| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
a Departments of Human Genetics and
b Pediatrics, University of Michigan, Ann Arbor, Michigan 48109-0618
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Hoxa13 null animals have been generated [9], but die in utero; no Hoxa13-/- mouse has survived past E16.5. Examination of these null mutants at earlier embryonic stages revealed reproductive deficiencies [7]. Specifically, genesis of the caudal portion of the Müllerian ducts is delayed, and the wolffian duct/ureter junction is abnormally displaced from the urogenital sinus, presumably because this structure is severely hypoplastic. However, given that the reproductive tract develops late in gestation and does not fully mature until after birth, it is unclear whether these defects would persist in the adult or whether fertility in these animals would be affected.
Hoxa13 is mutated in the spontaneous mouse mutation Hypodactyly (Hoxa13Hd) [10]. Mice homozygous for this mutation also exhibit a high degree of fetal lethality, but rare newborns that survive to adulthood are completely infertile [10, 11]. In addition to the infertility, mutant mice also exhibit a severe limb phenotype that has been described elsewhere [10, 12], and the cause of fetal lethality is under investigation. In this paper, we characterize the infertility of male and female adult Hoxa13Hd/Hd mice.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Hypodactyly mice were obtained from the Jackson Laboratory (West Grove, PA), where the mutation is carried on a BL6/C3HFe F1 hybrid. It has been carried in our laboratory by intercrossing Hoxa13Hd/+ mice; consequently, the contribution of BL/6J and C3HFe varies in each mouse. Despite this mixed background, the Hoxa13Hd phenotype in these animals remains very similar. Heterozygous matings were performed to generate liveborn Hoxa13Hd/Hd mice. All homozygotes were identified by the presence of a single digit on each paw and by genotype analysis as described [10]. All mouse experiments were carried out under approval of the University Committee on Use and Care of Animals at the University of Michigan (Ann Arbor, MI).
Skeletal Staining
Skeletons were stained with Alizarin red and alcian blue as described [13] for 4 days to 1 wk. The specimens were then cleared in 2% KOH for approximately 24 h and subsequently passed through increasing concentrations of glycerol before storage in 100% glycerol.
Tissue Sectioning
Ovaries and uterine tissues for longitudinal sectioning were fixed in 10% formalin, sectioned (5 µm), and stained with hematoxylin (Fisher Scientific, Pittsburgh, PA) and eosin (Sigma, St. Louis, MO). For cross sections through the female reproductive tract, the abdominal section (minus legs and tail) was fixed and then decalcified in Jenkin's solution [14] for several days. Tissues were then washed several times in absolute ethanol before sectioning (7 µm).
Whole-Mount In Situ Hybridization
Reproductive tracts from newborn +/+ females were dissected and fixed in 4% paraformaldehyde for 4 h, washed 3 times in PBS + 0.1% Tween-20, and dehydrated through increasing concentrations of methanol. Whole-mount in situ hybridization was carried out as described [15] with a digoxigenin-labeled antisense RNA probe corresponding to 900 base pairs (NcoI-PstI fragment) of the 3' untranslated region between the homeobox and first polyadenylation signal of Hoxa13.
Hysterograms
Tribromoethanol (10 g; Aldrich, Milwaukee, WI) in t-amyl alcohol (10 ml; Sigma) was diluted to a 2.5% working stock in PBS for use as anesthetic. Each mouse was anesthetized using an i.p. injection of the working stock (15 µl/g). The abdominal cavity was opened, and the entire reproductive tract was visualized. Methylene blue (0.05%) in PBS was injected from a syringe into the vaginal cavity using thin, flexible Tygon tubing (Norton Performance Plastics, Akron, OH), and the uterine horns were examined grossly for dye penetration.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Hypoplasia of the Proximal Os Penis in Hoxa13Hd/Hd Males
Seven homozygous males survived to weaning; however, four of these animals died before maturity. The three surviving males were placed with wild-type, fertile females, and although one male did produce a single copulation plug, no offspring were obtained. A second male had motile sperm in the epididymis, and sectioning of the testis revealed no obvious defects (data not shown). In addition to infertility, several animals exhibited urinary tract problems. Two males had cystic bladders, one had a hypoplastic bladder, and two had possible bowel obstructions, which may have contributed to their death. Four of these seven animals were subjected to skeletal staining. All had only a single digit on each paw; no axial skeletal defects were observed. The os penis is thought to be necessary for copulation in mice, and alteration in the shape of this bone contributes to infertility in other mutants [1, 6, 16]. The penian bones were dissected from stained skeletons and representative +/+, Hoxa13Hd/+, and Hoxa13Hd/Hd bones are shown in Figure 1. The proximal part of the penian bone from the Hoxa13Hd/Hd male is markedly smaller and misshapen compared to that of the controls. A similar malformation was observed in all 4 males examined (data not shown), suggesting that this structural defect may contribute to the infertility in Hoxa13Hd/Hd males.
|
Hoxa13 Expression in Female Reproductive Tissues
To identify potential areas within the female reproductive tract that may be affected by the Hoxa13Hd mutation, whole-mount in situ hybridization was performed to determine the normal RNA expression domain of Hoxa13 (Fig. 2). Expression was observed in the vaginal tissue with extension into the cervical canals (n = 4). The ovaries and uterine horns do not express Hoxa13 as determined by this assay.
Hoxa13Hd/Hd Females Exhibition of Genitourinary Defects
Eighteen females were obtained for analysis. Urinary tract defects were observed in one third of the females. These included cystic kidneys (three mice), hypoplastic kidney (one mouse), absence of a bladder (one mouse), and an abnormal fistula connecting the bladder and vaginal cavity (one mouse). Fifteen of the eighteen females were mated with fertile males and appeared to exhibit a normal estrous cycle (as determined by mounting behavior), but no offspring were produced. Ovaries of two +/+ and two Hoxa13Hd/Hd females were sectioned for histology; no defects were observed in either of the mutant tissues (data not shown).
Upon gross examination of the uterine tissue of Hoxa13Hd/Hd females, we observed that a significant proportion of them (5 of 18) exhibited enlarged, fluid-filled uterine horns. Fluid accumulation can occur due to excessive fluid production or inefficient clearance due to blockage along the tract. To determine whether the uterine horns of Hoxa13Hd/Hd females were patent, we performed hysterograms on six animals. Under anesthesia, 0.05% methylene blue was injected into the vaginal cavity, and the uterine horns were examined for dye entry. A positive hysterogram was defined as one in which the dye freely entered one or both uterine horns. The results of the hysterograms are shown in Table 1. Seven +/+ and eleven Hoxa13Hd/+ females were tested by this method. All +/+ and 10/11 Hoxa13Hd/+ females produced positive hysterograms. One horn of one Hoxa13Hd/+ female tested negative. Furthermore, none of the horns in these animals were fluid-filled. In contrast, two of six Hoxa13Hd/Hd females had bilateral fluid accumulation, and a third animal exhibited unilateral fluid accumulation. In addition, half of the mutant females had completely negative hysterograms while a fourth female was unilaterally negative. There was no correlation between fluid accumulation and hysterogram results. Two other striking abnormalities were observed in the Hoxa13Hd/Hd females during this assay. Five of the six Hoxa13Hd/Hd females demonstrated a reduced ano-vaginal distance (Table 1), and all six females exhibited a profoundly small vaginal cavity. The tubing used to insert the dye would freely enter the vaginal cavity of control animals for several millimeters. However, the vaginal cavities of the Hoxa13Hd/Hd females could support the tubing for a fraction of that distance. For several of the animals, internalization of the tubing was impossible. This structural defect probably is significant in the infertility of Hoxa13Hd/Hd female mice, and blockage or stenosis within the cervical or vaginal region is the likely mechanism for fluid accumulation.
|
To further explore the nature of the defects along the reproductive axis, cross sections of the entire uterovaginal canal were examined for one +/+ female and a Hoxa13Hd/Hd litter mate (Fig. 3). For both animals, sectioning began at the level of the ovaries and proceeded caudally. The physical distances (in micrometers) between the sections presented in Figure 3, A, C, and E (wild-type) are identical to those presented in Figure 3, B, D, and F (Hoxa13Hd/Hd). At proximal levels, the uterine horns (ut) appeared similar (Fig. 3, A and B). Defects were first apparent as the two uterine horns began to fuse together (Fig. 3, C and D). At this stage, the two horns in the +/+ female were relatively far apart but began to join through connective tissue from the cranial portion of the cervical canal (Fig. 3C). In the Hoxa13Hd/Hd tissue at this level, the two horns were unconnected but were much more closely apposed (Fig. 3D). In both animals, the uterine horns, but not the lumens, merged at the neck of the cervical canals (cc); however, the Hoxa13Hd/Hd stromal tissue surrounding the cervical canals was smaller (Fig. 3, E and F), and the cervical lumens had a simple architecture.
|
The remaining mutant sections presented (Fig. 3, H and J) were further caudal than the comparable sections in the wild-type female (Fig. 3, G and I). The lumens fused into the common cervical canal (ccc) dorsal to the urethra (ur) in Figure 3, G and H. In the Hoxa13Hd/Hd female, the common cervical canal was greatly reduced in size compared with that of the wild-type females, and the urethra was abnormally located within the same serosal layer (Fig. 3H). In addition, the level at which the common cervical canal was formed in the mutant was caudal to the comparable level in wild-type as evidenced by the appearance of the pubic bones (pu, Fig. 3H) in the Hoxa13Hd/Hd mouse. The physical distance between sections 3E and 3G (320 µm) compared with 3F and 3H (1288 µm) supports this conclusion. Finally, the common cervical canal expanded into vaginal tissue (va), with fornices evident cranial to the level of pubic bones in the wild-type female (Fig. 3I). However, in the Hoxa13Hd/Hd female, vaginal fornices were never observed, and the single lumen (combined vaginal and urethral opening) was much smaller than in the wild-type female (va*, Fig. 3J). These data demonstrate that the anterior portions of the female reproductive tract were normal, but that posterior sections including the cervix and upper vagina were markedly hypoplastic.
Anterior Transformation of Cervical Tissue
To further explore the nature of the tissue defects in the cervicovaginal region, the reproductive tract from several +/+, Hoxa13Hd/+, and Hoxa13Hd/Hd adult females (n = 8, 10, and 8, respectively) were fixed, sectioned longitudinally, and examined. As seen in the cross sections, Hoxa13Hd/+ and Hoxa13Hd/Hd females had normal anterior uterine morphology (except when dilated by fluid accumulation). Representative sections of the cervical region from each genotype are shown in Figure 4. The hypoplasia of the vaginal cavity in Hoxa13Hd/Hd females is clearly observed in low-magnification views of the distal reproductive tract (Fig. 4A). The defects in the cervical canal are more readily observed at higher magnifications. As the two horns fuse, the normally invaginated uterine stroma straighten into the cervical canals (Fig. 4B). These lumens merge into a common cervical canal cranial to the vagina. In Hoxa13Hd/+ mice, the straight, smooth lumens of the cervical canals seen in wild-type mice were replaced by invaginations into the surrounding tissue before the formation of the common canal (Fig. 4B). This change in structure, appearing more like uterine stroma, was even more pronounced in Hoxa13Hd/Hd females (Fig. 4B). Thus, another consequence of the Hoxa13Hd mutation is an apparent anterior transformation of the cervical tissue. Glands and other uterine-associated structures, including the columnar epithelium (see below), were not present in the transformed cervical region in heterozygous mutant females. In contrast, uterine glands were observed in two of the eight homozygous mutant females at the level of the common cervical canal (data not shown), suggesting that the anterior transformation is variable in some mice.
|
Maintenance of Cervicovaginal Epithelial Transition in Mutant Females
The epithelial lining of the wild-type cervicovaginal region undergoes a columnar-to-squamosal transformation postnatally [17, 18]. The transition between these two epithelia in the normal mature female occurs near the junction of the bifurcated cervical canal and the common cervical canal (Fig. 4C). We hypothesized that if the anterior transformation was complete, the transformed cervix would contain the columnar epithelium characteristic of uterine tissue. The cervicovaginal epithelial transition zone was observed in both Hoxa13Hd/+ and Hoxa13Hd/Hd females (Fig. 4C) in the region of the abnormal uterine stroma, suggesting that the transformation was incomplete, and that the epithelial transition occurred in this mutant independent of the stromal characteristics.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Hoxa13Hd is caused by a small deletion in the coding sequence of Hoxa13. Hoxa13 is expressed in the cloaca of the chick [19], in the murine genital tubercle (results of this study [data not shown]; [6, 7]), and in cervical/vaginal tissue in mice (Fig. 2, this paper; [20, 21]); thus, the defects described here are consistent with the known patterns of gene expression.
In males, the penian bone forms by a combination of intramembranous (distal portion) and endochondral (proximal portion) ossification [22]. In Hoxa13Hd/Hd males, the proximal portion of the penian bone is profoundly hypoplastic (Fig. 1). A mild alteration is seen in the penian bone of Hoxd13 null males and is thought to contribute to the infertility in these animals [16]. Removal of varying combinations of 5' Hox genes has a detrimental effect on the size of the penian bone as well. In the complete absence of Hoxd11, Hoxd12, and Hoxd13 (HoxDdel/del), the bone was smaller overall but normally proportioned [23]. When one copy of Hoxa13 was removed in this background, the penian bone was profoundly reduced in size, particularly the proximal segment, and resembled the malformed bone observed in Hoxa13Hd/Hd males (Fig. 1, this paper; [23]). Further, embryos homozygous for both Hoxa13Hd and the Hoxd13 null allele (Hoxa13Hd/Hd; Hoxd13-/-, [6]) exhibited a complete absence of external genitalia. These results suggest that the mutation in Hoxa13Hd may alter the function of one or more of the Hoxd genes leading to the severe phenotype; however, analysis of the penian bone from an adult Hoxa13-/- (created by homologous recombination) male would be necessary to confirm this. Interestingly, the distal segment of the penian bone that ossifies directly is only mildly affected by Hoxa13Hd or the other mutations just discussed. Therefore, the relative sparing of structures undergoing intramembranous ossification, such as the distal portion of the penian bone, suggests that different signals regulate this mechanism of bone formation.
Males homozygous null for other Hox genes are also infertile. Hoxd13-/- males exhibit a malformed penian bone [16] and alterations in morphogenesis of several accessory glands [24]. Hoxd11 null males are infertile for unknown reasons [25, 26]. Hoxa11 null males are able to produce copulation plugs, but spermatogenesis is altered; they also exhibit an anterior transformation of the proximal vas deferens to resemble the epididymis, and the lumen of the remaining vas deferens is narrow [2]. Hoxa10 null males exhibit cryptorchidism and an anterior transformation of the vas deferens to the cauda epididymis [3, 4]. In these mutants, complete loss of a single Hox gene has multiple effects on the reproductive tract. We did not examine these male genitourinary structures, primarily because of the difficulty in producing homozygous males, which are outnumbered by females at birth by a ratio of almost 3:1. Therefore, it is possible that Hoxa13Hd/Hd males have other abnormalities besides the altered penian bone that contribute to their infertility [27].
The female reproductive tract is derived from the Müllerian ducts (uterine horns, cervix, and upper vagina) and urogenital sinus (lower vagina). Those structures form late in gestation, and further differentiation occurs after birth. During normal differentiation, the columnar epithelium lining the entire lumen undergoes a squamosal transformation that extends cranially into the cervix [17]. The defects observed in Hoxa13Hd/Hd females support several conclusions. First, there is severe hypoplasia of vaginal and cervical tissue, both of which express Hoxa13, suggesting that growth of both the urogenital sinus and the distal Müllerian ducts requires Hoxa13. In addition, the bladder, which is also derived from the urogenital sinus, is affected in some animals. Several mutant females exhibited fluid-filled uterine horns. As Hoxa13 is not expressed in the anterior uterus and there is hypoplasia of the caudal lumens, we believe that this fluid accumulation is secondary to the vaginal/cervical hypoplasia. Hox genes play a role in mesenchymal proliferation; therefore, the absence of Hoxa13 in this region is consistent with a reduction in growth (Fig. 3, this paper; [7]). However, the Hoxa13Hd mutation, but not the Hoxa13-/- null allele, has been shown to lead to increased cell death of limb bud mesenchyme [12], and it is possible that the reproductive tract is similarly affected. Comparison of cell death in this region between wild-type and Hoxa13Hd/Hd females, and analysis of Hoxa13-/- adults would help distinguish between these mechanisms.
Second, the transformation appears to be incomplete. The columnar epithelium characteristic of uterine tissue is absent from the transformed cervical region. Instead, the columnar-to-squamosal transition of the epithelial lining in the cervical canal of normal animals is present in the transformed cervix of Hoxa13Hd/Hd females (Fig. 4). This transition appears to be independent of the underlying stroma of the cervix, which now resembles uterine tissue in these mutant mice. This transformation is apparent even in Hoxa13Hd/+ females (Fig. 4) but does not have an observable effect on fertility. Glands are present in +/+ females throughout the uterine horns but cannot be found caudal to the cervical canals, well anterior to the point of fusion into the common cervical canal. In heterozygous mutants, the transformed cervical tissue has a modest uterine-type stroma, and glands do not appear in this transformed region. In homozygotes, glands were observed in the transformed area to the level of the common cervical canal in two of the eight females examined (data not shown). We did not control for cycle-dependent uterine changes in our analyses, and this could account for the presence of glands in only two animals.
Finally, anterior transformations and infertility have been observed in mice with other Hoxa mutations. Hoxa10-/- females exhibit an anterior transformation of the proximal uterine horn into oviduct, and embryos fail to implant [4]; Hoxa10-/+ heterozygotes were not reported to have any abnormalities. Hoxa11-/- females also exhibit a failure at implantation because the uterine glands fail to produce leukemia inhibitory factor, a cytokine necessary for implantation [5]. In addition, Hoxa11-/+ heterozygotes have reduced litter sizes, suggesting a gene dosage effect of Hoxa11 in fertility [2]. Hoxa10 and Hoxa11 are expressed throughout the uterine horns, whereas Hoxa13 is expressed in the cervical and vaginal tissue, with minimal overlap with Hoxa11 in the cervix [20, 21]. Hoxa13 deficiency in limb buds of Hoxa13Hd/Hd mice leads to more distal expression of Hoxa11 in the autopod [12]. Therefore, absence of Hoxa13 in the cervix could lead to more posterior expression of Hox genes in these mutants.
It should be noted that the limb phenotypes of Hoxa13Hd and the Hoxa13 null mutations are not equivalent. The Hoxa13Hd mutation not only leads to loss of Hoxa13 function but also has an additional, negative effect on survival of the limb mesenchyme and alters the expression of Hoxd13 [12]. It is not known whether Hoxa13Hd has a similar, dominant-negative role on the development of the reproductive tract, but it should be considered as a potential contribution to the observed reproductive tract abnormalities.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Jeffrey W. Innis, Department of Human Genetics, University of Michigan, 3703 Medical Science II, Ann Arbor, MI 48109–0618. FAX: 734 763 3784; innis{at}umich.edu
Accepted: July 2, 1999.
Received: April 26, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  T. M. Williams, M. E. Williams, J. H. Heaton, T. D. Gelehrter, and J. W. Innis Group 13 HOX proteins interact with the MH2 domain of R-Smads and modulate Smad transcriptional activation functions independent of HOX DNA-binding capability Nucleic Acids Res., August 8, 2005; 33(14): 4475 - 4484. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. A. Morgan, S. B. Nguyen, V. Scott, and H. S. Stadler Loss of Bmp7 and Fgf8 signaling in Hoxa13-mutant mice causes hypospadia Development, July 15, 2003; 130(14): 3095 - 3109. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. de Santa Barbara and D. J. Roberts Tail gut endoderm and gut/genitourinary/tail development: a new tissue-specific role for Hoxa13 Development, January 2, 2002; 129(3): 551 - 561. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Zhao and S. S. Potter Functional specificity of the Hoxa13 homeobox Development, August 15, 2001; 128(16): 3197 - 3207. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
