Expression of adiponectin and leptin receptors in adrenal incidentaloma patients with subclinical hormone secretion

Abstract

BACKGROUND:

The role of adopokines in adrenal tumors’ hormonal activity remains unclear. Obesity may induce arterial hypertension, disorders of carbohydrate metabolism, and is a risk factor of cardiovascular disease. In patients with subclinical hormone secretion by the adrenal cortex or medulla the risk of metabolic disease is increased.

OBJECTIVE:

Authors of this retrospective study selected 78 patients with subclinical hormone secretion out of all adrenal incidentaloma patients hospitalized in the Department of Endocrinology and Internal Medicine between 1995 and 2014.

METHODS:

The analyzed group comprised of 38 subclinical Cushing’s syndrome (SCS), 40 incidentally discovered pheochromocytoma (PHEO) and 42 patients operated due to an adrenal tumor without pathological hormonal activity. Expression of adiponectin (AdipoR1, AdipoR2) and leptin (Ob-R) receptors in adrenal tumors was assessed in relation to body mass index (BMI) and hormonal activity.

RESULTS:

We found statistically significant negative correlations between BMI and expression of all examined receptors in SCS patients (AdipoR1: $p=$ 0.032; AdipoR2: 0.001; leptin Ob-R: $p=$ 0.001). In PHEOs, BMI correlated negatively only with AdipoR2 ($p=$ 0.014).

CONCLUSIONS:

Data obtained show that the most significant factor associated with the expression of AdipoR1, AdipoR2 and leptin Ob-R receptors in the adrenal tumor tissue is BMI, not their hormonal activity.

1. Introduction

Adipose tissue is considered the largest ‘endocrine gland’ of the human body. It synthesizes and secrets over 50 distinct hormones and cytokines, which play an important role in the metabolism, insulin sensitivity and cardiovascular disease [5, 21].

The adrenal glands, which are surrounded by adipose tissue, produce both gluco- and mineralocorticoids as well as catecholamines that also influence the cardiovascular system and glucose metabolism [2, 19]. In patients with subclinical hormonal secretion of the adrenal cortex or medulla risk of metabolic diseases is increased [2, 7, 8, 18, 22].

So far, few reports have been published on adipone-ctin receptor expression in the adrenal cortex [4] and medulla [12] tumor tissues. Adiponectin receptors demonstrated in the adrenal cortex tissue may help elucidate the role of a functional connection between BMI, arterial hypertension, and adrenal function. It has been suggested adipose tissue may regulate the function of the human adrenal cortex [19].

The authors of the present study attempted to answer the question whether the overproduction of glucocorticoid hormones and catecholamines by adrenal tumors affects the expression of adiponectin Adipo R1, Adipo R2 and leptin Ob-R receptors in resected tumors.

2. Materials and methods

The SCS group comprised of 38 patients; the PHEO group – 40 patients and the non-functioning adrenocortical tumors – control group (CONT) – 42 patients. Patients were aged 23 to 92 (mean age 53.3 years). In SCS group patients were aged 27 to 77; in PHEO group – 24 to 74 and CONT group patients were aged 25 to 75 (Table 1).

Table 1.

Age, sex and BMI of subclinical Cushing syndrome – SCS, incidentally discovered pheochromocytoma – PHEO patients and non – functioning adrenocortical tumors – CONT group

Variable	SCS	Incidentally discovered	Non – functioning adrenocortical	p
	$n=$ 38	PHEO $n=$ 40	tumors CONT $n=$ 42
Females/males	25/13	23/17	25/17	$p=$ 0.481
Mean age (years) females/males	53.3	48.6	58.5	$p=$ 0.083
	54.1/51.7	49.2/45.4	58.5/58.3
Mean BMI/$\pm$ SD kg/m${}^2$	27.1/2.3	25.8/1.7	26.8/2.1	$p=$ 0.308
BMI 25	5 (13%)	13 (32.5%)	2 (4.7%)	–
25–29.9	28 (74%)	26 (65%)	34 (81%)	–
$⩾$ 30	5 (13%)	1 (2.5%)	6 (14.3%)	–

Normal weight: BMI (body mass index) 18.5–24.9 kg/m${}^2$ , overweight: BMI 25.0–29.9 kg/m${}^2$; obesity BMI $⩾$ 30 kg/m${}^2$.

Both groups had comparable BMIs: mean BMI in SCS patients was 27.1 and in PHEO patients 25.8 (Table 1). Obtained results were compared with those adrenal tumor patients who did not present subclinical hormonal activity of the adrenal cortex or medulla. Mean BMI of patients with tumors without pathological hormonal activity of the adrenal glands were 26.8 (Table 1).

All patients were admitted to our Department. SCS was diagnosed in accordance with the definition in patients with an incidental adrenal mass found in an abdominal CT or MRI scan, who did not have typical physical features of Cushing Syndrome (CS) and exhibited endogenous hypercortisolemia in hormonal tests. Cortisol morning serum concentration of 140 nmol/L (5.04 $\mu$g/dl) and above in the overnight suppression test with 1 mg of dexamethasone was a sufficient criterion to diagnose SCS. Patients with serum cortisol values between 50 and 140 nmol/L (1.8–5.04 $\mu$g/dl) in this test must have met at least one other criterion: lack of circadian cortisol secretion rhythm (evening cortisol concentration accounts for more than 50% of morning cortisol), decreased morning ACTH concentration ($⩽$ 10 pg/ml), and/or increased free urinary cortisol excretion.

Obtained data were compared with those of 42 adrenal tumor patients who were not diagnosed with SCS. Hormonal characteristics of study samples are shown in Table 2.

Table 2.

Hormonal data of subclinical Cushing syndrome – SCS patients and non – functioning adrenocortical tumors (CONT)

Laboratory examination	SCS Mean $\pm$ SD	CONT Mean $\pm$ SD	Normal ranges
Serum cortisol: morning; nmol/l	408.9/181.9	396.9/212.7	101–536
Serum cortisol: evening; nmol/l	310/355.4	145.2/224.3	–
Serum cortisol in an overnight suppression test – 1 mg DXM;nmol/l	239.4/333.9	23/29.9	50
24-hour urinary free cortisol excretion; nmol/24 h	452.7/451.4	231.4/140.2	12–486
Plasma ACTH: morning; pg/ml	19.2/23.2	29.0/25.0	15–46

All hormone concentrations were determined in the same laboratory using freely available kits. The lab kits were not changed during the evaluated period. ACTH concentration was determined with a solid-phase, two-site sequential chemiluminescent immunometric assay Immulite 1000 ${}^{\mathrm{R}}$ACTH manufactured by Siemens. Serum and urinary cortisol concentrations were determined with a Chemiluminescent Microparticle Immunoassay Cortisol Reagent Kit on Abbott’s Architect analyzer. Blood was drawn in the morning, after fasting, from a cubital vein.

To evaluate metabolic disorders, body mass index (BMI) was calculated in all patients based on weight and height measurements. In the SCS subgroup 33 patients were either overweight or obese (Table 1).

The diagnosis of incidentally discovered PHEO was based on elevated 24-h urinary metanephrine and normetanephrines excretion, positive imaging study and lack of clinical symptoms indicative of a PHEO tumor (labile arterial hypertension, tachycardia, diaphoresis). Imaging phenotype of PHEO means: increased attenuation on nonenhanced CT $>$ 20 Housfield units; increased mass vascularity; delay in contrast medium washout (10 minutes after administration of contrast), an absolute contrast medium washout of less than 50 percent; high signal intensity on T2-weighted MRI and cystic or hemorrhagic changes.

The 24-h urinary metanephrine and normetanephr-ines excretion determined by means of high-perfor-mance liquid chromatography – HPLC on Bio-Rad. All medications that may interfere with urinary metan-ephrine and normetanephrine results were excluded. Hormonal characteristics of PHEO patients are shown in Table 3. Twenty seven patients with PHEO were either overweight or obese (Table 1).

Table 3.

Urinary metanephrine and normetanephrine excretion in incidentally discovered PHEO patients and non-functioning adrenocortical tumors – CONT

Laboratory examination; unit	PHEO IQR/Median	CONT IQR/Median	Normal range
24-hour urinary metanephrine; ug/24 h	2360,0/1074/	352.0/157.0	64–302
24-hour urinary normetanephrine; ug/24 h	373.0/707.0	256.5/398.0	162–527

IQR – interquartile range.

Figure 2.

Expression of AdipoR2 receptors in adrenal cortical tumor (A), PHEO (B), magnification 20x.

Due to subclinical hormonal activity all patients with SCS and incidentally discovered PHEOs underwent adrenalectomy. To eliminate possible BMI changes while waiting for surgery, BMI was evaluated 2–3 days before adrenalectomy. In the SCS group 23 patients were diagnosed with an adrenal cortex adenoma and 15 with nodular hyperplasia. In the incidentally found PHEO group, PHEO was in result confirmed histopathologically in all cases.

In the group of patients without pathological hormonal activity (CONT) indication for surgery was the size of tumor over 4 cm or suspicion of malignancy in MRI or CT. The lack of hormonal activity means the exclusion of SCS and overproduction of metanephrine or normetanephrine, but also the absence of testosterone, androstenedione, DHEAS or hydroxyprogesterone secretion. The adrenal adenoma was found in 16 cases and nodular hyperplasia in 26 in that control group.

Expression of adiponectin and leptin receptors was determined in the excised tumor tissues of the adrenal cortex or medulla.

Tumor tissue was preserved according to standard procedures (fixed for approximately 48 hours in 10% formaldehyde solution, later dehydrated using appropriate alcohol and xylene solutions, and, finally, embedded in low melting point paraffin). Tissue specimens of analyzed cases were re-assessed in routinely prepared hematoxin-eozin (H&E) stains to verify the histopathological diagnosis. From the studied specimens areas most representative for neoplasm were chosen; these areas contained neither necrotic lesions nor thermal damage artefacts.

2.1. Preparation of specimens

Immunochistochemical stains were made on microtome sections of 4 micrometer thickness. They were placed on Superfrost PLUS slides (by Surgipath) and later incubated overnight at 37${}^{\circ }$C. Antibodies used for detection along with applied methodology are presented in Table 4. To increase the reliability of the test results, two independent pathologists evaluated the results of immunohistochemistry in adrenal tumor tissues (Figs 1–3).

Table 4.

Antibodies used for detection along with applied methodology

Target	Catalog number	Dilution	Epitope retrieval	Incubation time	Control tissue	Method of evaluation
Adipo R1	ab126611	1:50	PTlink DAKO	60’	Endometrium	Semiquantitative
Adipo R2	LS-B9345	1:50	PTlink DAKO	Overnight	Small intestine	Semiquantitative
Leptin Ob-R receptor	ab2139	1:10	PTlink DAKO	60’	Placenta	Semiquantitative

Figure 1.

Expression of AdipoR1 receptors in adrenal cortical tumor (A), PHEO (B), magnification 20x.

Figure 3.

Expression of leptin Ob-R receptors in adrenal cortical tumor (A), PHEO (B), magnification 20x.

The expression of adiponectin and leptin receptors were supplemented with serum adiponectin and leptin concentration in 12 cases. Plasma adiponectin assay was performed using a Human HMW Quantikine Elisa Kit (Biokom). Serum leptin concentration were determined by commercial sandwich enzyme immunoassay EIA kit (DRG MedTek).

2.2. Statistical analysis

Standard descriptive statistics were calculated. Correlations were assessed using Spearman’s method. Multivariate analysis was performed using linear regression with robust standard errors. Some independent variables were log-transformed. p values lower than 0.05 were considered statistically significant. All calculations were performed using a PC computer and STATA v13.1 statistical software package (StataCorp LP, Texas, USA).

3. Results

3.1. Adiponectin and leptin receptors expression in adrenal cortex tumor tissue and subclinical hormonal activity

In univariate analysis no significant differences in AdipoR1 receptor expression were found between patients with and without SCS ($p=$ 0.901). AdipoR2 expression was significantly lower in SCS patients (interquartile range-IQR 50, median 100) compared to patients without SCS (IQR 100, median 150), $p=$ 0.007. In SCS patients compared to those without SCS lower leptin Ob-R receptor expression was found (IQR 100, median 100, versus IQR 190, median 100), $p=$ 0.054. However, in multivariate analysis, taking into account patients’ BMIs, these differences were not statistically significant ($p=$ 0.247 for AdipoR2 and $p=$ 0.835 for leptin Ob-R expression).

We analyzed adipokine receptors expression and cortisol levels after 1 mg dexamethasone suppression test in SCS patients.

In univariate analysis cortisol correlated with AdipoR1 ($R=$$-$0.19; $p=$ 0.049) and AdipoR2 ($R=$$-$0.33; $p=$ 0.001) receptors but not leptin Ob-R expression. In multivariate analysis, however, these relationships were not confirmed and became not significant after inclusion of the variable BMI.

3.2. Adiponectin and leptin receptors expression in incidentally discovered PHEO tissue.

In univariate analysis correlations were tested between expression of adipokine receptors and 24-h urinary metanephrine and normetanephrine excretion. A statistically significant, positive correlation was found only between AdipoR1 expression and metanephrine excretion ($R=$ 0.37, $p=$ 0.044). For normetanephrine, this correlation was not statistically significant ($R=$ 0.29; $p=$ 0.638). Also the correlation between AdipoR2 expression and metanephrine/normetanephrine excretion was not statistically significant (respectively for metanephrine $R=$ 0.29, $p=$ 0.119 and normetan-ephrine $R=$$-$0.22, $p=$ 0.718).

A positive correlation between leptin Ob-R receptor expression and 24-h urinary metanephrine excretion was found ($R=$ 0.36) with a p value of 0.053. This correlation was also not confirmed for normetanephrine ($R=$$-$0.36; $p=$ 0.559).

In multivariate analysis, similarly to adrenal cortex tumors, the relationships were not statistically significant after taking into account patients’ BMIs ($p=$ 0.0558 for AdipoR1 and $p=$ 0.725 for leptin Ob-R receptor).

3.3. AdipoR1, AdipoR2 and Leptin Ob-R receptors expression and BMI

Due to the confounding effect described above associations between BMI and adipokine receptor expressions were tested. Results were presented in Table 5.

Table 5.

Correlation coefficients between BMI, adiponectin and leptin receptors expression

	Adrenal cortical tumors	PHEO
	($n=$ 78)	($n=$ 40)
AdipoR1	$-$0.19${}^{*}$ ($p=$ 0.032)	$-$0.18 ($p=$ 0.285)
AdipoR2	$-$0.34${}^{**}$ ( 0.001)	$-$0.39${}^{*}$ ($p=$ 0.014)
Leptin Ob-R	$-$0.29${}^{**}$ ($p=$ 0.001)	$-$0.10 ($p=$ 0.555)

Spearman rank correlation coefficients. Statistically significant correlations were indicated: ${}^{*}$ 0.010 0.050; 0.001.

In adrenal cortex tumors statistically significant, negative correlations between BMI and expression of all examined receptors were found (AdipoR1: $p=$ 0.032; AdipoR2: 0.001; leptin Ob-R: $p=$ 0.001).

In PHEO, BMI correlated negatively with AdipoR2 ($p=$ 0.014); correlations between BMI and AdipoR1 ($p=$ 0.285), and leptin Ob-R ($p=$ 0.555) receptor expression were not statistically significant.

3.4. Serum adiponectin and leptin secretion and AdipoR1, AdipoR2 and leptin Ob-R receptors correlation

Only 12 cases have been evaluated for serum adip-onectin and leptin concentration. The obtained results were correlated with AdipoR1, AdipoR2 and leptin Ob-R receptors expression in human adrenal tumors. The authors demonstrated statistically significant positive correlations between serum adiponectin secretion and adrenal AdipoR1 expression only ($R=$ 0.7631, $p=$ 0.0039). Other results were not statistically significant.

Due to retrospective nature of the study, a small group of patients is significant limitation of obtained results.

4. Discussion

Interrelationship between human adipose tissue and the adrenal glands has lately been a subject of interest. Adipose tissue is the source of several adipokines and cytokines, which influence metabolism and neoformation [1, 5, 11, 14].

Obesity as a social problem may induce arterial hypertension, carbohydrate metabolism disorders and constitutes a risk factor for cardiovascular disease [5, 6, 20, 21].

Adiponectin is the best known adipokine which acts through two different receptors – AdipoR1 and AdipoR2. AdipoR1 is abundantly expressed in skeletal muscles and AdipoR2 predominates in the liver. Adiponectin stimulates phosphorylation and activation of AMP kinase, thus suppressing gluconeogenesis in liver and promoting fatty-acid oxidation and glucose uptake in muscle [9, 10]. Obesity decreases expression levels of AdipoR1 and AdipoR2, thereby reducing adiponectin sensitivity [9, 10, 12].

Concentration of circulating adiponectin is proportional to the amount of adipose tissue and is lower in patients with diabetes mellitus and/or cardiovascular diseases [5, 6, 8, 12, 11, 13, 16, 23]. Obesity is connected with reduced adiponectin levels even in healthy individuals [8, 12].

In patients with subclinical hormonal activity of the adrenal cortex or medulla the risk of metabolic disease is higher compared to patients with adrenal tumors without pathological hormonal activity [2, 18, 22, 24, 25].

Glucocorticoids participate in control of caloric intake and adipogenesis. Glucocorticoids stimulate food intake, synthesis and activity of adipose tissue lipoprotein lipase, adipocyte differentiation and distribution. Some of those actions are in synergy with insulin [24]. Glucocorticoids also participate in epinephrine synthesis in the adrenal medulla, both epinephrine and insulin influence energy metabolism [18].

Recent reports prove presence of adiponectin receptors in normal adrenal as well as adrenal tumor tissue [19]. In some studies, the authors have shown that adiponectin receptor expression is significantly higher in hormonally active adrenal tumors compared to normal adrenal tissue suggests that adipose tissue may regulate human adrenal cortex [8, 19]. Recent reports suggest also an active role of the fat depot surrounding the adrenal tumors with local secretion of adiponectin and leptin [15].

In the context of above-mentioned literature data, the current study analysed associations between adipo-nectin and leptin receptors expression in adrenal tumors and subclinical hypercortisolemia (SCS). Taking into account confounding, we showed that BMI increase was associated with decreased expression of AdipoR1, AdipoR2 and leptin Ob-R receptors in adrenal cortex tumor tissue. Subclinical glucocorticoid overproduction did not influence this expression. Higher risk of metabolic diseases in subclinical hypercorisolemic patients is most probably the result of higher incidence of overweight and obesity, and, in consequence, lower serum adiponectin concentrations present in these patients.

PHEO patients are characterized by decreased glucose tolerance and diabetes mellitus is present in one third of them [7]. The relationship between diabetes and PHEO involves suppression of insulin secretion by catecholamines and induction of insulin resistance [7]. Thus far, available data indicate that AdipoR1 and AdipoR2 receptors are also present in normal adrenal medulla tissue [12]. Isobe et al. reported that in examined patients adiponectin AdipoR1 receptor expression correlated with catecholamine oversynthesis by adrenal tumor tissue and an increase in the risk of diabetes [12]. In our research we showed that BMI was the only factor that negatively correlated with the expression of the AdipoR2 receptors in PHEOs.

In normal adrenal tissue the presence of leptin Ob-R receptor expression has also been demonstrated; however, its role is unclear [17]. Leptin inhibits adrenal steroid synthesis by influencing the cortex, and, enhances catecholamine release by the medulla. These observations allow putting forward a hypothesis that leptin has an activatory effect on the sympathetic-adrenal system [17]. Bottner and co-workers assessed serum leptin concentrations in PHEO patients. Despite catecholamine oversynthesis in PHEO adiponectin and leptin receptors’ expression was suppressed in this group of patients with increasing BMI [3].

Authors of the current study, after taking into account the confounding effect of BMI, found neither an association between catecholamine overproduction by adrenal medulla tumors and adiponectin nor leptin receptors’ expression. BMI was the only factor that negatively correlated with the expression of the AdipoR2 receptors in adrenal medulla tumors. This indicates that adiponectin and leptin receptors in PHEO are not under the control of inhibitory influence of the adrenergic system in patients with chronic hypercatecholaminemia. The decisive influence on their expression is mediated through BMI.

In the presented retrospective study, only 12 cases of serum adiponectin and leptin were evaluated. The authors demonstrated that higher serum adiponectin correlated positively with its AdipoR2 receptor in adrenal tumor tissue. Due to the small and heterogenous group of patients, the results could not be correlated with their hormonal activity or BMI. However, our research may suggest that adiponectin receptors mediate the effects of adiponectin in adrenal tumors.

5. Conclusions

Overweight and obesity are key in regulating hormonal activity of the adipose tissue. Our data suggest that adiponectin and leptin receptors expression in human adrenal tumors is mainly associated with BMI. Associations between serum concentrations of various adipokines, their receptors’ expression in the adrenal glands, and the influence of adipose tissue, including that one surrounding the adrenal glands, require further prospective research.

Conflict of interest

The authors declare that they have no conflicting interests.