| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Chemistry & Biochemistry and
b Department of Biological Sciences, University of South Carolina, Columbia, South Carolina 29208
ABSTRACT
Expression of the diphtheria toxin A-chain gene was directed to the male germ line by fusion to 1 kilobase of the 5'-flanking DNA of the rat histone H1t gene. Two independent lines of mice were established that expressed the toxic transgene. Female carriers were fertile; males were sterile although otherwise apparently normal. Adult transgenic males had very small testes that were virtually devoid of germ cells. A developmental study showed that germ cells survived until late fetal life but that testes of 3-day-old transgenic mice were severely depleted of prospermatogonia. During postnatal development of transgenic animals, remaining germ cells progressed to the pachytene stage of meiosis in 10% to 30% of tubular cross sections but degenerated before the completion of meiosis. By 3 mo of age the residual germ cells had almost completely disappeared. These transgenic lines demonstrate the complete tissue specificity of the H1t promoter and reveal a period of its activity just prior to formation of the definitive adult spermatogonial stem cell population. Whereas full expression of H1t occurs only in mid to late pachytene spermatocytes, one or more of the factors that impart tissue specificity to its expression must be transiently activated in the neonatal germ line. This report discusses the possibility that this genetic technique for eliminating germ cells may have practical application in making recipients for spermatogonial stem cell transplantation.
spermatogenesis, testes
INTRODUCTION
The expression of diphtheria toxin subunit A (DTA) in transgenic mice has been used to eliminate particular cell lineages and to thereby demonstrate, in a dramatic way, the exquisite specificity of promoters for a variety of genes that are transcribed only in particular differentiated tissues (see for example [1–5]). DTA expression leads to covalent modification of elongation factor 2 [6] and, thus, to a catalytic inhibition of protein synthesis. DTA is so effective that it is estimated that a single molecule of the natural protein is sufficient to kill a cell [7]. In this report we describe the consequences of DTA expression driven by the promoter region of histone H1t.
Histone H1t is an unusual H1 variant found only in spermatocytes and immature spermatids beginning in mid- to late-pachytene spermatocytes [8–15]. Although the reason a unique H1 variant is made at this point in spermatogenesis is not known, a number of studies have indicated that, in vitro, H1t is less able to condense chromatin than other members of the H1 family, and its function may be to promote a more open chromatin environment in late spermatocytes and round spermatids [16–18]. The H1t gene shares many cis-acting DNA regulatory elements with the five H1 histone variants that encode the bulk of the linker histone complement in somatic cells [10, 19–22], and it seems likely that H1t was derived from a standard H1 by evolutionary duplication and divergence. Because of the similarity between the H1t promoter region and those of the common somatic H1s, it has been possible to begin to differentiate the DNA elements that are important for the unique expression pattern of H1t in contrast to those that are shared among most or all H1 genes. Candidate DNA binding sites and protein factors have been identified for both presumptive transcriptional activation in spermatocytes [23–25] as well as transcriptional inhibition in somatic cells [26, 27].
We were prompted to examine the results of the transgenic expression of an H1t-DTA fusion gene for two reasons. The first was to produce a mouse model in which spermatogenesis would be arrested in the mid-pachytene stage. The second was to establish the absolute cell-type specificity of the H1t promoter. Because even leaky expression of the DTA transgene is lethal, a transgenic line showing just germ cell loss would be strong evidence for the absolute tissue specificty of the H1t promoter. Although a previous transgenic study showed that 1 kilobase (kb) of the 5'-flanking sequence led to correct expression of a lacZ fusion gene [28], absolute tissue specificity was not established.
Two transgenic lines have indeed been established with an H1t-DTA fusion gene, and they appear to be normal except that production of sperm is totally blocked. However, spermatogenesis was affected significantly earlier than we had expected, approximately at the point when prospermatogonia in fetuses are preparing to differentiate into the population of Type A stem cell spermatogonia, which will maintain the spermatogonial population in adult animals.
MATERIALS AND METHODS
Plasmid Construction and Transgenic Mice
The coding sequence for the DTA chain was taken from plasmid pIBI130-DT-A (obtained from I. Maxwell, University of Colorado Health Sciences Center, Denver, CO) as a BamHI fragment and was substituted for the BamHI fragment encompassing the lacZ coding sequence of the H1t-lacZ fusion plasmid described by Bartell et al. [28]. The resultant plasmid contained the 5'-flanking region of rat H1t from PvuII (-948) to Tth111I (+70), the DTA coding region; and the intron and poly(A) site of the mouse metallothionein II gene (+350 to +840; [29]) kindly provided to us in placI by R. Palmiter, University of Washington, Seattle, WA (Fig. 1). Cloning steps were based on techniques described by Crouse et al. [30].
|
The fragment for injection was excised from the vector backbone with HindIII, gel purified, and injected into the pronuclei of (C57BL/6 x DBA/2)F2 embryos as described previously [28, 31]. Transgenic founder animals were identified by Southern blotting [32] of BamHI-digested DNA obtained from tail samples [33] to the 32P-labeled [34] BamHI fragment of pIBI130-DT-A. Founders were subsequently mated to pure-strain C57BL/6 mice; positive offspring from this cross were likewise mated, and the line was propagated by female carriers through successive generations.
Histology
Tissues were fixed by immersion in Bouins solution and processed for production of routine paraffin-embedded sections that were stained with hematoxylin and eosin. Fetal testes are small; thus, to facilitate handling, they were embedded in agar prior to embedding in paraffin. Sections were photographed with a Zeiss Photomicroscope III using Kodak T-Max 100 film. For cell counts, the statistical estimate of the significance of differences between populations was performed by Student's t-test.
Incorporation of 5-bromodeoxyuridine (BRDU; Sigma Chemical Company, St. Louis, MO) into DNA was performed essentially as described by Eldridge et al. [35]. The thymidine analogue (20 mg/ml in Dulbeccos PBS) was loaded into Alzet Model 2001 osmotic pumps (1 µl/h, 200 µl capacity; Alza Corp., Palo Alto, CA). Each pump was implanted s.c. in the back region of an adult mouse for 4 days. At sacrifice, the testes and a sample of small intestine were fixed in ethanol/glacial acetic acid (3:1) overnight in the cold before routine paraffin embedding and sectioning. Dewaxed sections, mounted on aminopropylsilane-coated slides, were treated in 1 N HCl at 60°C for 10 min, neutralized in 50 mM Tris borate, 1 mM EDTA (pH 8), treated with 3% H2O2 in water for 5 min, and washed twice in PBS. For detection of BRDU, we used Becton Dickinson (San Jose, CA) monoclonal B44 (catalog no 347580) diluted 1:25 and the components of the Vector (Burlingame, CA) peroxidase-based Mouse on Mouse Immunodetection Kit, according to the protocol supplied. Final detection was with Pierce (Rockford, IL) Enhanced Diaminobenzidine. Slides were then stained briefly with Gills No 2 hematoxylin (Fisher Scientific, Pittsburgh, PA), dehydrated, and mounted with Cytoseal 60 (Stephens Scientific, Riverdale, NJ).
RESULTS
A 1-kb fragment of the rat H1t gene extending from +70 in the 5' nontranslated region to -948 relative to the cap site was shown to direct germ cell-specific transcription of a lacZ fusion gene in transgenic mice [28]. We substituted the DTA coding sequences for the ß-galactosidase structural sequences in the H1t-lacZ plasmid described by Bartell et al. [28] and used the H1t-DTA fusion construct (Fig. 1) to generate transgenic mice. Two female founder animals (H1t-DTA-1, and Hlt-DTA-2) were identified that passed the transgene to female and male offspring alike. Female descendants were fertile with no obvious phenotype resulting from the transgene. The male offspring that received the transgene were not fertile but otherwise appeared normal. Southern blot analysis of the transgenic lines indicated that H1t-DTA-1 had a multiple-copy tandem insert (Fig 2), whereas H1t-DTA-2 had a single-copy insert (results not shown). Visual examination of the reproductive tracts of mature males as well as the testis histology indicated that the phenotypes were very similar if not identical. A detailed histological analysis was carried out on the H1t-DTA-1 line.
|
Examination of the reproductive tract from a 2-mo-old adult male carrying the transgene showed it to have extremely small testes, epididymides of somewhat reduced size, and apparently normal vas deferens morphology (Fig. 3). The testes appeared to lie in the inguinal canal but were not in the normal scrotal position and were difficult to palpate. The seminal vesicles were of normal size and contained secretion, a good indicator that the androgen status of the animal was within the normal range (not shown). Histological examination of a 3-mo-old adult male transgenic revealed almost a complete lack of germ cells in the seminiferous tubules, which were typically lined with single layers of Sertoli cells (Fig. 4). In the entire cross section of this testis, there was only a single tubule in which a few cells might possibly have been entering meiotic prophase (Fig. 4A, insert). This complete lack of germ cells was unexpected because H1t first appears in late meiotic prophase.
|
|
We then examined younger animals, including fetuses, to determine the stage in testicular development at which germ cells became depleted. The testes of male fetuses at 17 days postconception normally contain prospermatogonia that have finished a period of multiplication and are entering a perinatal period of mitotic arrest [36, 37]. Both normal and transgenic fetuses at this stage of development had comparable numbers of large prospermatogonia (about 30 per 100 Sertoli cells) prominantly visible in the centers of tubule cross sections (Table 1; Fig. 5, A and B). At birth (Day 0), normal animals had about 15 prospermatogonia per 100 Sertoli cells (the decreased ratio of germ cells to Sertoli cells was presumably the result of continuing Sertoli cell proliferation in the face of mitotic arrest by the prospermatogonia). In contrast, the prospermatogonia in the transgenic animals had dropped sharply to about 2 per 100 Sertoli cells (Table 1). By 3 days of age, when prospermatogonia have normally reinitiated proliferation, a striking contrast was seen between a normal animal and its transgenic littermate. In the transgenic, the germ cell population was reduced to about a tenth that of the control (Table 1; Fig. 5, C and D). At 7 days of age, the same disparity between normal and transgenic animals was maintained, with the population of prospermatogonia/primitive type A spermatogonia only one-tenth of that of the control (Table 1; Fig. 5, E and F). Transgenics and controls were sacrificed at 11, 15, 16, 19, 21, 29, and 32 days of age. By 15 or 19 days of age, it was easier to detect transgenic tubules with germ cells because the remaining prospermatogonia had proliferated in some tubule regions, and limited spermatogenesis was progressing as far as early meiotic prophase. With all transgenics between 15 and 32 days of age, some 10% to 30% of tubule cross sections showed variable signs of developing germ cells, ranging from a small patch along part of a tubule wall to an almost normal pattern of meiotic cells (results not shown). In the older animals, some tubules contained cells that progressed to the late pachytene stage before degenerating, while neighboring tubules might have contained only Sertoli cells (Fig. 6A). In no case were haploid cells observed in this (or in any other) transgenic testis, although both round and elongated spermatids were observed in many tubules from a normal littermate by 30 days of age (Fig. 6B). Thus, although most germ cells degenerated before they became true spermatogonia, those that passed this stage managed to develop as far as the mid-pachytene stage of meiosis, the normal point of expression of H1t. No cells were observed to pass this point in development; however, in transgenic mice that were 3 to 4 mo old, only a few cells in a testis cross section showed any progression into meiosis, and virtually all tubules were devoid of germ cells (compare with Fig. 3).
|
|
|
To ensure that residual germ cells were not being missed in adult transgenics, two 6-mo-old mice along with normal controls were administered 5-BRDU via continuous infusion for 4 days to mark dividing cells. Testis sections were subsequently treated for immunological detection of BRDU. With control mice, numerous darkly stained spermatogonia and early spermatocytes were seen along the walls in the majority of tubules (Fig. 7A). As expected, nearly all tubules from each of the transgenics were devoid of labeled cells (Fig. 7, B and C). Despite the general absence of labeled cells, a rare tubule (8 out of 236 tubules in this cross section) contained a small patch of labeled cells (Fig. 7, C and D). Because 4 days is nearly half of the 8.6-day cycle of the seminiferous epithelium in the mouse, this result argues that there has been a major depletion of the stem cell spermatogonia, or that they are unable to proliferate. Because the population of radiation-resistant stem cell spermatogonia in the mouse has been estimated at just a few thousand [38], it is difficult to distinguish between their absence and inability to proliferate; however, it is clear that that the testes of animals older than a few months have lost almost all traces of active spermatogenesis.
|
DISCUSSION
We have produced transgenic mice in which the DTA chain is expressed under the control of the spermatocyte-specific H1t promoter. Whereas female animals appear completely normal, males that carry the transgene are infertile, have very small testes, and germ cells are deficient or absent, depending on their age. Most germ cells in transgenic males degenerated at about the time of birth. Remaining germ cells progressed to meiosis before degenerating during the initial waves of spermatogenesis. In older animals, germ cells completely disappeared from all but trace regions of testicular cross sections.
The restricted effect of this toxic transgene on the male germ line emphasizes the absolute specificity of the H1t promoter region. In usual tests of tissue specificity it is difficult to eliminate the possibility of some ectopic expression because it is impractical to examine every tissue for the trace presence of protein or mRNA. Given the extraordinary cytotoxic potential of the diphtheria toxin catalytic subunit [7], the phenotype of the H1t-DTA animals argues that expression of the transgene was effectively silenced in all cellular environments except the male germ line.
While the regulatory control of the H1t gene is under investigation in several laboratories, we do not fully understand the factors that so dramatically restrict its expression. The H1t gene has every DNA motif known to be important for the expression of ordinary somatic H1 genes [10, 19–22], which makes it surprising that expression is so completely suppressed in somatic lineages. The regulatory features that normally direct ubiquitous expression during S-phase are clearly subordinate to control elements that restrict somatic expression but promote germinal expression. Grimes and colleagues identified a testis-specific nuclear protein that binds to a short palindromic sequence found with two repeats in the immediate H1t promoter of many mammalian species [23, 39]. Although this factor is an attractive candidate to stimulate expression in spermatocytes, there is as yet no direct evidence for its role. Deletion of the promoter region encompassing its binding sites abolished expression of a rat H1t transgene [40]. However, in this experiment, the closely linked GC box binding site for the Sp1 family of factors, which is a common feature of H1 promoters, was also removed. This Sp1 site has been shown to be very important for H1t promoter function in vitro [25] and in transfected cells [26], which complicates interpretation of this transgenic study. There is also good evidence for features of the H1t gene that prevent its expression in somatic cells. Inhibitory regions have been identified both upstream of the core promoter [27, 41] and adjacent to the transcriptional start site [26]. However, the mechanisms by which these DNA sequences bring about silencing in somatic cells remain to be worked out.
The temporal pattern of germ cell loss in the H1t-DTA transgenic animals is intriguing because cell death occurred in at least three separate stages of germ cell development. The most pronounced was within the primitive spermatogonia present at about the time of birth. The second period was midway through meiosis, the point at which significant accumulation of H1t protein normally occurs and significant transgene expression was expected. Finally, to explain the extensive depletion of germ cells in mature animals, transgene expression must have occurred occasionally in the remaining spermatogonial stem cells so that, eventually, very few remained. The fact that some cells survive the neonatal period implies that the level of expression in early spermatogonia is so low as to be stochastic in nature. Expression in postnatal spermatogonial stem cells must be of such a probability that survival is possible for a period of time, but that destruction is virtually assured by 3 mo. The probablistic expression of transgenes in individual cells has been noted by others [42–44].
This multiphasic activation of the transgene was not anticipated because previous studies have reported H1t expression only relatively late in the first meiotic prophase. For example, H1t protein (identified electrophoretically) was not observed in extracts of 3-day-old or 15-day-old rat testes, but was readily detected in extracts of 21-day-old organs [8]. H1t was not recovered from spermatogonia and was first associated most convincingly with mid- to late-pachytene spermatocytes in extracts prepared from fractionated rat germ cells [9]. Antibody to H1t identified pachytene spermatocytes as the first antigenically positive cells in rats [14] and mice [15]. In agreement with detection of the protein, H1t mRNA was identified only in pachytene spermatocytes in the rat by in situ hybridization [11]. Recently, however, using a sensitive nuclease protection assay, H1t mRNA was detected in 9-day-old mice, and trace levels of protein were also detected in these young animals using Western blotting with chemiluminescent detection [45]. Furthermore, in a transgenic study in which the natural rat gene was introduced into mice, some animals showed premature expression of H1t protein in preleptotene spermatocytes, perhaps the result of a gene dosage effect [28].
Even in the light of the trace appearance of H1t in early spermatocytes or spermatogonia, transgene expression in the neonatal period is still surprising. The period of male germ cell development that extends from the appearance of the fetal testis to the postnatal establishment of type A spermatogonia has been designated as prespermatogenesis and the germ cells designated as prospermatogonia [36, 37]. The multiplying prospermatogonia enter a period of mitotic arrest about 3 days before birth and are then termed primary transitional (T1) prospermatogonia. This period of mitotic arrest lasts until about 3 days after birth in the rat [36, 37, 46]. However, in the mouse, the point of resumption of mitotic proliferation is strain-dependent and can occur within a day of birth [47–49]. The T1-prospermatogonia divide to become T2-prospermatogonia, which in turn divide to found the population of undifferentiated type A stem cell spermatogonia, that will serve as the germinal stem cell population in the adult [50].
In these transgenic mice, the principal loss of germ cells may correlate with the resumption of mitosis by the quiescent T1-prospermatogonia. This is an attractive hypothesis because it implies that one or more features of the tissue-specific transcriptional control of H1t are activated at the critical point in which the primitive spermatogonia differentiate to form the definitive spermatogonial stem cell population of adult animals. At least one of these features is therefore a marker for this important step in germ cell development. While expression from the 1 kb of H1t upstream sequence used as the transcriptional control region in our construct may not reproduce with complete accuracy the quantitative expression of the natural gene, some nuclear event must occur during maturation of the primitive spermatogonia that leads to transgene activation. This event may involve the first appearance of the binding factor for the H1t-specific palindrome [12], or it could be a partial lifting of the inhibitory controls that normally repress H1t expression [26, 27]. This issue is a difficult experimental question because of the small number of primitive spermatogonia that are available for biochemical analysis. It will be easier to address when the various relevant transcription factors are isolated and antibodies to them become available.
In older animals, virtually all traces of spermatogenesis disappeared, leaving a seminiferous tubule populated with Sertoli cells only. The very small testes in these animals may not have occupied the scrotum and so could, in part, reflect the impairment to spermatogenesis that occurs because of higher abdominal temperatures. In fact, in the long-term abdominal testis in the mouse, only Type A spermatogonia survive, although these cells can regenerate normal spermatogenesis if the testis is returned to its scrotal location [51]. Because spermatogonia were not evident in the older transgenic mice, it seems likely that their destruction was caused by transgene expression rather than by abdominal temperatures.
One of the most dramatic developments in reproductive biology in recent years is spermatogonial transplantation [52–55]. Recipients for transplantation have included both genetically sterile animals as well as animals in which spermatogenesis was temporarily interrupted by use of an alkylating drug. It would be interesting to know if the transgenic lines reported here would be useful recipients for such transplantation. If so, this transgenic approach may permit construction of male-sterile recipients in a variety of species. The small testes of the lines reported here could represent an experimental difficulty. In that case, it may be possible to delay the effects of toxin expression by using an attenuated version of the DTA gene [56] or by adding additional controls to the promoter that could be used to delay expression until meiosis [57].
ACKNOWLEDGMENTS
We thank Ian Maxwell for the gift of pIBI130-DT-A, Richard Palmiter for the gift of placI, Benny Davidson and Neda Osterman for histology assistance, and Richard Showman for making available his photomicroscope. We appreciate helpful discussions on germ cell histology with Rick Boockfor and Clarke Millette. We are grateful to an anonymous reviewer for suggesting the BRDU-incorporation experiment.
FOOTNOTES
First decision: 5 August 1999.
1 This work was supported by NIH Grant HD-10793 and by a grant from the Venture Fund of the University of South Carolina. 
2 Correspondence: W.S. Kistler, Department of Chemistry and Biochemistry, University of South Carolina, 730 S. Main St., Columbia, SC 29208. FAX: 803 777 9521; kistler{at}psc.sc.edu
3 Current address: MIDI Labs, Inc., Newark, DE 19713.
4 Current address: Department of Molecular Biology and Pharmacology, Washington University School of Medicine, St. Louis, MO 63110.
Accepted: March 14, 2000.
Received: July 1, 1999.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  G. C. Horvath, W. S. Kistler, and M. K. Kistler RFX2 Is a Potential Transcriptional Regulatory Factor for Histone H1t and Other Genes Expressed During the Meiotic Phase of Spermatogenesis Biol Reprod, November 1, 2004; 71(5): 1551 - 1559. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Kanatsu-Shinohara, S. Toyokuni, T. Morimoto, S. Matsui, T. Honjo, and T. Shinohara Functional Assessment of Self-Renewal Activity of Male Germline Stem Cells Following Cytotoxic Damage and Serial Transplantation Biol Reprod, May 1, 2003; 68(5): 1801 - 1807. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. C. Horvath, S. E. Clare, M. K. Kistler, and W. Stephen Kistler Characterization of the H1t Promoter: Role of Conserved Histone 1 AC and TG Elements and Dominance of the Cap-Proximal Silencer Biol Reprod, October 1, 2001; 65(4): 1074 - 1081. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
