REPRODUCTION-DEVELOPMENT
The Catalytic Function of Hormone-Sensitive Lipase is Essential for Fertility in Male Mice Shu Pei Wang, Jiang Wei Wu, Hugo Bourdages, Jean François Lefebvre, Stéphanie Casavant, Blair R. Leavitt, Damian Labuda, Jacquetta Trasler, Charles E. Smith, Louis Hermo, and Grant A. Mitchell Divisions of Medical Genetics (S.P.W., J.W.W., H.B., S.C., G.A.M.) and Hematology (J.F.L., D.L.), Department of Pediatrics, Centre Hospitalier Universitaire Sainte-Justine and Faculty of Medicine, Université de Montréal, Montréal, Québec, Canada, H3T 1C5; Valeant Cosméderme (H.B.), Laval, Québec, Canada, H7V 0A3; Centre for Molecular Medicine and Therapeutics (B.R.L.), Department of Medical Genetics, University of British Columbia, Vancouver, British Columbia, Canada, V5Z 4H4; Department of Pediatrics (J.T.), Human Genetics and Pharmacology and Therapeutics, McGill University and Research Institute of the McGill University Health Centre at the Montreal Children’s Hospital, Montréal, Québec, Canada, H3H 1P3; and Department of Anatomy and Cell Biology (C.E.S., L.H.), McGill University, Montréal, Québec, Canada, H3A 2B2
In male mice, deficiency of hormone sensitive lipase (HSL, Lipe gene, E.C.3.1.1.3) causes deficient spermatogenesis, azoospermia, and infertility. Postmeiotic germ cells express a specific HSL isoform that includes a 313 amino acid N-terminus encoded by a testis-specific exon (exon T1). The remainder of testicular HSL is identical to adipocyte HSL. The amino acid sequence of the testisspecific exon is poorly conserved, showing only a 46% amino acid identity with orthologous human and rat sequences, compared with 87% over the remainder of the HSL coding sequence, providing no evidence in favor of a vital functional role for the testis-specific N-terminus of HSL. However, exon T1 is important for Lipe transcription; in mouse testicular mRNA, we identified 3 major Lipe transcription start sites, finding numerous testicular transcription factor binding motifs upstream of the transcription start site. We directly explored two possible mechanisms for the infertility of HSL-deficient mice, using mice that expressed mutant HSL transgenes only in postmeiotic germ cells on a HSL-deficient background. One transgene expressed human HSL lacking enzyme activity but containing the testis-specific N-terminus (HSL⫺/⫺muttg mice). The other transgene expressed catalytically inactive HSL with the testis-specific N-terminal peptide (HSL⫺/⫺atg mice). HSL⫺/ ⫺muttg mice were infertile, with abnormal histology of the seminiferous epithelium and absence of spermatozoa in the epididymal lumen. In contrast, HSL⫺/⫺atg mice had normal fertility and normal testicular morphology. In conclusion, whereas the catalytic function of HSL is necessary for spermatogenesis in mice, the presence of the N-terminal testis-specific fragment is not essential. (Endocrinology 155: 3047–3053, 2014)
F
ertility in male mice requires hormone-sensitive lipase (HSL), a fatty acyl ester hydrolase expressed in adipose tissues, skeletal muscle, myocardium, adrenal cortex, macrophages, pancreas, and testis (1). In testis, HSL is predominantly located in postmeiotic germ cells although low-level HSL immunoreactivity is described in early germ cells, such as type B spermatogonia and primary spermato-
cytes and in Leydig and Sertoli cells (2, 3). HSL-deficient (HSL⫺/⫺) male mice are infertile (4, 5). This is due to a specific abnormality of spermatogenesis, with otherwise-normal masculinization in HSL⫺/⫺ males including normal mass of the seminal vesicles (4, 5). The abnormality of spermatogenesis occurs despite normal levels of testosterone, follicle stimulating hormone,
ISSN Print 0013-7227 ISSN Online 1945-7170 Printed in U.S.A. Copyright © 2014 by the Endocrine Society Received January 14, 2014. Accepted April 28, 2014. First Published Online May 5, 2014
Abbreviations: aHSL, adipose hormone-sensitive lipase; HSL, hormone-sensitive lipase; muttHSL, mutant testicular hormone-sensitive lipase.
doi: 10.1210/en.2014-1031
Endocrinology, August 2014, 155(8):3047–3053
endo.endojournals.org
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 April 2015. at 07:44 For personal use only. No other uses without permission. . All rights reserved.
3047
3048
Wang et al
HSL is Essential for Fertility in Male Mice
Endocrinology, August 2014, 155(8):3047–3053
and luteinizing hormone in HSL-deficient males (5). HSLdeficient males have normal frequency of copulation as shown by the equal prevalence of vaginal mucous plugs in females housed with HSL-deficient and with normal control males (4, 5). HSL⫺/⫺ mice that express a full length human testicular HSL transgene exclusively in postmeiotic germ cells are fertile and have normal germ cell morphology (6, 7). Therefore, expression of HSL within postmeiotic germ cells is in itself sufficient to confer fertility to HSL-deficient males, demonstrating a cell-autonomous mechanism of infertility in HSL deficiency. In testis, the transcription of the HSL gene, Lipe, differs from that of other organs. The resulting testicular isoform of HSL has a 313 amino acid N-terminal sequence encoded by an exon designated here as T1, a testis- specific exon located 16 kb upstream of the 5⬘-most exon of adipose HSL isoforms (8). A 1.4 kb region 5⬘ of exon T1 is sufficient to mediate testis-specific expression (7). Exon T1 splices to the same site in exon 1, as do the multiple other 5⬘ exons that are transcriptionally active in adipose tissue (9). The catalytic activity of purified full-length testicular HSL containing sequence from exon T1 is indistinguishable from that of adipose HSL (8). A testicular HSL mRNA with a transcription start site in a different exon (T2) is reported in human testis (10). It is predicted to encode a 775-residue protein, identical to the main adipose tissue HSL isoform. It is unknown why testicular HSL is necessary for spermatogenesis in mice. On general principals, the catalytic
a b
T1
A T2 B C
function of HSL may be essential, or HSL may play a noncatalytic, structural role in germ cells, or both. The catalytic properties of testicular HSL may enable normal fertility by hydrolyzing one or several of the fatty acyl ester substrates of HSL in postmeiotic germ cells, including diglycerides, triglycerides, or cholesteryl- and retinyl-esters (11). Alternatively, HSL might play a noncatalytic role in germ cells, unrelated to its catalytic function. Such “moonlighting” roles are known for several enzymes (12). The N-terminal testis-specific fragment of HSL accounts for a 28% peptide chain length of the molecular mass and has high contents of proline and glutamine, residues that are known to confer unique structural properties (8). Here, we explore the relative importance of HSL catalytic function and of the presence of the testis-specific sequence of HSL for the development of post meiotic germ cells in the testicle and for male fertility.
Materials and Methods Mouse exon T1 cloning and analysis of interspecies divergence The methods for sequencing the 129Sv mouse genomic Lipe clone p5⬘LipM (9), for 5⬘ rapid amplification of cDNA endsPCR (RACE-PCR) and RNase protection analysis, and for sequence alignment and prediction of potential transcription factor binding motifs are described in the Supplemental Materials. p5⬘LipM (Figure 1a) contains exon A plus approximately 2.8 kb of 5⬘ 129Sv mouse DNA (9) that we found to contain exon T1.
D
1
2 3 4 5
1 kb
67
8
9
Other HSL cDNAs
c
Testicular HSL cDNA BamHI
XhoI NdeI
d Prm 1
5’ UTR HSL cDNA 3 5 3’ UTR / PA BamHI
XhoI NdeI
S424A
HSL cDNA
3’ UTR / PA
e Prm 1
5’ UTR
Figure 1. Structure of the mouse Lipe gene and of human HSL cDNA-based transgenes. a, The mouse Lipe gene structure, showing the numerous upstream exons mediating its tissue-specific expression. b, All isoforms of HSL share the peptide sequences derived from exons 1–9. c, The testicular HSL cDNA includes a 5‘ coding sequence derived from exon T1. d and e, Structures of the HSL transgenes: (d) adipose (atg) and (e) mutant testicular (muttg), the latter showing the location of the inactivating p.Ser424Ala mutation.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 April 2015. at 07:44 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2014-1031
endo.endojournals.org
3049
We used the Translate Tool (http://www.expasy.ch/) to locate open reading frames within exon T1. The Psipred V2.2 program (http://www.psipred.net) was used to predict potential secondary structure motifs of mouse, rat (Genbank U40001), and human (Genbank U40002) testicular HSL peptides. Using the Conservation tracks of the Comparative Genomics group on the HG19 Assembly included in the UCSC table browser (http://genome.ucsc.edu), we compared conservation scores between exon T1 and the other nine Lipe exons. These tracks use two methods, phastCons (13) and PhyloP (14) from the PHAST package (http://compgen.bscb.cornell.edu/phast/ index.php), to study multiple alignments of 46 vertebrate species including 23 placental mammals and 9 primates. phastCons calculates the probabilities of a segment arising from a conserved or a nonconserved model, and assigns a log-odds score of conservation to each segment. PhyloP uses a phylogenetic model to estimate if elements are fully conserved or under lineage-specific selection.
Animal breeding and handling
Vector construction for transgenes
Genotyping of transgenic mice
Using full-length human HSL cDNAs for both adipose and testicular isoforms (6), we cloned the adipose HSL (aHSL) cDNA into pBluescript (SK) and the testicular HSL cDNA into pCDNA3.1. In this experiment we constructed two transgenic vectors, one expressing the aHSL cDNA and another expressing a mutant testicular HSL (muttHSL) cDNA. The aHSL vector was made by inserting the aHSL cDNA in the pPrCExV-1 vector (15), downstream of the mouse protamine-1 promoter that is specific for postmeiotic germ cells (6, 16). To create a catalytically-inactive human testicular HSL, we inserted the p.Ser424Ala mutation that was previously shown to produce a stable HSL protein with no detectable catalytic activity (17). The p.Ser424Ala muttHSL transgenic vector was assembled in three steps as follows. In testicular HSL cDNA, the adenine of the initiation methionine codon of the testicular exon is designated as position ⫹1. We first changed the serine 424 codon (AGT) to alanine in the full-length aHSL cDNA by changing the AG dinucleotide at positions 2173 and 2174 to GC, using site directed mutagenesis (QuikChange Site-Directed Mutagenesis Kit, Stratagene). The fragment between the SfiI site and 3⬘ end was verified by bidirectional sequencing in both constructs. Then the aHSL cDNA carrying the mutation and the full-length testicular HSL cDNA were digested with both SfiI and EcoRI, a restriction site in the polylinker of pBluescript (SK) and pcDNA3.1. The p.Ser424Ala mutation was introduced into a full-length testicular HSL cDNA by replacing the normal SfilEcoRI fragment with the mutant fragment. Finally, the fulllength muttHSL cDNA with the mutation was inserted downstream of the mouse protamine-1 promoter in the pPrCExV-1 vector (15).
To genotype the HSL⫺/⫺muttg transgenic mice, we used the same primers and PCR conditions as described for mice transgenic for human testicular HSL (6). To genotype HSL⫺/⫺atg mice, PCR was performed with the same conditions and the same sense primer but the antisense primer was LipH65 (5⬘-CGAAGAAGCACTCCTCCAGCG), complementary to nucleotides 1177 to 1198 of the human aHSL. Amplification of a 1.3 kb fragment was diagnostic for the presence of the human aHSL transgene. Founders detected by PCR were also genotyped by Southern blotting as described (6), using a human aHSL cDNA as probe. Hind III digests generated an 8 kb fragment for muttHSL transgenic mice and a 7 kb fragment for aHSL transgenic mice.
Production of transgenic mice expressing adipose or mutant testicular human HSL cDNA in postmeiotic germ cells Plasmid sequences of aHSL and muttHSL were removed by HindIII digestion and the transgenes were purified with a Gene Extraction kit (Qiagen). The purified aHSL and muttHSL vectors were microinjected into the pronucleus of FVB/N mouse embryos, and the resulting embryos implanted into pseudopregnant female mice. Founders were backcrossed 5 generations to
C57BL/6 mice, and then bred with HSL-deficient mice. The resulting male offspring have no endogenous HSL expression but express adipose (HSL⫺/⫺atg) or mutant testicular (HSL⫺/ ⫺muttg) human cDNAs. Mice were maintained on a 12-hour light, 12-hour dark photocycle and were provided with food and water ad libitum. Three-month-old mice were used in all experiments.
These studies were approved by the Animal Care Committee of CHU Sainte-Justine Research Center. All experiments were approved by the Canadian Council on Animal Care Committee of CHU Sainte-Justine accredited animal facility. Breeding and handling were performed in accordance with standard animal room practice, as described (6). Males were tested for fertility by housing each male with 2 CD1 females for 2 weeks. The birth of a litter and the number of offspring were recorded.
Northern blotting Total RNA was extracted from testes using Trizol (Invitrogen). Northern blotting was performed in ExpressHyb Hybridization Solution (Clontech) according to the manufacturer’s protocol using the probe described above for Southern blotting. -actin was used as a loading control.
Production of antihuman exon T1 antibody A peptide containing the entire peptide sequence of the human testicular exon was produced with the Bac-to-Bac baculovirus expression system (Invitrogen) according to the manufacturer’s protocol. Briefly, a 903 nt cDNA fragment of human testicular exon from the initiation methionine codon to the asparagine 301 codon (AAC) was amplified. In the amplification primer, a stop codon (TAA) was introduced immediately after asparagine 301. The amplicon was digested with BamHI and EcoRI, and then cloned into pFastBacHTA. The human exon T1 was in frame with upstream 6xHis tag. The recombinant bacmid was produced and purified from DH10Bac Escherichia coli cells, and then used to transfect SF9 insect cells. The human exon T1 peptide expressed in these cells, which included an N-terminal His tag, was purified using Ni sepharose 6 fast Flow (GE Healthcare). The His tag was removed by digestion with TEV protease. Rabbit polyclonal antihuman exon T1 antibody was produced (Chemicon International).
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 April 2015. at 07:44 For personal use only. No other uses without permission. . All rights reserved.
Endocrinology, August 2014, 155(8):3047–3053
embedded in paraffin. Sections were cut and examined in the light microscope. Details are as described previously (6).
HSL-/-ttg g
HSL-/-mu uttg2
HSL-/-mu uttg1
+/+
HSL is Essential for Fertility in Male Mice
HSL-/-atg g2
Wang et al
HSL-/-atg g1
3050
a Southern b Northern, Northern HSL
Statistical analysis kb 9.0 8.0 7.0
Data were analyzed by ordinary 1-way ANOVA followed by Turkey’s multiple comparison testing with GraphPad Prism 6 software. Differences of P ⬍ .05 were considered to be statistically significant.
44 4.4 2.4 2.4 1.4
c Northern, Actin
kDa
d Western, HSL
160 105 75 160 105
e Western, Exon T
75 Figure 2. HSL transgenes and their expression. a, Southern blot of different transgenic lines: atg, muttg and ttg (normal testicular transgene). b, Northern blot of HSL species. c, Northern blot of actin as a loading control. d, Western blot using anti-HSL antibody showing a 86 kDa protein in HSL⫺/⫺atg mice and a 117 kDa protein in HSL⫺/ ⫺muttg and HSL⫺/⫺ttg mice. e, Western blot using antihuman testicular exon antibody detects the expected 117 kDa protein only in HSL⫺/⫺muttg and HSL⫺/⫺ttg mice.
Western blotting and enzymatic studies Western blotting was performed as described (6) using the anti-HSL or antiexon T1 antibodies. For antiexon T1 antibodies, the procedure was the same but the dilution was 1:5000. Cholesteryl esterase activity was measured as described (18).
Histology Testes and epididymides of adult mice (wild type and transgenic, n ⫽ 3 for each) were removed from mice, immersed in Bouin’s fixative for 24 hours, and then dehydrated in ethanol and
Results Comparative analysis of mammalian LIPE genes In the genomic clone from 129Sv mice, exon T1 is 1.1 kb in length, contains a 313 codon open reading frame, and is located 16.4 kb upstream of exon 1 (Figure 1A). The overall amino acid sequence identity of mouse adipose HSL is 87% with rat and 75% with human HSL. Much of the divergence is in the testis-specific region. In the sequence encoded by the testicular exon, there is a predicted amino acid identity of 72% with rat and 46% with human testicular HSL, and 36% identity among all three (Supplemental Figure 1). Compared with this, the rest of the coding sequence of mouse HSL has a predicted amino acid identity of 96% with rat and 86% with human HSL. Identity was 83.9% among mouse, rat, and human HSLs. RNase protection and 5⬘ rapid amplification of cDNA ends (RACE)-PCR defined a cluster of transcription start sites 177 to 220 bp upstream of the first in-frame methionine initiation codon Supplemental Figure 2. The 849 bp upstream of this region contains potential binding motifs for germ cell-specific and general transcription factors, including SRY/SOX, GATA-1, SP-1, and C/EBP (Supplemental Figure 3). Analysis using phastCons showed that 75% of residues in exons 1–9 are predicted to be conserved, compared to
Table 1. Effect of HSL Transgene Expression on Tissue Mass and Testicular Cholesteryl Esterase Activity of 3Month-Old Male Mice Cholesteryl esterase
Mass (mg) HSL genotype
N
Testis (paired)
BAT
n
nmol/mg protein/min
HSL⫺/⫺atg1 HSL⫺/⫺atg2 HSL⫺/⫺mttg1 HSL⫺/⫺mttg2 HSL⫺/⫺ HSL⫹/⫹
5 4 7 7 7 8
205.6 ⫾ 9.2b 200.5 ⫾ 2.6b 160.0 ⫾ 9.3 177.1 ⫾ 1.8b 139.1 ⫾ 7.1d 187.8 ⫾ 8.3
181.8 ⫾ 20.1d 151.5 ⫾ 5.70d 144.9 ⫾ 27.7c 181.7 ⫾ 12.7d 161.5 ⫾ 22.7 d 72.8 ⫾ 4.8
4 4 4 4 4 4
10.99 ⫾ 2.42ac 5.98 ⫾ 0.92bc 0.10 ⫾ 0.01d 0.10 ⫾ 0.01d 0.13 ⫾ 0.01 1.95 ⫾ 0.14
P ⬍ .05 versus HSL ⫺/⫺. P ⬍ .01 versus HSL ⫺/⫺. c P ⬍ .05 versus HSL ⫹/⫹. d P ⬍ .01 versus HSL ⫹/⫹. BAT indicates brown adipose tissue. a
b
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 April 2015. at 07:44 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2014-1031
endo.endojournals.org
3051
Production and characterization of transgenic mice Vectors containing the human aHSL cDNA (Figure 1d) and human muttHSL cDNA (Figure 1e) were microinjected into single cell embryos from FVB/N mice. For aHSL, the vector was detected in 2/28 pups (7.1%), and for muttHSL, in 2/21 (9.5%). All four positive offspring were males. The presence of the transgene was confirmed on genomic Southern blots (Figure 2a). Both transgenes were transferred to a HSL-deficient background, creating mice that expressed HSL only in postmeiotic germ cells, either the stable catalytically-inactive mutant (HSL⫺/⫺muttg mice) or the adipose isoform (HSL⫺/⫺atg). Northern (Figure 2, b and c) and Western blotting (Figure 2, d and e) revealed that both adipose transgenic lines (HSL⫺/⫺atg1 and HSL⫺/⫺atg2) expressed the human HSL mRNA and protein, but HSL⫺/⫺atg1 mice had a higher level of expression than HSL⫺/⫺atg2 mice. The p.Ser424Ala HSL transgene, expressed in testes, produced easily-detectable levels of HSL mRNA and protein. As previously described for HSL⫺/⫺ mice, transgenic males had no obvious abnormality. At the time of killing at age 3 months, there was Figure 3. The effect of HSL genotype on histology of the testis and epididymis of control (a, e), no significant difference among the HSL⫺/-⫺ (b, f), HSL⫺/⫺atg1 (c, g) and HSL⫺/⫺muttg (d, h) mice. The seminiferous epithelium weights of males of each genotype: (SE) of a and c show a similar distribution and appearance of germ cells in the different tubules, unlike that seen for b and d that show abnormal tubules and absence of late spermatids. In the normal controls, 29.37 ⫾ 1.24g (n ⫽ epididymis, spermatozoa (S) are noted in the lumen (Lu) of e and g, whereas only immature 8); HSL⫺/⫺ mice, 27.59 ⫾ 1.69 g round germ cells (circles) are seen in the lumen of f and h. E indicates epididymal epithelium. (n ⫽ 7); HSL⫺/⫺ muttg1, 29.34 ⫾ Original magnification: b– d, ⫻300; a, e– h, ⫻250. 1.17 g (n ⫽ 7); HSL⫺/⫺ muttg2, 27.96 ⫾ 1.97 g (n ⫽ 7); HSL⫺/ only 12% in exon T1. Regarding the probability of neg- ⫺atg1, 28.03 ⫾ 0.65 g (n ⫽ 5), HSL⫺/⫺atg2, 30.78 ⫾ ative selection, a major cause of cross-species conserva- 0.01 g (n ⫽ 4). tion, 53% of residues in exons 1–9 achieve significance (P ⬍ .01), compared to only 3% in exon T1. Also, using Expression of human aHSL, but not of stable, phyloP, 39% of residues in exons 1–9 fulfilled the criteria enzymatically-inactive human testicular HSL, for conservation versus 5% in exon T1. Lineage-specific restores testicular cholesteryl esterase activity and selection was not detected (⬍ 3.2% in each), but in other testicular mass to HSLⴚ/ⴚ mice studies, we have found phyloP to be less sensitive for deCompared to normal control mice, cholesteryl esterase tection of lineage-specific selection than for conservation. activity was 5.6-fold greater in HSL⫺/⫺atg1 testicle ho-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 April 2015. at 07:44 For personal use only. No other uses without permission. . All rights reserved.
3052
Wang et al
HSL is Essential for Fertility in Male Mice
mogenates (P ⬍ .05) and 3-fold greater for HSL⫺/⫺atg2 homogenates (P ⬍ .05). As expected, cholesteryl esterase activity in HSL⫺/⫺muttg testicle was not distinguishable from that in HSL⫺/⫺ mice (Table 1). HSL⫺/⫺atg mice, expressing a catalytically-active aHSL transgene, had normal testis weight (Table 1). Interestingly, testicular masses of HSL⫺/⫺muttg mice were also significantly greater than those of HSL⫺/⫺ mice, despite their lack of cholesteryl esterase activity (Table 1). Examination of interscapular brown fat, a control for the specificity of transgenic HSL expression, showed the characteristics of HSL⫺/⫺ mice: increased mass (Table 1) and abnormal mono-vacuolar morphology, as previously-reported in HSL⫺/⫺ mice (6) (not shown). Testicular expression of normal human aHSL, which lacks the N-terminal testicle-specific sequence, confers normal testicular histology and normal fertility to HSLⴚ/ⴚ mice Wild type mice revealed a full complement of the 12 stages of the cycle of the seminiferous epithelium. All generations of germ cells were noted including the postmeiotic spermatids (Figure 3a). The epididymides of wild-type mice demonstrated an epididymal lumen full of spermatozoa (Figure 3e). In the HSL⫺/⫺mice, the seminiferous epithelium was devoid of postmeiotic germ cells, and some tubules were abnormal in appearance (Figure 3b), as published previously (6). Similarly, the lumen of the epididymis was devoid of spermatozoa but rather contained many small round cells (Figure 3f), shown previously to be immature germ cells (6). HSL⫺/⫺ atg1 mice presented a similar distribution of germ cells in the seminiferous epithelium to that of control mice (Figure 3c) and spermatozoa were abundant in the epididymal lumen (Figure 3g), as noted for controls. In contrast, the histology of the seminiferous epithelium of HSL⫺/⫺muttg mice was abnormal with an absence of postmeiotic germ cells and abnormal appearance of some tubules (Figure 3d). The epididymal lumen also demonstrated an absence of spermatozoa but abundance of round immature germ cells (Figure 3h). Expression of the aHSL transgene but not of the muttHSL transgene conferred normal fertility to HSL⫺/⫺ mice (Table 2). Seven out of eight HSL⫺/⫺atg males tested produced litters with each female, with a mean litter size of 10.29 ⫾ 1.28 offspring. In contrast, none of the 9 HSL⫺/⫺muttg males produced offspring.
Discussion HSL is a fatty-acyl esterase that is active against numerous substrates and was first identified in adipose tissue. Initial
Endocrinology, August 2014, 155(8):3047–3053
Table 2. Effect on Fertility of the Expression of HSL Transgenes on a HSL⫺/⫺ Background
Genotype
Males
Pregnancies/ females mated
HSL⫺/⫺atg1 HSL⫺/⫺atg2 HSL⫺/⫺muttg1 HSL⫺/⫺muttg2
4 3 4 5
7/8 4/4 0/8 0/9
Litter size 10.29 ⫾ 1.28 10.25 ⫾ 1.11 0 0
studies of HSL focused on its role in adipose tissue, in which it has a complex pattern of phosphorylation that correlates with highly-regulated shifts in location to and from the surface of the lipid droplet (1). Molecular studies and the creation of mice with modifications of the Lipe gene revealed the presence of HSL-independent lipolysis (19, 20) and that HSL plays important roles in nonadipose tissues including adrenal cortex (21) and testis (4 – 6). Whereas HSL-deficient male mice are infertile and show marked structural abnormalities in postmeiotic germ cells (4, 5), expression of a human testicular HSL transgene exclusively in postmeiotic germ cells confers normal fertility and normal testicular histology to these mice, showing that the role of HSL in normal male fertility in mice is cell-autonomous, and carried out within postmeiotic germ cells. Because of the small size and anatomical complexity of mouse testes, classical biochemical studies are not possible, and we pursued molecular and genetic dissection of the role of HSL in male fertility. First we examined the sequences of testicular HSL from mice, humans and rats, using several measures of amino acid conservation. By all measures, the conservation of exon T1 was strikingly less than that of the rest of the coding sequence. Such a lack of conservation argues against an important functional role of the N terminus of testicular HSL in male fertility. In contrast, exon T1 has clear importance as the major start site for Lipe transcription in mice (10). We identified major upstream start transcriptional sites and found several upstream transcription factor binding motifs. Together, these findings incited us to directly study the importance of the N-terminal region of the testicular isoform in transgenic mice. HSL⫺/⫺atg mice, which express a catalytically-active HSL that lacks the testes-specific N-terminal sequence, conclusively demonstrate that the N-terminal sequence of testicular HSL isoform is not necessary for fertility; they show normal fertility, normal testicular mass, and normal microscopic morphology. Conversely, in HSL⫺/⫺muttg mice, the presence of catalytically-inactive HSL that contained the sequence encoded by T1 did not produce male fertility. In studies of purified HSL, inclusion of the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 April 2015. at 07:44 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2014-1031
p.Ser424Ala mutation results in a stable HSL peptide that has no detectable enzymatic activity (17). In the adipocyte HSL isoform, p.Ser424Ala abolishes another function of HSL, translocation to lipid droplets after -adrenergic stimulation (22). Interestingly, HSL⫺/⫺muttg mice, expressing inactive testicular HSL containing p.Ser424Ala, showed some increase of testicular mass, although testicular histology was abnormal as in HSL⫺/⫺ mice. Therefore, the presence of a structurally-intact HSL testicular isoform may have some effect upon the testicle, but this is insufficient to confer fertility or normal histology in the absence of HSL activity. These experiments do not reveal the precise substrates of HSL in postmeiotic germ cells. HSL can cleave fatty acyl esters of glycerol (diglycerides and triglycerides), retinoic acid, and cholesterol (11); other substrates may yet be unidentified. The infertility of HSL-deficient male mice could potentially arise from the noncleavage of any of these substances. Some circumstantial evidence is consistent with a cholesterol-related mechanism. HSL accounts for all of the measureable cholesteryl esterase activity in testis (5, 6), suggesting that testicular cholesterol metabolism may be severely affected in HSL deficiency. Male mice with disruption of the Dhcr24 gene that are deficient in cholesterol biosynthesis are infertile (23), and in HSL ⫺/⫺ mouse testes, expression of the cholesterol- and lipoprotein-related transcripts SR-BI, SR-BII, and LIMP II are increased (3). Although the exact mechanism is unclear by which HSL is necessary for male fertility in mice, this study clearly shows that the presence of catalytically-active HSL in postmeiotic germ cells can confer normal fertility even in the absence of the testis-specific N-terminal peptide.
Acknowledgments Address all correspondence and requests for reprints to: Grant A. Mitchell, Divisions of Medical Genetics and Hematology, Department of Pediatrics, CHU Sainte-Justine and Faculty of Medicine, Université de Montréal, 3175 Côte Sainte-Catherine, Montréal, Québec, Canada, H3T 1C5. Email: [email protected]. This work was supported in part by a Canadian Institutes of Health Research grant to G.M. The creation of the transgenic mice described in this article was financed by a Canadian Genetic Diseases Network collaborative project between B.L. and G.M. The Research Institutes of CHU Sainte-Justine and of the McGill University Health Centre at the Montreal Children’s Hospital are supported in part by the Fonds de la Recherche du Québec-Santé. Disclosure Summary: The authors have nothing to disclose.
References 1. Holm C, Osterlund T, Laurell H, Contreras JA. Molecular mechanisms regulating hormone-sensitive lipase and lipolysis. Ann Rev Nutr. 2000;20:365–393.
endo.endojournals.org
3053
2. Arenas MI, Lobo MV, Caso E, Huerta L, Paniagua R, Martin-Hidalgo MA. Normal and pathological human testes express hormone-sensitive lipase and the lipid receptors CLA-1/SR-BI and CD36. Hum Pathol. 2004;35:34 – 42. 3. Casado ME, Huerta L, Ortiz AI, et al. HSL-knockout mouse testis exhibits class B scavenger receptor upregulation and disrupted lipid raft microdomains. J Lipid Res. 2012;53:2586 –2597. 4. Chung S, Wang SP, Pan L, Mitchell G, Trasler J, Hermo L. Infertility and testicular defects in hormone-sensitive lipase-deficient mice. Endocrinology. 2001;142:4272– 4281. 5. Osuga J, Ishibashi S, Oka T, et al. Targeted disruption of hormonesensitive lipase results in male sterility and adipocyte hypertrophy, but not in obesity. Proc Natl Acad Sci USA. 2000;97:787–792. 6. Wang SP, Chung S, Soni K, et al. Expression of human hormonesensitive lipase (HSL) in postmeiotic germ cells confers normal fertility to HSL-deficient mice. Endocrinology. 2004;145:5688 –5693. 7. Vallet-Erdtmann V, Tavernier G, Contreras JA, et al. The testicular form of hormone-sensitive lipase HSLtes confers rescue of male infertility in HSL-deficient mice. J Biol Chem. 2004;279:42875– 42880. 8. Holst LS, Langin D, Mulder H, et al. Molecular cloning, genomic organization, and expression of a testicular isoform of hormonesensitive lipase. Genomics. 1996;35:441– 447. 9. Laurin NN, Wang SP, Mitchell GA. The hormone-sensitive lipase gene is transcribed from at least five alternative first exons in mouse adipose tissue. Mamm Genome. 2000;11:972–978. 10. Mairal A, Melaine N, Laurell H, et al. Characterization of a novel testicular form of human hormone-sensitive lipase. Biochem Biophys Res Comm. 2002;291:286 –290. 11. Kraemer FB, Shen WJ. Hormone-sensitive lipase: control of intracellular tri-(di-)acylglycerol and cholesteryl ester hydrolysis. J Lipid Res. 2002;43:1585–1594. 12. Sriram G, Martinez JA, McCabe ER, Liao JC, Dipple KM. Singlegene disorders: what role could moonlighting enzymes play? Am J Hum Genet. 2005;76:911–924. 13. Siepel A, Bejerano G, Pedersen JS, et al. Evolutionarily conserved elements in vertebrate, insect, worm, and yeast genomes. Genome Res. 2005;15:1034 –1050. 14. Siepel APK, Haussler D. New methods for detecting lineage-specific selection. Proc 10th International Conference on Research in Computational Molecular Biology, RECOMB, Venice, Italy, 2006, pp190 –205 (Abstract). 15. Braun RE, Behringer RR, Peschon JJ, Brinster RL, Palmiter RD. Genetically haploid spermatids are phenotypically diploid. Nature. 1989;337:373–376. 16. Zambrowicz BP, Harendza CJ, Zimmermann JW, Brinster RL, Palmiter RD. Analysis of the mouse protamine 1 promoter in transgenic mice. Proc Natl Acad Sci USA. 1993;90:5071–5075. 17. Holm C, Davis RC, Osterlund T, Schotz MC, Fredrikson G. Identification of the active site serine of hormone-sensitive lipase by site-directed mutagenesis. FEBS Lett. 1994;344:234 –238. 18. Holm C, Osterlund T. Hormone-sensitive lipase and neutral cholesteryl ester lipase. Methods Mol Biol. 1999;109:109 –121. 19. Wang S, Soni KG, Semache M, et al. Lipolysis and the integrated physiology of lipid energy metabolism. Mol Genet Metab. 2008;95: 117–126. 20. Zechner R, Zimmermann R, Eichmann TO, et al. FAT SIGNALS– lipases and lipolysis in lipid metabolism and signaling. Cell Metab. 2012;15:279 –291. 21. Li H, Brochu M, Wang SP, Rochdi L, Cote M, Mitchell G, Gallo-Payet N. Hormone-sensitive lipase deficiency in mice causes lipid storage in the adrenal cortex and impaired corticosterone response to corticotropin stimulation. Endocrinology. 2002;143:3333–3340. 22. Su CL, Sztalryd C, Contreras JA, Holm C, Kimmel AR, Londos C. Mutational analysis of the hormone-sensitive lipase translocation reaction in adipocytes. J Biol Chem. 2003;278:43615– 43619. 23. Wechsler A, Brafman A, Shafir M, et al. Generation of viable cholesterol-free mice. Science. 2003;302:2087.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 April 2015. at 07:44 For personal use only. No other uses without permission. . All rights reserved.
The Catalytic Function of Hormone-Sensitive Lipase is Essential for Fertility in Male Mice Shu Pei Wang, Jiang Wei Wu, Hugo Bourdages, Jean François Lefebvre, Stéphanie Casavant, Blair R. Leavitt, Damian Labuda, Jacquetta Trasler, Charles E. Smith, Louis Hermo, and Grant A. Mitchell Divisions of Medical Genetics (S.P.W., J.W.W., H.B., S.C., G.A.M.) and Hematology (J.F.L., D.L.), Department of Pediatrics, Centre Hospitalier Universitaire Sainte-Justine and Faculty of Medicine, Université de Montréal, Montréal, Québec, Canada, H3T 1C5; Valeant Cosméderme (H.B.), Laval, Québec, Canada, H7V 0A3; Centre for Molecular Medicine and Therapeutics (B.R.L.), Department of Medical Genetics, University of British Columbia, Vancouver, British Columbia, Canada, V5Z 4H4; Department of Pediatrics (J.T.), Human Genetics and Pharmacology and Therapeutics, McGill University and Research Institute of the McGill University Health Centre at the Montreal Children’s Hospital, Montréal, Québec, Canada, H3H 1P3; and Department of Anatomy and Cell Biology (C.E.S., L.H.), McGill University, Montréal, Québec, Canada, H3A 2B2
In male mice, deficiency of hormone sensitive lipase (HSL, Lipe gene, E.C.3.1.1.3) causes deficient spermatogenesis, azoospermia, and infertility. Postmeiotic germ cells express a specific HSL isoform that includes a 313 amino acid N-terminus encoded by a testis-specific exon (exon T1). The remainder of testicular HSL is identical to adipocyte HSL. The amino acid sequence of the testisspecific exon is poorly conserved, showing only a 46% amino acid identity with orthologous human and rat sequences, compared with 87% over the remainder of the HSL coding sequence, providing no evidence in favor of a vital functional role for the testis-specific N-terminus of HSL. However, exon T1 is important for Lipe transcription; in mouse testicular mRNA, we identified 3 major Lipe transcription start sites, finding numerous testicular transcription factor binding motifs upstream of the transcription start site. We directly explored two possible mechanisms for the infertility of HSL-deficient mice, using mice that expressed mutant HSL transgenes only in postmeiotic germ cells on a HSL-deficient background. One transgene expressed human HSL lacking enzyme activity but containing the testis-specific N-terminus (HSL⫺/⫺muttg mice). The other transgene expressed catalytically inactive HSL with the testis-specific N-terminal peptide (HSL⫺/⫺atg mice). HSL⫺/ ⫺muttg mice were infertile, with abnormal histology of the seminiferous epithelium and absence of spermatozoa in the epididymal lumen. In contrast, HSL⫺/⫺atg mice had normal fertility and normal testicular morphology. In conclusion, whereas the catalytic function of HSL is necessary for spermatogenesis in mice, the presence of the N-terminal testis-specific fragment is not essential. (Endocrinology 155: 3047–3053, 2014)
F
ertility in male mice requires hormone-sensitive lipase (HSL), a fatty acyl ester hydrolase expressed in adipose tissues, skeletal muscle, myocardium, adrenal cortex, macrophages, pancreas, and testis (1). In testis, HSL is predominantly located in postmeiotic germ cells although low-level HSL immunoreactivity is described in early germ cells, such as type B spermatogonia and primary spermato-
cytes and in Leydig and Sertoli cells (2, 3). HSL-deficient (HSL⫺/⫺) male mice are infertile (4, 5). This is due to a specific abnormality of spermatogenesis, with otherwise-normal masculinization in HSL⫺/⫺ males including normal mass of the seminal vesicles (4, 5). The abnormality of spermatogenesis occurs despite normal levels of testosterone, follicle stimulating hormone,
ISSN Print 0013-7227 ISSN Online 1945-7170 Printed in U.S.A. Copyright © 2014 by the Endocrine Society Received January 14, 2014. Accepted April 28, 2014. First Published Online May 5, 2014
Abbreviations: aHSL, adipose hormone-sensitive lipase; HSL, hormone-sensitive lipase; muttHSL, mutant testicular hormone-sensitive lipase.
doi: 10.1210/en.2014-1031
Endocrinology, August 2014, 155(8):3047–3053
endo.endojournals.org
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 April 2015. at 07:44 For personal use only. No other uses without permission. . All rights reserved.
3047
3048
Wang et al
HSL is Essential for Fertility in Male Mice
Endocrinology, August 2014, 155(8):3047–3053
and luteinizing hormone in HSL-deficient males (5). HSLdeficient males have normal frequency of copulation as shown by the equal prevalence of vaginal mucous plugs in females housed with HSL-deficient and with normal control males (4, 5). HSL⫺/⫺ mice that express a full length human testicular HSL transgene exclusively in postmeiotic germ cells are fertile and have normal germ cell morphology (6, 7). Therefore, expression of HSL within postmeiotic germ cells is in itself sufficient to confer fertility to HSL-deficient males, demonstrating a cell-autonomous mechanism of infertility in HSL deficiency. In testis, the transcription of the HSL gene, Lipe, differs from that of other organs. The resulting testicular isoform of HSL has a 313 amino acid N-terminal sequence encoded by an exon designated here as T1, a testis- specific exon located 16 kb upstream of the 5⬘-most exon of adipose HSL isoforms (8). A 1.4 kb region 5⬘ of exon T1 is sufficient to mediate testis-specific expression (7). Exon T1 splices to the same site in exon 1, as do the multiple other 5⬘ exons that are transcriptionally active in adipose tissue (9). The catalytic activity of purified full-length testicular HSL containing sequence from exon T1 is indistinguishable from that of adipose HSL (8). A testicular HSL mRNA with a transcription start site in a different exon (T2) is reported in human testis (10). It is predicted to encode a 775-residue protein, identical to the main adipose tissue HSL isoform. It is unknown why testicular HSL is necessary for spermatogenesis in mice. On general principals, the catalytic
a b
T1
A T2 B C
function of HSL may be essential, or HSL may play a noncatalytic, structural role in germ cells, or both. The catalytic properties of testicular HSL may enable normal fertility by hydrolyzing one or several of the fatty acyl ester substrates of HSL in postmeiotic germ cells, including diglycerides, triglycerides, or cholesteryl- and retinyl-esters (11). Alternatively, HSL might play a noncatalytic role in germ cells, unrelated to its catalytic function. Such “moonlighting” roles are known for several enzymes (12). The N-terminal testis-specific fragment of HSL accounts for a 28% peptide chain length of the molecular mass and has high contents of proline and glutamine, residues that are known to confer unique structural properties (8). Here, we explore the relative importance of HSL catalytic function and of the presence of the testis-specific sequence of HSL for the development of post meiotic germ cells in the testicle and for male fertility.
Materials and Methods Mouse exon T1 cloning and analysis of interspecies divergence The methods for sequencing the 129Sv mouse genomic Lipe clone p5⬘LipM (9), for 5⬘ rapid amplification of cDNA endsPCR (RACE-PCR) and RNase protection analysis, and for sequence alignment and prediction of potential transcription factor binding motifs are described in the Supplemental Materials. p5⬘LipM (Figure 1a) contains exon A plus approximately 2.8 kb of 5⬘ 129Sv mouse DNA (9) that we found to contain exon T1.
D
1
2 3 4 5
1 kb
67
8
9
Other HSL cDNAs
c
Testicular HSL cDNA BamHI
XhoI NdeI
d Prm 1
5’ UTR HSL cDNA 3 5 3’ UTR / PA BamHI
XhoI NdeI
S424A
HSL cDNA
3’ UTR / PA
e Prm 1
5’ UTR
Figure 1. Structure of the mouse Lipe gene and of human HSL cDNA-based transgenes. a, The mouse Lipe gene structure, showing the numerous upstream exons mediating its tissue-specific expression. b, All isoforms of HSL share the peptide sequences derived from exons 1–9. c, The testicular HSL cDNA includes a 5‘ coding sequence derived from exon T1. d and e, Structures of the HSL transgenes: (d) adipose (atg) and (e) mutant testicular (muttg), the latter showing the location of the inactivating p.Ser424Ala mutation.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 April 2015. at 07:44 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2014-1031
endo.endojournals.org
3049
We used the Translate Tool (http://www.expasy.ch/) to locate open reading frames within exon T1. The Psipred V2.2 program (http://www.psipred.net) was used to predict potential secondary structure motifs of mouse, rat (Genbank U40001), and human (Genbank U40002) testicular HSL peptides. Using the Conservation tracks of the Comparative Genomics group on the HG19 Assembly included in the UCSC table browser (http://genome.ucsc.edu), we compared conservation scores between exon T1 and the other nine Lipe exons. These tracks use two methods, phastCons (13) and PhyloP (14) from the PHAST package (http://compgen.bscb.cornell.edu/phast/ index.php), to study multiple alignments of 46 vertebrate species including 23 placental mammals and 9 primates. phastCons calculates the probabilities of a segment arising from a conserved or a nonconserved model, and assigns a log-odds score of conservation to each segment. PhyloP uses a phylogenetic model to estimate if elements are fully conserved or under lineage-specific selection.
Animal breeding and handling
Vector construction for transgenes
Genotyping of transgenic mice
Using full-length human HSL cDNAs for both adipose and testicular isoforms (6), we cloned the adipose HSL (aHSL) cDNA into pBluescript (SK) and the testicular HSL cDNA into pCDNA3.1. In this experiment we constructed two transgenic vectors, one expressing the aHSL cDNA and another expressing a mutant testicular HSL (muttHSL) cDNA. The aHSL vector was made by inserting the aHSL cDNA in the pPrCExV-1 vector (15), downstream of the mouse protamine-1 promoter that is specific for postmeiotic germ cells (6, 16). To create a catalytically-inactive human testicular HSL, we inserted the p.Ser424Ala mutation that was previously shown to produce a stable HSL protein with no detectable catalytic activity (17). The p.Ser424Ala muttHSL transgenic vector was assembled in three steps as follows. In testicular HSL cDNA, the adenine of the initiation methionine codon of the testicular exon is designated as position ⫹1. We first changed the serine 424 codon (AGT) to alanine in the full-length aHSL cDNA by changing the AG dinucleotide at positions 2173 and 2174 to GC, using site directed mutagenesis (QuikChange Site-Directed Mutagenesis Kit, Stratagene). The fragment between the SfiI site and 3⬘ end was verified by bidirectional sequencing in both constructs. Then the aHSL cDNA carrying the mutation and the full-length testicular HSL cDNA were digested with both SfiI and EcoRI, a restriction site in the polylinker of pBluescript (SK) and pcDNA3.1. The p.Ser424Ala mutation was introduced into a full-length testicular HSL cDNA by replacing the normal SfilEcoRI fragment with the mutant fragment. Finally, the fulllength muttHSL cDNA with the mutation was inserted downstream of the mouse protamine-1 promoter in the pPrCExV-1 vector (15).
To genotype the HSL⫺/⫺muttg transgenic mice, we used the same primers and PCR conditions as described for mice transgenic for human testicular HSL (6). To genotype HSL⫺/⫺atg mice, PCR was performed with the same conditions and the same sense primer but the antisense primer was LipH65 (5⬘-CGAAGAAGCACTCCTCCAGCG), complementary to nucleotides 1177 to 1198 of the human aHSL. Amplification of a 1.3 kb fragment was diagnostic for the presence of the human aHSL transgene. Founders detected by PCR were also genotyped by Southern blotting as described (6), using a human aHSL cDNA as probe. Hind III digests generated an 8 kb fragment for muttHSL transgenic mice and a 7 kb fragment for aHSL transgenic mice.
Production of transgenic mice expressing adipose or mutant testicular human HSL cDNA in postmeiotic germ cells Plasmid sequences of aHSL and muttHSL were removed by HindIII digestion and the transgenes were purified with a Gene Extraction kit (Qiagen). The purified aHSL and muttHSL vectors were microinjected into the pronucleus of FVB/N mouse embryos, and the resulting embryos implanted into pseudopregnant female mice. Founders were backcrossed 5 generations to
C57BL/6 mice, and then bred with HSL-deficient mice. The resulting male offspring have no endogenous HSL expression but express adipose (HSL⫺/⫺atg) or mutant testicular (HSL⫺/ ⫺muttg) human cDNAs. Mice were maintained on a 12-hour light, 12-hour dark photocycle and were provided with food and water ad libitum. Three-month-old mice were used in all experiments.
These studies were approved by the Animal Care Committee of CHU Sainte-Justine Research Center. All experiments were approved by the Canadian Council on Animal Care Committee of CHU Sainte-Justine accredited animal facility. Breeding and handling were performed in accordance with standard animal room practice, as described (6). Males were tested for fertility by housing each male with 2 CD1 females for 2 weeks. The birth of a litter and the number of offspring were recorded.
Northern blotting Total RNA was extracted from testes using Trizol (Invitrogen). Northern blotting was performed in ExpressHyb Hybridization Solution (Clontech) according to the manufacturer’s protocol using the probe described above for Southern blotting. -actin was used as a loading control.
Production of antihuman exon T1 antibody A peptide containing the entire peptide sequence of the human testicular exon was produced with the Bac-to-Bac baculovirus expression system (Invitrogen) according to the manufacturer’s protocol. Briefly, a 903 nt cDNA fragment of human testicular exon from the initiation methionine codon to the asparagine 301 codon (AAC) was amplified. In the amplification primer, a stop codon (TAA) was introduced immediately after asparagine 301. The amplicon was digested with BamHI and EcoRI, and then cloned into pFastBacHTA. The human exon T1 was in frame with upstream 6xHis tag. The recombinant bacmid was produced and purified from DH10Bac Escherichia coli cells, and then used to transfect SF9 insect cells. The human exon T1 peptide expressed in these cells, which included an N-terminal His tag, was purified using Ni sepharose 6 fast Flow (GE Healthcare). The His tag was removed by digestion with TEV protease. Rabbit polyclonal antihuman exon T1 antibody was produced (Chemicon International).
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 April 2015. at 07:44 For personal use only. No other uses without permission. . All rights reserved.
Endocrinology, August 2014, 155(8):3047–3053
embedded in paraffin. Sections were cut and examined in the light microscope. Details are as described previously (6).
HSL-/-ttg g
HSL-/-mu uttg2
HSL-/-mu uttg1
+/+
HSL is Essential for Fertility in Male Mice
HSL-/-atg g2
Wang et al
HSL-/-atg g1
3050
a Southern b Northern, Northern HSL
Statistical analysis kb 9.0 8.0 7.0
Data were analyzed by ordinary 1-way ANOVA followed by Turkey’s multiple comparison testing with GraphPad Prism 6 software. Differences of P ⬍ .05 were considered to be statistically significant.
44 4.4 2.4 2.4 1.4
c Northern, Actin
kDa
d Western, HSL
160 105 75 160 105
e Western, Exon T
75 Figure 2. HSL transgenes and their expression. a, Southern blot of different transgenic lines: atg, muttg and ttg (normal testicular transgene). b, Northern blot of HSL species. c, Northern blot of actin as a loading control. d, Western blot using anti-HSL antibody showing a 86 kDa protein in HSL⫺/⫺atg mice and a 117 kDa protein in HSL⫺/ ⫺muttg and HSL⫺/⫺ttg mice. e, Western blot using antihuman testicular exon antibody detects the expected 117 kDa protein only in HSL⫺/⫺muttg and HSL⫺/⫺ttg mice.
Western blotting and enzymatic studies Western blotting was performed as described (6) using the anti-HSL or antiexon T1 antibodies. For antiexon T1 antibodies, the procedure was the same but the dilution was 1:5000. Cholesteryl esterase activity was measured as described (18).
Histology Testes and epididymides of adult mice (wild type and transgenic, n ⫽ 3 for each) were removed from mice, immersed in Bouin’s fixative for 24 hours, and then dehydrated in ethanol and
Results Comparative analysis of mammalian LIPE genes In the genomic clone from 129Sv mice, exon T1 is 1.1 kb in length, contains a 313 codon open reading frame, and is located 16.4 kb upstream of exon 1 (Figure 1A). The overall amino acid sequence identity of mouse adipose HSL is 87% with rat and 75% with human HSL. Much of the divergence is in the testis-specific region. In the sequence encoded by the testicular exon, there is a predicted amino acid identity of 72% with rat and 46% with human testicular HSL, and 36% identity among all three (Supplemental Figure 1). Compared with this, the rest of the coding sequence of mouse HSL has a predicted amino acid identity of 96% with rat and 86% with human HSL. Identity was 83.9% among mouse, rat, and human HSLs. RNase protection and 5⬘ rapid amplification of cDNA ends (RACE)-PCR defined a cluster of transcription start sites 177 to 220 bp upstream of the first in-frame methionine initiation codon Supplemental Figure 2. The 849 bp upstream of this region contains potential binding motifs for germ cell-specific and general transcription factors, including SRY/SOX, GATA-1, SP-1, and C/EBP (Supplemental Figure 3). Analysis using phastCons showed that 75% of residues in exons 1–9 are predicted to be conserved, compared to
Table 1. Effect of HSL Transgene Expression on Tissue Mass and Testicular Cholesteryl Esterase Activity of 3Month-Old Male Mice Cholesteryl esterase
Mass (mg) HSL genotype
N
Testis (paired)
BAT
n
nmol/mg protein/min
HSL⫺/⫺atg1 HSL⫺/⫺atg2 HSL⫺/⫺mttg1 HSL⫺/⫺mttg2 HSL⫺/⫺ HSL⫹/⫹
5 4 7 7 7 8
205.6 ⫾ 9.2b 200.5 ⫾ 2.6b 160.0 ⫾ 9.3 177.1 ⫾ 1.8b 139.1 ⫾ 7.1d 187.8 ⫾ 8.3
181.8 ⫾ 20.1d 151.5 ⫾ 5.70d 144.9 ⫾ 27.7c 181.7 ⫾ 12.7d 161.5 ⫾ 22.7 d 72.8 ⫾ 4.8
4 4 4 4 4 4
10.99 ⫾ 2.42ac 5.98 ⫾ 0.92bc 0.10 ⫾ 0.01d 0.10 ⫾ 0.01d 0.13 ⫾ 0.01 1.95 ⫾ 0.14
P ⬍ .05 versus HSL ⫺/⫺. P ⬍ .01 versus HSL ⫺/⫺. c P ⬍ .05 versus HSL ⫹/⫹. d P ⬍ .01 versus HSL ⫹/⫹. BAT indicates brown adipose tissue. a
b
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 April 2015. at 07:44 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2014-1031
endo.endojournals.org
3051
Production and characterization of transgenic mice Vectors containing the human aHSL cDNA (Figure 1d) and human muttHSL cDNA (Figure 1e) were microinjected into single cell embryos from FVB/N mice. For aHSL, the vector was detected in 2/28 pups (7.1%), and for muttHSL, in 2/21 (9.5%). All four positive offspring were males. The presence of the transgene was confirmed on genomic Southern blots (Figure 2a). Both transgenes were transferred to a HSL-deficient background, creating mice that expressed HSL only in postmeiotic germ cells, either the stable catalytically-inactive mutant (HSL⫺/⫺muttg mice) or the adipose isoform (HSL⫺/⫺atg). Northern (Figure 2, b and c) and Western blotting (Figure 2, d and e) revealed that both adipose transgenic lines (HSL⫺/⫺atg1 and HSL⫺/⫺atg2) expressed the human HSL mRNA and protein, but HSL⫺/⫺atg1 mice had a higher level of expression than HSL⫺/⫺atg2 mice. The p.Ser424Ala HSL transgene, expressed in testes, produced easily-detectable levels of HSL mRNA and protein. As previously described for HSL⫺/⫺ mice, transgenic males had no obvious abnormality. At the time of killing at age 3 months, there was Figure 3. The effect of HSL genotype on histology of the testis and epididymis of control (a, e), no significant difference among the HSL⫺/-⫺ (b, f), HSL⫺/⫺atg1 (c, g) and HSL⫺/⫺muttg (d, h) mice. The seminiferous epithelium weights of males of each genotype: (SE) of a and c show a similar distribution and appearance of germ cells in the different tubules, unlike that seen for b and d that show abnormal tubules and absence of late spermatids. In the normal controls, 29.37 ⫾ 1.24g (n ⫽ epididymis, spermatozoa (S) are noted in the lumen (Lu) of e and g, whereas only immature 8); HSL⫺/⫺ mice, 27.59 ⫾ 1.69 g round germ cells (circles) are seen in the lumen of f and h. E indicates epididymal epithelium. (n ⫽ 7); HSL⫺/⫺ muttg1, 29.34 ⫾ Original magnification: b– d, ⫻300; a, e– h, ⫻250. 1.17 g (n ⫽ 7); HSL⫺/⫺ muttg2, 27.96 ⫾ 1.97 g (n ⫽ 7); HSL⫺/ only 12% in exon T1. Regarding the probability of neg- ⫺atg1, 28.03 ⫾ 0.65 g (n ⫽ 5), HSL⫺/⫺atg2, 30.78 ⫾ ative selection, a major cause of cross-species conserva- 0.01 g (n ⫽ 4). tion, 53% of residues in exons 1–9 achieve significance (P ⬍ .01), compared to only 3% in exon T1. Also, using Expression of human aHSL, but not of stable, phyloP, 39% of residues in exons 1–9 fulfilled the criteria enzymatically-inactive human testicular HSL, for conservation versus 5% in exon T1. Lineage-specific restores testicular cholesteryl esterase activity and selection was not detected (⬍ 3.2% in each), but in other testicular mass to HSLⴚ/ⴚ mice studies, we have found phyloP to be less sensitive for deCompared to normal control mice, cholesteryl esterase tection of lineage-specific selection than for conservation. activity was 5.6-fold greater in HSL⫺/⫺atg1 testicle ho-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 April 2015. at 07:44 For personal use only. No other uses without permission. . All rights reserved.
3052
Wang et al
HSL is Essential for Fertility in Male Mice
mogenates (P ⬍ .05) and 3-fold greater for HSL⫺/⫺atg2 homogenates (P ⬍ .05). As expected, cholesteryl esterase activity in HSL⫺/⫺muttg testicle was not distinguishable from that in HSL⫺/⫺ mice (Table 1). HSL⫺/⫺atg mice, expressing a catalytically-active aHSL transgene, had normal testis weight (Table 1). Interestingly, testicular masses of HSL⫺/⫺muttg mice were also significantly greater than those of HSL⫺/⫺ mice, despite their lack of cholesteryl esterase activity (Table 1). Examination of interscapular brown fat, a control for the specificity of transgenic HSL expression, showed the characteristics of HSL⫺/⫺ mice: increased mass (Table 1) and abnormal mono-vacuolar morphology, as previously-reported in HSL⫺/⫺ mice (6) (not shown). Testicular expression of normal human aHSL, which lacks the N-terminal testicle-specific sequence, confers normal testicular histology and normal fertility to HSLⴚ/ⴚ mice Wild type mice revealed a full complement of the 12 stages of the cycle of the seminiferous epithelium. All generations of germ cells were noted including the postmeiotic spermatids (Figure 3a). The epididymides of wild-type mice demonstrated an epididymal lumen full of spermatozoa (Figure 3e). In the HSL⫺/⫺mice, the seminiferous epithelium was devoid of postmeiotic germ cells, and some tubules were abnormal in appearance (Figure 3b), as published previously (6). Similarly, the lumen of the epididymis was devoid of spermatozoa but rather contained many small round cells (Figure 3f), shown previously to be immature germ cells (6). HSL⫺/⫺ atg1 mice presented a similar distribution of germ cells in the seminiferous epithelium to that of control mice (Figure 3c) and spermatozoa were abundant in the epididymal lumen (Figure 3g), as noted for controls. In contrast, the histology of the seminiferous epithelium of HSL⫺/⫺muttg mice was abnormal with an absence of postmeiotic germ cells and abnormal appearance of some tubules (Figure 3d). The epididymal lumen also demonstrated an absence of spermatozoa but abundance of round immature germ cells (Figure 3h). Expression of the aHSL transgene but not of the muttHSL transgene conferred normal fertility to HSL⫺/⫺ mice (Table 2). Seven out of eight HSL⫺/⫺atg males tested produced litters with each female, with a mean litter size of 10.29 ⫾ 1.28 offspring. In contrast, none of the 9 HSL⫺/⫺muttg males produced offspring.
Discussion HSL is a fatty-acyl esterase that is active against numerous substrates and was first identified in adipose tissue. Initial
Endocrinology, August 2014, 155(8):3047–3053
Table 2. Effect on Fertility of the Expression of HSL Transgenes on a HSL⫺/⫺ Background
Genotype
Males
Pregnancies/ females mated
HSL⫺/⫺atg1 HSL⫺/⫺atg2 HSL⫺/⫺muttg1 HSL⫺/⫺muttg2
4 3 4 5
7/8 4/4 0/8 0/9
Litter size 10.29 ⫾ 1.28 10.25 ⫾ 1.11 0 0
studies of HSL focused on its role in adipose tissue, in which it has a complex pattern of phosphorylation that correlates with highly-regulated shifts in location to and from the surface of the lipid droplet (1). Molecular studies and the creation of mice with modifications of the Lipe gene revealed the presence of HSL-independent lipolysis (19, 20) and that HSL plays important roles in nonadipose tissues including adrenal cortex (21) and testis (4 – 6). Whereas HSL-deficient male mice are infertile and show marked structural abnormalities in postmeiotic germ cells (4, 5), expression of a human testicular HSL transgene exclusively in postmeiotic germ cells confers normal fertility and normal testicular histology to these mice, showing that the role of HSL in normal male fertility in mice is cell-autonomous, and carried out within postmeiotic germ cells. Because of the small size and anatomical complexity of mouse testes, classical biochemical studies are not possible, and we pursued molecular and genetic dissection of the role of HSL in male fertility. First we examined the sequences of testicular HSL from mice, humans and rats, using several measures of amino acid conservation. By all measures, the conservation of exon T1 was strikingly less than that of the rest of the coding sequence. Such a lack of conservation argues against an important functional role of the N terminus of testicular HSL in male fertility. In contrast, exon T1 has clear importance as the major start site for Lipe transcription in mice (10). We identified major upstream start transcriptional sites and found several upstream transcription factor binding motifs. Together, these findings incited us to directly study the importance of the N-terminal region of the testicular isoform in transgenic mice. HSL⫺/⫺atg mice, which express a catalytically-active HSL that lacks the testes-specific N-terminal sequence, conclusively demonstrate that the N-terminal sequence of testicular HSL isoform is not necessary for fertility; they show normal fertility, normal testicular mass, and normal microscopic morphology. Conversely, in HSL⫺/⫺muttg mice, the presence of catalytically-inactive HSL that contained the sequence encoded by T1 did not produce male fertility. In studies of purified HSL, inclusion of the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 April 2015. at 07:44 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2014-1031
p.Ser424Ala mutation results in a stable HSL peptide that has no detectable enzymatic activity (17). In the adipocyte HSL isoform, p.Ser424Ala abolishes another function of HSL, translocation to lipid droplets after -adrenergic stimulation (22). Interestingly, HSL⫺/⫺muttg mice, expressing inactive testicular HSL containing p.Ser424Ala, showed some increase of testicular mass, although testicular histology was abnormal as in HSL⫺/⫺ mice. Therefore, the presence of a structurally-intact HSL testicular isoform may have some effect upon the testicle, but this is insufficient to confer fertility or normal histology in the absence of HSL activity. These experiments do not reveal the precise substrates of HSL in postmeiotic germ cells. HSL can cleave fatty acyl esters of glycerol (diglycerides and triglycerides), retinoic acid, and cholesterol (11); other substrates may yet be unidentified. The infertility of HSL-deficient male mice could potentially arise from the noncleavage of any of these substances. Some circumstantial evidence is consistent with a cholesterol-related mechanism. HSL accounts for all of the measureable cholesteryl esterase activity in testis (5, 6), suggesting that testicular cholesterol metabolism may be severely affected in HSL deficiency. Male mice with disruption of the Dhcr24 gene that are deficient in cholesterol biosynthesis are infertile (23), and in HSL ⫺/⫺ mouse testes, expression of the cholesterol- and lipoprotein-related transcripts SR-BI, SR-BII, and LIMP II are increased (3). Although the exact mechanism is unclear by which HSL is necessary for male fertility in mice, this study clearly shows that the presence of catalytically-active HSL in postmeiotic germ cells can confer normal fertility even in the absence of the testis-specific N-terminal peptide.
Acknowledgments Address all correspondence and requests for reprints to: Grant A. Mitchell, Divisions of Medical Genetics and Hematology, Department of Pediatrics, CHU Sainte-Justine and Faculty of Medicine, Université de Montréal, 3175 Côte Sainte-Catherine, Montréal, Québec, Canada, H3T 1C5. Email: [email protected]. This work was supported in part by a Canadian Institutes of Health Research grant to G.M. The creation of the transgenic mice described in this article was financed by a Canadian Genetic Diseases Network collaborative project between B.L. and G.M. The Research Institutes of CHU Sainte-Justine and of the McGill University Health Centre at the Montreal Children’s Hospital are supported in part by the Fonds de la Recherche du Québec-Santé. Disclosure Summary: The authors have nothing to disclose.
References 1. Holm C, Osterlund T, Laurell H, Contreras JA. Molecular mechanisms regulating hormone-sensitive lipase and lipolysis. Ann Rev Nutr. 2000;20:365–393.
endo.endojournals.org
3053
2. Arenas MI, Lobo MV, Caso E, Huerta L, Paniagua R, Martin-Hidalgo MA. Normal and pathological human testes express hormone-sensitive lipase and the lipid receptors CLA-1/SR-BI and CD36. Hum Pathol. 2004;35:34 – 42. 3. Casado ME, Huerta L, Ortiz AI, et al. HSL-knockout mouse testis exhibits class B scavenger receptor upregulation and disrupted lipid raft microdomains. J Lipid Res. 2012;53:2586 –2597. 4. Chung S, Wang SP, Pan L, Mitchell G, Trasler J, Hermo L. Infertility and testicular defects in hormone-sensitive lipase-deficient mice. Endocrinology. 2001;142:4272– 4281. 5. Osuga J, Ishibashi S, Oka T, et al. Targeted disruption of hormonesensitive lipase results in male sterility and adipocyte hypertrophy, but not in obesity. Proc Natl Acad Sci USA. 2000;97:787–792. 6. Wang SP, Chung S, Soni K, et al. Expression of human hormonesensitive lipase (HSL) in postmeiotic germ cells confers normal fertility to HSL-deficient mice. Endocrinology. 2004;145:5688 –5693. 7. Vallet-Erdtmann V, Tavernier G, Contreras JA, et al. The testicular form of hormone-sensitive lipase HSLtes confers rescue of male infertility in HSL-deficient mice. J Biol Chem. 2004;279:42875– 42880. 8. Holst LS, Langin D, Mulder H, et al. Molecular cloning, genomic organization, and expression of a testicular isoform of hormonesensitive lipase. Genomics. 1996;35:441– 447. 9. Laurin NN, Wang SP, Mitchell GA. The hormone-sensitive lipase gene is transcribed from at least five alternative first exons in mouse adipose tissue. Mamm Genome. 2000;11:972–978. 10. Mairal A, Melaine N, Laurell H, et al. Characterization of a novel testicular form of human hormone-sensitive lipase. Biochem Biophys Res Comm. 2002;291:286 –290. 11. Kraemer FB, Shen WJ. Hormone-sensitive lipase: control of intracellular tri-(di-)acylglycerol and cholesteryl ester hydrolysis. J Lipid Res. 2002;43:1585–1594. 12. Sriram G, Martinez JA, McCabe ER, Liao JC, Dipple KM. Singlegene disorders: what role could moonlighting enzymes play? Am J Hum Genet. 2005;76:911–924. 13. Siepel A, Bejerano G, Pedersen JS, et al. Evolutionarily conserved elements in vertebrate, insect, worm, and yeast genomes. Genome Res. 2005;15:1034 –1050. 14. Siepel APK, Haussler D. New methods for detecting lineage-specific selection. Proc 10th International Conference on Research in Computational Molecular Biology, RECOMB, Venice, Italy, 2006, pp190 –205 (Abstract). 15. Braun RE, Behringer RR, Peschon JJ, Brinster RL, Palmiter RD. Genetically haploid spermatids are phenotypically diploid. Nature. 1989;337:373–376. 16. Zambrowicz BP, Harendza CJ, Zimmermann JW, Brinster RL, Palmiter RD. Analysis of the mouse protamine 1 promoter in transgenic mice. Proc Natl Acad Sci USA. 1993;90:5071–5075. 17. Holm C, Davis RC, Osterlund T, Schotz MC, Fredrikson G. Identification of the active site serine of hormone-sensitive lipase by site-directed mutagenesis. FEBS Lett. 1994;344:234 –238. 18. Holm C, Osterlund T. Hormone-sensitive lipase and neutral cholesteryl ester lipase. Methods Mol Biol. 1999;109:109 –121. 19. Wang S, Soni KG, Semache M, et al. Lipolysis and the integrated physiology of lipid energy metabolism. Mol Genet Metab. 2008;95: 117–126. 20. Zechner R, Zimmermann R, Eichmann TO, et al. FAT SIGNALS– lipases and lipolysis in lipid metabolism and signaling. Cell Metab. 2012;15:279 –291. 21. Li H, Brochu M, Wang SP, Rochdi L, Cote M, Mitchell G, Gallo-Payet N. Hormone-sensitive lipase deficiency in mice causes lipid storage in the adrenal cortex and impaired corticosterone response to corticotropin stimulation. Endocrinology. 2002;143:3333–3340. 22. Su CL, Sztalryd C, Contreras JA, Holm C, Kimmel AR, Londos C. Mutational analysis of the hormone-sensitive lipase translocation reaction in adipocytes. J Biol Chem. 2003;278:43615– 43619. 23. Wechsler A, Brafman A, Shafir M, et al. Generation of viable cholesterol-free mice. Science. 2003;302:2087.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 April 2015. at 07:44 For personal use only. No other uses without permission. . All rights reserved.
