HSL-knockout mouse testis exhibits class B scavenger receptor upregulation and disrupted lipid raft microdomains.

There is a tight relationship between fertility and changes in cholesterol metabolism during spermatogenesis. In the testis, class B scavenger receptors (SR-B) SR-BI, SR-BII, and LIMP II mediate the selective uptake of cholesterol esters from HDL, which are hydrolyzed to unesterified cholesterol by hormone-sensitive lipase (HSL). HSL is critical because HSL knockout (KO) male mice are sterile. The aim of the present work was to determine the effects of the lack of HSL in testis on the expression of SR-B, lipid raft composition, and related cell signaling pathways. HSL-KO mouse testis presented altered spermatogenesis associated with decreased sperm counts, sperm motility, and infertility. In wild-type (WT) testis, HSL is expressed in elongated spermatids; SR-BI, in Leydig cells and spermatids; SR-BII, in spermatocytes and spermatids but not in Leydig cells; and LIMP II, in Sertoli and Leydig cells. HSL knockout male mice have increased expression of class B scavenger receptors, disrupted caveolin-1 localization in lipid raft plasma membrane microdomains, and activated phospho-ERK, phospho-AKT, and phospho-SRC in the testis, suggesting that class B scavenger receptors are involved in cholesterol ester uptake for steroidogenesis and spermatogenesis in the testis.

testis ( 26 ) and both 3.3 kb and 3.9 kb HSL mRNAs in human testis ( 27,32 ). HSL is localized to elongated spermatids and Sertoli cells guinea pig and mink testis ( 8,33 ). In the rat, HSL is only located in the seminiferous tubules (Sertoli and spermatogenic cells) ( 34 ), whereas in the normal human testis, HSL is detected in seminiferous tubules and Leydig cells ( 35 ).
The role of HSL in testis was revealed by the phenotype of HSL-defi cient mice ( 4,36 ). Male mice homozygous for the mutant allele (HSL Ϫ / Ϫ ) are sterile and show abnormalities in spermatogenesis, resulting in profound alterations in spermatid maturation and oligospermia ( 4 ). Because HSL catalyzes the hydrolysis of CE, it is reasonable to speculate that cholesterol released by the action of HSL is required for spermatogenesis. The lack of HSL is accompanied by a 2-to 4-fold increase in the accumulation of CE in the testis ( 4,36 ) and decreased testis weights ( 4 ). The expression of the testicular form of HSL (HSLtes) in postmeiotic germ cells rescues the male infertility observed in HSLdefi cient mice ( 37 ). These results and the presence of HSL in spermatids ( 4,35 ) revealed the rate-limiting role of HSLtes and the importance of CEH activity in the action of HSL on spermatogenesis and, more specifi cally, on spermiogenesis.
In membranes, cholesterol is concentrated in lipid rafts, which are also enriched in sphingolipids ( 38 ). Lipid rafts serve as major platforms for the initiation, propagation, and maintenance of signal transduction events ( 39 ). Thus, it is not surprising that dysfunction in cholesterol synthesis, storage, transport, and removal in mice testis may lead to infertility. Many key signaling molecules have been shown to function via lipid rafts; examples of such molecules are SR-BI and SR-BII. Several authors reported that these receptors localize to cholesterol-rich plasma membrane caveolae ( 40,41 ), which infl uence cholesterol homeostasis by acting as conduits for cellular fl ux ( 38 ) and which concentrate signaling proteins for signal transduction ( 39 ). In this regard, the MAPK signaling pathway has been shown to be activated when SR-BI is bound by its ligand ( 42,43 ). However, in different cell types, Peng et al. ( 44 ) showed that SR-BI is localized primarily in clusters on microvillar extensions of the plasma membrane, whereas it is nearly absent from caveolae.
The main objective of the present work was to study the changes in the expression of SR-B (SR-BI, SR-BII, and LIMP II), as well as their association with lipid rafts and relevant signaling pathways in HSL Ϫ / Ϫ mouse testis, to gain insight on the mechanisms mediating the alterations in spermatogenesis and, ultimately, infertility in these animals.
Considerable amounts of cholesterol are present inside seminiferous tubules ( 11 ). The cholesterol used for steroidogenesis or spermatogenesis is derived from different sources ( 6 ), as follows: de novo cholesterol synthesis; mobilization of stored cholesterol esters (CE) via the action of neutral cholesteryl ester hydrolase activities (CEH or hormone-sensitive lipase [HSL]); lipoprotein-derived cholesterol via the LDL receptor-mediated endocytic pathway; and class B scavenger receptor (SR-B) (SR-BI, SR-BII, and LIMP II)-mediated "selective" HDL cholesteryl ester uptake pathway. Sertoli cells possess the capacity to synthesize cholesterol from acetate in vitro ( 12 ), but it seems unlikely that this source is suffi cient to sustain steroidogenesis in vivo. Actually, at least in rodents, it appears that HDL is the major source of cholesterol for Sertoli cells ( 13 ).
A striking correlation has been demonstrated between the enhanced lipid uptake from HDL, promoted by the induction of SR-BI expression and steroid biosynthesis in steroidogenic cells (14)(15)(16). The vast majority of lipoproteinderived cholesterol utilized for murine steroidogenesis in testes is obtained via SR-B ( 7 ). The SR-B are multiligand membrane proteins highly expressed in tissues with a strong cholesterol demand for steroidogenesis, such as the adrenals, ovaries, and testes ( 14,17 ). SR-B are responsible of the "selective uptake of cholesteryl esters" from HDL by steroidogenic cells (18)(19)(20). In the testis, SR-BI is most abundantly expressed in steroidogenic Leydig cells ( 7,14 ), with lesser amounts in Sertoli cells ( 21 ). Experimental deficiency of SR-BI causes abnormalities in the development of female gametes. A signifi cant proportion of ovulated oocytes die soon after ovulation, all resulting in sterility in SR-BI gene-knockout (KO) female mice ( 22 ).
HSL is an intracellular neutral lipase that hydrolyzes triacylglycerols (TAG), diacylglycerols (DAG), monoacylglycerols (MAG), CE, and retinyl esters (RE) ( 23 ). In fact, the relative maximal hydrolysis rates are in the range of 1:10:1:4:2 for TAG, DAG, MAG, CE, and RE, respectively. These data suggest that HSL has more DAG and CE hydrolase activities than TAG hydrolase activity. HSL is expressed in white and brown adipose tissues, and in several nonadipose tissues, including placenta, macrophages, heart, skeletal muscle, mammary glands, pancreatic ␤ -cells, adrenal glands, ovaries, and testes (24)(25)(26)(27)(28)(29). This multifunctional enzyme has emerged as a key regulator of cholesterol metabolism in steroidogenic tissues (adrenals, ovaries, and testes) ( 30 ) as responsible for the hydrolysis of intracellular CE, including those coming from CE selective uptake mediated by SR-B receptors ( 30 ). The direct interaction of HSL with steroidogeneic acute regulatory protein (StAR) in cytosol increases the hydrolytic activity of HSL and facilitates cholesterol movement from stored lipid droplets to mitochondria for steroidogenesis ( 30,31 ).
The testis expresses several protein species of HSL with apparent molecular masses ranging from 26 to 130 kDa, which are encoded by a single 3.9 kb HSL mRNA in rat 30 min at 14.000 g to obtain a Golgi-free supernatant and disrupted by sonication for three bursts of 30 s each. Equal amounts of membrane proteins were then adjusted to 42.5% sucrose prepared in MBS [25 mM Mes (pH 6.5) and 0.15 M NaCl]. A 5-35% discontinuous sucrose gradient was formed above, and the tube was centrifuged at 190,000 g for 18-20 h in a SW41 rotor (Beckman Instruments, Palo Alto, CA). Twelve 1 ml fractions were collected from the top of the gradient and processed for caveolin-1 (Cav-1) (Santa Cruz Biotechnology, Santa Cruz, CA) and anti-transferrin receptor (TfR; Zymed, Invitrogen, San Francisco, CA) analysis by Western blotting. The fractions were concentrated with 10% trichloroacetic acid precipitation prior to electrophoretic analysis for SR-BI and SR-BII (Novus Biologicals, Littleton, CO) by Western blotting. The sample proteins were measured using a bicinchoninic acid (BCA) protein assay (Pierce Biotechnology).

Western blotting
Western blot analysis was performed as previously described ( 35 ). The samples were homogenized in 10 mM Tris-HCl buffer (pH 7.4) containing 1 mM EDTA, 12 mM 2-mercaptoethanol, 1 mM benzamidine, and 1 mM phenylmethylsulphonyl fl uoride, with the addition of a cocktail of protease inhibitors (10 µg/ml of leupeptin and 1 µg/ml of aprotinin) and phosphatase inhibitors (10 mM sodium fl uoride and 1 mM sodium orthovanadate) in the presence of 0.5% Triton X-100. An equal amount of protein from each gradient fraction or whole-cell lysate (20-50 µg) were subjected to 8-12% SDS-PAGE and transferred to polyvinylidene difl uoride (PVDF) or nitrocellulose membranes (GE Millipore, Bedford, MA). After blocking, the blots were probed with specifi c antibodies followed by incubation with horseradish peroxidaseconjugated secondary antibodies (Santa Cruz Biotechnology). The immunoblots were incubated with the enhanced chemiluminescence reagent (Bio-Rad) and exposed to X-ray fi lm or for equal times on a VersaDoc MP 4000 System (Bio-Rad) and images were obtained. The density of the bands was quantifi ed by using the Quantity One 4.5.2 version program (Bio-Rad). The specifi city of the procedure was assessed by means of negative controls that were performed omitting the primary antibody and incubating only with the secondary antibody at optimal titers. Phospho-Akt-S473, Akt, phospho-p38, and p38 antibodies were obtained from Cell Signaling Biotechnology (Danvers, MA), diphospho-ERK1/2 and ERK1/2 (Sigma-Aldrich, St. Louis, MO), phospho-SRC (BD Biosciences, San Jose, CA), caveolin-1 and GADPH (Santa Cruz Biotechnology), SR-BI and SR-BII (Novus Biologicals, Littleton, CO), HSL (Chemicon, Billerica, MA), and TfR (Zymed, Invitrogen).

Immunohistochemistry
Streptavidin-peroxidase immunostaining was performed, as previously described ( 28,35 ). Briefl y, the sections were incubated for 12 h at 4°C with the primary antibody [anti-SR-BI and SR-BII (1/1000), anti-LIMP II (1/700), and anti-HSL 1:800; chicken polyclonal antiserum directed against HSL from white rat adipose tissue diluted in 0.3% normal goat serum, 0.001% Triton X-100, and 0.01% glycine in TBS (pH 7.6)]. The sections were washed twice in TBS to remove unbound primary antibody, then incubated with the secondary antibody for 1 h at room temperature. The secondary biotin-conjugated antibodies were anti-rabbit IgG (1/500) for SR-BI and SR-BII (Novus Biologicals, Littleton, CO), anti-goat IgG (1/400) for LIMP II (R and D Systems, Minneapolis, MN ) , and goat anti-chicken IgY (1:200) for HSL (Vector Labs, Burlingame, CA), diluted in 0.3% normal rabbit or goat serum, 0.001% Triton X-100, and 0.01% glycine in TBS. Sections were rinsed in TBS and incubated with the streptavidin-peroxidase complex (Zymed Labs, San Francisco CA) for 30 min, and washed were fed ad libitum and had continuous access to tap water. The experimental protocol was approved by the Animal Research Ethics Committee (CEBA) of the Hospital Universitario Ramón y Cajal (Madrid, Spain).
The generation of mice with a targeted disruption of the Lipe gene encoding HSL (HSL Ϫ / Ϫ ) has been reported; these mice have no detectable HSL activity or protein in any tissue, including the testis ( 4,36 ). HSLþ/F1 mice were bred for more than 5 generations into a C57BL/6 background. Given that homozygote males are infertile, to maintain the transgenic mice colony, heterozygote females were mated with heterozygote males. The pups were maintained with the mothers until the end of the lactating phase. Two weeks after weaning, males and females were separated in individual cages. Mice were genotyped by PCR analysis of tail DNA, as described previously ( 4 ).
At least seven mice from each experimental group were analyzed when 4 months old. Mice had free access to food and were euthanized by cervical dislocation between 10:00 and 11:00 h. Whole-animal body and testicular weights were recorded. Mouse testes were removed, one of which was frozen in liquid nitrogen and stored at Ϫ 80°C until processing, and another testis was fi xed in 4% paraformaldehyde for 6-24 h at 4°C. The fi xed tissues were embedded in paraffi n following conventional methods ( 28 ) and sectioned at 5 µm, either to be stained with hematoxylin and eosin or to be used for immunohistochemistry.

Plasma testosterone, total cholesterol, and triglycerides
Plasma testosterone was measured by ELISA with the kit from Cayman Chemical. Total cholesterol and TAG were measured by using the kits Cholesterol-Liquids CHO-PAD and GPO-PAP liquid, respectively, from Bradford Diagnostics (Kemia Científi ca).

Mice sperm collection for spermatozoa counts and motility
After animals were euthanized, cauda epididymis and vasa differentia were retrieved and placed in M2 medium (Sigma Aldrich) after removing all excess adipose tissue and blood vessels to minimize the risk of contamination ( 45 ). Spermatozoa were released from the epididymis by slicing with a 30 gauge needle and left in the medium for 5 min to recover motility.

Determination of sperm characteristics
For the determination of sperm concentration, a Bürker hemocytometer was used. The number of spermatozoa was counted under a light microscope. For determination of motile sperm, semen was loaded into a prewarmed (37°C) slide and placed on the heated (37°C) microscope stage, and the percentages of motile spermatozoa were assessed by a Sperm Class Analyzer (SCA 2002; Microptic, Barcelona, Spain).

Viability assessment of spermatozoa
The proportions of living and dead sperm were assessed by a living-cell nucleic acid stain (SYBR-14), in combination with the conventional dead-cell nucleic acid stain propidium iodide ( 41 ), using the staining protocol of a live/dead sperm viability kit (Molecular Probes, Eugene, OR). At least 500 sperm were counted for each treatment.

Sucrose gradient fractionation of membranes
The separation of membrane fractions was performed as previously reported ( 46 ) with same modifi cations. Frozen testis samples from WT and HSL-KO mice were homogenized in 2 ml of 500 mM sodium carbonate (pH 11) and protease inhibitor cocktail (Pierce Biotechnology, Rockford, IL) using a Dounce homogenizer. Then, the homogenates were centrifuged for the signifi cance of difference was assessed using a paired Student t -test. Data were considered signifi cant at a P < 0.05 (* P < 0.05, ** P < 0.01, and *** P < 0.001).

Body weight and plasma levels of lipids and testosterone in wild-type (HSL +/+ ) and HSL KO (HSL
HSL is a lipolytic enzyme, and the lack of this protein could affect body weight. We observed that body weight was signifi cantly reduced (27%) in adult HSL Ϫ / Ϫ mice compared with HSL +/+ mice ( Fig. 1A ). No differences in plasma testosterone concentration between the two groups of animals were found ( Fig. 1B ). Plasma total cholesterol levels were signifi cantly higher in HSL Ϫ / Ϫ mice than in controls, whereas the triglyceride concentration did not differ between groups ( Fig. 1C ).

Testis weight, number of spermatozoa, motility per epididymidis, and histological study in wild-type (HSL +/+ ) and HSL-KO (HSL ؊ / ؊ ) mice at 16 weeks of age
The testis weight was reduced in HSL Ϫ / Ϫ mice compared with HSL +/+ mice ( Fig. 2A ). HSL Ϫ / Ϫ mice are sterile, and we sought to know whether the reduction in the testis weight of the HSL Ϫ / Ϫ mice has an effect in the number and motility of the spermatozoids. Sperm counts in epididymis preparations revealed that the number of spermatozoa per epididymis in HSL Ϫ / Ϫ mice was 1.5 ± 1.8 spermatozoids/ml and 16.6 ± 1.2 × 10 6 spermatozoids/ml in HSL +/+ mice. HSL Ϫ / Ϫ spermatozoa in the epididymis were not motile and were dead, whereas 50% of the spermatozoa from HSL +/+ mice were motile and 60% were alive ( Fig. 2A ). In summary, HSL Ϫ / Ϫ mouse testes have a very low number of spermatozoids, and the spermatozoids were not motile. Testis histology after hematoxylin/eosin staining showed that adult HSL Ϫ / Ϫ mouse testis (16 weeks) were atrophic ( Fig. 2B ). Semi-thin cross-sections of testes revealed a 40% decrease in the average diameter of seminiferous tubules (ST), exhibiting reduction and disorganization of the multilayered seminiferous epithelium. The thickness of the epithelial layer of the HSL Ϫ / Ϫ seminiferous tubules was greatly reduced, from 12 to 5-7 layers. Together with the complete absence of mature spermatozoa, the number of mature spermatids was also markedly reduced. Extensive vacuolation was observed in the epithelial cells. The vacuoles may contain CE because it has been described that CE content is increased 2.3-fold in HSL Ϫ / Ϫ mice ( 4 ). The intertubular space of the HSL Ϫ / Ϫ testis was occupied by proliferating Leydig cells (LC) in clusters of all developmental stages ( Fig. 2B ). We observed an increased number of LCs in HSL Ϫ / Ϫ testis mice compared with HSL +/+ mice. In HSL +/+ , the lamina propia of the ST is composed of a continuous layer of fl at elongated myoid cells (MC) separated by a rather thin basal lamina from ST and the endothelium of lymphatic capillaries. In HSL Ϫ / Ϫ mouse testis, the lamina propia increases in thickness and proliferation of myoid cells correlates with thickening of basement membranes. in TBS followed by Tris-HCl buffer (pH 7.6). Peroxidase activity was developed using 3-3 ′ diaminobenzidine tetrahydrochloride (DAB) as the chromogen (Sigma-Aldrich). The sections were counterstained with Carazzi ′ s hematoxylin. Thereafter, the sections were dehydrated in ethanol, cleared with xylene, mounted in DePeX, and observed under a light microscope.
The specifi city of the immunohistochemical procedures was assessed by means of negative controls that were performed as follows: omitting the primary antibodies; using nonimmune serum instead of the primary antibodies; and incubating with an inappropriate secondary antibody after the incubation with the primary antibodies at optimal titers.
The staining intensity was qualitatively evaluated and classifi ed as absent ( Ϫ ), weak (+), moderate (++), or intense (+++). The assessment of the grade of staining was performed in a blinded way by two independent investigators under a high-power fi eld (×400) using standard light microscopy. When there was a disagreement between observers on scores, the observers were blinded to the last score and consensus judgement was reached through discussion.

Neutral lipid staining with Oil Red O
Tissue sections were immersed in isopropanol for 3-5 s and incubated in a solution of 5 g Oil Red O/100 ml of isopropanol diluted 2/3 in H 2 O for 20-30 min in dark at room temperature. Subsequently, the sections of tissues were washed with isopropyl alcohol, then with water. The sections were contrasted with Carazzi ′ s hematoxylin (Panreac) and embedded for 10 min, washed with distilled water, and mounted in aqueous medium (glycerol/PBS). The slides were visualized on a confocal microscope (Nikon C1 Eclipse Ti-e, Japan).

Free cholesterol staining with fi lipin
Tissue sections were incubated in a solution of fi lipin (Sigma-Aldrich) (0.1 µg/µl) during 1 h in dark at room temperature. Subsequently, the sections of tissues were washed three times with PBS and mounted in NonFade medium (Invitrogen). The slides were visualized on a confocal microscope (Nikon C1 Eclipse Ti-e, Japan).

RNA isolation and real-time quantitative RT-PCR
Total RNA from testis was extracted using TriReagent (Sigma) and reverse transcribed using M-MLV reverse transcriptase enzyme (Promega, WI) in the presence of the RNase inhibitor RNAsin (2 U/µl) (Promega). Real-time quantitative PCR (qRT-PCR) amplifi cation was performed on a LightCycler 480 using the SYBR Green I Master kit (Roche Applied Science). The thermocycle protocol was 95°C for 5 min, followed by 45 cycles of 95°C for 10 s, 60°C for 10 s, and 72°C for 10 s. We evaluated the melting curves for each gene and separated the PCR reaction products on 2% agarose gels to confi rm the presence of a single product. We analyzed the effi ciency of the reaction by amplifying serial dilutions of cDNA (1:10, 1:100, 1:1000, and 1:10000). We ensured that the relationship between the threshold cycle (Ct) and the log (RNA) was linear ( Ϫ 3.6 < slope < 3.2). All analyses were performed in triplicate, and the target gene copy number was normalized against the housekeeping gene CypB (encoding cyclophilin B). The primers used in the real-time PCR are shown in supplementary Table I.

Statistical analysis
Data are presented as the mean ± SEM unless otherwise stated. All statistical analyses were performed with Graph Pad Prism, version 4 (GraphPad Software, San Diego, CA). For comparison between WT and KO animals, two-way ANOVA was conducted and whether the location or quantity was affected by the lack of HSL.

Free-cholesterol staining with fi lipin and neutral-lipids staining with Oil Red O of wild-type (HSL +/+ ) and HSL KO (HSL ؊ / ؊ ) mouse testis
We observed positive staining of free cholesterol with fi lipin ( Fig. 3A ) and neutral lipids with Oil Red O ( Fig. 3B ) in the Leydig cells from HSL +/+ and HSL Ϫ / Ϫ mouse testis, seemingly more abundant in the testes that lack HSL, which indicates an accumulation of neutral lipids in the Leydig cells of mice lacking the HSL compared with controls. No staining was noted in any of the negative controls (data not shown). HSL +/+ mouse testis showed HSL-positive staining in the cytoplasm of the Leydig cells and in the elongated spermatids of the ST. However, as expected, HSL Ϫ / Ϫ mouse testis was immunonegative for HSL.
We then analyzed the protein expression by Western blotting in total lysates of mouse testes. HSL +/+ mouse testis showed the main immunoreactive HSL protein to be 130 kDa in this tissue (supplementary Fig. I). HSL Ϫ / Ϫ mouse testis showed no immunoreactive HSL protein (supplementary Fig. I).
We then studied the expression and cellular localization of key proteins in lipid metabolism of the testis and the histologic analysis of neutral lipids in the testis to determine  mouse testis, SR-BI was localized to the surfaces of Leydig and spermatid cells. SR-BII was localized to the membranes of germinal cells (spermatocytes and spermatids) from the ST, Leydig and Sertoli cells were immunonegative for SR-BII. LIMP II was mainly localized in the membranes of the Sertoli cells and less so in Leydig cells. The lack of HSL in mouse testis induces an increased SR-B expression that could contribute to increase CE uptake from the plasma lipoproteins by these receptors.

Class B scavenger receptor distribution onto lipid raft and nonraft plasma membrane microdomains and signaling cascades activation in wild-type (HSL +/+ ) and HSL-KO (HSL ؊ / ؊ ) mouse testis
To determine whether the decreased availability of free cholesterol in HSL-KO (HSL Ϫ / Ϫ ) mouse testis could compromise the amount of cholesterol in the plasma membrane and, specifi cally, the formation and functionality of the lipid raft structures, we isolated plasma membrane fractions from mouse testis in a sucrose density gradient analyzing protein expression by Western blot. In HSL +/+ mouse testis, immunoreactive caveolin-1 was mainly detected in fractions 2 and 3, indicating that these fractions contain lipid rafts, whereas TfR, a marker of the nonraft membrane domain, was only detected in fractions 8-11 ( Fig. 5A ). The lack of HSL in mouse testis resulted in a signifi cant decrease in caveolin-1 content in lipid raft fractions, whereas there was an increase in caveolin-1 in nonraft fractions compared with the control ( Fig. 5A ). Thus, HSL led to a shift in the distribution of caveolin-1 from lipid rafts to denser membrane domains.
We then studied whether the lipid raft disruption affected or altered SR-B (SR-BI and SR-BII) plasma membrane localization and expression. In the testes of the control (HSL +/+ ) and HSL Ϫ / Ϫ mice, the mature form of SR-BI (82 kDa) was present in both plasma membrane domains (lipid raft and nonraft; Fig. 5B ). In contrast, the immature form of SR-BI (57 kDa) was localized only in nonraft plasma membrane fractions. No differences in the localization of both forms in the plasma membrane fractions between the two groups of animals were found ( Fig. 5B ).
In HSL Ϫ / Ϫ mice, we observed an increase in the expression of the mature and immature forms of the SR-BI receptor in both microdomains compared with controls ( Fig. 5B ). SR-BII was present in both plasma membrane domains ( Fig. 5C ), and there were no differences in the distribution or expression in the testes of HSL Ϫ / Ϫ mice compared with controls.
It has been reported that the binding of HDL to SR-BI activates the signaling cascades of the SRC family kinases [phosphatidylinositol 3-kinase (PI3K)/AKT and MAPK] in endothelial cells ( 47 ). Thus, we next sought to determine whether these signaling pathways were affected by the lack of HSL. The testes of HSL +/+ mice expressed total and phosphorylated forms of the proteins ERK1/2, AKT, and Fyn (SRC) ( Fig. 6 ). There was an increase in the expression of p-ERK, p-AKT, and p-SRC in HSL Ϫ / Ϫ mouse testis lysates versus HSL +/+ mouse testis, without any change in the total form of the proteins studied ( Fig. 6 ).
of the Leydig cells and in some spermatids. HSL Ϫ / Ϫ testis shows a great increase in SR-BI-positive staining in the Leydig cells ( Fig. 4A ). To confi rm these results, SR-BI protein levels were determined by Western blotting. HSL +/+ testis expressed two SR-BI immunoreactive proteins. The apparent molecular masses of the inmunoreactive bands were 82 kDa (mature protein, intense band) and 57 kDa (immature protein, light band). HSL Ϫ / Ϫ testis showed an increased intensity of these two SR-BI proteins compared with HSL +/+ mouse testis ( Fig. 4A ).
The results of SR-BII localization and protein expression in HSL-WT (HSL +/+ ) and HSL-KO (HSL Ϫ / Ϫ ) mouse testis are shown in Fig. 4B . HSL +/+ mouse testis showed SR-BII-positive staining in the membrane of the germinal cells (spermatocytes and round spermatids) and negative staining in elongated spermatids, and in Sertoli and Leydig cells. HSL Ϫ / Ϫ testis showed increased SR-BII-positive staining in the membranes of the germinal cells from the ST. By Western blotting, an increase in the expression of SR-BII (82 kDa) was observed in HSL Ϫ / Ϫ mouse testis ( Fig. 4B ).
LIMP-II staining was positive in the membrane of the Sertoli cells and slightly more in Leydig cells from HSL +/+ mouse testis ( Fig. 4C ). HSL Ϫ / Ϫ testis showed increased LIMP-II-positive staining in Sertoli and Leydig cells compared with HSL +/+ mouse testis ( Fig. 4C ).
Finally, we measured the SR-B mRNA levels. SR-BII and LIMP-II mRNA expression was signifi cantly increased in HSL Ϫ / Ϫ mouse testis compared with HSL +/+ mouse testis ( Fig. 4D ). No changes were observed in SR-BI mRNA levels among the studied groups.
Taken together, these results indicate a specifi c cell-type localization of the SR-B in mouse testis. Thus, in HSL +/+ ( 4, 36 ). The absence of CEH activity in HSL Ϫ / Ϫ mouse testis may induce a decrease in the hydrolysis of CE that causes a reduction in the availability of free cholesterol, which is involved in cell proliferation ( 48,49 ) and, therefore, in spermatogenesis. The aim of the present work was to study the SR-B cellular expression in HSL Ϫ / Ϫ mouse testis for their potential importance in cholesterol provision, as well as their localization to lipid rafts and the activity of related signaling cascades, which ultimately may be involved in infertility in HSL Ϫ / Ϫ male mice.
In the present study, we found profound alterations in the morphology of HSL Ϫ / Ϫ mouse testis, affecting both the intertubular space, with increased number of Leydig and myoid cells, and the seminiferous tube, with disrupted epithelium and reduced number of mature spermatids, all These results show that HSL Ϫ / Ϫ mouse testis has increased expression of the SR-BI mature isoform in plasma membrane lipid raft fractions and increased expression of the SR-BI immature isoform in the plasma membrane nonraft fractions with respect to the HSL +/+ mouse testis. This could be associated with the activation of different signaling pathways (AKT, ERK, and SRC) mediated by the SR-B.

DISCUSSION
The absence of HSL in testes results in severe oligospermia and infertility, underscoring the importance of this enzyme in spermatogenesis ( 4,36 ). The lack of HSL disturbs tissue lipid composition and cholesterol metabolism of free cholesterol for steroidogenesis may be operating in the HSL Ϫ / Ϫ mouse, including the endogenous cholesterol biosynthesis and the hydrolysis of lipoprotein-derived CE within the lysosomes. The lack of HSL resulted in the augmented deposit of lipid droplets observed in Leydig cells, confi rming the inability of HSL-lacking cells to utilize esterifi ed lipids. This observation is in accordance with previous results by others showing the accumulation of both CE ( 4 ) and DAG ( 51 ) in the HSL Ϫ / Ϫ testis. The absence of CE hydrolase activity may compromise the availability of free cholesterol, which is required for cell proliferation ( 48,49 ) and, therefore, for spermatogenesis.
In the testis, the main pathway to obtain cholesterol is the selective uptake of CE from HDL mediated by SR-BI ( 18,20,55 ). The SR-BI receptor also mediates the bidirectional fl ow of unesterifi ed cholesterol and phospholipids between HDL and cells ( 56 ). SR-BII, which is an isoform of the SR-BI gene formed by alternative splicing of the transcript ( 41 ), also binds HDL with high affi nity, but it confi rming previous results reported by others ( 4,36 ). These morphological changes were accompanied by severe oligospermia, with the decreased sperm count and reduced sperm motility shown by HSL Ϫ / Ϫ male mice (Refs. 4, 50, 51 and present results). The acquisition of sperm motility and fertility have been associated with changes in cholesterol content of male gametes during the transit through the epididymis ( 52 ) and during capacitation ( 53 ). Inadequate cholesterol levels in the sperm apparently cause infertility in humans ( 54 ).
Testes from HSL +/+ mice mainly express the immunoreactive HSL protein of 130 kDa. We found that HSL is located in Leydig cells, which suggests that HSL may be involved in CE hydrolysis to provide free cholesterol for the synthesis of steroid hormones in physiological conditions. However, oligospermia in the HSL Ϫ / Ϫ male mouse does not appear to result from hypogonadism. Actually, HSL Ϫ / Ϫ mice have normal plasma levels of folliclestimulating hormone, luteinizing hormone ( 4,51 ), and testosterone (Refs. 4, 51 and present results). Other sources ( 61 ) described that the induction of SR-BI protein expression by both tamoxifen and 4-hydroxytamoxifen was independent of transcription. Actually, these authors demonstrated that SR-BI mRNA levels and promoter activity were not infl uenced by any of these selective estrogenreceptor modulators; instead, these drugs increased the stability of SR-BI protein ( 61 ). It may be speculated that HSL defi ciency also increases SR-BI protein stability. SR-BII is located in the membranes of germinal cells (spermatocytes and spermatids) from the ST (Ref. 60 and present results), whereas Leydig and Sertoli cells were found to be immunonegative for SR-BII. Mice lacking HSL showed an increased expression of SR-BII in testicular cells as observed by immunohistochemistry, which corresponded with augmented SR-BII protein and mRNA levels in the whole tissue. Finally, LIMP II was located in the membranes of Sertoli and Leydig cells, and similar to the other SR-B, testes from HSL Ϫ / Ϫ mice had increased expression of LIMP II as indicated by the increased staining in both cell types (Sertoli and Leydig cells) as well as by the higher mRNA levels in HSL Ϫ / Ϫ mice testis compared with the WT. Taken together, the results show an increased expression of SR-B (SR-BI, SR-BII, and LIMP II) in testes in the HSL Ϫ / Ϫ mouse, which is likely a response to the inability to obtain free cholesterol from CE due to the lack of HSL. This response certainly might facilitate the entry of HDL-CE into the cells, contributing to the intracellular accumulation of CE rather than to alleviating the cholesterol demands.
One of the fates of cellular cholesterol is the formation of lipid rafts structures. We thus studied the distribution of caveolin-1 and SR-B in membrane fractions of HSL Ϫ / Ϫ mouse testis. The HSL defi ciency in mouse testis resulted in an alteration in lipid raft architecture, as evidenced by the shift of caveolin-1 to denser fractions. It has been reported that the reduction of free cholesterol availability impairs the formation of lipid rafts, leading to a redistribution of caveolin-1 to nonraft fractions ( 46,62 ). Therefore, it is suggested that the lack of HSL results in the impairment of lipid-raft/caveolae formation, likely due to the defi ciency of free cholesterol.
We found that SR-BI was distributed onto both raft and nonraft fractions from testis plasma membranes. In other cell types, others reported that SR-BI is associated with caveolae. Thus, in steroidogenic cells, Babitt et al. ( 40 ) observed that SR-BI copurifi ed with caveolin-1 as isolated by density gradient fractionation and that SR-BI colocalized with caveolin-1 in punctuate microdomains across the cell surface by immunofl uorescence microscopy. Moreover, by studying the fate of radiolabeled HDL in CHO cells expressing SR-BI, Graf et al. ( 63 ) showed that HDL-CE rapidly associates with caveolin-1-containing fractions and that it is then translocated to an intracellular compartment, thus suggesting that caveolae are acceptors for HDL-derived CE. In contrast, by electron microscopy in different cell types, Peng et al. ( 44 ) concluded that SR-BI is localized primarily in clusters on microvillar extensions of the plasma membrane and that it is nearly absent from caveolae. We did not address the functional site of SR-BI. The coisolation differs from SR-BI in that it mediates HDL endocytosis through a clathrin-dependent, caveolae-independent pathway ( 57 ) and does not mobilize the intracellular stores of CE ( 58 ).
In the present study, in the HSL +/+ control mouse testis, we show that SR-BI is located in the membrane of Leydig cells and on elongated spermatids, which agrees with previous studies in other species ( 57,59,60 ), including humans ( 35 ). Unlike other species ( 35, 60 ), we could not detect SR-BI in Sertoli cells. Of note, testis of HSL Ϫ / Ϫ mice showed an increased expression of SR-BI as determined by immunostaining, particularly in Leydig cells. This was confi rmed by Western blot, showing a high increase in the protein expression of the two forms of SR-BI, the mature (82 kDa) and the immature (57 kDa) forms, in testis from mice lacking HSL. To further analyze the changes in SR-BI expression, we measured mRNA levels by qRT-PCR. Intriguingly, there were no differences in SR-BI mRNA levels between the two groups. This latter result indicates that the increased SR-BI protein expression observed in HSL Ϫ / Ϫ mouse testis is not due to increased transcription. Recently, Dong et al. of the SR-BI mature form (82 kDa) and caveolin-1 in lowdensity membrane fractions does not necessarily indicate that both proteins are within the same morphological structure in the cell. The possibility exists that SR-BI localizes to lipid rafts not containing caveolin-1. Nevertheless, the fact that the immature form of SR-BI (57 kDa) was excluded from these structures, whether lipid rafts or caveolae, fi rmly suggests that they are physiologically relevant in the function of SR-BI in the testis. In plasma membranes from HSL Ϫ / Ϫ mice, the levels of SR-BI, both the mature and the immature forms, were highly increased compared with the WT, which is in accordance with the augmented expression observed in whole tissue; however, the distribution between the raft and nonraft fractions was not altered. Regarding SR-BII, it was localized in both the lighter and, mostly, in the denser plasma membrane fractions considered, and this distribution was not affected by the HSL status.
Lipid rafts/caveolae not only infl uence cholesterol homeostasis by acting as conduits for cellular fl ux ( 38 ) but also concentrate signaling proteins for signal transduction ( 39 ). It has been reported that HDL binding to SR-BI leads to the activation of a tyrosine kinase (SRC), which causes the activation of PI3K and subsequent activation of AKT and MAPK in endothelial cells ( 47 ). We report herein that different signaling pathways (AKT, ERK, and SRC) are activated in the HSL Ϫ / Ϫ mouse testis. The signifi cance of these results is unknown; however, the activation of MAPK cascades is involved in differentiation, proliferation, and other cellular processes during spermatogenesis (64)(65)(66)(67). The exact mechanism through which HSL defi ciency produces male infertility remains elusive. The observed alterations in SR-B expression, lipid raft composition, and signaling cascades, however, may provide some clues. First, infertility of HSL-KO mice is not associated with hypogonadism or adrenal insuffi ciency, as no changes in plasma testosterone levels in HSL-KO were observed (present results and Ref. 4 ). On the other hand, the increased expression of SB-BI and SR-BII in HSL Ϫ / Ϫ testis may favor the uptake of CE from HDL and the activation of cell signaling pathways (ERK, AKT, and SRC) involved in differentiation, proliferation, and other cellular processes during spermatogenesis. In this sense, SR-BI overexpression may be contemplated as an attempt to provide cells with cholesterol. However, due to the lack of HSL, the internalized CE cannot be hydrolyzed, and hence, it accumulates in the cytoplasm, contributing to steatosis. This inadequate supply of free cholesterol could be involved in the observed lipid-raft disruption. Finally, functional cholesterol defi ciency may compromise cell division ( 68 ), thereby interfering with spermatogenesis. Obviously, this is a speculation that requires further studies in vitro with selected cell types to confi rm the proposed mechanisms.
In summary, present results indicate that the lack HSL in HSL Ϫ / Ϫ mice dramatically disrupts cholesterol homeostasis in the testis, with augmented expression of SR-B (SR-BI, SR-BII, and LIMP II), altered lipid-raft microdomains, and activation of different signaling pathways (p-ERK, p-AKT, and p-SRC) with relevance in spermatogenesis.