| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Laboratories of Veterinary Ethology3 and Veterinary Physiology,4 The University of Tokyo, Tokyo 113-8657, Japan Laboratory of Neurobiology,5 National Institute of Agrobiological Sciences, Tsukuba 305-8602, Japan
ABSTRACT
The male effect is a well-known phenomenon in female sheep and goats whereby a pheromone-induced activation of reproductive function occurs. However, the molecule(s) involved in this phenomenon are unknown. We investigated gene expression profiles for the induction of male effect pheromone synthesis using a PCR-based cDNA subtraction strategy. We constructed two subtracted cDNA libraries using mRNA from the skin of the head or rump region of orchidectomized male goats with or without pheromone induction using testosterone or dihydrotestosterone (DHT). Both libraries were assumed to contain genes whose expression increases with pheromone induction. Clones (n = 480) from each library were sequenced and identified using BLAST to reveal 115 and 239 types of sequences in the libraries of the head and rump region, respectively. Among these, 12 genes were expressed in both libraries. We conducted real-time PCR to further analyze their expression using cDNA samples derived from pheromone-producing or nonproducing skin from the head of an ovariectomized female goat with or without DHT implantation, respectively. For nine genes, we observed significantly increased expression in samples following DHT implantation. Among these, stearoyl-CoA desaturase 1 (SCD1) and elongation of long chain fatty acids family member 5 (ELOVL5) genes showed more than 100-fold higher expression levels in pheromone-positive samples, suggesting that the products of these genes may be important in pheromone synthesis.
gene regulation, implantation, pheromones, seasonal reproduction, testosterone
Pheromones play important roles in various social interactions among individuals of some mammalian species [1]. In sheep [2] and goats [3], the male effect is a well-known pheromone action in which the introduction of a male induces out-of-season ovulation in the female. The central target of this pheromone is thought to be the hypothalamic GnRH pulse generator, which regulates the intermittent GnRH discharge into the pituitary portal circulation and thereby modulates the pulsatile secretion of LH to thus act as a key determinant of reproductive function in mammals [4]. We previously developed a method to monitor the activity of the GnRH pulse generator in goats (Capra hircus) via an electrode inserted in the hypothalamus [5], and confirmed that the hair of sexually mature male goats induces the electrophysiological manifestation of the GnRH pulse generator in females [6]. Production of the male effect pheromone is testosterone (T)-dependent and occurs in the skin of the head to shoulder region, but not the flank, back, or rump regions [7]. Immunohistochemical observations showed that this difference in pheromone production among body regions is correlated with the expression pattern of 5
-reductase, an enzyme that converts T to dihydrotestosterone (DHT) [8]. In orchidectomized (ODX) goats, DHT induces production of the male effect pheromone in skin of both the head and rump regions; however, sebaceous glands develop much less in the rump than in the head region, and pheromone synthesis does not necessarily occur along with the development of sebaceous glands [9]. Interestingly, DHT also induces male effect pheromone production in the skin of the head region in ovariectomized (OVX) female goats [10]. Thus, it is thought that the male effect pheromone is produced in a steroid-dependent manner in the skin of specific regions on the goat.
The purification of the pheromone molecule(s) responsible for the male effect has not yet been successful, despite several biochemical analyses using ram's fleece [11] and hairs from the head and neck regions of male goats [12, 13]. Information concerning the genes involved in pheromone synthesis and the chemical structure of this pheromone may help to advance the purification procedure. The rationale for the strategy we took herein was based on ideas derived from insect studies. Although few articles are available regarding the mechanisms of primer pheromone synthesis in mammals, more studies have been published on insects. The synthesis of certain pheromones in beetles, for example, is also under endocrine regulation as is seen in goats. Juvenile hormone III (JH III) is known to induce the process of pheromone synthesis. By implanting JH III, studies have shown that the expression of genes encoding key enzymes involved in the pheromone synthetic pathway are enhanced in certain body regions responsible for pheromone production in the beetle [14–16]. In goats as well, we expect the genes that increase their expression concurrent with pheromone production may be associated. For identifying the male pheromone in goats, we consider this method to be a practical way of searching for candidate genes of key enzymes in the pheromone synthesis cascade, as we lack information about the chemical identity of responsible molecules.
In this study, we tested the hypothesis that the expression of genes encoding enzymes involved in the pathway of pheromone synthesis would increase in response to androgen treatment known to induce "male pheromone" production. If this is correct, we would be able to identify those candidate genes by searching for similarly enhanced gene expression patterns across different experimental models (i.e., male and female goat models for inducing pheromone production). For this purpose, we took a stepwise approach. First, we investigated gene expression profiles of skin associated with the induction of male effect pheromone synthesis using a PCR-based cDNA subtraction strategy. Two subtracted cDNA libraries were constructed using the suppressive subtractive hybridization (SSH) technique [17, 18]. One library was constructed using skin from the head of an ODX goat, taken prior to and after T treatment. The other was prepared using skin from the rump of another ODX goat, taken with and without DHT treatment. Because T implantation raised plasma T levels to approximately that of gonadally intact males [19], similar gene expression in the head region of an intact male was anticipated for the former model. However, the size (major axis) of sebaceous glands in the head region was enlarged more than seven times by T implantation [9]; thus, the subtractive library would also include genes involved in the development of the sebaceous glands. It would be difficult to separate these two sets of genes in the gene expression lists. In addition, DHT implantation may increase plasma DHT to levels greater than those in intact males and induce unexpected or inappropriate expression of genes in the skin, also making it difficult to detect genes involved only in pheromone synthesis using this library. However, compared to T transplantation, DHT implantation induced the development of sebaceous glands to a much lesser extent in the rump region, and the subtractive gene expression library of the rump region may contain fewer genes involved in the development of sebaceous glands. Both libraries may also contain genes that are not involved in pheromone synthesis; however, genes found in both libraries were considered candidates for genes involved in pheromone synthesis. The reason why we used only one male goat for constructing these libraries was to minimize the chance of confusion caused by pooling mRNAs derived from several subjects. This is based upon a preceding study in which subtractive libraries were constructed from only one subject [20].
From each library, 480 clones were sequenced and examined using BLAST to produce lists of genes that were expressed, in order to identify key genes involved in pheromone synthesis. We identified genes that showed increased expression in both ODX models. In the second step, the candidate genes were tested to determine whether their expression rates would actually increase along with pheromone synthesis in the third model—namely, in DHT-treated OVX female goats. For quantitative analyses of the expression rate of each candidate gene, six OVX goats were used. The rationale for using female goats was to further narrow down the candidate genes by excluding any male-specific, but pheromone-unrelated, genes. Quantitative analyses of 12 genes that occurred in both libraries were conducted using real-time PCR for cDNA samples derived from the skin of the head of OVX goats treated with or without DHT.
All adult male and female goats were obtained from a closed colony at the experimental station of the University of Tokyo, Japan. Males were castrated at least 6 months prior to the experiment. They were housed under conditions of natural day length and temperature and fed daily with sufficient amounts of dry hay and formula feed. Water was available ad libitum.
For study 1, we used two ODX goats. Each goat was implanted with capsules containing either T or DHT, as described previously [7, 9]. Skin samples were taken immediately before (ODX sample) and 28 days after implantation (ODX+T or ODX+DHT samples). Goats were anesthetized using ketamine hydrochloride and xylazine hydrochloride, and two squares of skin (1 x 1 cm for the bioassay and 2 x 2 cm for the RNA extraction) were cut using a scalpel. Skin samples were taken from the head region between the horns for the goat treated with T and from the rump region for the goat treated with DHT. Skin samples for the bioassay were stored at 4°C, whereas those for RNA extraction were stored in RNAlater (Takara Bio Inc., Shiga, Japan) according to the manufacturer's instructions.
For study 2, six OVX goats were implanted with capsules containing DHT. The first skin sample (OVX+DHT sample) was collected 28 days after DHT implantation and the DHT capsules were removed at the same time. The second skin sample (OVX sample) was collected 28 days later. Two skin squares (1 x 1 cm) were collected from the head region and processed similar to those for the ODX goat.
This experiment was approved by the Animal Care and Use Committee of the Faculty of Agriculture, the University of Tokyo.
Bioassay for Pheromone Activity
Pheromone activity was examined using previously reported methods [6, 7, 21]. Briefly, each skin sample was homogenized using an ultrasonic homogenizer (Tomy Co., Tokyo, Japan) and extracted with diethyl ether. A gauze pad was soaked in the ether extract and then completely dried. The dried gauze sample was placed in a plastic cup in front of the nose of an ovariectomized "detector" goat, into which arrays of bilateral recording electrodes were implanted at the medial basal hypothalamus for monitoring hypothalamic multiple-unit activity (MUA) volleys, which reflect hypothalamic GnRH pulse generator activity. The methods for stereotaxically implanting the recording electrodes and for continuously monitoring the hypothalamic GnRH pulse generator activity in the goat have been previously described in detail [5, 21]. Exposure to each testing sample was timed midway between spontaneous GnRH pulses, and when a MUA volley occurred within 2 min of the sample presentation, the sample was determined to contain male effect pheromone activity [6, 7, 21].
Study 1: Subtraction Analysis of Gene Expression Induced by T or DHT
RNA extraction. Total RNA was extracted from the skin samples for the subtraction analysis. Each sample was trimmed using a scalpel and extracted with 5 ml of Trizol reagent (Gibco BRL, Burlington, ON, Canada). Poly(A)+ RNA was then isolated from the total RNA using the Oligotex-dT30 mRNA purification kit (Takara Bio Inc.). The extracted RNA was dissolved in water and quantified spectrophotometrically at 260 nm.
Suppressive subtractive hybridization. SSH was performed using the PCR-select cDNA subtraction kit (BD Biosciences Clontech, Mountain View, CA), as described by the manufacturer, to construct two libraries. For the first library, 2.5 µg of mRNA from the head skin of the ODX+T sample was used as the "tester" and the same amount from the ODX sample was used as the "driver." For the second library, the tester and the driver were mRNA from the rump skin of the ODX+DHT and ODX samples, respectively. PCR was performed with TaKaRa Ex Taq (Takara Bio Inc.) in a TaKaRa PCR thermal cycler GP (Takara Bio Inc.). The subtracted PCR products generated by SSH were cloned into the EcoRV site of the pT7Blue T-vector (Novagen, Madison, WI) and then transformed into ElectroMAX DH10B cells (Invitrogen, Carlsbad, CA) by electroporation using a Gene Pulser II (Bio-Rad Laboratories, Hercules, CA).
DNA sequencing and data analysis. Clones (n = 480) were obtained from each subtracted library and sequenced using a MegaBACE1000 DNA sequencer (Amersham Biosciences Co., Piscataway, NJ). Sequencing reactions were carried out using a BigDye Terminator v1.1 cycle sequencing kit (PE Applied Biosystems, Foster City, CA) and the T7 primer (5'-TAATACGACTCACTATAGGG-3'). Similarity searches were performed on sequences obtained from GenBank using BLAST software with the filter set to default. A match was considered significant if a similarity of more than 75% occurred for a continuous overlapping region of at least 70 bp. When a clone sequence matched the caprine mitochondrial sequence (accession number: AF533441), we identified the gene based on the study by Parma et al. [22]. Complementary DNA sequences were categorized according to their sequence identity with the GenBank database: known genes, mitochondrial genes, uncharacterized sequences, and unknown sequences.
Study 2: Quantitative Analysis of Gene Expression Using Real-Time PCR
Real-time PCR. Twelve genes from the gene expression lists generated in study 1 were found in both libraries, and an increase in their expression was confirmed quantitatively using real-time PCR. Total RNA (1 µg) extracted from the OVX and OVX+DHT samples was reverse transcribed into cDNA using SuperScript III reverse transcriptase (Invitrogen) according to the manufacturer's instructions. Primers were designed based on the sequence of the cloned cDNA segments (Supplemental Table 1, available online at www.biolreprod.org) using the Primer 3 program (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi). Real-time PCR amplification was performed in duplicate in 25-µL reaction volumes consisting of 400 nM of each primer, 15 µL of QuantiTect SYBR Green PCR Master Mix (Qiagen, Amsterdam, The Netherlands), 1 µL of 25 mM MgCl2 (Roche, Basel, Switzerland), and 5 µL of cDNA. Cycling conditions were 15 min at 95°C; 35 cycles of 15 sec at 95°C, 20 sec at 60°C, and 20 sec at 72°C; followed by a chain elongation period at 72°C for 10 min. Amplification, detection, and data analysis were performed using a LightCycler 480 system (Roche). Samples were normalized using one of two reference genes that were selected based on the gene expression levels, namely beta actin (ACTB; AF481159) or aminolevulinate delta-synthase 1 (ALAS1; AB232536). To confirm the greatly increased expression of the two genes revealed using real-time PCR, we designed new primer pairs for the coding region (Supplemental Table 1) and carried out real-time PCR using the same cDNA samples. Data from real-time PCR were analyzed using StatView 5.0J (Abacus Concepts, Berkeley, CA). The effects of the DHT treatment (OVX vs. OVX+DHT) on the mRNA levels of the 12 genes were analyzed separately using Wilcoxon's signed-ranks tests. Results are presented as the mean ± standard error and were considered significant at P < 0.05.
|
Sequencing of ELOVL5. Among the 12 genes examined, real-time PCR analysis showed that the expression of stearoyl-CoA desaturase (SCD) and elongation of long chain fatty acids family member 5 (ELOVL5) increased markedly with DHT treatment. Whereas the full sequence of the caprine SCD gene (AF325499) on caprine chromosome 26 has already been reported [23, 24], to our knowledge, that of the ELOVL5 gene has not. Therefore, we determined the full sequence of the caprine ELOVL5 gene by RT-PCR, using consensus primers based on sequences from other mammals such as humans (NM_021814), rats (NM_134382), and mice (NM_134255) and the rapid amplification of cDNA ends (RACE) method using a BD SMART RACE cDNA amplification kit (BD Biosciences Clontech). The obtained result was then used to design new primer pairs for the coding region for a second real-time PCR analysis.
Bioassay for Pheromone Activity
The exposure of the "detector" goat to an ether extract of the ODX+T, ODX+DHT, or OVX+DHT samples resulted in the induction of an abrupt increase in MUA volleys within 2 min of exposure, as shown in our previous report [6, 7, 21]. These results indicate that the samples were pheromone-positive. In contrast, no changes were observed in response to an ether extract of the OVX or ODX sample, indicating that no synthesis of the male effect pheromone occurred at the time of sampling.
Study 1: Subtraction Analysis of Gene Expression Induced by T or DHT
Two subtracted cDNA libraries were constructed. The subtraction efficiency was evaluated using a housekeeping gene, glyceraldehyde-3-phosphate dehydrogenase (GAPDH). PCR products were observed after 18 PCR cycles in the unsubtracted sample, whereas 20 additional PCR cycles were required for detection by agarose gel electrophoresis in the subtracted sample in both libraries (data not shown). This indicates a marked decrease in GAPDH abundance in the subtracted samples. After cloning and transforming the subtracted cDNAs, 480 clones from each subtracted library were amplified using PCR. Inserts were found for 452 clones from the head library and 430 clones from the rump library, and all 882 cDNA inserts were sequenced. The sequences were identified using BLAST. In the head library, 401 sequences consisting of 87 known genes, 29 sequences of 6 mitochondrial genes, 9 uncharacterized sequences of 9 types, and 13 unknown sequences of 13 types were identified (Fig. 1A; Supplemental Table 2, available online at www.biolreprod.org). In the rump library, 346 sequences consisting of 186 known genes, 1 sequence of 1 mitochondrial gene, 38 uncharacterized sequences of 20 types, and 45 unknown sequences of 32 types were identified (Fig. 1B; Supplemental Table 3, available online at www.biolreprod.org).
|
|
Twelve genes occurred in both libraries (Table 1): acyl-CoA synthetase long-chain family member 1 (ACSL1), adipose differentiation-related protein (ADFP), HLA-B associated transcript 1 (BAT1), biliverdin reductase B (BLVRB), ectodysplasin A (EDA), ELOVL5, ferritin heavy polypeptide 1 (FTH1), prohibitin 2 (PHB2), ribophorin I (RPN1), SCD, solute carrier family 15 (oligopeptide transporter) member 1 (SLC15A1), and testis enhanced gene transcript (BAX inhibitor 1; TEGT). All sequences were very similar to those of other mammals and were submitted to GenBank (AB231299–AB231310). None of the uncharacterized and unknown sequences occurred in both libraries.
Study 2: Quantitative Analysis of Gene Expression Using Real-Time PCR
Real-time PCR. To confirm the differential expression of the 12 genes identified, we performed real-time PCR using specific primer pairs (Supplemental Table 1, available online at www.biolreprod.org). Among the 12 candidate genes, nine showed significantly higher expression in the OVX+DHT samples than in the OVX samples (Table 2). The BAT1 gene showed a diverse expression pattern. Because the SCD and ELOVL5 genes showed more than 100-fold higher expression levels in pheromone-positive samples, new primer pairs for these two genes were designed based on sequences in the open reading frame. Additional real-time PCR was carried out using the same cDNA to confirm the large increases in expression observed in the first real-time PCR, and increases greater than 100-fold for the expression of both genes were confirmed (data not shown).
Sequencing of ELOVL5. We determined the full sequence of ELOVL5 using RT-PCR and RACE methods. The full sequence of caprine ELOVL5 consisted of 2764 bp, with a 900-bp open reading frame encoding 299 amino acids (AB232151). The goat ELOVL5 sequence showed high homology with known amino acid sequences from humans (91.3%), rats (90.6%), and mice (90.3%). Furthermore, the goat sequence contained all residues identified by Leonard et al. [25] as important in the elongation of enzymes.
We generated expression profiles for genes involved in the induction of pheromone synthesis by T or DHT. Twelve genes occurred in both gene expression libraries; these were analyzed quantitatively, and nine genes were confirmed to be expressed at significantly increased levels from 1.73 to more than 100-fold control levels (Table 2). These nine genes met at least one criterion to be considered candidate genes involved in the male effect pheromone biosynthesis; they showed increased expression in a manner that parallels pheromone production in regions where the pheromone is synthesized. Among them, the SCD and ELOVL5 genes were given priority as candidate genes because increases in their rates of expression were much greater than those of the other seven genes. However, we cannot exclude the possibility that these candidate genes have nothing to do with pheromone synthesis but are simply expressed for some other unknown function. To address this issue we need to investigate the role of these genes by conducting experiments such as RNA interference mediated loss-of-function studies, which were successfully applied to the study of the Bombyx mori sex pheromone [26].
SCD catalyzes a critical step in the biosynthesis of monounsaturated fatty acids from saturated fatty acids through the introduction of a cis double bond at the delta-9 position. In addition, it affects a variety of key physiological activities, including insulin sensitivity, metabolic rate, and adiposity [27]. Divergence in the number of SCD genes occurs among mammals; SCD1 and SCD5 genes occur in the human genome [28, 29], two Scd genes have been cloned in the rat [30], and four have been cloned in the mouse [31, 32]. In contrast, only one SCD gene has been detected to date in the cow [33], pig [34], sheep [35], and goat [24]. The evolutionary lineage suggests that the Scd2, Scd3, and Scd4 genes appeared after the divergence of small rodents from other mammals, and SCD5 may have resulted from a single gene duplication event that occurred before the divergence of rodents and other higher mammals [29]. Because SCD5-like genomic DNA fragments are found in the dog [29] and cow, an ortholog of human SCD5 may occur in the goat. However, human SCD5 is expressed abundantly in the adult brain and pancreas [29], and we detected only one band with an estimated size of 5 kb in mRNA derived from head skin of a gonadally intact male goat using Northern blot analysis (data not shown), as was reported in lactating goat mammary glands by Bernard et al. [24]. These results suggest that the partial sequence obtained from both libraries may actually be part of the SCD1 gene, rather than SCD5. Increased expression as a result of T administration has also been observed in murine Scd1 and Scd3 [36] and hamster Scd1 [37]. In these animals, the induction of SCD by T may contribute to various androgen dependent physiological phenomena that are currently unknown (e.g., chemical communication).
BLAST analysis and real-time PCR suggested that the expression of the ELOVL5 and SCD1 genes increased by similar amounts upon induction of the male effect pheromone by treatment with T or DHT. However, a full gene sequence has not been reported for ELOVL5. We determined that the nucleotide sequence of the caprine ELOVL5 consists of 2764 bp with a 900-bp open reading frame encoding 299 amino acids (AB232151) and that it shares high homology with other mammalian ELOVL5 sequences. The predicted amino acid sequence includes the conserved "histidine box" and all identical and highly conserved residues in the 22 members of the ELO family of elongating enzymes characterized by Leonard et al. [25]. Thus, the caprine ELOVL5 may be involved in the elongation of various polyunsaturated long-chain fatty acids of C18–C20, as is the human ELOVL5 [38].
SCD1 catalyses the delta-9 desaturation of various fatty acyl-CoA substrates [39], and ELOVL5 is involved in the elongation of various polyunsaturated long-chain fatty acids of C18–C20 [38]. If these enzymes are important in pheromone synthesis as expected, the pheromone molecule(s) may well be unsaturated long-chain fatty acids or metabolic product(s) thereof. Among pheromone molecules detected in mammals [1], dodecyl propionate secreted from the preputial glands of pups that evokes anogenital licking in rats [40], and (Z)-7-dodecen-1-yl acetate found in preovulatory urine of female elephants, which signals their readiness to mate [41], have large molecular weights. Thus, polyunsaturated long-chain fatty acids may be responsible for the male effect in goats. The molecule 4-ethyloctanoic acid (4-EOA) is secreted from the sebaceous glands of male goats and is the main constituent causing the strong caprine odor [42]. This molecule has a "releaser" pheromone effect whereby the female goat shows interest in 4-EOA [13], but not a "primer" pheromone effect that elicits the female LH response [43]. When considering the function of the nine genes detected in our analyses and the structure of 4-EOA [42, 44], none of these genes were considered to be directly associated with the specific structure of 4-EOA.
In summary, we produced gene expression profiles for two independent inductions of male effect pheromone synthesis to identify genes involved in pheromone synthesis and investigated the synthesis pathway for pheromone molecule(s) responsible for the male effect in the goat. We considered SCD1 and ELOVL5 genes to be associated with this pheromone synthesis with the highest probability because their expression increased remarkably in a manner that paralleled pheromone biosynthesis. Therefore, we expect the pheromone molecule(s) to be an unsaturated long-chain fatty acid or a metabolic product thereof. Our gene expression profiles may be useful in understanding the genes involved in pheromone synthesis and regulation once the pheromone molecule is identified, and our analysis strategy may aid in the identification of primer pheromone molecules that have not been identified in certain mammals because of limitations in applying the bioassay method.
ACKNOWLEDGMENTS
We thank the staff of the experimental station of the University of Tokyo. We also thank Eri Iwata, Satoru Tomimatsu, Hiromi Ohara, and Ryuji Sako for technical assistance.
FOOTNOTES
1Supported by Grants-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (15GS0306) and by Grants-in-Aid for Animal Behavior Program from the Ministry of Agriculture, Forestry, and Fisheries, Japan. 
Correspondence: 2Yukari Takeuchi, Laboratory of Veterinary Ethology, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan. FAX: 81 3 5841 8190; e-mail: aytake{at}mail.ecc.u-tokyo.ac.jp
Received: 20 November 2006.
First decision: 19 December 2006.
Accepted: 23 March 2007.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
