Testosterone replacement therapy improves erythrocyte membrane lipid composition in hypogonadal men

Abstract

Aim: The aim of this study was to investigate the effects of testosterone replacement therapy (TRT) on erythrocyte membrane (EM) lipid composition and physico-chemical properties in hypogonadal men. Methods: EM isolated from three patients before and after TRT with injectable testosterone undecanoate or testosterone gel were used for analysis of the phospholipid and fatty acid composition, cholesterol/phospholipid ratio, membrane fluidity, ceramide level and enzyme acivities responsible for sphingomyelin metabolism. Results: TRT induced increase of phosphatidylethanolamine (PE) in the EMs and sphingomyelin. Reduction of the relative content of the saturated palmitic and stearic fatty acids and a slight increase of different unsaturated fatty acids was observed in phosphatidylcholine (PC). TRT also induced decrease of the cholesterol/total phospholipids ratio and fluidization of the EM. Discussion: The TRT induced increase of PE content and the reduction of saturation in the PC acyl chains induced alterations in the structure of EM could result in higher flexibility of the erythrocytes. The increase of the SM-metabolizing enzyme neutral sphingomyelinase, which regulates the content of ceramide in membranes has a possible impact on the SM signaling pathway. Conclusion: We presume that the observed effect of TRT on the composition and fluidity of the EM contributes for improvement of blood rheology and may diminish the thrombosis risk. Larger studies are needed to confirm the findings of this pilot study.

Keywords::

• Ceramide
• erythrocyte membrane
• hypogonadal men
• membrane fluidity
• phosphatidylcholine
• phosphatidylethanolamine
• phospholipid and fatty acid composition
• sphingomyelin
• Testosterone replacement therapy

Introduction

After the peak during the third decade, serum total testosterone level in men decreases gradually by about 1% (0.5–2%) per year. Free testosterone decreases steeper because of the increase in sex hormone-binding globulin (SHBG) [1–3]. This androgen deficient state has been defined as late-onset hypogonadism in males (LOH [4,5]). The diagnosis of LOH is made by the concomitant presence of symptoms and objectified hypotestosteroneaemia. Currently, there is a consensus that determination of serum total testosterone (TT) should be the preferred method for laboratory diagnosis of LOH and levels below 8 nmol/l (231 ng/dl) usually require TRT [6].

The unfavorable impact of LOH on health has been recognized in large-scale observational and interventional studies, and TRT on metabolism and cardiovascular system have been reported [7–17]. One of the few known adverse effects of the TRT is the increase of the hematocrite, which is expected to elevate the thrombotic risk.

Several studies support the involvement of EM properties in development of cardiovascular pathologies and hypertension [18,19]. Although the link between hypogonadism, metabolic syndrome and cardiovascular diseases has been proved in many observational and interventional studies, there is still insufficient information about the intimate sub-cellular mechanisms of action of testosterone on the composition and structural organization of EM lipids.

The aim of this pilot study was to investigate the effects of TRT on EM lipid composition and physico-chemical properties in hypogonadal men.

Patients and methods

This study consists of three Caucasian male patients, who represent typical cases, but differ in their type of hypogonadism and/or treatment.

Patient #1 (AS) is a 47-year-old male, married, has one child, with diabetes type 2 diagnosed ten months ago. He has longstanding obesity with body mass index (BMI) = 35.2 kg/m² at admission. From the diagnosis of diabetes, he has been treated with a stable dose of 1000 mg metformin TID with good glucemic control – HbA1c is 6.8% when hospitalized. No history for arterial hypertension, but BP in the clinic was 140/90 mmHg. He takes 160 mg fenofibrate since the diagnosis of diabetes. No clinical data for microvascular diabetic complications. Several dietetic trials to lose weight have been unsuccessful in long-term aspect. Total testosterone (TT) on two consecutive tests was determined as 6.61 and 7.29 nmol/l (normal 8.4–28.7). Other causes for hypogonadism were excluded and the diagnosis of LOH in a man with metabolic syndrome has been set. Treatment with testoster-one undecanoate (Nebido, Bayer Shering Pharma AG, Berlin, Germany) has been proposed, agreed by the patient and started in the standard dose of 1000 mg im, followed by a second injection after 6 weeks. The third injection was made after a 3 months interval, following the second one. The investigations were performed before and 3 months after the first injection.

Patient #2 (DA) is a 45-year-old male, married, has one child, with diabetes type 2, diagnosed 4 months ago, when metformin has been started in escalating dose to 3 g daily. At admission his BMI was 31.1 kg/m², HbA1c – 7.8%, BP – 120/80 mmHg without medication, no lipid-lowering drugs were used, no clinical data for microvascular complications were present. Unsuccessful weight loss was found in the medical history of the patient. Due to detected low testosterone levels (7.0 and 7.8 nmol/l) and complaints of erectile dysfunction and astheno-adinamia, after exclusion of other causes for hypogonadism, he was advised to initiate substitution with testosterone gel (Androgel, Laboratoires Besins International, Paris, France) in a standard dose of 50 mg daily (5g gel). Before and after three months of treatment, the investigations were performed.

Patient #3 (KG) is 32-year-old, with normal pubertal development. In 2006, he visited our center complaining on low libido and erectile dysfunction. He is single and not immediately interested in fatherhood. The endocrine investigations established hypogonadotropic hypogonadism with TT levels of 2.28 ng/ml (normal 2.8–8.0) and 5.32 nmol/l (normal 8.4–28.7); LH 1.86 mIU/ml; FSH 3.08 mIU/ml; prolactin 137 940 mIU/l (normal 98–456) and 146 555 mIU/l. MRI showed a pituitary macroadenoma 24/31 mm, not sight threatening. He denied surgical intervention and a bromocryptine treatment has been initiated with escalating dose to 15 mg/day, decreases the prolactin level to 1478 IU/l at the end of the first year, when the MRI showed a decrease of the dimensions of the prolactinoma to 16/25 mm. The therapy was continued with cabergoline in ascending dose from 2 to 5 mg weekly for 3 more years with a very good tolerance and compliance. The lowest level of prolactin – 583 IU/l was reached in 2009. During the next years, the condition was stable with small variations in the hormonal concentrations. The results in 2010 demonstrated prolactin 756 IU/l and continuing hypogonadotropic hypogonadism LH – 1.32 IU/l; FSH – 5.94 IU/l; TT – 6.93 nmol/l. On MRI, the tumor has diminished to 4/14 mm (last result in 2011 – 4/10 mm). At this point, a substitution with testosterone gel (Androgel, Solvay Pharmaceuticals) has been started. The complete battery of tests was performed at baseline and after 3 months of TRT.

The anthropometric indicators of the patients: height, weight, body mass index (BMI) and body composition – fat content; fat mass; fat-free mass (ffm); total body water (tbw); basal metabolic rate – (bmr) were measured by bioelectrical impedance with the leg-to-leg body composition analyzer Tanita TBF-215 (Tanita Corporation, Tokyo, Japan) before and after the testosterone treatment. Arterial pressure was measured after 15-min rest with a calibrated sphygmomanometer.

The clinical laboratory tests were performed in the Central biochemical laboratory of the University hospital “Alexandrovska”, which is the reference laboratory for the country. The samples for testosterone were drawn between 8 and 10 AM.

Preparation of EMs [20]

Human erythrocytes were washed three times as a 25%(v/v) suspension in ice-cold phosphate buffered saline (PBS, containing 137 mM NaCl; 2.7 mM KCl; 10.6 mM Na₂HPO₄; 8.5 mM KH₂PO₄, pH 7.4) by centrifugation (1000×g, 5 min, 4°C). The washed cells were lysed by dilution with eight volume of ice-cold lysis buffer (20 mM Tris/HCl pH 7.4) for 10 min at 4°C and the ghosts were pelleted by centrifugation at 20 000×g for 15 min. Essentially haemoglobin-free membranes were obtained after six washes in 20 mM Tris/HCl (pH 7.4) and were resuspended in the same buffer for further investigations.

Lipid extraction and analysis

Lipid extraction was performed with chloroform/methanol according to the method of Bligh and Dyer [21]. The organic phase obtained after extraction was concentrated and analyzed by thin layer chromatography. The phospholipid fractions were separated on silica gel G 60 plates in a solvent system containing chloroform/methanol/2-propanol/triethylamine/0.25% KCl (30:9:25:18:6 v/v) running standards in parallel [22]. The location of the separate fractions was determined either by spraying the plates with 2′,7′-dichlorofluorescein or by iodine staining. The spots were scraped and quantified by the determination of the inorganic phosphorus [23]. Cholesterol content was assayed by gas chromatography using a medium polarity RTX-65 capillary column (0.32 mm internal diameter, length 30 m, thickness 0.25 µm). Calibration was achieved by a weighted standard of cholestane.

Fatty acid analysis

The phospholipid extracts were saponified with 0.5 N methanolic KOH and methylated with boron trifluoride–methanol complex (Merck [24]). The fatty acid methyl esters were extracted with hexane and separated by gas chromatography on a capillary column coated with Supelcowax 10-bound phase 9 (i.d. 0.32 mm, length 30 m, film thickness 0.25 µm; (Supelco, Bellafonte, PA, USA) fitted in a Perichrom (France) gas chromatograph. Quantification was referred to an internal standard of heptadecanoic methyl ester.

Sphingomyelinase activity assay

Sphingomyelinase activity was determined by the method of Nikolova-Karakashian et al. [25] with minor modifications. Briefly, aliquots of membranes were lysed in 0.2% Triton X-100 in 100 mM Tris pH 7.4 buffer supplemented with 25 µM genestein for 10 min on ice. The samples were homogenized with three passes through a 25-gauge needle and 10 µl aliquots were taken for protein assay. NBD-sphingomyelin was added to the lysates to a final concentration of 20 µM and incubations were performed for 10 min at 4°C. Aliquots of this mixture containing 0.1 mg protein and 3 μM substrate were added to 5 mM MgCl₂, 10 mM Tris pH 7.4 to a final volume of 0.3 ml. All buffers contained 0.2% Triton X-100. After incubation for 1 h at 37°C, the reaction was stopped by the addition of 1 ml chloroform–methanol 2:1 (v/v). The samples were evaporated and separated in a system containing diethyl ether:methanol (99:1v/v) and the spots corresponding to ceramide were scraped and eluted. After addition of hexane, the fluorescence of the samples was measured at 455 nm (excitation) and 530 nm (emission).

Ceramidase assay

The activity of ceramidase was measured using the fluorescent analog of ceramide C12-NBD-Cer as substrate [26]. The reaction mixture contained 550 pmol of C12-NBD-Cer and 112 µg membrane protein in 100 µl of 25 mM Tris–HCl buffer pH 7.5, containing 1% sodium cholate. Following incubation at 37°C for 30 min, the reaction was terminated by the addition of 100 µl of chloroform/methanol (2/1, v/v), dried in a Speed Vac concentrator, redissolved in 30 µl of chloroform/methanol (2/1, v/v) and applied to a TLC plate, which was developed with chloroform/methanol/25% ammonia (90/20/0.5, v/v). NBD-dodecanoic acid released by the action of the enzyme and the remaining C12-NBD-Cer were separated by TLC, then analyzed, and quantified (excitation 470 nm, emission 525 nm).

Determination of ceramide content [27]

The ceramide mass was measured using E. coli diacylglycerol kinase, which phosphorylated ceramide to ceramide-1-phosphate and the product was quantified by densitometry on Konica X-ray film.

Fluorescence assay

1, 6-Diphenyl-1,3,5-hexatriene (DPH) was used as a fluorescent probe for determination of membrane fluidity. The lipid structural order parameter (SDPH) was calculated by the empirical method described by Van Blitterswijk et al [28]. The fluorescence experiments have been performed at 355 nm (excitation beam) and 425 nm (emission beam).

Statistical analysis

Statistical processing of the data was made by one-way analy-sis of variance (ANOVA), using the scientfic statistics software GraphPad InStat 3.0 (GraphPad Software Inc., San Diego, California, USA).

Results

Data about the dynamics in anthropometric and body composition indicators, and laboratory tests before treatment and after three months of TRT are presented on Table I. The anthropometric changes were different – patient #1 did not change his weight, but the other two patients tend to increase the weight by 1 and 2 kg. The fat content decreased in patient #1, but was unchanged in #2 and increased in #3. Interestingly, the fat-free mass increased in diabetic patients, and did not change in patient #3. Both diabetic men showed different changes in their HbA1c.

Table I.  Basic and after 3 months of testosterone replacement therapy data of the patients.

Patient	#1		#2		#3
Time of measurement	Before TRT	After TRT	Before TRT	After TRT	Before TRT	After TRT
Age	47		45		32
Diabetes	DM2		DM2		No
Height (cm)	185		168		182
Weight (kg)	106.5	106.1	99.5	100.5	100.6	102.5
BMI (kg/m^2)	31.1	31.0	35,3	35,6	30.4	30.9
Fat content (%)	31.2	27.7	41.8	40.7	26.1	27.7
Fat mass (kg)	33.2	29,4	41.6	40.9	26.3	28,4
ffm (kg)	73.3	76.7	57,9	59.6	74.3	74.1
tbw (kg)	53.7	56.1	42.4	43.6	54.5	54.2
bmr (kj)	9004	8973	8209	8245	8650	8760
bmr (kcal)	2152	2145	1962	1971	2067	2094
HbA1c (%)	6.8	7.2	7.8	7.0	–
TC (mmol/l)	6.0	6.85	5.7	4.4	4.2	3.8
HDL (mmol/l)	0.85	0.73	1.75	1.25	–	–
LDL (mmol/l)	3.38	4.7	3.01	2.11	–	–
TG (mmol/l)	3.83	3.81	1.89	1.84	0.89	0.83
LH (U/l)	3.62	–	2.89	–	1.32	–
FSH (U/l)	3.94	–	2.25	–	5.94	–
Prolactin (mIU/l)	126.1	–	161.1	–	756.8	–
Testosterone I meas.	6.61	14.54	7.01	17.1	4.72	10.9
Testosterone II meas.	7.29		7.82		6.93

ffm, fat-free mass; tbw, total body water; bmr, basal metabolic rate; BMI, body mass index, TC, total cholesterol, HDL, high-density lipoproteins; LDL, low-density lipoproteins; TG, triglycerides; LH, luteinizing hormone; FSH, follicle-stimulating hormone; TRT, testosterone replacement therapy.

The dynamics of the lipids was also different. There was an unfavorable trend in patient #1 – the TC/HDL ratio increased from 7.1 to 9.4. Patient #2 decreased the total, LDL- and HDL-cholesterol, resulting in a 3.3–3.5 TC/HDL ratio difference. Patient #3 also decreased the total cholesterol level. The triglycerides did not change significantly in all patients.

Further studies involved analysis of the phospholipid and fatty acid composition of EMs isolated from the three patients before and after TRT. Table II shows the alterations in the major membrane phospholipids after 3 months of testosterone administration. Apparently, testosterone treatment induced statistically significant elevation in the levels of sphingomyelin (SM) and phosphatidylethanolamine (PE) in the EMs of the first two patients. However, in the membranes of the third patient was observed increase only in the content of PE, whereas SM remained almost unchanged. In addition, phosphatidylcholine (PC), which is the major membrane phospholipid, was reduced in the EMs of all three patients (Table II).

Table II.  Changes in erythrocyte membranes observed during the TRT.

	Patient # 1		Patient # 2		Patient # 3
	Before TRT	After TRT	Before TRT	After TRT	Before TRT	After TRT
Phospholipid composition of erythrocyte membranes^a (mol%)
Sphingomyelin	16.6	18.1*	24.3	25.7*	17.4	17.9
Phosphatidylcholine	38.9	30.4**	34.7	30.2*	40.5	40.2
Phosphatidylserine	16.7	15.1*	12.5	12.2	15.9	14.7*
Phosphatidylinositol	9.8	10.6	8.6	10.5*	8.7	7.5
Phosphatidylethanolamine	19.4	25.9**	19.7	22.9**	17.7	22.8**
Fatty acid composition of phosphatidylcholine in erythrocyte membranes^b (mol%)
C16:0	32.95 ± 1.17	27.63 ± 1.32**	25.71 ± 1.17	23.83 ± 1.32*	15.66 ± 1.05	13.63 ± 1.12*
C18:0	15.51 ± 0.93	12.57 ± 1.36*	21.28 ± 0.93	19.37 ± 1.36*	38.21 ± 0.64	36.15 ± 1.25*
C18:1	14.32 ± 1.04	15.32 ± 1.37	14.72 ± 1.04	15.72 ± 1.37	13.87 ± 1.22	14.56 ± 1.17
C18:2	13.23 ± 1.36	14.24 ± 1.09	18.28 ± 1.36	18.14 ± 1.09	10.53 ± 1.06	12.74 ± 1.29*
C20:4	11.39 ± 1.42	11.39 ± 0.51	16.79 ± 1.42	20.39 ± 0.51**	15.89 ± 1.32	15.11 ± 0.41
C22:6	2.09 ± 0.37	6.19 ± 0.78**	3.19 ± 0.37	2.58 ± 0.18*	5.79 ± 0.31	7.89 ± 0.68**
Structural order parameter (SDPH) and cholesterol/total phospholipids (CH/TPL) ratio in erythrocyte membranes
SDPH	0.7149	0.6948*	0.7034	0.6875*	0.7172	0.6932*
CH/TPL	0.61	0.57	0.54	0.46*	0.56	0.42*
Activity of neutral sphingomyelinase (nmol·min^−1·mg^−1), ceramidase and content of ceramide in erythrocyte membranes
SMase	2.4	1.7*	2.1	1.5*	1.9	2.1*
Ceramide	3.8	2.3**	4.6	3.7*	3.6	3.8
Ceramidase	100	102	100	98	100	98

TRT, testosterone replacement therapy ^aResults are means of five separate measurements. ^bValues are means ± SD of three separate measurements. ^cCeramide is expressed as nmol/mol membrane phospholipids. SMase-Sphingomyelinase. Control values of ceramidase are expressed as 100%. *P < 0.01; **P < 0.001.

To analyze in detail the alterations in the lipid molecules comprising the membrane bilayer, further studies were focused on the fatty acid composition of the most abundant membrane phospholipid – PC (Table II). As evident for patient #1, the mol% of the two major saturated fatty acids, C16:0 and C18:0, was reduced after a 3-month long TRT. Among the unsaturated fatty acids C18:1 and C18:2 showed certain increase, which was not statistically significant. However, C22:6, which is a minor acyl chain fraction was augmented almost three-fold after the 3-month of testosterone administration in patient # 1. In the PC molecules of patient #2 was observed a reduction of the two saturated fatty acids mentioned above, as well as a significant elevation of the polyunsaturated C20:4. In the PC of patient #3 was found a decrease of the saturated fatty acids and augmentation of C18:1 and C22:6. So what was common for the three patients was a decrease of the saturated fatty acids for all three of them and an increase in some of the unsaturated ones.

As the acyl chains of the phospholipid molecules play an essential role in maintenance of the physico-chemical properties of the membrane bilayer, studies were carried out to determine the structural organization of the erythrocyte ghosts. For this purpose was used the non-polar fluorescent probe 1,6-diphenyl - 1,3,5- hexatriene (DPH). Table II shows the differences in the structural organization of the EMs before and after testosterone treatment. Apparently, the structural order parameter was lower in EMs of all three testosterone-treated patients, implying that testosterone therapy induced fluidization of the EM.

In addition, the cholesterol/total phospholipids (CH/TPL) ratio, which is also an important factor affecting membrane fluidity, was lower in the EMs obtained from patients # 2 and 3 as a result of testosterone treatment.

Studies were also performed in order to elucidate the molecular mechanisms, underlying the increase in the content of SM, observed especially in patients # 1 and 2 after testosterone treatment. For this purpose was determined the activity of membrane-bound neutral sphingomyelinase (SMase), which regulates the content of SM and ceramide in cellular membranes (Table II). SMase activity was lower in EMs from patients # 1 and 2, but not in patient #3, which implies that the alterations in SMase activity could be the reason for the observed elevation of SM content. The content of the main product of SMase, ceramide, was reduced in the membranes of patients # 1 and 2 (Table II). As this lipid, besides being a product of SMase, is also substrate of the enzyme ceramidase, studies were carried out on the changes in ceramidase activity. The obtained results will provide information which one of these two enzymes, SMase and/or ceramidase, is/are responsible for the altered content of ceramide, the latter showing high physiological activity when accumulated in membranes. Ceramidase activity was not altered in the tested membrane (Table II).

Discussion

Still many aspects of the testosterone effect in the male body remain to be elucidated. It is well known that elevated hematocrit increases the risk for thrombosis. The most frequent adverse effect of TRT is the increase of hematocrit [29]. A recent systematic review and meta-analysis found that testosterone-treated men were at higher risk of developing erythrocytosis than the placebo/nonintervention group (RR, 3.15; 95% CI, 1.56–6.35), but there were no significant differences in the rates of death, myocardial infarction, revascularization procedures, or cardiac arrhythmias between the testosterone and the placebo/nonintervention groups [30]. Searching for an explanation of this discrepancy, we aimed to investigate the effect of TRT on erythrocytes membrane lipid composition.

Two of the patients were relatively young men with metabolic syndrome and LOH. They have been chosen because of the recently diagnosed diabetes type 2, before the development of advanced late diabetic complications. This category of patients is quite common and this is the stage in the development of the metabolic syndrome, when everything possible must be attempted for improvement of the cardiovascular prognosis. The third patient was a typical case of hypogonadotropic hypogonadism. The therapeutic intervention was a standard one, but different for patient #1 (intramuscular) compared to the other patients (gel). The effects of TRT on anthropometric parameters, HbA1c, and lipid profile were incoherent. No conclusions could be made about any differences between the specific treatment options on metabolic aspects based on these cases. This was also our aim – to present different cases with different conventional treatment as a background for the further analyses.

The purpose of this study was to analyze the impact of TRT on the composition and structural organization of the lipids comprising EMs. The obtained results showed that testosterone treatment induced alterations mainly in three phospholipid fractions – SM, PE and PC. In patients # 1 and 2 were observed changes in all three phospholipid fractions, whereas in patient # 3 was established an increase only of PE and a decrease of PC. The synthesis of these two lipids is closely related because methylation of PE induces accumulation of PC in membranes. It is possible that the methylation of PE is inhibited due to testosterone treatment, thus leading to accumulation of PE and corresponding reduction of the PC level. The elevation of PE in the EMs was common for the three patients under investigation. This phospholipid is of particular importance for the structure and functions of the membrane bilayer due to the specific shape of its molecule, known as inverted cone. It has a small polar head and highly unsaturated acyl chains, which underlies its predominant positioning in the curvature of cellular membranes. Thus, the abundance of this phospholipid in membranes makes them more flexible, which, especially in the case with blood cells, would improve blood rheology.

In addition, the effect of the phospholipid alterations on the physico-chemical properties of the EM becomes more pronounced due to the observed changes in the fatty acid composition of the phospholipid molecules. It is well known that the increase of the relative content of unsaturated fatty acids induces fluidization of the membrane bilayer, whereas accumulation of saturated acyl chains makes membranes more rigid and less flexible. As evident from Table II, testosterone administration induced alterations in the acyl chains of the major membrane lipid – PC. Interestingly, in all three patients was observed a reduction of the two major saturated fatty acids – stearic and palmitic. This is a very important observation, because a decrease in the relative content of the saturated acyl chains in the membrane phospholipids is associated with fluidization of the membrane bilayer, which affects various membrane-related processes like transport, signal transduction etc. However, this finding, taken together with the reported elevation of unsaturated fatty acids (although different for the three patients) is a significant prerequisite for fluidization of the erythrocyte bilayer as a result of TRT. This hypothesis was tested using DPH as a fluorescent probe for monitoring of the structural organization of the membrane bilayer of erythrocytes collected from the three patients after testosterone administration. Apparently, the structural order parameter was lower for all tested membrane, implying that the lipid bilayer has undergone a different degree of fluidization (Table II). What is more, the cholesterol/total phospholipids (CH/TPL) ratio, which is essential for maintenance of the membrane physico-chemical properties, was also lower in all three cases, but this decrease was statistically significant only for patients # 2 and 3.

The alterations in the fatty acid composition, induced by testosterone administration need special attention, because the activity of the enzymes, responsible for fatty acid desaturation has been reported to be insulin-dependent [31] and desaturase enzymes are sensitive to hormonal imbalance [32]. In addition, erythrocytes in diabetics are less flexible due to specific lipid changes, resulting in membrane rigidity, which has been considered as a part of the pathological processes accompanying diabetic microvascular complications [33]. Our results demonstrate that testosterone administration is able to influence erythrocyte flexibility due to the elevation of PE and the degree of fatty acid unsaturation, thus improving blood rheology and different processes, dependent on it like erection and a global decrease of CV risk.

The other membrane phospholipid, sphingomyelin, which was also affected by TRT, especially in patients # 1 and 2, is a molecule with marked functional activity.

It is noteworthy that the content of ceramide, which is a product of SMase-induced hydrolysis of SM, was lower in the membranes of patients # 1 and 2 (Table II). The increased SM content and the lower level of ceramide observed in EMs of these patients may be of special physiological importance. Insulin resistance has been associated with the SM signaling pathway [34,35]. Ceramide molecules might impair the insulin activity through maintaining protein kinase B in inactive state and/or could decrease insulin-stimulated glucose uptake [36]. So it is possible that the reduced ceramide level induced by TRT could lead to a recovery of the insulin receptor functions. The obtained results showed that the reduced level of ceramide was more likely due to its lower production by SMase, rather than to its activated catabolism, performed by ceramidase (Table II). In addition, the SM signaling pathway is also related to the activity of certain inflammatory cytokines such as TNF-α as well as interleukins [37,38]. So it could be speculated that the altered level of ceramide might be a factor regulating as well inflammatory processes in the organism.

The small number of patients represents the main weakness of the study and it should be considered as a pilot one. All described changes in the EMs could not be explained with the diverse systemic lipid or carbohydrate alterations and with changes in body composition, described in Table I. Each case is unique, though some of the mentioned membrane changes became common for all three patients. We were not able to find analogous data in the literature. The improvement in EM quality could be one of the reasons for the lack of an unfavourable effect of elevated hematocrit.

In conclusion, the erythrocytes are a suitable sample for lipid analysis due to the lack of desaturases, implying that the fatty acids are taken from the plasma [39]. This makes the membrane of red blood cells an adequate tool for evaluation of the lipid alterations in the developing pathology and the therapeutic effect. At this point the molecular mechanisms underlying the observed changes in the lipid molecules in the EMs are not clear. The reported alterations in the phospholipid and fatty acid composition as well in some lipid-metabolizing enzymes and possibly in the signal transduction pathways need more profound investigations in order to elucidate the intimate biochemical mechanisms underlying the effect of testosterone on the EMs structure and functions.

Acknowledgement

The authors would like to thank Prof. Farid Saad for his valuable comments on the manuscript.

Declaration of Interest: PA, AM, DP, GS and RP have no conflict of interests. ZK has held lectures about TRT for Bayer-Schering and Solvay Pharmaceuticals.