| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Gamete Biology |
Laboratory of Electron Microscopy,3 Department of Cell Biology, Faculty of Sciences, National Autonomous University of Mexico (UNAM), Mexico D.F., Mexico
Center of Electron Microscopy,4 University of Lausanne, CH-1005 Lausanne, Switzerland
Department of Molecular Genetics and Cell Biology,5 University of Chicago, Chicago, Illinois 60637
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
gametogenesis, meiosis, spermatogenesis, testis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The searching for homology implicates long-distance actions in which one or several types of macromolecules, proteins, DNA, and/or RNA may be involved. However, most authors consider that DNA is the primary agent of the initial homology search, recognition, and pairing [1, 2]. The most accepted hypothesis for recognition of homologous sequences implies the formation of duplexes of complementary nonsister DNA molecules [1, 3–6]. Double-strand breaks and their repair are considered to be the most probable mechanisms of nonsister DNA-DNA interactions. The alignment of chromosomes may result from homology search initiated by the formation of double-strand breaks. Rad51 and Dmc1 proteins are frequently detected in leptotene and zygotene chromosomes of lily and mammals [7–10]. Plug et al. [8] found foci of Rad51 of similar size and spacing in asynapsed homologous chromosomes in the leptotene and zygotene stage, and they concluded that Rad51 plays a role in homology search. However, Barlow et al. [9] have not found Rad51 before leptotene or a symmetric distribution of Rad51-containing foci in pairing chromosomes, and they interpret these data as indications that Rad51 is not involved in homology search. One of the aims of the present work was to localize the recombination proteins during the process of homology recognition to shed some light on this question.
RNA may also mediate long-distance homologous sequence recognition. Transcription and chromosome alignment are related phenomena. Evidence suggests that chromosomes only pair when they are transcriptionally active [11]. Effective pairing systems are primarily associated with coding sequences, which exclude heterochromatin and nonheterochromatic repetitive DNA sequences from initial pairing [2]. In male Drosophila melanogaster, the major sex chromosome pairing sites correspond to the intergenic spacer repeats of the rDNA [12]. These sequences contain a copy of RNA polymerase I promoter, from which transcripts are initiated in vivo and in vitro [13]. Without promoters and flanking sequences, the rDNA is rarely involved in pairing.
The aim of the present study was to characterize the long-distance probe that mediates homologous recognition and to explore the relationships between structures involved in recognition, alignment, pairing, and SC formation.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The standard procedure for preparing the samples involved fixation in 2.5% glutaraldehyde in 0.2 M phosphate buffer (pH 7.3), dehydration with ethanol, and embedding in an epoxy resin (Glycide Ether 100; Merck, Darmstadt, Germany). Sections (thickness, 1 µm) were stained with toluidine blue. The cellular associations seen in seminiferous tubules were compared with the classical descriptions of the spermatogenic cycle of the guinea pig [15].
To stain ribonucleoprotein (RNP) structures, the uranyl acetate-EDTA-lead citrate procedure [16] was used. The DNA was stained using the osmium-ammine procedure [17].
Immunolocalization
Samples of seminiferous tubules from 33 animals were fixed in 2%–4% paraformaldehyde in 0.15 M phosphate buffer (pH 7.3) and embedded in Lowicryl K4M. Polymerization was carried out using ultraviolet light at -20°C. The ultrathin sections were mounted on formar-coated nickel grids (Polysciences, Warrington, PA) and processed for postembedding immunolocalization as described previously [18]. The following antibodies were used for immunolocalization: anti-RNA polymerase II rabbit antibody [19]; anti-DNA monoclonal antibody (mAb) purchased from Progen (Heidelberg, Germany); anti-heterogeneous nuclear (hn) RNP core protein chicken antibody [20]; anti-Sm mAb recognizing an epitope common to particles containing U1, U2, U5, and U6 small nuclear (sn) RNA [21]; anti-trimethyl-guanosin cap structure rabbit antibody (Oncogene, New York, NY); anti-Dmc1 mAbs; and anti-Rad51 polyclonal antibody. Negative controls were carried using the preimmune serum of the animal in which the primary antibody was developed or by suppressing the first antibody. All negative controls showed a stochastic distribution of gold grains with a background numerical density of labeling. The background intensity of labeling was estimated at the intercellular spaces of the preparations (Table 1).
|
Ultrastructural In Situ Hybridization
Seminiferous tubules from eight specimens were fixed in 2% paraformaldehyde with 0.2% glutaraldehyde added and embedded in Lowicryl K4M. Genomic DNA from guinea pig liver and spleen was labeled with digoxigenin-11-dUTP by nick translation according with the manufacturer's protocol (Boehringer-Mannheim, Mannheim, Germany). The probes were precipitated with ethanol, centrifuged, and resuspended in the presence of 50% formamide. The in situ hybridization was performed on the grids with the denatured probe with or without preannealing for 24–48 h at 37–45°C. The probe was used at a final concentration of 10 ng/µl in a hybridization solution containing 10% deionized formamide, 10% dextran sulfate, and 1x SSC (0.15 M sodium chloride and 0.015 M sodium citrate; pH 7.0). The DNA in the specimen was not denatured. After hybridization, the grids were washed successively in 5x SSC, 2x SSC, Tris-saline buffer (TBS) containing 0.05% Tween 20 (Sigma, St. Louis, MO), and TBS containing 0.5% blocking reagent (Boehringer-Mannheim). The hybrids were detected by incubation of the grids on a drop of mouse anti-digoxigenin (Boehringer-Mannheim) diluted 1:5 in TBS containing 0.5% blocking reagent for 30 min at room temperature. Grids were washed in TBS-Tween several times. Several antibodies were used to increase the in situ hybridization signal. The mouse anti-digoxigenin was detected by a rabbit anti-mouse immunoglobulin G (Aurion, Wageningen, The Netherlands), which in turn was revealed with 10-nm gold particle-conjugated Protein A (Aurion). Possible unrevealed epitopes of the mouse anti-digoxigenin antibody were visualized by goat anti-mouse antibody associated with 12-nm gold particles (Jackson Immunoresearch Laboratories, West Grove, PA), and the rabbit anti-mouse antibody was detected with goat anti-rabbit serum labeled with 12-nm gold. As a control, some specimens were incubated in the same way but with the molecular probe omitted. To confirm the RNA nature of the hybridization signal, some sections were pretreated with 2 mg/ml of ribonuclease (type IA; Sigma). Other sections were treated with 0.2 mg/ml of Proteinase K (Sigma) to improve the target accessibility to the probes.
Autoradiography
Small fragments of the testis from 12 guinea pigs were incubated at 37°C in Eagle minimum essential medium supplemented with 100 µCi/ml of 5,6-tritiated uridine (48 Ci/mmol; Amersham, Piscataway, NJ) for 120 min. Samples were fixed with 2% glutaraldehyde in Sörensen phosphate buffer (0.1 M, pH 7.3) for 1 h at 4°C; after three rinses in the buffer, the samples were dehydrated in acetone and embedded in Epon (Ilford, Ltd., Knutsford, UK).
Light-microscopic autoradiography was performed by coating sections (thickness, 1 µm) by immersion in Ilford K2 emulsion (Ilford, Ltd.). The slides were exposed for 20 days at 4°C. The preparations were stained with toluidine blue-borax for 1 min at 60°C. Imagenia 5000 software from Biocom (Paris, France) was used to digitize and analyze the images.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Nuclei of spermatogonia B exhibit, in addition to the usual structural components of interphase nuclei, pairs of elongated parallel structures (Fig. 1A, inset). These structures originate in the chromatin located in contact with the nuclear envelope and may be longer than a half-nuclear diameter.
|
Osmium-ammine staining specific for DNA contrasts axes of extended chromatin surrounded by filaments, resembling small lampbrush DNA-containing structures in spermatogonia type B (Fig. 1A). These structures are associated with RNP fibrils and granules, as shown by uranyl acetate-EDTA-lead citrate staining preferential for nuclear RNP structures (Fig. 1B). Immunolocalization of DNA corroborates that lampbrush structures are formed by extended chromatin (Fig. 2A). The RNA polymerase II is located in lampbrush structures and also in fibrils dispersed in the nucleoplasm, suggesting associated transcriptional activity (Fig. 2B). Lampbrush structures are also labeled by an anti-hnRNP antibody (Fig. 2C), indicating that they also contain pre-mRNA. The Sm epitopes of proteins associated with snRNAs (Fig. 2D) and trimethyl-guanosin cap of snRNAs (Fig. 2E) are scarcely present in lampbrush structures and nucleoplasmic RNP fibrils and granules. The components of pre-mRNA splicing are less abundant than pre-mRNPs and RNA polymerase II. A quantitative estimation shows that the numerical density of labeling of the trimethyl-guanosin cap in the Sertoli cell and Leydig cell nuclei is much higher than in spermatocytes in the leptotene stage (Table 1).
|
Meiotic Interphase Retains Unusual Structures
During meiotic interphase (preleptotene stage), the clumps of compact chromatin in contact with the nuclear envelope are progressively smaller, and nucleoli migrate to the periphery of the nucleus (data not shown). At the same time, the lampbrush structures of extended chromatin associated with RNP particles become more numerous. Frequently, pairs of lampbrush structures are aligned in parallel and related by radial filaments and loops forming structures such as microrope ladders (Fig. 3). At the end of the preleptotene stage, extended chromatin and RNP structures pervade the nuclear space, and transcription is ubiquitous.
|
Most Chromosomes Are Already Aligned at the Beginning of the Formation of Axial Elements at the Leptotene-Zygotene Stage
The beginning of the leptotene-zygotene stage is marked by the appearance of denser regions in the axis of lampbrush structures. These regions are generally short and discontinuous, and most appear as pairs in contact with the nuclear envelope. Frequently, the formation of chromosomal axis occurs at the same time in the corresponding regions of aligned lampbrush structures (Fig. 4). Continuities of lampbrush structures and the nascent chromosomal axes, which keep the lampbrush structure, are frequently seen (Fig. 4). Most of chromosomal axes are formed as pairs in register, in continuity with lampbrush structures connected by transverse fibrils and loops. Short, isolated chromosomal axes are also present. Shortly after the beginning of the appearance of axial densities, which is when these structures are approximately 1 µm in length, short stretches of the SC can be noticed near the nuclear envelope or in the internal region of the nucleus.
|
Quantitative Autoradiography Reveals Intense Transcription in Spermatogonia and in the Early Stages of Meiotic Prophase
The quantitative study of the incorporation of [3H]uridine in the nuclei and its export to the cytoplasm was carried out at the light-microscopic level. The nuclei of spermatogonia type B and spermatocytes in the leptotene-zygotene stage, as well as initial and medial pachytene cells, were strongly labeled with [3H]uridine (Fig. 5). The types of germ cells were recognized comparing the cellular associations seen in the seminiferous tubules of the autoradiograms stained with toluidine blue with the pioneer descriptions of the spermatogenic cycle in the guinea pig [15]. The cytoplasm of any of these cells is barely labeled over the background noise. Thus, the nucleus:cytoplasm (N:C) ratio is high (Table 2), indicating a very slow transport of newly synthesized RNA to the cytoplasm, contrary to Sertoli cells.
|
|
In Situ Hybridization Demonstrates the Presence of RNA in Structures Related to Sequence Recognition and Pairing
The axial element as well as the lateral loops of the lampbrush structure are labeled by the hybridization of a guinea pig DNA genomic probe with the RNA of nondenatured preparations (Fig. 6A). When pairs of lampbrush elements associate, forming rope-ladder structures, the RNA can be revealed in both aligned axes and in the transverse filaments joining the axes (Fig. 6B). The lateral branches surrounding leptotene-zygotene chromosomal axes also contain abundant RNA. The lateral elements of forming SCs in leptotene-zygotene are intensively labeled (Fig. 6B, inset). However, in the pachytene stage, mature SCs are seldom labeled.
|
Temporal and Spatial Distribution of Dmc1 and Rad51
Most of the immunocytochemical signal is grouped in discrete foci in all the cells studied. These clusters of gold grains are associated with nodular structures or with filaments of extended chromatin. Anti-Dmc1 labeling is scarce in spermatogonia type B. During meiotic interphase, round, elliptic, and irregularly shaped, dense nodules, having a rather homogenous structure, are present. Most of these nodules are intensely labeled by the anti-Dmc1 mAb (Fig. 7). Gold grains are predominantly located at the periphery of the nodules (Fig. 7B). However, nodules without labeling were also found. Two adjacent labeled nodules are seldom found (Fig. 7A, inset). Most lampbrush structures are devoid of anti-Dmc1 labeling (Fig. 7A, inset). Small groups of gold grains also occur, dispersed in regions rich in filaments of extended chromatin (Fig. 7), which are frequently intermingled with RNP structures. The clusters of interchromatin granules are devoid of labeling.
|
At the beginning of leptotene-zygotene stage, anti-Dmc1 labeled nodules occur near the growing chromosomal axes (Fig. 7B). No direct contact between labeled nodules with cores was found. Groups of gold grains not associated with nodules are located in short regions of the chromosomal axes, whereas long stretches of the axes remain unlabeled. As the leptotene-zygotene stage advances, the number of clusters of gold grains diminishes. In the advanced leptotene-zygotene stage the synapsing structures are generally unlabeled. Nodules labeled by anti-Dmc1 mAb are frequently present in the vicinity (100–250 nm) of the growing SC, but not in the pairing space (Fig. 7B, inset). This disposition is also found in already-formed SC in the early pachytene stage. In addition, groups of gold grains are also associated with nodules in contact with the thick axes of sex chromosomes located in the XY body.
The pattern of distribution of Rad51 is similar to that of Dmc1 in all stages studied.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Previous use of fluorescence in situ hybridization, in studies related to the distribution of homologous chromosomes in mouse and human spermatogonia, showed that homologous chromosomes occupy separated territories [22, 23]. Our present results demonstrate the occurrence of connections between distant structures containing compact and extended chromatin. These connections are established by long, lateral, radial loops of transcribing extended chromatin, some of them longer than a half-nuclear diameter, that cannot be traced along their entire length, because they seldom extend in the plane of the section. These loops, which are well under the resolution of light microscopy, apparently correspond to the long-range probes needed for recognition of homologous sequences and probably relate to the progressive alignment of distant chromosomes. This alignment may result from the stabilization of successful homologous sequence recognition events and from the instability of the interactions of nonhomologous sequences. Thus, homologous lampbrush structures become progressively related by numerous connecting filaments (rope-ladder structures) during meiotic interphase and the leptotene-zygotene stage. This sort of extensive recognition obviously excludes, or at least diminishes, the interlocking of chromosomes in the posterior stages. This interpretation of our results is in accordance with previous theoretical views on the role of unstable interactions in pairing [5]. The ample distribution of the associations of radial filaments probably corresponds to the connections at multiple interstitial loci, as postulated in studies of yeast meiotic prophase [6], and also to the general homology search, as put forward by Smithies and Powers [3] and further developed by Kleckner et al. [4]. However, the pairing connections described here begin to develop in spermatogonia type B—that is, much earlier than those previously proposed.
The present results indicate that most of the recognition and alignment of homologous chromosomes occur before the formation of the bouquet in the guinea pig. In this moment, the few chromosomes not yet aligned may begin successful recognition of sequences. Thus, one of the possible functions of the bouquet might be to assure the recognition of still-unaligned homologous chromosomes by bringing them together—a backup mechanism as proposed by Zickler and Kleckner [24].
Rad51 has previously been immunolocalized to chromosomes as discrete foci during meiotic prophase I. These foci were frequently related with early recombination nodules [25, 26]. The distributions of Dmc1 and Rad51 are similar to those reported previously [10]. The gold grains occur in clusters associated with dark-stained nodules and extended chromatin. However, the space between pairs of aligned lampbrush structures related by transverse filaments, which represents the site of homologous recognition, is generally unlabeled. It is important to note that during the leptotene-zygotene stage, the labeling is practically absent from the pairing space in the forming SC. These results strongly suggest that Dmc1 and Rad51 are not involved in early search, recognition of homology, and final pairing.
The observation that effective paring is associated with active [11] coding sequences [2, 27] suggested that transcription and homology recognition may be related phenomena [11]. Furthermore, a previous study using uranyl acetate-EDTA-lead citrate preferential staining for RNPs [16] revealed the presence of RNP in the fibrils bridging the pairing axial elements in the zygotene stage [28]. The present study shows, by means of different cytochemical procedures, that connections between lampbrush structures of extended chromatin are composed by fibrils containing RNA and DNA. Furthermore, the immunolocalization of DNA, RNA polymerase II, and pre-mRNP core proteins demonstrate that connecting fibrils correspond to extended chromatin transcribing hnRNA. Moreover, high-resolution in situ hybridization also demonstrates the presence of RNA in lampbrush structures, rope ladders, and leptotene-zygotene SC in the process of formation, which are related to homology searching, homology recognition, and pairing, respectively.
Our autoradiographic data demonstrate quantitatively an intense incorporation of [3H]uridine and a high N:C ratio of labeling in spermatogonia type B and in spermatocytes in the leptotene-zygotene stage, as well as in initial and medial pachytene stages, in accordance with the qualitative observations of Taylor [29] and Söderström [30]. The high N:C ratio can be explained by an unusually slow export of RNA from nucleus. This N:C ratio of labeling is higher than that of cells with experimentally decreased transport of RNA to the cytoplasm [31]. The importance of the RNA in the homology search has seldom been taken into account [11]. However, several reports on intriguing features of transcription during meiotic prophase have appeared. Studies of transcription in yeast suggest that during meiosis and sporulation, a loosening of the regulation of gene expression occurs. A significant fraction of the transcription could be noise rather than useful RNA [32]. This process may also occur in the meiocytes of other organisms, although to a lesser extent than in fungi [33]. By the same token some peculiarities of gene expression have been observed in mammalian spermatogenic cells. A large number of genes are expressed at high levels in these cells, but most of the mRNA never arrives to ribosomes [34]. The immunolocalization of the Sm domain of snRNPs and of trimethyl-guanosin cap of snRNAs suggests that a small number of fibers connecting lampbrush structures undergo cotranscriptional splicing. This slow rate of splicing may prevent, or at least slow down, the transport of the mRNA to the cytoplasm. The intense transcription occurring in these stages and the small rate of export suggest that most of the pre-mRNA synthesized is involved in a nuclear function. Interestingly, these phenomena occur in the same cells in which the processes of homology recognition, alignment, and pairing are taking place. Thus, pre-mRNA is in the right place (in the pairing space) at the right moment to be involved in the sequence recognition, alignment, and pairing of homologous chromosomes.
Our results show that in the guinea pig, most SC formation takes place by pairing of the growing lateral elements. In the same nucleus coexist the structural and functional features of the leptotene stage (growing of axial structures) and of the zygotene stage (formation of SC) in a single leptotene-zygotene stage. Very short SCs without connection to axial elements but continuous with rope-ladder structures or lampbrush elements are frequent at the beginning of the leptotene-zygotene stage. These observations indicate that the growth of lateral elements of the SC is probably centered in aligned lampbrush structures related by DNA- and RNA-containing fibrils.
During the early leptotene-zygotene stage, some isolated axial elements develop, demonstrating that preleptotene recognition of homologies and alignment are not always successful. These elements go through a classical leptotene stage, which is characterized by the development of long axial cores, followed by a zygotene stage, which is distinguished by the pairing of long cores, that begins at the onset of the bouquet.
The formation and development of the SC are complex processes that require further three-dimensional study. The precise structural reorganization of chromatin is another aspect of final pairing that is not yet clear.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: S. Fakan, Centre of Electron Microscopy, University of Lausanne, Bugnon 27, CH-1005 Lausanne, Switzerland. FAX: 41 21 692 50 55; sfakan{at}cme.unil.ch
Received: 18 February 2003.
First decision: 4 March 2003.
Accepted: 28 May 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  Y. S. Cho, N. Iguchi, J. Yang, M. A. Handel, and N. B. Hecht Meiotic Messenger RNA and Noncoding RNA Targets of the RNA-Binding Protein Translin (TSN) in Mouse Testis Biol Reprod, October 1, 2005; 73(4): 840 - 847. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
