAT) is also a target organ of GC action, where they exert pleiotropic effects (Lee et al. 2011, 2012; Patsouris et al. 2009, Peckett et al. 2011), among which lipid mobilization is included. Although the effect of these hormones in inducing lipolysis is well documented (Baxter & Forsham 1972, Slavin et al. 1994; Villena et al. 2004), the result of its action is complex involving both fat degradation or storage and appears to depend on physiological context. GCs influence on fat accumulation was clearly demonstrated in previous studies. In humans, cortisol excess at a systemic level results in two to five-fold increase in central fat, particularly visceral fat, in patients with Cushing’s syndrome (Geer et al. 2014, Rockall et al. 2003, Wajchenberg et al. 1995). Rosmond et al. (1998) found that individuals under chronic stress (with increased circulating cortisol) are more likely to increase visceral fat. According to the authors, one possible explanation is the fact that cortisol and insulin synergistically increase lipoprotein lipase (LPL) activity.However, most studies developed with GC chronic use, intending to investigate its effects on WAT, uses high doses of these hormones. As described by Lee et al. (2013) in an extensive review, works seeking to investigate the effects of chronic body exposure to lower concentrations are lacking. In this context, the aim of this study was to analyze some effects of a chronic, continuous and steady infusion of a low-dose hydrocortisone and the relationship with lipid accumulation in rats. For this purpose, male adult Wistar rats were subjected to infusion of 0.6 mg/kg/day of hydrocortisone over six weeks. The dose was established based on previous definition of high (Bell et al. 2000 [~ 120 mg/kg/day], Campbell et al. 2011 [~ 33mg/kg/day]) and low (Bouillon & Berdanier 1980 [0.75 mg/kg/day]) glucocorticoid dose and considering the pump capacity. We investigated possible changes in the activity of lipogenic enzymes; molecular changes of enzymes involved in lipid metabolism, besides the expression of 11βHSD1 enzyme.

Different from the most studies found in the literature, our results showed an increase in central subcutaneous fat depot, while visceral depots remained unchanged.The experiments were carried out on 19 adults, male Wistar rats, weighing 300-350 g, housed individually under standard environmental conditions (lights on from 7 PM to 7 AM [inverted cycle], temperature 21 ± 1 ºC, water and food ad libitum – Nuvital® balanced standard chow pellets, Nuvital SA, Colombo, Brazil). Allexperimental procedures were reviewed and approved by the Ethical Committee for Animal Research (CEEA) (nº 010/10) of the Institute of Biomedical Sciences of University of São Paulo. After 2 weeks of acclimatization they were randomly divided into 2 groups: control (CON, n=11) and cortisol (CORT, n=8). Then, the animals received intraperitoneal anesthesia with ketamine and xylazine (Anasedan®, Seropédica, RJ, Brazil – 0,15 mL/100 g of body mass), prior to a surgical subcutaneous implant (dorsal interescapular region) of an osmotic minipump (Alzet®, Cupertino, CA, United States – model 2006 with 200 µl capacity) containing a 0,9% saline solution for CON group, while CORT group received hydrocortisone (hydrocortisone 21- hemisuccinate sodium salt – Sigma Aldrich Co®, São Paulo, SP, Brazil) diluted in distilled water at the dose of 0,6 mg/kg/day (or 0,18 mg/day). The treatment lasted for 6 weeks during which food and water intakes and body mass were measured weekly.Ex vivo experiments. At the end of the 6th week, after 12 h of fasting, CON and CORT animals were decapitated under sodium thiopental anesthesia (Tiopentax®, São Paulo, SP, Brazil) (3%, 5 mg/100 g bw, ip) at 7 AM (early environmental dark period, ZT=0). This time was defined based on the circadian rhythm of corticosterone release in the animals, when the serum concentrations are supposed to be highest in controls.

Trunk blood was collected for glucose, insulin, corticosterone and lipids determinations. Median laparotomy was performed to excise the following adipose tissues: subcutaneous (SC) (inguinal), retroperitoneal (RP) and mesenteric (ME). Adrenal glands were also excised, cleaned from the surrounding fat and weighted. Subcutaneous, RP and ME fat pads were weighed, fragments were collected for adipocyte isolation and samples were stored at – 80 °C.Adipocytes isolation and morphometric analysis. Adipocytes from each fat pad were isolated by collagenase tissue digestion as described by Rodbell (1964). Cell suspension aliquots were photographed and evaluated under an optical microscope (100x magnification) coupled to the digital microscope camera 1.3 MP (Moticam 1000®; Motic, Richmond, British Columbia, Canada). Mean adipocyte diameters were determined by measuring 100 cells using MOTIC-IMAGES Plus 2.0® software.Indirect calculation of cellularity. The number of cells in each fat pad was indirectly calculated dividing the total fat pad mass by the mean adipocytes mass. Adipocyte mass was calculated using the following formula: d = m/v, where d is the mean density of an adipocyte (0,91 g/mL), m is adipocyte mass (in pg [10-9 mg]) and v is the mean adipocyte volume (in pL [10-9 mL]).Hormones, glucose and lipids determinations. Serum glucose, triglycerides (TAG), total cholesterol and its fractions was determined as previously described (Chimin et al. 2014). Plasma corticosterone was determined by ELISA method using specific commercial kit for rats (IBL International, reference RE52061) following the manufacturer instructions (the kit percentage of cross reactivity to hydrocortisone is 0,3%).

Serum insulin was also measured by ELISA method, using commercial kit (EMD Milipore Corporation, Cat. #EZRMI-13K). Insulin resistance was calculated using the homoeostasis model of assessment insulin resistance (HOMAir) index, as defined by the equation HOMAir = (fasting glucose [mmol/L] × fasting insulin [µU/L])/22.5.Oral glucose tolerance test. One week before euthanasia the animals underwent an oral glucose tolerance test (oGTT) at 7 AM (early dark environmental phase in inverted cycle) after 12 hours of fasting. Tail blood was collected at time 0, then a glucose load was offered by gavage (75 mg/100 g of body mass) and new blood samples were collected at 5, 10, 20, 30, 60 and 90 minutes afterwards. Glucose concentration was determined using a glucometer (One Touch Ultra®, Johnson & Johnson®, São Paulo, SP, Brazil).Maximal enzyme activity of the de novo fatty acid synthesis in adipose tissue samples. The activities of malic enzyme, glucose-6-phosphate dehydrogenase (G6PDH), fatty acid synthase (FAS) and ATP citrate lyase (ACL) that belong to the de novo lipogenesis (DNL) pathway were analyzed following the methods described in Chimin et al. (2014). Enzyme activities were expressed as millimoles per minute per total fat pad mass (mmol/min/total fat pad mass).Tissue samples were homogenized in buffer at 4 oC (Triton-X 100 1%, 100 mM Tris (pH 7,45), 100 mM sodium pyrophosphate, 100 mM sodium fluoride, 10 mM EDTA, 10 mM sodium orthovanadate, 2 mM PMSF and aprotinin 0.01 mg/ml) in Polytron (PT 3100, Kinematica AG, Littau-Lucerne, Switzerland). Tissues extracts were centrifuged at 12,000 rpm at 4 °C for 20 minutes to remove insoluble material. Sample’s protein content (from the supernatant portion) was quantitated using Bradford reagent (Bio-Rad) and treated with Laemmli buffer containing 100 mM DTT. Aliquots (50 µg) of total protein were subjected to polyacrylamide gel electrophoresis (SDS- PAGE 10%, Invitrogen) and transferred to a nitrocellulose membrane. The antibodies used for obtaining the blottings are detailed in Table 1.Statistical analysis. Results were expressed as mean ± SEM, using the StatGraphics Centurion XVI® program. Mann-Whitney test was used for comparisons between groups for enzyme activity analysis, adrenal mass and blood lipids, corticosterone, insulin, glucose and HOMAir. For body mass, food and water intake Student’s t test was applied. The oGTT test was analyzed using ANOVA two-way (group vs. time) as a factor for repeated measures, followed by Bonferroni post-test. The fiducial level of significance was set at 95% (p <0.05).

RESULTS
Model characterization. The effects of GC chronic infusion at low dose were different from those commonly found in long-term use of high doses of these hormones and which characterize Cushing’s syndrome. As shown in Table 2, after six weeks infusing0.6 mg/kg/day of hydrocortisone, no difference was seen between control and treated groups for food intake, total body mass gain and fasting blood glucose and insulin.Also, the hypothalamic-pituitary-adrenal axis (HPA) was not inhibited by the treatment since the serum corticosterone content and adrenal mass in CORT group remained similar to control group.Values represent groups means ± SEM (CON n = 11; CORT n= 8). (*) = statistically significant difference (P<0,05). Analysis: Mann-Whitney for food/water intake, body weight and body mass gain. Student’s t test for the others.Fat pad mass. Once finished the experimental period, the mass of SC, RP and ME fat pads from both groups were verified (g tissue). Although there was no differencebetween groups for total body mass gain and food intake, SC fat pad from CORT group increased 26,8% compared to control. No difference was seen in other fat pads (Figure 1).Cellularity. Glucocorticoids are necessary for adipogenesis process, comprising the formation of new fat cells. The estimated number of cells in each fat pad was indirectly determined by the cellularity method, considering the total mass of the respective fat pad. There was no difference between CON and CORT to this parameter in any of the analyzed fat pads (Figure 2A), suggesting that the observed increase in SC may not have occurred by stimuli to formation of new adipocytes.

Cell volume. In order to verify if the pharmacological intervention would cause hypertrophy of adipocytes, the mean cell volume of each tissue was analyzed measuring100 cells per fat pad. Although the statistical analysis did not reveal differences between groups, SC adipocytes from CORT increased 27.6% comparing to CON, following the increase of 26.8% in the mass of the same tissue. No difference was seen in the remaining tissues (Figure 2B).Serum lipids. Plasma triglycerides, total cholesterol and its fractions were verified at the end of the experimental period. There was an increase in triglycerides, VLDL and HDL and a decrease in LDL in CORT animals. No changes were observed for total cholesterol (Table 2).Oral glucose tolerance test. To verify whether CORT animals had impaired glucose tolerance, one week before sacrifice they were submitted to oral glucose tolerance test(oGTT). In Figure 3 is possible to see that the glycemic response following the oral glucose load does not differ between the groups. The curves were overlayed.Maximum activity of enzymes involved in lipogenesis. The maximum activity of enzymes involved in de novo lipogenesis (DNL) (ACL, FAS, G6PDH and malic enzyme) was determined in adipocytes from SC, RP and ME tissues in order to verify whether the increase in SC mass from CORT animals or the lack of change in control group could result from modifications in these enzymes kinetics. As shown in Figure 4, the activities of FAS, ACL and G6PDH were increased only in the SC fat pad from CORT group. No changes were observed either in visceral tissues (RP and ME) from the same group, or in any fat pad from control group. These results suggest higher capability to generate fatty acids in SC depot from animals treated with hydrocortisone, compared to control. Data are presented as mmol/min/total fat pad mass.Protein expression. The analysis of the content of proteins involved in lipogenesis (LPL, CD36 and GPAT), AMPKα and 11βHSD1 was restricted to SC fat pad, since this was the only depot to show a significant increase in mass. The results show decreased expression of AMPKα enzyme, suggesting a possible facilitation of the lipogenic path. There was also a significant decrease in 11βHSD1 content. There was no difference between groups for the enzymes involved in fatty acid uptake (LPL and CD36), as for GPAT enzyme, which is involved in fatty acid esterification (Figure 5).

DISCUSSION
The effects of chronic GC treatment reported in literature frequently result from use of high doses of these hormones, whose symptoms feature Cushing’s syndrome. Differently, in this study we verified some effects of chronic low-dose administration of hydrocortisone on white adipose tissue (WAT). As a consequence, we observed some metabolic changes, although a different picture from that found when excessive doses are used came out. The alterations commonly seen in Cushing’s syndrome, such as changes in food intake, body mass evolution, high fasting glucose and insulin, insulin resistance and inhibition of the HPA axis did not occur in CORT group reported here. However, we observed that even a low-dose GC affects serum concentrations of triacylglycerol (TAG) and lipoproteins, and increases subcutaneous fat in abdominal region.Dyslipidemia has been identified as one of the consequences of chronic use of excessive doses of GC, although their causative role has not been fully proven and the results found in the literature are controversial (Auvinen et al. 2013). In our model, the increased serum TAG and VLDL in CORT group corroborates previous studies in which GC increase their synthesis and secretion by the liver (Arnaldi et al. 2003; 2010, Drolet et al. 2004). CORT animals showed also increased HDL and decreased LDL lipoproteins. However, total cholesterol remained unchanged, as observed in Auvinen et al. (2013) and Kumari et al. (2003) works, performed in mice. Although in humans chronic GC excess can cause reduction in serum HDL, its raise in our model was also reported by Wang et al. (2012), in healthy individuals under daily low doses of dexamethasone for three weeks. The same was also seen in the study of Choi & Seeger (2005).

Chronic GC excess increases fat deposition in central depots in detriment of periphery (Dallman et al. 2003, Macfarlane et al. 2009). In Cushing’s syndrome the centripetal redistribution of WAT is a common signal, where peripheral subcutaneous fat is wasted and visceral fat is increased (Wajchenberg et al. 1995). However, abdominal SC fat from region shows a characteristic that differs from peripheral SC fat of arms and legs, and is also increased in Cushing’s syndrome as well as visceral fat (Rebuffé-Scrive et al. 1988). By measuring the mass of central fat depots, our model revealed an unusual characteristic of GC chronic exposure: the abdominal SC fat pad of CORT animals was larger than those of CON group, while the masses of visceral fat pads remained unchanged, contrasting to what is seen in GC excess. Subcutaneous fat has smaller adipocytes comparing to visceral depots. Small adipocytes are more insulin- sensitive and have high avidity for lipids uptake preventing their deposition in non- adipose tissues, and the latter feature is even more pronounced in abdominal SC adipocytes compared to those of gluteofemoral region (Ibrahim 2010). Thus, subcutaneous fat accumulation represents a normal physiological buffer, storing excess circulating lipids to prevent metabolic disorder. Considering all these statements, a possible explanation for the increase in SC depot in detriment of visceral fat pads in our model could be as follows: the raise of circulating TAG in CORT group, added to the high capacity of abdominal SC to store lipids, may have contributed to the enlargement seen in this tissue. In the same group, the protein content of LPL and CD36 (important for lipid uptake) was unchanged. However, we can speculate that perhaps the content of these proteins in that tissue was sufficient to handle and take up the elevated levels of circulating lipids, since it was shown (Rebuffé-Scrive et al. 1988) that GC increases the activity of LPL. To confirm this hypothesis, however, it would be necessary to evaluate the intrinsic activity of free fatty acids transport kinetics of these proteins.

Adipose tissue grows when new adipocytes are formed by increasing the number of cells (hyperplasia), or when the pre-existing cells has its volume increased(hypertrophy) (Drolet et al. 2008). The analysis of cell content (figure 2A) suggests that there was no difference between CON and CORT for any fat pad, indicating that the increase observed in SC tissue of CORT animals may have not occurred by hyperplasia. On the other hand, adipocyte volume of SC fat pad increased 27.6 % in comparison to control. In the same direction, the difference of SC fat pad mass of CORT animals in relation to CON ones represents 26.8 %. Therefore, we can infer that the increase in SC may have occurred by adipocytes hypertrophy, as its mean cellular volume showed a growth percentage similar to that of its mass. Also, it has been shown that CG interferes in the activity of enzymes involved in lipogenesis (Diamant & Shafrir 1975, Volpe & Marasa 1975, Wang et al. 2004). Our data showed an increment in the activity of the lipogenic enzymes FAS, ACL and G6PDH of CORT group, corroborating previous work of our group (Chimin et al. 2014). However, in our study this alteration was seen only in SC tissue. Simultaneously, there was a significant decrease of AMPKα enzyme expression in the same tissue from CORT animals, resembling studies which revealed that GC reduces AMPKα activity in human and rodent WAT (Kola et al. 2008; Christ- Crain et al. 2008). Inactivation of AMPKα in WAT releases de novo lipogenic pathway, increasing lipids storage in adipocytes. Together, the impairment of AMPKα content and the reinforcement of the FAS, ACL and G6PDH enzymes activity observed in our study indicate that the de novo lipogenesis pathway may have been more vigorous in CORT animals, suggesting a further explanation for the increase in SC mass. On the other hand, our study showed that there were no changes in the activity of these same enzymes in RP and ME depots in both groups, as well as in the expression of AMPKα in visceral tissues.

These data infer that the treatment did not influenced lipogenic pathway in these tissues and supports the absence of change observed in their masses, reinforcing the hypothesis that the pharmacological intervention applied in this study disturbs only the function of central subcutaneous adipose tissue, as discussed above.Long term GC treatment is classically referred to as causing glucose intolerance and systemic insulin resistance, being this feature time and dose-dependent (Angelini et al. 2010, Rafacho et al. 2008, Severino et al. 2002). Our fasting blood glucose and insulin, and oGTT assay data show very similar results in both groups, suggesting that the treatment was not sufficient to induce glucose intolerance. Studies reveal differences between visceral and subcutaneous fat tissues for glucose uptake under GC treatment (Geer et al. 2014, Lundgren et al. 2004), being it more intense in the SC fat depot. An in vivo study in humans given hydrocortisone overnight resulted in systemic insulin resistance but increased its action in subcutaneous adipose tissue (Hazlehurst et al. 2013). Thereby, abdominal SC could be highly absorbing the available glucose, and we can assume that this tissue is contributing to avoid hyperglycemia and systemic glucose intolerance in our model.Studies conducted with chronic GC administration have been developed long ago seeking to unveil the causes of their effects on WAT. However, this is the first study to our knowledge that aimed to verify the changes caused by its chronic use at low dose, contrasting to classic hypercortisolism. We demonstrate here some effects of this drug intervention, showing that such a scenario can also be potentially harmful to the body, especially considering the increase in TAG. Further studies should be developed from our model to check in deeper details the mechanisms responsible for the changes observed.

CONCLUSION
The effects of long term use of low-dose GC differ from those commonly observed in chronic excess of these hormones. In low dose, there are no changes in body mass gain, food intake, fasting glucose and insulin or inhibition of the HPA axis. However, there is an increase in triglycerides, a change in lipoprotein profile and an increase in subcutaneous fat from abdominal region of rats, while the mass of visceral fat does not change. The increase in subcutaneous tissue appears to involve an increment in lipid storage and in DNL enzymes MEDICA16 activity.