Molecular Cloning, Genomic Organization, and Expression of a Testicular Isoform of Hormone-Sensitive Lipase

GENOMICS

35, 441–447 (1996) 0383

ARTICLE NO.

Molecular Cloning, Genomic Organization, and Expression of a Testicular Isoform of Hormone-Sensitive Lipase LENA STENSON HOLST,*,1 DOMINIQUE LANGIN,† HINDRIK MULDER,‡ HENRIK LAURELL,* JACQUES GROBER,† ANDERS BERGH,§ HARVEY W. MOHRENWEISER,Ø GUDRUN EDGREN,\ AND CECILIA HOLM* *Section for Molecular Signalling, Department of Cell and Molecular Biology, Lund University, P.O. Box 94, S-22100 Lund, Sweden; †INSERM 317, Institut Louis Bugnard, Faculte´ de Me´decine, Universite´ Paul Sabatier, Toulouse, France; ‡Department of Physiological Sciences, Lund University, Lund, Sweden; §Department of Pathology, University of Umea˚, Umea˚, Sweden; ØHuman Genome Center, Biology and Biotechnology Research Program, Lawrence Livermore National Laboratory, Livermore, California 94550; and \Section for Cell and Matrix Biology, Department of Cell and Molecular Biology, Lund University, Sweden Received February 1, 1996; accepted May 7, 1996

By catalyzing the rate-limiting step in adipose tissue lipolysis, hormone-sensitive lipase (HSL) is an important regulator of energy homeostasis. The role and importance of HSL in tissues other than adipose are poorly understood. We report here the cloning and expression of a testicular isoform, designated HSLtes . Due to an addition of amino acids at the NH2-termini, rat and human HSLtes consist of 1068 and 1076 amino acids, respectively, compared to the 768 and 775 amino acids, respectively, of the adipocyte isoform (HSLadi). A novel exon of 1.2 kb, encoding the human testis-specific amino acids, was isolated and mapped to the HSL gene, 16 kb upstream of the exons encoding HSLadi . The transcribed mRNA of 3.9 kb was specifically expressed in testis. No significant similarity with other known proteins was found for the testis-specific sequence. The amino acid composition differs from the HSLadi sequence, with a notable hydrophilic character and a high content of prolines and glutamines. COS cells, transfected by the 3.9-kb human testis cDNA, expressed a protein of the expected molecular mass (Mr É 120,000) that exhibited catalytic activity similar to that of HSLadi . Immunocytochemistry localized HSL to elongating spermatids and spermatozoa; HSL was not detected in interstitial cells. q 1996 Academic Press, Inc. INTRODUCTION

Hormone-sensitive lipase (HSL)2 is expressed at high levels in adipocytes, where it catalyzes the rate-limThe HSLtes sequences reported in this paper are deposited in the GenBank data base under Accession nos. U40001 (rat) and U40002 (human). 1 To whom correspondence should be addressed. Telephone: /46462228587. Fax: /46462224022. 2 Abbreviations used: HSL, hormone-sensitive lipase; PCR, polymerase chain reaction; RACE, rapid amplification of cDNA ends; ECL, enhanced chemiluminiscence; PAGE, polyacrylamide gel electrophoresis; kb, kilobases; nt, nucleotides; bp, basepairs; HSL-LI, hormone-sensitive lipase-like immunoreactivity.

iting step in lipolysis (Fredrikson et al., 1981). Due to the importance of free fatty acids as energy substrates, HSL is a critical regulator of energy homeostasis in human and other mammals. In tissues other than adipose, the role and importance of HSL are not fully understood. Significant levels of HSL have been found in steroid-producing tissues, whereas heart and skeletal muscle express low levels of the enzyme (Holm et al., 1987). Interestingly, the size of HSL mRNAs is variable. In rat, heart, skeletal muscle, placenta, and ovaries express slightly larger HSL mRNAs (3.5 kb) than adipose tissue (3.3 kb). Testis is characterized by expression of a large mRNA species of 3.9 kb (Holm et al., 1988a; Stenson Holst et al., 1994) and a protein of 130 kDa (Stenson Holst et al., 1994), compared to the 84-kDa adipose tissue protein. Since HSL was found to be identical with the neutral cholesterol ester hydrolase of the adrenal glands and the ovaries (Cook et al., 1982, 1983), it was hypothesized that HSL could function, in steroidogenic tissues, as a cholesterol ester hydrolase to release free cholesterol for steroid hormone biosynthesis (Cook et al., 1982). However, we have shown that, in rat, testis HSL mRNA is expressed in the seminiferous tubuli and not in steroid-producing Leydig cells (Stenson Holst et al., 1994). cDNAs for HSL have been obtained from rat (Holm et al., 1988b), mouse (Li et al., 1994), and human (Langin et al., 1993) adipose tissues. Human and rat HSL show 83% identity at the amino acid level, while the mouse and rat enzymes share 94% identity. The organization of the mouse and human HSL genes is remarkably similar. Both genes span 10–11 kb and contain nine exons encoding the adipocyte protein. The human gene, designated LIPE, has been mapped to chromosome 19q13.1–q13.2 (Schonk et al., 1990; Levitt et al., 1995; Laurell et al., 1995). It is clear from previous data that testicular HSL differs from that of adipose tissue with regard to struc-

441

AID

Genom 4227

/

6r1b$$$381

06-27-96 22:31:02

0888-7543/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

gnma

AP: Genomics

442

STENSON HOLST ET AL.

ture, regulation during development (Stenson Holst et al., 1994; Kraemer et al., 1991), and possibly also function. In this study, we show that testicular HSL is expressed as a distinct isoform to the adipocyte HSL. The adipocyte and testicular isoforms of HSL are referred to as HSLadi and HSLtes , respectively, and we propose these abbreviations for future use. MATERIALS AND METHODS Rapid amplification of cDNA ends. 5*-ends of human and rat testis HSL cDNAs were obtained using a 5*-RACE kit (Life Technologies). Poly (A)/ RNA was either from Clontech (human testis) or isolated from total RNA (rat testis) as described (Chomczynski and Sacchi, 1987). Primers from the human cDNA (Langin et al., 1993) were the following: sequence complementary to nt 693–713 for reverse transcription and nt 627–646 for amplification together with the anchoring primer included in the kit. Primers designed from rat HSL cDNA (Holm et al., 1988b) were sequences complementary to nt 790–815 for reverse transcription and to nt 764–790 for PCR. RACE products were subcloned in the pCR-Script SK(/) vector (Stratagene). In addition, the Marathon RACE kit (Clontech) was used to amplify various regions of human adipose tissue and testis cDNAs and the 5*-end of human testis HSL cDNA. QUICK-clone cDNAs (Clontech) were used for PCR amplification of different regions of rat adipose tissue and testis HSL. Sequence determination of both strands was carried out using either Sequenase (US Biochemicals Corp.) or the Taq Dyedeoxy Terminator Cycle Sequencing kit and a Model 373A DNA sequencer (Applied Biosystems). cDNA library screening. Two testis lgt11 cDNA libraries (Clontech) were screened on nitrocellulose filters, using as probes 32Plabeled (Feinberg and Vogelstein, 1983) full-length HSL cDNAs from either human or rat adipose tissue or an 850-bp testis-specific human HSL cDNA. The latter was generated by PCR using the oligonucleotide 5* CATGAGGAATCAATGAGAG 3* as the sense primer and 5* CATCGTGGCTGGAGAATCT 3* as the antisense primer. Hybridization was performed under standard conditions (Sambrook et al., 1989). Stringent washes were carried out at 607C in 0.21 SSC and 0.1% SDS. Positive clones were subcloned into pBluescript SK (Stratagene). Mapping of a testis-specific exon into human genomic DNA. Cosmids mapping to chromosome 19 region q13.1–q13.2 were analyzed for the presence of the HSL gene (Levitt et al., 1995). Fragments obtained by digestion with EcoRI, BamHI, or XhoI were analyzed by Southern blot analysis using a full-length human adipose tissue HSL cDNA or the 850-bp testis-specific fragment (see above) as probes. Hybridization was performed in Quik-hybrid solution (Stratagene) according to the manufacturer’s instructions. Hybridizing fragments were isolated and subcloned into pBluescript, then further digested until a complete map was obtained. Northern blot analysis. Human poly(A)/ RNAs (Clontech or Promega) were electrophoresed and blotted as described (Stenson Holst et al., 1994). The blots were probed with either a full-length human adipose tissue HSL cDNA or an 850-bp cDNA specific for human testis HSL. Tissue sources and enzyme preparation. Rat tissues were obtained from adult Sprague–Dawley males (B&K Universal, Stockholm, Sweden). Human testes were ablatios from prostate cancer patients, kindly provided by Dr. Arne M. Olsson at the University Hospital, Lund, Sweden. Human subcutaneous adipose tissue was obtained from patients undergoing elective surgery at the University Hospital, Lund, Sweden. Biopsies from four different patients were pooled for the analysis described in this report. Tissues were homogenized as described (Holm et al., 1987) using a glass homogenizer (adipose tissue) or a Polytron Pt 10-35. Fat-

AID

Genom 4227

/

6r1b$$$381

06-27-96 22:31:02

depleted infranatants were obtained by centrifugation at 110,000g for 60 min at 47C. Enzyme assays. Diacylglycerol lipase and cholesteryl ester hydrolase activities were measured as described (Tornqvist et al., 1978; Fredrikson et al., 1981). Immunoinhibition experiments were performed by preincubation with 10-fold diluted polyclonal chicken antiHSL (whole plasma), directed against recombinant rat HSL (Holm et al., 1994), for 60 min at 377C. In vitro translation. A full-length testis HSL cDNA was constructed using testis-specific sequence from nt /104 through adipose tissue sequence corresponding to nt 1040 in exon 1 (Langin et al., 1993), subcloned into pBluescript. The insert was digested using BglII, which cuts in the testis-specific 5*-untranslated sequence, and KpnI, uniquely cutting adipose tissue cDNA in the exon 1-derived region. The resulting 1200-bp piece was ligated into a BamHI/KpnIdigested expression vector pcDNA3 (Invitrogen), containing a human HSL–adipose tissue cDNA construct, to obtain pcDNA3/HSLtes . The pcDNA3/HSLadi construct, containing full-length coding sequence and 6 nt of the 5*-noncoding sequence, was used to translate the adipose tissue protein. A total of 1.5 mg of each plasmid construct was used for transcription/translation with the Single Tube Protein system (Novagen) according to the manufacturer’s instructions. Expression in COS cells. COS cells on 60-mm dishes were transfected with 5 mg of the pcDNA3/HSL constructs using Lipofectin (Life Technologies) as described (Holm et al., 1991). Immunoprecipitation and Western blot analysis. Aliquots of fatdepleted tissue infranatants and COS cell homogenates, corresponding to approximately 25 mU of anti-HSL-inhibited diacylglycerol lipase activity, were incubated with affinity-purified polyclonal chicken anti-HSL antibodies for 2 h at room temperature, followed by incubation with immobilized anti-chicken IgY (Promega) for 1 h at 07C. For competition experiments, 200 ng of homogeneous recombinant rat HSL was included in the incubations. After washing the precipitates five times with 1 ml of 10 mM Tris–HCl, pH 7.4, 0.9% NaCl, they were subjected to SDS–PAGE and electroblotting to nitrocellulose membranes. Western blot analysis was performed using the ECL system (Amersham) with chicken anti-HSL as primary antibody and a horseradish peroxidase-conjugated anti-chicken IgG (Sigma) as secondary antibody. Immunocytochemistry. Testes and epididymi from adult rats were investigated with indirect immunofluorescence as previously described (Mulder et al., 1993), using affinity-purified chicken antiHSL antibodies (dilution 1:640). Preabsorption of the HSL antibodies with homogeneous HSL protein (100 mg/ml in antibody solution at working dilution) blocked the immunoreaction.

RESULTS

Cloning of testis-specific 5*-end HSL cDNA sequences. Using rat testis and adipocyte cDNAs as templates, PCR amplifications of various regions downstream of the HSLadi translation start site generated products identical in size (data not shown). We then employed the 5*-RACE procedure and obtained PCR products of 1.3 kb (rat testis) and 1.2 kb (human testis). The 3*ends of these sequences were identical to those of the previously published adipocyte cDNAs (Holm et al., 1988; Langin et al., 1993) from 20 nt upstream of the translation initiation codon. In-frame ATG codons were identified in the testis-specific sequences, followed by 900 (rat) and 903 (human) coding nt in frame with the adipose tissue sequences. Thus, rat HSLtes is a 1068amino-acid-long protein with a calculated molecular mass of 116,811 Da, and the human equivalent has 1076 amino acids and a molecular mass of 116,533 Da.

gnma

AP: Genomics

A TESTICULAR ISOFORM OF HORMONE-SENSITIVE LIPASE

443

FIG. 1. Regions of homology in the testis-specific parts of human and rat testicular HSL. An alignment of the amino acids encoding the testis-specific regions and the first 47 amino acids of HSLadi is shown. The arrow indicates the HSLadi start codon. The underlined amino acids upstream of the arrow are encoded from nucleotides of the 5*-untranslated region of HSLadi mRNA. The single-letter code for amino acids is given, and identical amino acids are boxed. The identity of the testis-specific part is 37%.

Features of the testis-specific sequence. None of the two testis-specific sequences showed sequence similarity with HSLadi or significant similarities with other known proteins (Swiss-Prot EMBL, January, 1996). The identity between the testis-specific sequences was only 37% (Fig. 1), although the rat and human HSLadi show an identity of 83% at the amino acid level. The testis-specific sequences are markedly hydrophilic compared to the rest of the protein, as predicted from hydropathy profiles according to Kyte and Doolittle (1982) (data not shown). The most remarkable feature of the testis-specific polypeptides is their high content in prolines and glutamines, most notable in the first half of the human sequence, where these residues constitute 16.7 and 18.0%, respectively. Most of them are conserved between rat and human. In searches of the Swiss-Prot PROSITE database (January, 1996), no motifs, consensus sequences, or protein localization signals were found in any of the testis-specific sequences, except for a potential motif for binding to SH3 domains (GPGEPPPAQQ in the human sequence and GPAEPPPATE in the rat sequence). Secondary structure predictions according to Predict-Protein (EMBL, Heidelberg) suggest that the testis-specific NH2-terminal part consists mostly of loop structure, interrupted by some helical regions. None of the helices were predicted to be membrane-spanning or amphipathic. Characterization of the sequence shared by HSLtes and HSLadi . Sequence analysis of a number of cDNA clones from two human testis cDNA libraries revealed no differences between testis and adipose sequence

AID

Genom 4227

/

6r1b$$$381

06-27-96 22:31:02

downstream of the adipocyte translation start site. These results were also verified using cDNAs prepared from adipocyte and testis poly(A)/ RNAs for PCR amplification with primers located at various positions from the testis-specific sequence to exon 9 (data not shown). Isolation of a testis-specific exon from human genomic DNA. Five cosmids containing overlapping DNA sequences, which have been mapped to 19cent.q13.1– q13.2 (Laurell et al., 1995), were hybridized with HSL cDNA probes. One cosmid contained the previously localized HSL exons (Langin et al., 1993) and an upstream flanking sequence of approximately 20 kb. The organization of exons 1–9 was confirmed using a fulllength adipose tissue cDNA probe. An 850-bp human testis-specific cDNA probe was used to map a testisspecific exon(s). This probe hybridized to two BamHI fragments of 5 and 7 kb (Fig. 2), which were subcloned and sequenced. A unique exon was found to contain the entire testis-specific cDNA sequence. The genomic sequence was identical with the cDNA sequence. The testis-specific exon was located 16 kb upstream of the first exon common to adipocyte and testis HSL. Consensus sequences for 5*-and 3*-splice junctions were found at the 3*-end of the testis exon and at the 5*-end of exon 1, respectively, 20 bp upstream of the translation initiation codon. Expression of HSL mRNA in human tissues. Northern blot analysis of human testis poly(A)/ RNA using a full-length human HSLadi cDNA probe revealed ex-

gnma

AP: Genomics

444

STENSON HOLST ET AL.

FIG. 2. Genomic organization of the human HSL gene including the testis-specific exon. Coding exons are shown as black boxes and untranslated sequences as open boxes above a restriction map (E, EcoRI; X, XhoI; B, BamHI). The novel testis-specific exon is located 16 kb upstream of the nine previously described exons encoding the 3.3-kb adipose tissue HSL mRNA.

pression of two different messages, one of approximately 3.3 kb and a larger mRNA of 3.9 kb (Fig. 3A). The 3.9-kb species was not detected in any of the other tissues analyzed using either an 850-bp human testisspecific cDNA probe (Fig. 3B) or a 20-mer oligonucleotide, complementary to the 5*-untranslated region of the testis-specific exon (data not shown). Expression of testicular HSL protein. Upon immunoprecipitation of fat-depleted infranatants from rat and human testes with an anti-HSL antibody, we isolated proteins with apparent molecular masses of approximately 130 kDa (rat) and 116 kDa (human) (Figs. 4A and 4B). A longer ECL exposure time was needed to visualize the weaker immunoreactivity in human testis (Fig. 4B). In vitro transcription/translation experiments from a human HSLtes cDNA construct revealed a protein of approximately 120 kDa (data not shown). This construct was also expressed in COS cells, and a protein of approximately the same size as that translated in vitro was immunoprecipitated from homogenates of these cells (Fig. 4A). Preincubation of anti-HSL with an excess of recombinant rat adipose

FIG. 3. Northern blot analysis of HSL mRNA in human tissues. One microgram of poly(A)/ RNA from each tissue was electrophoresed in a 2.2 M formaldehyde, 1% agarose gel and blotted onto nylon membrane. (A) The blot hybridized with a 32P-labeled, full-length human adipose tissue HSL cDNA probe. (B) Hybridization of the same blot with an 850-bp human testis-specific cDNA probe (for details see Materials and Methods).

AID

Genom 4227

/

6r1b$$$381

06-27-96 22:31:02

tissue HSL before the immunoprecipitation almost completely abolished the 116- to 130-kDa signals (data not shown). In several transfection experiments, enzyme activity measurements of COS cell homogenates demonstrated that HSLtes had an activity that was similar to HSLadi activity after correction for expression of anti-HSL-immunoprecipitated protein. In agreement with the known substrate specificity of HSL (Fredrikson et al., 1981), the enzyme activity was severalfold lower when cholesterol oleate instead of diacylglycerol was used as substrate, but the relative activities between the two isoforms were the same (data not shown). However, a proper determination of the specific enzyme activity must await the availability of a purified preparation of HSLtes protein. Immunocytochemistry. Using an affinity-purified anti-HSL antibody, HSL-like immunoreactivity (LI) in the rat testis was found exclusively in maturing germ cells of the seminiferous tubules (Fig. 5). Due to the evaginating/invaginating nature of the lateral mem-

FIG. 4. Expression of testicular HSL. Infranatants from rat and human testes and adipose tissue as well as homogenates from COS cells transfected with human HSL cDNAs were enriched for HSL protein by immunoprecipitation, followed by SDS–PAGE and Western blot analysis. The amount of infranatant used was standardized according to HSL activity. Immunoreactive proteins were detected by ECL using chicken anti-HSL as primary antibody and a horseradish peroxidase-conjugated anti-chicken IgG as secondary antibody. Indicated molecular weight markers, in order of decreasing Mr , are a2macroglobulin, b-galactosidase, and fructose-6-phosphate kinase. In A, lanes 1–3 contain COS homogenates: lane 1, from transfection with pcDNA3 alone; lane 2, with pcDNA3/HSLadi ; and lane 3, with pcDNA3/HSLtes . Lanes 4–7 contain tissue infranatants: lane 4, human testis; lane 5, rat testis; lane 6, human adipose tissue; and lane 7, rat adipose tissue. In B, lane 1 contains infranatant of human testis and lane 2 infranatant of human adipose tissue. The gel in B represents a different experiment from the one in A, and the exposure time was several times longer.

gnma

AP: Genomics

A TESTICULAR ISOFORM OF HORMONE-SENSITIVE LIPASE

445

FIG. 5. Immunocytochemistry. Immunofluorescence with anti-HSL antibodies in rat testis. There is a prominent supply of HSL-like immunoreactivity (LI) in most seminiferous tubules (A and B), although the intensity and extent are subject to intertubule variation. Note that some tubules are devoid of HSL-LI (A, arrow) and that interstitial cells lack HSL-LI. At a higher magnification (B), it is seen that immunoreactive HSL occurs in the adluminal part of the tubules, representing principally spermatid cytoplasm. Bars, 100 mm.

brane of the Sertoli cells, the occurrence of HSL-LI in the lateral portions of the Sertoli cells cannot be ruled out. However, Leydig and myoid cells lacked immunoreactive HSL. The intensity and extent of HSL-LI varied between tubules and was even absent from some tubules. HSL-LI was observed in the adluminal parts of tubules only in stages XIII–VIII, corresponding to the localization of the cytoplasm of elongating spermatids and of adluminal parts of Sertoli cells (Fig. 5). In the epididymis, HSL-LI was present in spermatozoa as well as in adipocytes (not shown). DISCUSSION

Tissue-specific expression and heterogeneities with regard to the size of HSL mRNA and protein have been known for many years (Holm et al., 1987, 1988a; Stenson Holst et al., 1994), but the molecular structure of HSL from tissues other than adipose has not been described. We report here the presence of a novel isoform of HSL, distinct from the human adipocyte form, although derived from the same gene. A large primary transcript, which includes a unique testis-specific coding exon plus the exons encoding adipocyte HSL, is produced and processed to a final mRNA of 3.9 kb in testis. The testicular exon, located 16 kb upstream of the first exon of the adipocyte form, encodes 301 amino acids that are in frame with the adipocyte translation initiation codon. A full-length testicular polypeptide of 1076 amino acids with a calculated molecular mass of approximately 116 kDa is translated from the 3.9-kb mRNA. The addition of 16 kb to the previously described 11-kb gene is substantial and should include new and valuable information for HSL gene regulation studies. Although upstream located tissue-specific exons are not unusual, the splicing of a coding exon into the non-

AID

Genom 4227

/

6r1b$$$381

06-27-96 22:31:02

translated sequence of another exon, thus accomplishing new coding sequence between the upstream exon and the initiation codon of the downstream exon, is not common. Several genes encoding gonadal proteins with 5*-additions due to upstream exons have been recently reported. Examples are the murine GATA-1 gene (Ito et al., 1993), the rat vasopressin gene (Foo et al., 1994), and the human CYP19 gene (Harada et al., 1993; Jenkins et al., 1993). However, in all these cases the upstream exons are noncoding. Several immunoreactive HSL proteins in rat testis have been reported (Holm et al., 1987; Stenson Holst et al., 1994; Kraemer et al., 1993). Although the 130kDa size seems to be consistent in all investigations, various smaller immunoreactive components often appear (Fig. 4A, lane 5), most likely due to proteolysis since testis is a tissue rich in proteinases. Whether the greater mobility of HSLtes from the human tissue (Fig. 4B) is also reflecting proteolysis is uncertain. Differences in posttranslational maturation between the rat and the human proteins cannot be ruled out, since the molecular mass deduced from the primary structure is 116 kDa for both proteins. However, the size of the major product recognized by immunoprecipitation with anti-HSL of transfected COS cell infranatants is close to the rat HSLtes size (Fig. 4A, lane 3). The lower molecular mass component of approximately 116 kDa in this lane might result from proteolysis during the processing of homogenates and immunoprecipitation. Proteolysis might also be the cause of the weaker HSL band obtained from adipose tissue (Fig. 4A, lane 6). In contrast to rat, where a single HSL mRNA of 3.9 kb is expressed, two different HSL mRNAs, 3.3 and 3.9 kb, are found at approximately equal abundancy in human testis (Fig. 3). Despite this, the 116-kDa protein was the only protein that was immunoprecipitated from human testis infranatants using anti-HSL antibodies,

gnma

AP: Genomics

446

STENSON HOLST ET AL.

indicating that the smaller mRNA is either not translated or translated only to a very low extent (Fig. 4B). The unusually high representation of proline and glutamine residues causes a weak resemblance between the testis-specific sequences and transcription factors that are characterized by glutamine- and/or proline-rich activation domains. Proline-rich domains are found also in many other biologically important proteins and peptides (Williamson, 1994; Vanhoof et al., 1995). For instance, proline-rich regions are often involved in protein–protein interactions. Indeed, a putative motif for SH3 binding, in accordance with a consensus SH3-binding motif (Williamson, 1994; Alexandropoulos et al., 1995), was identified in both the human and the rat sequences. A proline-rich O-glycosylated repeat unit has been identified in the C-terminal part of bile-salt stimulated lipase/carboxyl ester lipase (BSSL/ CEL) of different species (Reue et al., 1991; Nilsson et al., 1990). The function of the proline-rich region is yet unknown (Bla¨ckberg et al., 1995). Despite the fact that HSL and BSSL/CEL recently have been suggested to be members of the same esterase subfamily (Petersen and Drablos, 1994), the proline-rich sequences of HSLtes and BSSL/CEL show no sequence similarity. The presence of HSL mRNA in seminiferous tubules and its absence from Leydig cells were recently demonstrated by us using in situ hybridization (Stenson Holst et al., 1994). The present immunohistochemistry data strongly support this and, in addition, indicate that HSL is principally expressed in late spermatids and spermatozoa, although some expression in Sertoli cells cannot be ruled out. The in situ hybridization data demonstrated that the highest mRNA expression occurred in the cytoplasm of elongated spermatids and/ or adluminal parts of Sertoli cells in stages X–XIV. We now demonstrate that the HSL protein is apparently expressed in the succeeding stages (XIII–VIII), suggesting that the protein is present several days after the mRNA has disappeared. The most likely interpretation of the present data is that the protein is expressed in haploid germ cells. As elongating spermatids are transcriptionally inactive (Penttila¨ et al., 1995), this means that the mRNA expression must already start in round spermatids. When our in situ hybridization data are reexamined, this appears to be the case. The functional role of HSL in the testis is unknown. As HSL appears to be absent from the steroid hormoneproducing Leydig cells, it is unlikely that it serves as a cholesterol ester hydrolase involved in hormone biosynthesis. The present study instead suggests that it could be important for spermiogenesis and sperm function. The data in this paper provide a good starting point for studies aimed at establishing structure–function relationships for HSLtes as well as its regulation during spermiogenesis. ACKNOWLEDGMENTS We thank Dr. Frank Sundler (Lund University, Sweden) for valuable discussions. Ann-Helen Thore´n is greatly acknowledged for ex-

AID

Genom 4227

/

6r1b$$$381

06-27-96 22:31:02

cellent technical assistance. This work was supported by grants from the following foundations: The Swedish Medical Research Council (Grants 3362 to Per Belfrage, 11284 to C.H., 5935 to A.B., and 8638 to G.E.); the Swedish Society of Medicine, Stockholm, the A. Pa˚hlsson foundation, the E. and W. Cornell’s foundation, the Medical Faculty of the Lund University, and the Ministe`re de l’Enseignement Supe´rieur et de la Recherche (grant to D.L.). Work performed by the Lawrence Livermore National Laboratory was under the auspices of the US Department of Energy (Contract No. W-7405-Eng-48). L.S.H., D.L., H.L., J.G., and C.H. are members of the BIOMED I Concerted Action EUROLIP supported by the European Union.

REFERENCES Alexandropoulos, K., Cheng, G., and Baltimore, D. (1995). Prolinerich sequences that bind to Src homology 3 domains with individual specificities. Proc. Natl. Acad. Sci. USA 92: 3110–3114. Bla¨ckberg, L., Stro¨mqvist, M., Edlund, M., Juneblad, K., Lundberg, L., Hansson, L., and Hernell, O. (1995). Recombinant human-milk bile-salt-stimulated lipase: Functional properties are retained in the absence of glycosylation and the unique proline-rich repeats. Eur. J. Biochem. 228: 817–821. Chomczynski, P., and Sacchi, N. (1987). Single-step method of RNA isolation by acid guanidinium thiocyanate–phenol–chloroform extraction. Anal. Biochem. 162: 156–159. Cook, K. G., Colbran, R. J., Snee, J., and Yeaman, S. J. (1983). Cytosolic cholesterol ester hydrolase from bovine corpus luteum: Its purification, identification, and relationship to hormone-sensitive lipase. Biochim. Biophys. Acta 752: 46–53. Cook, K. G., Yeaman, S. J., Stra˚lfors, P., Fredrikson, G., and Belfrage, P. (1982). Direct evidence that cholesterol ester hydrolase from adrenal cortex is the same enzyme as hormone-sensitive lipase from adipose tissue. Eur. J. Biochem. 125: 245–249. Feinberg, A. P., and Vogelstein, B. (1983). A technique for radiolabeling DNA restriction endonuclease fragments to high specific activcity. Anal. Biochem. 132: 6–13. Foo, N-C., Funkhouser, J. M., Carter, D. A., and Murphy, D. (1994). A testis-specific promoter in the rat vasopressin gene. J. Biol. Chem. 269: 659–667. ¨ ., and Belfrage, P. (1981). Fredrikson, G., Stra˚lfors, P., Nilsson, N. O Hormone-sensitive lipase of rat adipose tissue: Purification and some properties. J. Biol. Chem. 256: 6311–6320. Harada, N., Utsumi, T., and Takagi, Y. (1993). Tissue-specific expression of the human aromatase cytochrome P-450 gene by alternative use of multiple exons 1 and promoters, and switching of tissuespecific exons 1 in carcinogenesis. Proc. Natl. Acad. Sci. USA 90: 11312–11316. Holm, C., Belfrage, P., and Fredrikson, G. (1987). Immunological evidence for the presence of hormone-sensitive lipase in rat tissues other than adipose tissue. Biochem. Biophys. Res. Commun. 148: 99–105. Holm, C., Belfrage, P., Osterlund, T., Davis, R. C., Schotz, M. C., and Langin, D. (1994). Hormone-sensitive lipase: Structure, function, evolution and overproduction in insect cells using the baculovirus expression system. Prot. Eng. 7: 537–541. Holm, C., Davis, R. C., Fredrikson, G., Belfrage, P., and Schotz, M. C. (1991). Expression of biologically active hormone-sensitive lipase in mammalian (COS) cells. FEBS Lett. 285: 139–144. Holm, C., Kirchgessner, T. G., Svenson, K. L., Fredrikson, G., Nilsson, S., Miller, C. G., Shively, J. E., Heinzmann, C., Sparkes, R. S., Mohandas, T., Lusis, A. J., Belfrage, P., and Schotz, M. C. (1988a). Hormone-sensitive lipase: Sequence, expression and chromosomal localization to 19 cent–q13.3. Science 241: 1503–1506. Holm, C., Kirchgessner, T. G., Svenson, K. L., Lusis, A. J., Belfrage, P., and Schotz, M. C. (1988b). Nucleotide sequence of rat adipose tissue hormone-sensitive lipase cDNA. Nucleic Acids Res. 16: 9879. Ito, E., Toki, T., Ishihara, H., Ohtani, H., Gu, L., Yokoyama, M.,

gnma

AP: Genomics

A TESTICULAR ISOFORM OF HORMONE-SENSITIVE LIPASE Engel, J. D., and Yamamoto, M. (1993). Erythroid transcription factor GATA-1 is abundantly transcribed in mouse testis. Nature 362: 466–468. Jenkins, C., Dodson, M., Mahendroo, M., and Simpson, E. (1993). Exon-specific Northern analysis and rapid amplification of cDNA ends (RACE) reveal that the proximal promoter II (PII) is responsible for aromatase cytochrome P450 (CYP19) expression in human ovary. Mol. Cell. Endocrinol. 97: R1–R6. Kraemer, F. B., Patel, S., Saedi, M. S., and Sztalryd, C. (1993). Detection of hormone-sensitive lipase in various tissues. I. Expression of an HSL/bacterial fusion protein and generation of anti-HSL antibodies. J. Lipid Res. 34: 663–671. Kraemer, F. B., Tavangar, K., and Hoffman, A. R. (1991). Developmental regulation of hormone-sensitive lipase mRNA in the rat: Changes in steroidogenic tissues. J. Lipid Res. 32: 1303–1310. Kyte, J., and Doolittle, R. F. (1982). A simple method for displaying the hydropathic character of a protein. J. Mol. Biol. 157: 105–132. Langin, D., Laurell, H., Stenson Holst, L., Belfrage, P., and Holm, C. (1993). Gene organization and primary structure of human hormone-sensitive lipase: Possible significance of a sequence homology with a lipase of Moraxella TA144, an antarctic bacterium. Proc. Natl. Acad. Sci. USA 90: 4897–4901. Laurell, H., Grober, J., Stenson Holst, L., Holm, C., Mohrenweiser, H. W., and Langin, D. (1995). The hormone-sensitive lipase (LIPE) gene located on chromosome 19q13.1–13.2 is not duplicated on 19p13.3. Int. J. Obesity 19: 590–592. Levitt, R. C., Liu, Z., Nouri, N., Meyers, D. A., Brandriff, B., and Mohrenweiser, H. W. (1995). Mapping of the gene for hormonesensitive lipase (LIPE) to chromosome 19q13.1–q13.2. Cytogenet. Cell Genet. 69: 211–214. Li, Z., Sumida, M., Birchbauer, A., Schotz, M. C., and Reue, K. (1994). Isolation and characterization of the gene for mouse hormone-sensitive lipase. Genomics 24: 259–265. Mulder, H., Lindh, A. C., and Sundler, F. (1993). Islet amyloid polypeptide gene expression in the endocrine pancreas of the rat: A combined in situ hybridization and immunocytochemical study. Cell Tissue Res. 274: 467–474.

AID

Genom 4227

/

6r1b$$$381

06-27-96 22:31:02

447

Nilsson, J., Bla¨ckberg, L., Carlsson, P., Enerba¨ck, S., Hernell, O., and Bjursell, G. (1990). cDNA cloning of human-milk bile-saltstimulated lipase and evidence for its identity to pancreatic carboxylic ester hydrolase. Eur. J. Biochem. 192: 543–550. Penttila¨, T-L., Yuan, L., Mali, P., Ho¨o¨g, C., and Parvinen, M. (1995). Haploid gene expression: Temporal onset and storage patterns of 13 novel transcripts during rat and mouse spermiogenesis. Biol. Reprod. 53: 499–510. Petersen, S. B., and Drablos, F. (1994). A sequence analysis of lipases, esterases and related proteins. In ‘‘Lipases—Their Structure, Biochemistry and Application’’ (P. Wolley and S. B. Petersen, Eds.), pp. 23–48, Cambridge Univ. Press, Cambridge. Reue, K., Zambaux, J., Wong, H., Lee, G., Leete, T. H., Ronk, M., Shively, J. E., Sternby, B., Borgstro¨m, B., Ameis, D., and Schotz, M. C. (1991). cDNA cloning of carboxyl ester lipase from human pancreas reveals a unique proline-rich repeat unit. J. Lipid Res. 32: 267–276. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989). ‘‘Molecular Cloning: A Laboratory Manual,’’ 2nd ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Schonk, D., van Dijk, P., Riegmann, P., Trapman, J., Holm, C., Willcocks, T. C., Sillekens, P., van Venrooij, W., Wimmer, E., Geurts van Kessel, A., Ropers, H.-H., and Wieringa, B. (1990). Assignment of seven genes to distinct intervals on the midportion of human chromosome 19q surrounding the myotonic dystrophy gene region. Cytogenet. Cell Genet. 54: 15–19. Stenson Holst, L., Hoffmann, A. M., Mulder, H., Sundler, F., Holm, C., Bergh, A., and Fredrikson, G. (1994). Localization of hormonesensitive lipase to rat Sertoli cells and its expression in developing and degenerating testes. FEBS Lett. 355: 125–130. Tornqvist, H., Bjo¨rgell, P., Krabisch, L., and Belfrage, P. (1978). Monoacylmonoalkyl-glycerol as a substrate for diacylglycerol hydrolase activity in adipose tissue. J. Lipid Res. 19: 654–656. Vanhoof, G., Goossens, F., De Meester, I., Hendriks, D., and Scharpe´, S. (1995). Proline motifs in peptides and their biological processing. FASEB J. 9: 736–744. Williamson, M. P. (1994). The structure and function of proline-rich regions in proteins. Biochem. J. 297: 249–260.

gnma

AP: Genomics