Reduced Levels of Testosterone Induce LDL Oxidation and Atherosclerotic Lesions Involving Inflammatory Imbalance and Reduced Macrophage Apoptosis

Corresponding Author: Tadeu Uggere de Andrade Pharmaceutical Sciences Graduate Program, University Vila Velha, Vila Velha, ES, Brazil Tell: +552734212001 Fax: +552734212000 Email: tadeu.andrade@uvv.br Abstract: Atherosclerosis is among the major cardiovascular diseases that cause death in the world. Men who are deficient in androgen hormones are more susceptible to this disease. The aim of this study was to evaluate the influence of reduced levels of testosterone on the lipid profile, pro and anti-inflammatory cytokines, apoptosis of peritoneal macrophages and the development of atherosclerosis in LDLr-/-mice, as well as on the levels of OxLDL in healthy male humans. Tissues and organs of the experimental animals were used to evaluate lipid deposition in aorta and apoptosis of peritoneal macrophages. In humans the levels of testosterone and OxLDL were analyzed in blood samples from healthy male subjects, who were divided into two groups: low level and high level of testosterone. Animals with decreased testosterone did not present changes in the lipid profile or the levels of IL-6 and TNF-α. However, a decrease in IL-10 and an increase in the IL-6/IL-10 and TNF-α/IL-10 ratios were observed. An increase in the aorta lipid deposition was observed and a decrease in macrophage apoptosis was also demonstrated. In humans, there was an increase of OxLDL in the group with lower levels of testosterone. The findings suggest that decreasing endogenous testosterone levels may induce the development of atherosclerosis, with participation of inflammatory response and macrophage apoptosis. This is reinforced by our human studies, where individuals with low testosterone levels had higher LDL oxidation, which may predict a high atherosclerotic risk for these men.


Introduction
The leading causes of death in the world are related to cardiovascular diseases (Usman et al., 2015). Some metabolic conditions may act as protectors and make some individuals less likely to develop these diseases. In this context, the protective effects of androgens on the cardiovascular system have been reported previously (Wu and von Eckardstein, 2003). A prospective study showed an increase in cardiovascular risk in men whose serum testosterone levels were decreased, which was independent of other risk factors (Kelly and Jones, 2013). Additionally, androgen deprivation therapy also increases the risk of cardiovascular disease (Rojas et al., 2015). Androgen replacement in the population that is deficient in these hormones has been associated with a reduction in mortality (Frostegård et al., 1999).
Atherosclerosis is among the leading cardiovascular diseases in men with low testosterone levels (Araujo et al., 2011). Despite the higher incidence of cardiovascular disease in men compared to women, studies suggest that androgens, of which the major one is testosterone, protect ■■ man against atherosclerosis (Wu and Von Eckardstein, 2003). Such protection is due to the fact that testosterone plays an important role in the regulation of carbohydrate, lipid and protein metabolism and positively affect the control of glucose, liver fat and cardiac biomarkers (Kirby et al., 2019;Ng Tang Fui et al., 2017;Rezanezhad et al., 2018), however, the mechanisms are not completely elucidated.
Atherosclerosis is an inflammatory and progressive disease characterized by endothelial dysfunction, vascular inflammation and accumulation of lipids in the intimal layer of medium-and larger-caliber arteries (Moore et al., 2013), where changes promote cardiac events, such as myocardial infarction and stroke, as a response to the rupture of atherogenic plaque (Usman et al., 2015).
Low density lipoprotein (LDL) is an acute phase mediator of atherosclerotic disease. At high levels, it is involved in the organization of the lipid nucleus and the polarization of circulating monocytes. In plaques, LDL is especially susceptible to oxidation, forming oxidized LDL (OxLDL) (Rojas et al., 2015). As a response to these molecules, T cells are activated and release adhesion molecules that provide recruitment of monocytes from the circulation to tissue, which is differentiated into macrophages (Frostegård et al., 1999;Ilhan and Kalkanli, 2015;Rojas et al., 2015). As a result, an increase in inflammation can be observed by the activation of T helper-1 (Th1) cells; the factors released by these cells contribute to the development of atherosclerosis (Frostegård et al., 1999).
Inflammatory mediators seem to initiate the phonotypical modulation of macrophages on the arterial wall during atherogenesis (Ley et al., 2011). Activated macrophages can increase plaque vulnerability, while an alternative phenotype can promote stability and regression of plaque for minimizing inflammation (Viola and Soehnlein, 2015). Cytokines are closely related to the progression of atherosclerotic lesions. According to Ait-Oufella et al. (2011) all cells involved with atherosclerosis are able to produce and respond to cytokines, which play a dual role in the disease, with pro-or anti-inflammatory responses. Moore et al. (2013) reported that macrophages are important in inflammatory responses, which is attributed to their capacity to sustain pro-inflammatory factors and to their apoptosis, which contributes to a necrotic nucleus prone to rupture. Macrophage apoptosis has been identified as a key characteristic of atherosclerotic plaques in all phases of the lesion (Tabas, 2007). Macrophage apoptosis can be triggered by a combination of factors, according to the phase of the disease (Tabas, 2005) and this can affect plaque progression in different ways (Gautier et al., 2009). In early atherosclerotic lesions, macrophage apoptosis is associated with the reduction of plaque progression. However, in advanced atherosclerotic lesions, it promotes an increase in injury and an increased risk of plaque rupture (Lammers et al., 2011).
Among the in vivo experimental models that are used to study atherosclerosis, LDL receptor knockout (LDLr-/-) mice are widely used. This is due to the fact that these animals present metabolic changes and spontaneous lipid deposition in the aorta induced by aging or through interventions that characterize risk factors, such as the use of atherogenic diet (Hatch et al., 2012), castration (orchiectomy and ovariectomy) (de Oliveira et al., 2013;Hopmans et al., 2014) or treatment with supraphysiological doses of androgenic anabolic steroids (de Andrade et al., 2018). Moreover, the LDLr-/-are chosen for this study because mice present a lipoprotein profile similar to hyperlipidemic humans, since the cholesterol concentrates on the LDL fraction (Zadelaar et al., 2007).
Usually, many studies are performed using the induction of risk factors associated with the use of atherogenic diet (Hatch et al., 2012;Merat et al., 2000). However, atherogenic diets alone are responsible for metabolic changes that lead to the development of other pathogenic processes that may be confused with the object of study (Hopmans et al., 2014). Thus, the study of the development of atherosclerosis in experimental animals fed normal diet is justified, aiming to eliminate possible interferences that may be caused by the high fat diet (Hopmans et al., 2014).
Therefore, the aim of the present study was to evaluate the lipid profile, pro-and anti-inflammatory cytokines, apoptosis of peritoneal macrophages and the development of atherosclerosis in LDLr-/-and C57BL/6 mice subjected or not to orchiectomy. Additionally, we translationally evaluated the blood levels of testosterone and OxLDL in healthy, middle-aged men with different blood level patterns of testosterone.

Animals and Experimental Groups
Eight-week-old LDLr-/-and C57BL/6 Wild-Type (WT) mice were used, with a corporal mass between 20-30 g. The animals were separated into the following four groups (n = 8 each): WT and LDLr-/-(controls subjected to sham surgery); and WTOrx and LDLr-/-Orx (WT and LDLr-/-submitted to orchiectomy). The animals were obtained by the Research Laboratory of the Faculty of Pharmaceutical Sciences of USP/SP and were kept in a living room of the University Vila Velha (UVV). The animals were maintained in controlled conditions with a light-dark cycle of 12 h and a temperature of 22C. The individual cages provided free access to water and food (Standard chow, Probiotério, Moinho Primor, S.A). All the experimental procedures were performed in accordance with the USA National Institutes of Health (Garber et al., 2010) and were approved by the Committee on Ethics in the Use of Animals of the University Vila Velha (CEUA-UVV, Protocol no. 191/2011).

Surgery Procedure
The animals subjected to orchiectomy were anesthetized with a mixture of ketamine and xylazine (11.5/1.15 mg/kg of body weight, intraperitoneal-i.p.) and after deep anesthesia, the scrotum was cleaned and a small incision was made. The tunic was opened and the testis, epididymis syrup, deferent duct and blood vessels were exteriorized. The blood vessels and deferent ducts were cauterized and the testis and epididymis were removed. The remaining tissue was returned to the scrotal sac and a suture was made. This procedure was repeated on the other testicle.
The local site of the incision was cleaned with iodopovidone and an injection of benzathine penicillin G and procaine penicillin G (60.000 U each, intramuscular) was applied in each mouse. After the surgical procedure, the animals were maintained in individual cages with free access to water and food until complete recovery (Zhang et al., 1998).

Experimental Protocol
Four weeks after the surgical procedure, all the animals were anesthetized with a mixture of ketamine and xylazine (11.5/1.15 mg/kg of body weight, i.p.) and then euthanized. The blood was collected by a cardiac puncture to determine the serum levels of testosterone and cytokines as well as the lipid profile and the aorta was collected to quantify the lipid deposition. The other group of animals was separated to perform the isolation of the peritoneal cells to determine the level of macrophage apoptosis.

Biochemistry Analysis
The blood was collected in plastic dry tubes from the cardiac puncture. The plasma was separated by centrifugation (3500 rpm, 4C, 15 min) and maintained at -20C until used. The total cholesterol (TC) and HDL cholesterol (HDL-C) were determined using commercial kits (BIOCLIN, Minas Gerais, Brasil). In addition, the non-HDL cholesterol was determined by calculating the difference between the TC and the HDL-C (Daleprane et al., 2012).

Quantification of Serum Testosterone
The determination of serum testosterone was made by ELISA using commercial kits in agreement with the manufacturing instructions (In vitro diagnostica, Minas Gerais, Brasil). The assay was made in duplicate and read at 450 nm using an ELISA Reader (Filter Max F5, Molecular Devices LLC, Austria).

Quantification of Pro-and Anti-Inflammatory Cytokines
Interleukin-6 (IL-6), Tumor Necrosis Factor alpha (TNF-α) and interleukin-10 (IL-10) were simultaneously quantified using the Cytometric Bead Array Mouse commercial kit (BD Biosciences, San Diego, CA, USA) in agreement with the description by Menezes et al. (2012). The analysis was made by flow cytometry using the animal's plasma. The ratio between pro-inflammatory (TNF-α and IL-6) and anti-inflammatory (IL-10) cytokines was obtained by division of the correspondent mean.

Determination of Apoptosis of Peritoneal Cells
Male LDLr-/-mice and wild-type C57BL/6 mice (WT) subjected to orchiectomy or sham surgery (sham) received an intraperitoneal injection with 1 mL of sterile thioglycolate (3% p/v. Sigma-Aldrich, St. Louis, MO, EUA) four days before their use. After, the animals were euthanized and the peritoneal cells were collected by washing the peritoneal cavity, using sterile cold PBS (Pedrosa et al., 2010). To determine the apoptotic cells, anexin V (FITC) and Propidium Iodite (PI) were used as markers, which was followed by a flow cytometer analysis. Ten thousand events were registered from each sample and the macrophage populations were determined by morphological parameters, such as size and granularity. The procedure was performed in agreement with Tonini et al. (2013); briefly, 110 6 cells/mL were used in the assay and received 2 μL of anexin V and PI; after incubation, an analysis was performed using a flow cytometer (FACS Canto II; Becton Dickinson) at 585 nm and 530 nm. On the dot plot graph, the quadrants Q1 contains cells positive only to PI; Q2 contains necrotic or late apoptotic cells; Q3 contains live cells; and Q4 shows early apoptosis. Apoptosis was calculated as the sum of Q2 and Q4. For the positive control, H2O2 was used.

The Morphological Analysis of Lipid Deposition
For the analysis of lipid deposition, briefly, the aorta was completely dissected and maintained in a solution of PBS-formalin (10%; 15C) for 24 h. After cleaning, a symmetric cut was performed and the aorta was fixed onto a surface of ethyl vinyl acetate with steel pins. After, the aorta was stained with Oil-Red-O (Sigma-Aldrich, St. Louis, MO, EUA) for "en face" analysis (Paigen et al., 1987;Xu, 2006). The lipid deposition was evaluated by images obtained with a high-resolution camera (Canon T5 18-55; Canon, USA) and analyzed using the public software "ImageJ" (National Institute of Health, USA), by a researcher blind to experimental groups.

Human Testosterone and Oxidized LDL Analyses
To perform the analyses of testosterone and oxidized LDL on human serum, male healthy subjects were invited to participate. The research was performed following all the instructions of the Helsinki Declaration (WMA, 2008) and was approved by the Human Ethics Committee of University Vila Velha (#399751). All the participants provided informed consent and signed the consent form. The inclusion criteria were heathy men between ages 40-55 who had no apparent physical or physiological ■■ problems. The exclusion criteria were previous use of alcohol, smoking or drug abuse, anabolic androgenic steroid use and other health problems. The patients' blood was collected and separated. The serum was maintained frozen (-80C) until use. The serum testosterone and oxidized LDL were determined using ELISA commercial kits (Elabscience ® Houston, Texas, USA), following the manufacturer's instructions. The subjects were divided into two groups according to the level of testosterone as follows: Low level of testosterone (LT; below 350 ng/dL) and high level of testosterone (HT; above 350 ng/dL). This cut point of testosterone is in accordance with English et al. (2000), that showed men with arterial cardiovascular disease were under low total testosterone levels (around 12 nmol/L, which means 346 ng/dL). In both groups, the subjects were under the normal range of the testosterone levels, but with lower or higher levels.

Statistical Analyses
The results were expressed as the mean ± Standard Error of the Mean (S.E.M). The data were analyzed using a one-way Analysis of Variance (ANOVA) followed by the Tukey post hoc test, adjusted for multiple comparisons; significance was accepted when p<0.05. The statistical analysis was performed using Prism software (Prism6, GraphPad Software, Inc., San Diego, CA, USA), with the exception of the flow cytometry data, which was analyzed using DIVA software and FCS Express Flow Research Edition. Figure 1 shows the results of the cholesterol profile of the groups (TC, HDL-C and nHDL-C). The LDLr-/group had an increase in TC compared to the others (WT and WTOrx) and the orchiectomy did not change the total cholesterol (Fig. 1

Aorta Lipid Deposition
As expected, the LDLr-/-animals showed an increase in lipid deposition when compared to the WT group. In addition, orchiectomy promoted an increase in lipid deposition in the orchiectomized groups when compared to the controls (

Human Testosterone and Oxidized Lipid Analyses
Forty-eight subjects signed the consent form and had their levels of testosterone and oxidized LDL measured. Eight subjects were excluded (one did not attend the day to collect blood sample; and 7 had declare previous health problem). The mean age was 47.5±0.7 years and the testosterone levels were considered normal in all the participants (361±23 ng/dL; reference values: 166 to 923 ng/dL). However, the testosterone levels between the two groups were different (LT: 232.7±12.16 (N = 23); HT: 398.1±16.41 ng/dL; (N = 17); P<0.05), which reflects the OxLDL levels, when two groups were compared (LT: 65.25±4.28; HT: 51.97±4.49 mU/L; p<0.01). The subjects from the group of low levels of testosterone showed higher levels of OxLDL compared with those from the HT group.

Discussion
The results of the present study demonstrated, for the first time, the influence of testosterone levels on atherosclerosis development in young male LDL receptor knockout mice that did not receive an atherogenic diet. Additionally, an imbalance was observed in pro-and antiinflammatory cytokines, with a decrease in IL-10 levels and a decrease in apoptosis of macrophage cells. No change in the lipid profile was observed. Additionally, an increase in the oxidation of LDL was observed in human blood samples under low levels of testosterone.
The role of androgens on the pathogenesis of atherosclerosis has been the aim of many clinical (Khazai et al., 2016) and experimental investigations (Huang et al., 2014). However, the results are still controversial (Kapoor and Jones, 2008;von Dehn et al., 2001). Traish et al. (2009) reviewed the role of androgens in atherosclerosis and demonstrated that changes in the lipid profile promote oxidative stress and endothelial dysfunction, with increased release of pro-inflammatory factors, culminating in atherosclerosis development. Bourghardt et al. (2010) suggested that the effects on the lipid profile are not a prerequisite to the antiatherogenic effect of testosterone in LDLr --/-mice. Additionally, Nathan et al. (2001) used orchiectomized animals and observed no difference in the total cholesterol and triglyceride levels, but HDL levels were decreased. In that research (Nathan et al., 2001), an occidental diet was used; thus, the decrease in HDL cholesterol could be attributed to the diet and not to testosterone levels.
However, our results demonstrated no changes in the lipid profiles. It seems reasonable to infer that the reduction in testosterone levels is more important than changes in the lipid profile to induce atherosclerotic lesions in orchiectomized animals. Our data corroborates the results found by Hatch et al. (2012), which demonstrated an increase in lipid deposition in LDLr-/-animals with a decrease in testosterone levels but under a lipid-rich diet.
The experimental model used in the present study indicates that levels of plasmatic cholesterol slightly increased when compared to wild-type animals and that there is a slowly developing atherosclerosis when using the chow diet (Merat et al., 2000), which mimics the development in humans (Emini Veseli et al., 2017). Therefore, the reduction in the male hormone, testosterone, should be as important as a lipid-rich diet for atherosclerosis development.
The inflammatory process associated with atherosclerosis has been studied for the last few years and has demonstrated importance in the physiopathology of the disease. Pro-inflammatory cytokines, such as IL-6 and ■■ TNF-α, are produced by macrophages and are related to the transmigration of leukocytes (Ait-Oufella et al., 2011), which contribute to lesion progression. Additionally, IL-6 is considered to have a dual character (Ait-Oufella et al., 2011;Kleemann et al., 2008) and has been involved in endothelial dysfunction, macrophage transition to foam cells and the acute response to inflammation (Keidar et al., 2001;Khan et al., 2015). In contrast, some cytokines are able to suppress inflammatory responses, such as IL-10, which is also secreted by macrophages and is considered atheroprotective (Kleemann et al., 2008). Additionally, data in the literature has demonstrated that the balance between pro-and anti-inflammatory cytokines is crucial for determining the development of the lesion (Kleemann et al., 2008). Therefore, in the present study, we investigated the classical markers involved in atherosclerotic processes, such as IL-6, TNF-α and IL-10, which can be modulated by testosterone. Additionally, this was demonstrated by in vitro studies that showed the capacity of testosterone to suppress pro-inflammatory cytokines and stimulate anti-inflammatory cytokine production, especially IL-10 (Malkin et al., 2004). Liva and Voskuhl (2001) observed that testosterone can act directly on CD4+ lymphocytes to promote an increase in the amount of IL-10 released and demonstrated that macrophages express the androgenic receptor, which does not exclude the direct action of the hormone in these cells. Moreover, the imbalance between IL-10 and TNF-α, with an increase in TNF-α, has been proposed as having an important role in the development of atherosclerotic lesions (Waehre et al., 2002).
Our study demonstrated no differences in proinflammatory cytokines; however, the decrease in testosterone produced a negative effect on IL-10 levels. Additionally, the TNF-α/IL-10 and IL-6/IL-10 ratios were increased in orchiectomized animals. Based on the inflammatory response observed, we investigated macrophage viability, since these cells are crucial for plaque development.
Macrophage apoptosis was evaluated by the isolation of peritoneal cells from mice. Zhang et al. (2008) demonstrated that among the peritoneal cells stimulated with thioglycolate, approximately 70% are macrophages. This technique is largely used in studies to evaluate molecular modulations involved with atherosclerosis lesion development (Hasegawa et al., 2016;Liu et al., 2016). Our data demonstrated that animals with decreases in testosterone levels present a decrease in macrophage apoptosis simultaneously with an observed increase in lipid deposition. The development of plaques seems to have a paradoxical relationship with apoptosis that depends on the phase of the lesion (Lammers et al., 2011).
The initial phase of atherosclerosis seems to be closely linked to live macrophages, since many macrophages that are present at the site of the lesion are responsible for lipid and cell debris phagocytosis at the beginning of lesion development and macrophage apoptosis at this moment could have an anti-atherogenic effect (Ley et al., 2009;Tabas, 2005). Gautier et al. (2009) demonstrated the importance of time on plaque development; according to their findings, a five-week lesion can be considered to be at an early stage and a 15week lesion considered at an advanced stage. Therefore, we can infer that the lesion observed in the present study can be considered early, since it was evaluated in a period of four weeks without an atherogenic diet. Together, these results support the role of testosterone and macrophage activation in atherosclerosis development.
Additionally, Jin et al. (2006) verified the relationship between testosterone and apoptosis, demonstrating that testosterone is capable of promoting macrophage apoptosis in cells derived from bone marrow. Lammers et al. (2011) demonstrated an inverse relationship between atherosclerotic lesions and macrophage apoptosis in LDLr-/-mice. However, in both studies (Gautier et al., 2009), (Lammers et al., 2011) a western diet was used, which was enriched with fat and no hormonal evaluation was performed. Our results demonstrated for the first time the effect of testosterone on spontaneous plaque development in LDLr --/-mice and showed that in LDLr-/-mice, the decrease in the male hormone, testosterone, is just as important as diet in plaque development.
Atheroprotective effects of testosterone are attributed to many mechanisms and some authors suggest changes in the lipid profile (Malkin et al., 2004), inflammatory reduction (Li et al., 2008), the role of the aromatase enzyme (Nettleship et al., 2007) androgenic receptors and others. In a review, Vasconsuelo et al. (2011) showed that sexual hormones promote their action by apoptosis, which is an important cell event. Furthermore, they demonstrated the dual effect of testosterone, which can induce both apoptotic and antiapoptotic effects that will depend on the cell type, hormone concentration and cellular environment.
Therefore, we can assume that the macrophages observed after induction with thioglycolate could reflect an inflammatory status of the orchiectomized animals, since the macrophages are probably maintaining the Th1 response, which contributes to plaque progression independently of lipid changes (Frostegård et al., 1999).
Together with inflammation, oxidation is also a mechanism reported in the development of atherosclerosis (Kattoor et al., 2017). As a translational study, we performed the evaluation of OxLDL in healthy middle-aged men under low and high levels of testosterone (both in the normal range of testosterone levels). It was possible to observe an increase in OxLDL when testosterone levels were decreased. The literature has demonstrated a high correlation of OxLDL levels ■■ and atherosclerosis development (Gao and Liu, 2017;Linna et al., 2008). A negative correlation between testosterone levels and oxidized LDL, highlighting the atheroprotective effect of testosterone under physiological conditions, has been demonstrated (Barud et al., 2002;Linna et al., 2008). Therefore, our study with human blood samples supports our findings with animals and macrophage cells, indicating that, under physiological conditions, testosterone can be a protective factor against plaque development involving antioxidant and antiinflammatory mechanisms.

Conclusion
In summary, our results show that the reduction of endogenous testosterone levels can induce early atherosclerotic lesions in male LDLr-/-mice, which occurs with participation of inflammatory response and macrophage apoptosis, indicating that testosterone may play an atheroprotective role. This is reinforced by our studies in humans where individuals with low testosterone levels had higher LDL oxidation, which may predict a high atherosclerotic risk for these men. However, further studies are needed to further investigate these mechanisms.