| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular article |
a Israel Oceanographic and Limnological Research, National Institute of Oceanography, Tel-Shikmona, Haifa 31080, Israel
b The Israel Institute for Biological Research, Ness-Ziona, Israel
c Departments of Biochemistry and Molecular Biology and Medicine and the Howard Hughes Medical Institute, University of Chicago, Chicago, Illinois 60637
ABSTRACT
Expression of insulin-like growth factor (IGF)-I, IGF-II, and IGF type I receptor (IGF-1R) genes was studied in gonads at different developmental stages of the protandrous hermaphroditic species the gilthead seabream (Sparus aurata) by reverse transcription-polymerase chain reaction and Northern blot analysis. Both IGF-I and IGF-II mRNA levels were highest in bisexual gonads and decreased during gonadal development. Regardless of the stage of gametogenesis, IGF-II mRNA levels exceeded those of IGF-I. Transcripts for IGF-1R RNA were detected in gonads at all stages studied. A major transcript of 11 kb was found in gonads and in gill arch and brain, but it was not found in liver and muscle. Distribution of the two types of IGF-1R and IGF-I in gonads was studied by immunohistochemistry. Immunoreactive IGF-I was found in the granulosa and theca cells of follicles at different vitellogenic stages and in oocytes at the chromatin-nucleolus and perinucleolus stage. In the testis, immunoreactive IGF-I was found in somatic cells of the cyst wall, interstitial cells, and spermatogonia A. In addition, IGF-1R was detected in the membrane of previtellogenic oocytes and in the theca and granulosa cells of vitellogenic and late vitellogenic follicles. In the testis, a positive reaction was identified in spermatogonia A and spermatids for the germ cells and in somatic cells of the cyst walls and interstitial cells. Local expression and production of IGFs and their receptors in fish gonads support a role for the IGF system in fish gonadal physiology.
gene regulation, growth factors, IGF receptor, ovary, testis
INTRODUCTION
Numerous studies have described the involvement of insulin-like growth factors (IGFs) in ovarian and testicular physiology among mammals [1–4]. Less attention has been paid to these growth factors in fish reproduction; however, substantial evidence supports the existence of an intragonadal IGF system in teleosts. Expression of IGF-I and IGF-II genes has been shown in the ovaries and testes of salmonids and tilapia [5–11], and immunoreactive IGF-I has been reported in red seabream ovaries [12].
Biological actions of IGFs are mediated by transmembrane receptors. Results of binding studies have suggested the presence of IGF type 1 receptors (IGF-1R) in teleost gonads [9, 13–15]. Levels of IGF-I binding in carp ovaries changed with follicular development [15]. In the trout testis, IGF-I binding was higher in Sertoli cell-enriched populations and in spermatogonia with primary spermatocytes [9].
The function of IGF-I, and particularly of IGF-II, in fish gonads has not been fully elucidated. Stimulatory effects of IGF-I on cell proliferation and DNA synthesis in premeiotic germ cells have been reported in trout [8, 14, 16] and in dogfish testis [17]. In the ovary, IGF-I has affected steroidogenesis [18, 19], and both IGF-I and IGF-II have induced oocyte maturation [20] and DNA synthesis by isolated follicles [21]. Much less information is available regarding the hormonal regulation of IGF expression in fish gonads. Recent in vivo and in vitro studies have demonstrated that expression of IGF-I and IGF-II genes was increased by growth hormone (GH) and gonadotropins in immature trout testis [10].
These studies strongly suggest the existence of an IGF system in fish gonads, including the presence of IGF-1R, but no information is available regarding the expression of an IGF receptor gene in gonads. Similarly, to our knowledge, no data exist regarding cellular localization of the corresponding proteins during fish gonadal development. Recently, two distinct IGF receptor cDNAs were cloned from coho salmon [22]. Cloning studies also have revealed that the gilthead seabream (Sparus aurata) expresses two different IGF receptors (unpublished results).
Sexual development in fish can be divided into gonochorism and hermaphroditism. During gonochorism, undifferentiated gonads differentiate into ovary or testis during ontogenesis. Hermaphroditism in a species is defined as a substantial proportion of individuals among a population functioning as both sexes, either simultaneously or sequentially, at some time during their life. Hermaphroditism is categorized in three ways: 1) protogyny, in which some or all individuals function first as females and, later in life, exclusively as males; 2) protandry, in which the sex change is from male to female; and 3) simultaneous hermaphroditism, in which individuals function simultaneously as both male and female. Many fishes are sequential hermaphrodites.
The possible involvement of the IGF system in fish reproduction has been investigated mainly in salmonids and carps, both of which are gonochoristic species. Among the hermaphroditic teleosts, the sparid family has both types of sequential hermaphroditism: protogyny, and protandry. These two sexual modalities are possible because of the special structure of the sparid ovotestis, with a dorsal ovarian part and a ventral testicular part, which are separated by a connective tissue. In our studies, we have used S. aurata, which is a protandrous hermaphroditic species. In captivity, and until the age of 8 mo, the dorsal ovarian part of the bisexual gonad in this species is dominant [23, 24]. Toward the first breeding season, at the end of the first year of life, the ventral testicular part of the gonad proliferates and forms a mature testis, and from the second year onward, approximately 80% of the population undergo sex reversal and become females. During the reproductive season, females undergo daily cycles of final oocyte maturation, ovulation, and spawning. During this time, the ovaries contain oocytes at varying stages of growth. This asynchronous type of ovarian development as well as the hermaphroditic nature of this species make the gilthead seabream an interesting comparative model for studying the involvement of IGFs in reproduction.
The aim of this study was to characterize the expression of IGF-I, IGF-II, and IGF-1R genes and to determine the cellular localization of IGF-I and the two types of IGF-1R proteins during the gonadal development of S. aurata.
MATERIALS AND METHODS
Chemicals
Oligonucleotides were prepared by Universal DNA, Inc. (Tigard, OR). Radionucleotides were obtained from Dupont NEN (Boston, MA).
Fish and Tissues
Sparus aurata fingerlings obtained from Mevo'ot Yam School (Michmoret, Israel) were maintained at Israel Oceanographic Limnological Research (IOLR) in 1000-L tanks supplied with running seawater at 25°C, under natural photoperiod conditions. Additional fish were obtained from Kibbutz Ma'agan Michael and from The Salt Company (Atlit, Israel) and maintained at IOLR as described earlier.
Fish were killed by decapitation, and gonads were excised at different stages of the reproductive cycle. In some experiments, other tissues were excised as well, including liver, muscle, brain, and gill arch. Gonads were weighed to determine the gonadosomatic index (GSI = gonadal weight x 100/BW). For each sample, a transverse section from the middle part of the gonad was fixed in Bouin's solution and processed for histology (to determine the gonadal stage) and for immunohistochemical studies. The tissues were frozen in liquid nitrogen and kept at -70°C until RNA extraction.
Histological Analysis of Gonads
Pieces of gonads fixed in Bouin's solution for 1 wk were dehydrated through a series of graded concentrations of ethanol, cleared with xylene, and embedded in paraffin wax. Sections were cut at a thickness of 4–5 µm using a hand microtome. Histological analysis and immunohistochemistry were systematically performed on adjacent sections. For histological examination, sections were stained with hematoxylin-and-eosin. The proportion of testicular to ovarian parts in the gonads was determined using a camera lucida (i.e., drawing tube) attached to a stereomicroscope. Projections of the appropriate gonadal parts were traced on drawing paper, cut, and weighed. The ratio between male and female parts was calculated as being the average of 10 projections of each gonad obtained at different levels and expressed as a percentage.
For gene expression studies, four groups representing four developmental stages of gonads were used. Stage 1 included fish collected during September 1997 (body weight [BW] 175–195 g; gonads, 0.14–0.33 g; GSI, 0.054%–0.077%) in which the ratio of female to male parts in the gonads was 80:20. Stage 2 included fish collected during November 1997 (BW, 265–276 g; gonads 3.4 g; GSI, 1.2%) in which the ratio of testicular to ovarian parts was 90:10. Stage 3 included functional males with mature testes during the spawning season collected in February 1998 (BW, 370 g; gonads, 13 g; GSI, 3.7). Stage 4 included mature females during the spawning season collected in February 1998 and February 1999 (BW, 350–410 g; gonads, 16–33 g; GSI, 4.75%–8.1%). For immunohistochemical studies, additional fish at the age of approximately 1.5 to 2 yr were killed during the winter and spring of the second year of life. These fish were collected during January 1998 (BW, 181 g; gonads, 0.48 g; GSI, 0.27%), March 1998 (BW, 197 g; gonads, 1.14 g; GSI, 0.58%), and June 1998 (BW, 560 g; gonad, 1.90 g; GSI, 0.34%).
The terminology used in the histological description of oocyte stages is that described by Mayer et al. [25]. We distinguished the following periods and stages during oocyte development: 1) period of oogonial divisions and beginning of meiotic transformations (not included in the present study), 2) period of previtellogenic stages (chromatin-nucleolus stage, perinucleolus stage, and completion of cytoplasmic growth and formation of cortical and lipid vesicles), 3) period of vitellogenesis (beginning of accumulation of yolk granules and globules, and completion of accumulation of yolk globules), and 4) period of maturation.
Immunohistochemistry
For IGF-I, polyclonal antibodies against recombinant S. aurata IGF-I [26] were raised in mice and used at a dilution of 1:100. The specificity of this antibody was checked by immunoblot and ELISA (data not shown). For IGF-1R, two polyclonal antibodies were raised in rabbits against two synthetic peptides corresponding to the C-terminals of the two S. aurata IGF-1R (termed SpIR5 and SpIR6) and used at dilutions of 1:100 and 1:50, respectively (characterization to be published elsewhere). Antiserum against native S. aurata. GH was raised in rabbit by Dr. M. Schartl (Wurzburg, Germany) and used at a dilution of 1:100. These dilutions gave the optimal staining during preliminary experiments, in which serial dilutions were used (data not shown).
Immunohistochemistry was performed according to the method described by Perrot et al. [27]. Briefly, paraffin sections were dewaxed and hydrated in decreasing concentrations of ethanol. Sections were first incubated with 0.3% hydrogen peroxide to eliminate endogenous peroxidase activity. Nonspecific binding sites were blocked by incubation in PBS/0.1% saponin/0.1% BSA and 4% normal serum for 30 min. Sections were then incubated with specific antibodies overnight at 4°C. After a brief wash in PBS/saponin, sections were incubated with antirabbit or antimouse immunoglobulin G (DakoA/S, Glostrup, Denmark) for 1 h at 25°C. After being rinsed, the sections were incubated with rabbit or mouse peroxidase-antiperoxidase (PAP) (DakoA/S) for 30 min at 25°C. The reaction was visualized using 3,3'-diaminobenzidine (Aldrich, Milwaukee, WI).
Specificity of the immunohistochemical reactions was verified using the following controls: 1) nonimmune serum instead of specific antiserum, 2) incubation of primary antibody with an excess of antigen (340 µg/ml), and 3) omission of primary or secondary antibodies. Adsorption of either antisera with an excess of antigen before the incubation resulted in a reduction of immunostaining. Omission of the primary antibody or use of nonimmune serum abolished the staining, indicating the specificity of the reaction. In each assay, a few sections were also incubated with a homologous anti-GH as a negative control.
Gonadal sections were counterstained with Ehrlich's hematoxylin to identify the cells of the immunostaining.
Isolation of RNA
Total RNA was prepared from gonads and other tissues by the guanidinium-rapid method [28] using Tri Reagent (Molecular Research Center, Inc., Cincinnati, OH). Poly(A)+ RNA was isolated using affinity chromatography on oligo(dT)-cellulose columns (5'-3', Inc., Boulder, CO). The RNA was quantified based on absorbance at 260 nm and appeared to be undegraded on a 1% agarose gel containing formaldehyde and ethidium bromide. The RNA samples were used for Northern blot and reverse transcription-polymerase chain reaction (RT-PCR) analyses.
Northern Blot Analysis
Equal amounts of total RNA (30 µg) from gonads at different stages of the reproductive cycle or from different tissues and 5 µg of gonadal poly(A)+ RNA were electrophoresed on denaturing 1% agarose-9% formaldehyde gel and blotted onto nylon membrane (Hybond-N; Amersham Pharmacia Biotech (Little Chalfont, UK). Hybridization was performed for 16–20 h at 42°C in a solution containing 50% formamide. Inserts of cDNAs coding for S. aurata IGF-I and IGF-II [29], plasmid pSpIR5-340 coding for the 5'-region of S. aurata IGF-1R cDNA (S. J. Chan, unpublished data), and S. aurata ß-actin cDNA (kindly provided by Dr. M. Tom) were labeled by random priming with [32P]-dCTP. Hybridization with the different probes was performed sequentially. Membranes were washed to maximal stringency and then exposed to Kodak XAR-5 film (Eastman Kodak, Rochester, NY) at -70°C using intensifying screens for 2 and 10 days. Filters were dehybridized between each hybridization.
Reverse Transcription-Polymerase Chain Reaction
Total RNA (10 µg) from gonads and liver were reverse transcribed using random hexamers (200 ng) or oligo(dT)12–18 (500 ng) and 200 U Superscript II reverse transcriptase (Gibco-BRL, Gaithersburg, MD) according to the manufacturer's recommendations.
Oligonucleotides primers were designed according to the sequences of S. aurata IGF-I and IGF-II [29] and S. aurata IGF-1R (unpublished data). Primers were 5'-AGTGCGATGTGCTGTATC-3' (sense) and 5'-CAGCTCACAGCTTTGGAAG-3' (antisense) for IGF-I, 5'-GGCCCTGACGCTCTACGT-3' (sense) and 5'-GATCAGAGGCCTGTGGAAGAT-3' (antisense) for IGF-II, and 5'-CCGGGGATGGATATTCGTAA-3' (sense) and 5'-CAGGGTTCTTCTCAATGCGG-3' (antisense) for IGF-1R.
The S. aurata ß-actin [30] was also amplified in each assay as a control for using equal amounts of RNA in the RT-PCR reaction. The primers used were 5'-CGACGGACAGGTCATCACCA-3' (sense) and 5'-AGAAGCATTTGCGGTGGACG-3' (antisense).
For each set of primers, the reaction was optimized for amount of primers, cDNA, and number of cycles (data not shown). All PCR reactions were performed using an aliquot of the same RT reaction. The specificity of the RT-PCR assay for IGFs was also confirmed using Southern blot analysis with radiolabeled S. aurata IGF-I and IGF-II cDNAs. The IGF-I PCR product hybridized specifically to IGF-I cDNA, whereas the IGF-II PCR product hybridized specifically to IGF-II cDNA. The PCR amplification was performed in 25 µl reaction using 1.25 U of AmpliTaq DNA polymerase (Appligene-Oncor, Parc d'innovation, France), 1x buffer (10 mM Tris-HCl [pH 9], 50 mM KCl, 1.5 mM MgCl2, 0.1% Triton X-100, and 0.2 mg/ml BSA), 15–100 pmol of each primer, and 200 µM of each dNTP. The PCR was performed for 30–35 cycles, with each cycle consisting of denaturation at 94°C for 1 min, annealing at 55–60°C for 1 min and 30 sec, and extension at 72°C for 2 min. Exceptions to this method involved the first cycle, which used 4 min of denaturation, and the last cycle, which used an extension time of 5 min. The PCR products were electrophoresed on 1.8% agarose gel and visualized by ethidium bromide staining before being blotted onto nylon membrane (Hybond-N) and being hybridized with specific probes.
RESULTS
General Histology
The stages of the gonads used in the study were systematically determined by histological analysis. Light microscopy confirmed the observations reported by Zohar et al. [23, 24] and Brusle-Sicard and Fourcault [31]. Figure 1 illustrates the developmental stages of the gonads used for the gene expression studies. At stage 1 (Fig. 1A), the dorsal part of the bisexual gonad has the structure of a young ovary, which dominates the gonad. It consists of ovigerous folds containing previtellogenic oocytes at the chromatin-nucleolus and perinucleolus stages. The testicular part contains spermatogonia organized in cysts (Fig. 1A), and rare groups of spermatocytes are visible among spermatogonia. Subsequently, in stage 2, the ovarian part regresses, and the testicular one develops (Fig. 1B). Degenerating oocytes can be seen at the central part of the gonad, whereas the rest is a functional testis, in which spermatogenesis has been nearly completed and lobules with spermatozoa as well as released spermatozoa can be identified (Fig. 1B). During the second year, approximately 80% of the population undergo sex reversal [23, 24] and become females (stage 4, Fig. 1D), whereas a small proportion remains as functional males (stage 3, Fig. 1C). Ovarian development in the gilthead seabream is asynchronous, and oocytes of different developmental stages can be simultaneously found. As shown in Fig. 1D, previtellogenic oocytes can be found side by side with oocytes during the early and late vitellogenic stages. Follicles are comprised of the oocyte surrounded by a layer of granulosa cells (i.e., follicular cells). External to the granulosa layer lies the basement membrane and the vascularized thecal cell layer. Occasionally, atretic follicles can be seen between the normally developing oocytes (Fig. 1D).
|
Northen Blot Analysis of IGF-I, IGF-II, and IGF-1R Expression
Steady-state levels of IGF-I and IGF-II mRNAs in S. aurata gonads collected at different stages of the reproductive cycle (Fig. 1) are shown in the Northern blots depicted in Figure 2, A and B. Two transcripts of approximately 6.4 and 3.7 kilobases (kb) were found for IGF-I when gonadal poly(A)+ RNA or liver total RNA were used (Fig. 2A, lanes 8 and 9, respectively). Two transcripts were also found for IGF-II with the estimated size of 4.8 and 3.3 kb (Fig. 2B, lanes 8 and 9, respectively). Expression of IGF-I in liver was much higher than in gonads and could be detected after 15 h of exposure (Fig. 2A, lane 9), whereas expression of IGF-II was similar to that in young gonads (see Fig. 4B, and compare lane 9 and lanes 1–3). Both IGF-I and IGF-II mRNAs appeared to be higher in young gonads (stage 1) than in other stages (Figs. 2, A and B). Amounts of RNA loaded onto the gel are demonstrated by the ethidium bromide staining of the ribosomal RNA (Fig. 2C). Similar results were obtained in an additional experiment. A long exposure of 10 days was necessary to detect IGF-I and IGF-II in gonads.
|
|
Expression of IGF-1R gene in gonads with different GSIs, as analyzed by Northern blot, is shown in Figure 3A. On hybridization of total RNA with a cDNA clone coding for the 5'-end of S. aurata IGF-1R, which was performed under high-stringency conditions, a major band with the estimated size of 11 kb was detected after 10 days of exposure (Fig. 3A). An additional transcript of approximately 7 kb (not as sharp as that of 11 kb) could be seen (Fig. 3A). The IGF-1R mRNA was detected in all gonads studied; however, the levels of expression varied between gonads with different GSIs. Amounts of RNA loaded onto the gel are demonstrated by the ethidium bromide staining of the ribosomal RNA (Fig. 3B). The 11-kb transcript detected in total RNA from gonads was also identified in brain and gill cartilage (Fig. 4A) but not in muscle and liver (not shown). The amount of RNA loaded onto the gel is shown in Fig. 4B.
|
RT-PCR Analysis of IGF-I, IGF-II, and IGF-1R Gene Expression
Levels of IGF-I, IGF-II, and IGF-1R mRNAs in gonads are low. To increase the sensitivity of detecting these mRNAs in gonads undergoing developmental changes, the highly sensitive method of RT-PCR was employed. As shown in Figure 5, PCR products of the expected sizes (IGF-I, 237 base pairs [bp]; IGF-II, 466 bp) were obtained after amplifications of 30 and 35 cycles (Fig. 5, upper and lower panels, respectively). Several controls were included, such as cDNA synthesized from RNA previously treated with DNase (Fig. 5, lane 9) as well as RNA and H2O amplification (Fig. 5, Bl I and Bl II, respectively). Both IGF-I and IGF-II mRNA levels appeared to be higher in young gonads (stage 1) than in other stages (Figs. 5 and 6, A and B), but this pattern was more pronounced for IGF-I. Levels of IGF-II mRNA in the gonads were higher than those of IGF-I, as shown by the intensity of the bands obtained after 30 or 35 cycles (Fig. 5). A single PCR product corresponding to IGF-1R was obtained in all gonads examined (Fig. 6A). ß-Actin was used as a control for using equal amounts of RNA in the RT-PCR reaction. Specificity of the RT-PCR products was confirmed by hybridization with the homologous cDNAs (Fig. 6B).
|
|
Localization of the Two Types of IGF Receptors and IGF-I by Immunohistochemistry
IGF-I In the testis, IGF-I immunoreactivity was found in somatic cells delineating the germinal cysts (Fig. 7, A and C), which are also referred to as Sertoli cells by many investigators [32], as well as in interstitial cells localized between the cysts. In addition, IGF-I was found in the cytoplasm of some spermatogonia A, especially in the perinuclear region (Fig. 7, A–C). In the ovary, IGF-I immunoreactivity was found in the granulosa cell layer of follicles enclosing oocytes at different developmental stages, such as completion of cytoplasmic growth and formation of cortical and lipid vesicles and different vitellogenic stages (Fig. 7, D–G). Sparse positive cells could also be seen in the thecal cell layer (Fig. 7, F and G), and IGF-I immunoreactivity was also found in the cytoplasm of oocytes at the chromatin-nucleolus and perinucleolus stage (Fig. 7, D, E, and G).
|
SpIR5 In the testis, strong immunoreactivity was found in somatic cells bordering spermatogenic cysts and lobules, interstitial cells localized between the cysts, and in connective tissue (Fig. 8, A, B, and D–F). Immunoreactivity was also detected in the cellular membrane of spermatogonia A and in spermatids (Fig. 8, D and E). Other types of germ cells did not react with IGF receptor antiserum. In the ovary, SpIR5 immunoreactivity was found around previtellogenic oocytes (Fig. 8H) and in the cellular membrane of granulosa and theca cells of vitellogenic oocytes (Fig. 8, I and J). The zona radiata externa was positive as well (Fig. 8J).
|
SpIR6 The reaction with SpIR6 antibody resembled that of SpIR5, though with a few exceptions (i.e., spermatids and zona radiata). In the testis, membranes of spermatogonia A were immunoreactive (Fig. 9A). A strong immunoreactivity was found in somatic cells of the cyst wall and in interstitial cells localized between the lobules (Fig. 9A). In the ovary, immunoreaction with SpIR6 antibody also resembled that obtained with SpIR5. Both granulosa and theca cells reacted positively (Fig. 9, C and D).
|
DISCUSSION
The present study analyzed the expression of three components of the IGF system during gonadal development in a protandrous hermaphroditic species, the gilthead sea bream (S. aurata). Transcripts for IGF-I, IGF-II, and IGF-1R could be demonstrated using Northern blot and RT-PCR analyses. However, mRNA levels were low, and a long exposure of the Northern blots was necessary. Generally, the pattern of expression of IGF-I and IGF-II seen by Northern blot analysis was similar to that obtained using the more sensitive RT-PCR. The size of IGF-I and IGF-II transcripts in gonads was identical to that in liver. Northern blot analysis revealed lower levels of IGF-I mRNA in gonads at all stages compared with those in liver. By contrast, gonadal IGF-II mRNA levels were only slightly lower than in liver and, at all stages studied, exceeded those of IGF-I. Similar findings regarding higher expression of IGF-II than of IGF-I in gonads were reported recently by us in immature rainbow trout testis [10] and by others in testes and ovaries of sexually mature rainbow trout [33]. Interestingly, higher expression of IGF-I was noted in young gonads with a dominated ovarian part, consisting mainly of previtellogenic oocytes, than in later stages. Differences in IGF-II levels were less pronounced. Although our RT-PCR method is only semiquantitative and few samples were used for each developmental stage, these findings suggest an important role for both IGF-I and IGF-II during the early stages of gonadal development.
The presence of IGF-1R in fish gonads had been suggested by others based on the results of binding studies [9, 13, 15]; however, to our knowledge, the present study is the first demonstration of IGF receptor mRNA in fish gonads. On hybridization with a cDNA clone coding for the N-terminal of S. aurata IGF receptor, a major transcript of approximately 11 kb was found in gonads. A similar size of IGF-1R transcript was reported for avian IGF-1R [34] and mammalian IGF-1R [35]. A second transcript of approximately 7 kb, which was not always as clear as the 11-kb transcript, has been reported for human IGF-1R but appears to be absent in the rat [35]. Furthermore, we demonstrate here for the first time, again to our knowledge, the presence and size of IGF-1R transcripts in two tissues: the brain, which in fish displays high levels of IGF binding [26, 36]; and gill cartilage, which in fish responds to IGF-I by sulfate incorporation [26, 37]. The size of IGF-1R transcripts in these tissues was similar to that found in gonads. The inability of the partial cDNA clone used in this study to detect IGF-1R transcripts in skeletal muscle and liver RNA, although known to bind IGF-I [36, 38], could be explained by the expression of several IGF-1Rs in fish, as was suggested recently in coho salmon [22] and in S. aurata (unpublished results). It could also result from extremely low levels of IGF-1R mRNA that were not detected by Northern blot analysis. That levels of IGF-1R RNA in gonads exceeded those in brain or gill cartilage, and that transcripts for IGF-1R could be detected in gonads when analyzed by Northern blot, are noteworthy. These findings are supported by recent observations reported by Maestro et al. [39] that among the fish tissues studied so far, the highest affinity of the IGF receptors has been found in granulosa and theca-interstitial cell layers from salmon ovary.
Previous studies, including our own, have used enriched populations of testicular cells from rainbow trout for identifying cellular expression of IGF-I and IGF-II employing RT-PCR and of IGF receptors using binding assays [9, 10, 14]. However, as these authors recognized, this approach does not ensure purity of the cell populations or exclude the possibility of cross-contamination. Binding assays also have been used to demonstrate the presence of IGF receptors in semipurified receptors from isolated theca-interstitial layers and granulosa cells of coho salmon and brown trout ovaries at the preovulatory stage [19, 39]. Again, the latter study could not exclude the possibility of cross-contamination between isolated cell types. Moreover, although IGF mRNAs were detected in S. aurata gonads, the sites of synthesis are not known, and the origin of IGF involved in the control of germ cells activity remains to be determined. To avoid these problems and provide conclusive evidence regarding the identity of cells expressing IGF-I and its receptor (or receptors), we performed a detailed immunohistochemical study using homologous antibodies. Our results provide direct evidence for the presence of IGF-I and two types of IGF receptors in the testicular and of ovarian cells of S. aurata.
In the testis, IGF-I immunoreactivity was found mainly in somatic cells lining the spermatogenic cysts, which are also known as Sertoli cells [32], but such reactivity was also found in spermatogonia A germ cells. These observations are consistent with the presence of IGF-I mRNA in Sertoli cell-enriched populations reported in rainbow trout [8]. Similarly, immunoreactive IGF-I was localized to Sertoli cells in human [40] and IGF-I mRNA detected in rat Sertoli cells by in situ hybridization [41]. Others have reported that the pattern of IGF-I immunoreactivity in rat testis differs during development. Whereas during early postnatal life IGF-I was demonstrated in spermatogenic cells, Sertoli cells, and Leydig cells, IGF-I-like immunoreactivity in mature rat testis was seen only in spermatocytes and not in spermatogonia [42]. Positive IGF-I immunoreactivity in rat spermatocytes but not in spermatogonia was also reported by Tres et al. [43].
In the present study, high immunoreaction for the two types of IGF receptors was found in the testis in somatic cells lining the cysts (i.e., Sertoli cells), in cells localized between lobules, and in the cellular membrane of spermatogonia A. This novel observation regarding the presence of putative receptors for IGFs in the membrane of germ cells supports a role for IGF-I (and probably also for IGF-II) produced locally, or provided from the circulation, in germ cell proliferation. Indeed, both IGF-I and IGF-II stimulated the proliferation of rainbow trout early germ cells in vitro [16], and IGF-I had a proliferative effect in early germ cells of the dogfish [17]. Furthermore, high levels of IGF-I and IGF-II binding have been found in premeiotic germ cells [9]. One of the antibodies to IGF receptor (SpIR5) has displayed immunoreaction with spermatids as well. In human testis, IGF-receptor-like immunoreactivity was localized in secondary spermatocytes and early spermatids [40]. The presence of intense IGF receptor immunoreactivity in Sertoli cells, as found in the present study, is compatible with the presence of IGF-I and IGF-II binding found in these cells in trout [9] and with the stimulatory effect of IGF-I on DNA synthesis in cultured Sertoli cells from the dogfish [17]. Considerable information has been accumulated regarding Sertoli cell-germ cell interactions in mammalian testis, but relatively less is known regarding such interactions in teleosts. In several teleost species, Sertoli cells, in addition to being a structural part of the germinal cyst, phagocytize spermatid residual bodies [44–46]. The organization of the germinal compartment in fish testis is less complex than that in mammals, and several Sertoli cells maintain a clone of isogenic germ cells throughout the entire spermatogenic process. This simplified organization of the fish testis permits a unique experimental approach to the paracrine/autocrine regulation of testicular function. The effect of Sertoli cells on mitotic germ cells is exerted via both diffusible molecules and plasma membrane molecules. Attempts to clarify the function of fish Sertoli cells have shown their importance for the survival of cultured trout spermatocytes [47]. Regardless of the maturational stage of the trout testes from which the Sertoli cells were used, they always inhibited, at least partly, the mitogenic effect of IGF-I on germ cells [16]. In mammals, IGF-1R has been found on Sertoli cells [43, 48], Leydig cells [48], and spermatocytes [43, 48], but no IGF receptor has been reported on mammalian spermatogonia. Thus, IGF-I cannot interact directly with spermatogonia. By contrast, we demonstrated that in fish, both Sertoli cells and germ cells (i.e., spermatogonia A) express IGF-1R, leading to a potential direct effect of IGF-I on these germ cells that stimulates their proliferation. It has been hypothesized that in fish, IGF-I could be involved in stem cell/spermatogonial proliferation [17]. This hypothesis is supported by the recent findings of IGF-I acting together with the spermatogenesis-inducing hormone (11-KT) to stimulate spermatogonia proliferation [49].
Immunoreactivity for both IGF-I and IGF-1R was also detected in cells between spermatogenic lobules. At this stage, it is not possible to identify unequivocally whether these are Leydig cells, but it has been demonstrated that mammalian Leydig cells express IGF-I [40, 41, 50, 51].
In the ovary, IGF-I was found primarily in the granulosa cell layer, as observed previously in the red seabream [12]. However, we also detected sparse IGF-I immunoreactive cells in the theca cell layer. The IGF-I immunoreactivity was detected in follicular cells of follicles at different developmental stages (including previtellogenic and vitellogenic stages) present in S. aurata ovaries collected a few hours before spawning. Our results differ from those reported in the red seabream, in which IGF-I was found only in granulosa cells of the early developmental stages (lipid stage and primary yolk globule) and diminished substantially in the secondary and tertiary globule stages, until complete disappearance in maturational stages [12]. The presence of immunoreactive IGF-I in both granulosa and theca cells of S. aurata is supported by in vitro studies that demonstrated IGF-I in 24-h incubation medium of salmon granulosa and theca-interstitial cells [39]. The theca cell layer in teleost follicles is composed of "special" theca cells, which are steroid-producing cells [52, 53], and a network of capillaries within a connective tissue stroma. In situ hybridization studies have revealed that granulosa cells of developing rat follicles are the major site for IGF-I expression, and IGF-I mRNA transcripts have been found primarily in the antral cell layers and in the cells of the cumulus oophorus [54]. Solution hybridization ribonuclease protection assays have shown that in the rat, IGF-I transcripts are found exclusively in the granulosa cells [55], whereas IGF-II transcripts were expressed solely in the theca-interstitial cells [56]. In addition, IGF-II has been suggested as a general stimulator in the proliferation and differentiation of mammalian granulosa cells, including steroid production [57].
Immunolocalization of IGF receptors in both granulosa and theca cells of follicles at various developmental stages, as found in S. aurata, provide conclusive evidence for the potential role of IGF-I (and probably for IGF-II as well) in the physiology of these cells. Furthermore, these results corroborate those of earlier reports using binding assays that demonstrated specific receptors for insulin and IGF-I in salmon granulosa and theca-interstitial cell layers [39]. In both follicular layers, the number and affinity of IGF-I receptors were higher than those of insulin receptors [19, 39], and IGF-I produced locally by the granulosa cells may elicit its steroidogenic effects on both granulosa and theca-interstitial cell layers [19]. The potential importance of IGF-II in fish gonadal function, as suggested in our study, is further supported by the recent report of its mRNA and protein localization in granulosa cells of the late follicle stage from tilapia [58].
Several functions for IGF-I in fish granulosa and theca cells have been proposed. For example, IGF-I stimulated the incorporation of thymidine into goldfish vitellogenic follicles and had an additive effect when tested with hCG [21]. In addition to promoting proliferation, IGF-I has been attributed other functions in fish ovary during the preovulatory stages. It enhanced estradiol and 17
,20ß-P (i.e., the maturation-inducing hormone) production by granulosa cells in coho salmon but inhibited testosterone and 17OH-P by theca-interstitial layer [19]. IGF-I also stimulated estradiol-17ß production by ovarian fragments from preovulatory white perch [59]. In addition, IGF-I induced final maturation of isolated oocytes in red seabream [20] and acted together with maturation-inducing steroid (20ß-S) to increase germinal vesicle migration and breakdown in the white perch [59], suggesting its possible role during final oocyte maturation. This hypothesis is also supported by the recent report on the increased formation of gap junctions in the ovarian follicles of red seabream [60].
Interestingly, IGF-I immunoreactivity was also found in the cytoplasm of oocytes at the chromatin-nucleolus and perinucleolus stage. What the function of IGF-I might be in these oocytes is not clear; however, IGF-I may affect oocyte growth by an as-yet-unknown mechanism. No such reaction, however, was reported in the red seabream [12].
In conclusion, we have shown that teleost gonads express IGF-1R mRNA throughout gonadal development, and that its levels are higher in gonads than in gill cartilage or brain. The immunohistochemical study showed that IGF-I and its receptors were detected in somatic cells lining cysts and in interlobular cells. The IGF receptor was localized to spermatogonia. In the ovary, IGF-I was localized mainly to granulosa but also to theca cells, whereas IGF receptors were found in both granulosa and theca cells in S. aurata ovary. These results suggest that IGFs are potential autocrine/paracrine regulators in teleost gonads and can act directly on germ cells and/or somatic cells to stimulate their proliferation or participate in other physiological functions, such as steroidogenesis or responsiveness to other hormones. Furthermore, our findings support an important role for both IGF-I and IGF-II in fish gonadal development. Protandrous S. aurata provides a unique model for studying a possible role for the IGF system during sex reversal. Involvement of the endocrine system in natural sex reversal has been suggested, but the mechanism is not known. Studies of sex reversal in the protandrous black porgy (of same family as S. aurata) point to a possible role of aromatase activity and E2 production in this process [61]. It is tempting to speculate that IGF might be involved in this process as well. Morphologically, changes that occur during sex reversal involve degeneration of germ cells of the initial sex, proliferation of germ cells of the second sex, and degeneration of the somatic components (i.e., testicular tissue in protandrous gonads, and proliferation and vitellogenesis of the ovarian tissue). These changes are accompanied by increased vascularization and appearance of phagocytes. Positive immunoreaction for IGF receptor found in these phagocytes (data not shown) suggests that in addition to the role shown by other investigators in gametogenesis and steroidogenesis, IGF might have a unique role in sex reversal.
ACKNOWLEDGMENTS
We thank the team of NCM, Eilat, The Salt Company, and Kibbutz Ma'agan Michael for kindly providing us with S. aurata fish. Adult fish were raised at IOLR (Haifa, Israel) from fingerlings kindly provided by Mevo'ot Yam School (Michmoret, Israel). Thanks are also due to the Academie d'Agriculture de France.
FOOTNOTES
First decision: 15 October 1999.
1 Supported in part by the U.S.-Israel Binational Agricultural Research and Development Fund (BARD, Project IS-2769-96CR). This study represents a portion of a dissertation submitted in partial fulfillment of a Ph.D. degree (V.P.) 
2 Correspondence: Bruria Funkenstein, Israel Oceanographic and Limnological Research, Tel-Shikmona, P.O. Box 8030, Haifa 31080, Israel. FAX: 972 4 8511911; bruria{at}ocean.org.il
Accepted: March 1, 2000.
Received: September 10, 1999.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  N. Shved, G. Berishvili, H. D'Cotta, J.-F. Baroiller, H. Segner, E. Eppler, and M. Reinecke Ethinylestradiol differentially interferes with IGF-I in liver and extrahepatic sites during development of male and female bony fish J. Endocrinol., December 1, 2007; 195(3): 513 - 523. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. K Davis, N. Hiramatsu, K. Hiramatsu, B. J Reading, T. Matsubara, A. Hara, C. V Sullivan, A. L Pierce, T. Hirano, and E. G. Grau Induction of Three Vitellogenins by 17beta-Estradiol with Concurrent Inhibition of the Growth Hormone-Insulin-Like Growth Factor 1 Axis in a Euryhaline Teleost, the Tilapia (Oreochromis mossambicus) Biol Reprod, October 1, 2007; 77(4): 614 - 625. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Campbell, J. Dickey, B. Beckman, G. Young, A. Pierce, H. Fukada, and P. Swanson Previtellogenic Oocyte Growth in Salmon: Relationships among Body Growth, Plasma Insulin-Like Growth Factor-1, Estradiol-17beta, Follicle-Stimulating Hormone and Expression of Ovarian Genes for Insulin-Like Growth Factors, Steroidogenic-Acute Regulatory Protein and Receptors for Gonadotropins, Growth Hormone, and Somatolactin Biol Reprod, July 1, 2006; 75(1): 34 - 44. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T.-T. Wong and Y. Zohar Novel Expression of Gonadotropin Subunit Genes in Oocytes of the Gilthead Seabream (Sparus aurata) Endocrinology, November 1, 2004; 145(11): 5210 - 5220. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
