Absence of the prion protein homologue Doppel causes male sterility

Abstract

The agent that causes prion diseases is thought to be identical with PrP^Sc, a conformer of the normal prion protein PrP^C. PrP^C-deficient mice do not exhibit major pathologies, perhaps because they express a protein termed Dpl, which shares significant biochemical and structural homology with PrP^C. To investigate the physiological function of Dpl, we generated mice harbouring a homozygous disruption of the Prnd gene that encodes Dpl. Dpl deficiency did not interfere with embryonic and postnatal development, but resulted in male sterility. Dpl protein was expressed at late stages of spermiogenesis, and spermatids of Dpl mutants were reduced in numbers, immobile, malformed and unable to fertilize oocytes in vitro. Mechanical dissection of the zona pellucida partially restored in vitro fertilization. We conclude that Dpl regulates male fertility by controlling several aspects of male gametogenesis and sperm–egg interaction.

Introduction

PrP^Sc, which is derived by a poorly understood post-translational mechanism from the normal host-encoded protease-sensitive PrP^C (Oesch et al., 1985; Basler et al., 1986), is a major component of the infectious agent that causes transmissible spongiform encephalopathies (TSEs) (Prusiner, 1982). Prion diseases are fatal neurodegenerative disorders that occur naturally in humans and in a variety of animals. A wealth of evidence indicates that PrP^C is essential for the development of prion disease (Aguzzi et al., 2001). Most importantly, mice lacking PrP^C are resistant to experimental scrapie inoculation (Bueler et al., 1993), and familial cases of human TSEs are characterized by mutations of Prnp, the gene that encodes PrP^C (Prusiner, 1998).

Sequence analysis of a cosmid containing Prnp revealed that 16 kb downstream of the murine Prnp gene, an open reading frame (ORF) is present that encodes a protein, which shares significant homology with PrP^C (Moore et al., 1999). The new protein of 179 residues was christened Dpl for ‘downstream of the Prnp locus’ (or ‘doppel’, German for ‘double’) (Moore et al., 1999). The Prnd gene coding for Dpl is evolutionarily conserved from humans to sheep and cattle, suggesting an important function of Dpl (Tranulis et al., 2001). Dpl has a lower molecular weight than PrP^C because it lacks the N-terminally located octameric repeats and transmembrane region present in PrP^C. The predicted Dpl protein showed roughly 25% identity with the carboxy-proximal two thirds of PrP^C, and both PrP^C and Dpl are GPI-anchored membrane proteins (Moore et al., 1999; Silverman et al., 2000). Dpl may contain three α helices like PrP^C and two disulfide bridges between the second and third helix, and the structure of Dpl has significant similarity to PrP^C (Lu et al., 2000; Silverman et al., 2000; Mo et al., 2001). Dpl mRNA is expressed at high levels in testes, less in other peripheral organs and at very low levels in brain of adult wild-type mice. However, significant Prnd mRNA was detected during embryogenesis (Li et al., 2000b).

The Dpl protein resembles an N-terminally truncated PrP^C protein lacking the octamer repeats. The latter version of PrP^C is capable of PrP^Sc propagation (Flechsig et al., 2000). We have previously investigated the role of Dpl in prion pathogenesis using a neural grafting paradigm. Embryonic stem (ES) cells carrying a homozygous null mutation of the Prnd locus, but a normal Prnp locus, were found to undergo normal neurogenic differentiation and were capable of giving rise to all neural cell lineages when transplanted into host brains. After inoculation with scrapie prions, Dpl-deficient neural grafts showed spongiosis, gliosis and unimpaired accumulation of PrP^Sc and infectivity similar to wild-type neuroectodermal grafts (Behrens et al., 2001). Four polymorphisms in the human Prnd gene were detected, but no strong linkage disequilibrium was found between any of these polymorphisms and human prion diseases (Mead et al., 2000; Peoc’h et al., 2000). Taken together, these findings suggest that Dpl is dispensable for prion disease progression and generation of PrP^Sc.

Although it was not realized at the time, Dpl was first identified accidentally by generating knock-out mice lacking PrP^C. To elucidate the physiological function of PrP^C, several mouse lines with targeted disruptions of Prnp were independently generated. All mutant mouse lines lacked significant portions of the Prnp ORF and did not produce PrP^C protein, but showed two strikingly different phenotypes. Zrch Prnp^0/0 and Edbg Prnp^–/– (named after their cities of origin, Zürich and Edinburgh, respectively), were generated by replacing parts of the Prnp ORF with a neomycin phosphotransferase gene and showed only minor electrophysiological and circadian rhythm defects (Bueler et al., 1992; Collinge et al., 1994; Manson et al., 1994; Tobler et al., 1996). In contrast, Ngsk Prnp^–/– (which were generated in Nagasaki) developed cerebellar Purkinje cell degeneration causing late-onset ataxia. In the Ngsk mice, not only the ORF but also 0.9 kb of intron 2, the 5′ non-coding region and 0.45 kb of the 3′ non-coding region of Prnp were deleted (Sakaguchi et al., 1996). Subsequently, the observations made in the Ngsk mice were independently confirmed in two additional Prnp knockout lines, Zürich II and Rcm0, in which the Prnp ORF and its flanking regions were replaced by a lox sequence and an HPRT cassette, respectively (Moore et al., 1999; Rossi et al., 2001). This enigma was solved through the realization that in the brain of ataxic but not of healthy Prnp-mutant mice, Prnd mRNA was upregulated (Moore et al., 1999; Li et al., 2000a). An intergenic splicing event places the Dpl locus under the control of the Prnp promoter, probably due to the deletion of the Prnp intron 2 sequence including its splicing acceptor (Moore et al., 1999). This intergenic splicing event could also be detected at very low levels in wild-type mice, but was greatly enhanced by the absence of the intron 2 splice acceptor (Moore et al., 1999; Behrens and Aguzzi, 2002). The Prnp promoter, but not the Prnd promoter, is highly active in neuronal cells (Moore et al., 1999; Li et al., 2000a,b; Rossi et al., 2001). Therefore, Prnd transcription off the Prnp promoter results in overproduction of Dpl in the brain (Moore et al., 1999; Li et al., 2000a; Rossi et al., 2001). Further experiments have demonstrated an inverse correlation between the mRNA levels of Prnd and the onset of ataxia. Disease progression was accelerated by increasing Prnd levels, supporting the notion that ectopic Dpl expression, but not functional loss of PrP^C, may be responsible for neuronal degeneration in ataxic Prnp-deficient mice (Rossi et al., 2001). Interestingly, Ngsk and Zürich II Prnp^0/0 mice were rescued from neuronal degeneration by introduction of a Prnp transgene (Nishida et al., 1999; Moore et al., 2001; Rossi et al., 2001) suggesting that PrP^C can antagonize Dpl neurotoxicity and that the absence of PrP^C is necessary for Dpl to induce cell death. Therefore, the structural similarity between PrP^C and Dpl argues for a similar biological function, but the genetic data demonstrate that they can also have antagonistic function in the regulation of neuronal cell death.

To clarify the physiological function of the Dpl protein, we have generated mice with a homozygous targeted disruption of the Prnd gene. Whereas females lacking Dpl were viable and fertile, males without Dpl were sterile. We found that, in the absence of Dpl, spermatids were malformed and showed a defect in acrosome function, which resulted in the inability to achieve oocyte fertilization. Therefore, mice lacking Prnd identify Dpl as a critical regulator of male gametogenesis.

Results

Generation of Prnd-mutant mice

We addressed the physiological role of Dpl through targeted inactivation of the Prnd gene in the mouse. The complete ORF encoding Dpl was replaced by a neomycin resistance gene in ES cells, as described previously (Behrens et al., 2001), and mutant mice were generated (Figure 1A). Heterozygous mice carrying the mutated Prnd allele (Prnd^neo) were intercrossed to obtain homozygous mutant mice lacking Dpl. Both male and female homozygous Prnd^neo/neo mice were born with Mendelian frequencies (data not shown) and displayed normal growth and survival, demonstrating the absence of obvious detrimental effects of Dpl deficiency on development and growth. To ascertain the successful functional inactivation of the Prnd locus, northern blot analysis was performed. Prnd mRNA was detected in the testes of Prnd^+/+ and Prnd^+/neo mice, but was absent in homozygous mutant animals confirming that the Prnd^neo allele was null (Figure 1B). The levels of Prnp mRNA, which is also expressed in testes (Shaked et al., 1999), was unaffected by the absence or presence of Dpl (Figure 1B). Similarly, Dpl protein was present in the testes of Prnd^+/+ and Prnd^+/neo mice, but was absent in homozygous mutants, whereas PrP^C protein was detected in all samples (Figure 1C). As Prnd^neo/neo mice showed no apparent gross morphological or behavioural abnormalities, we investigated the development of the hematopoietic system in Dpl mutant mice. Prnd mRNA was found to be expressed in the spleen, suggesting a potential role for Dpl in lymphoid cells (Li et al., 2000b). However, the immune system appeared to be normally developed. CD4⁺ and CD8⁺ T cell subsets as well as B220⁺ B cells in peripheral blood revealed no obvious alterations. B-lymphocyte development in bone marrow, as analyzed by anti-B220, anti-CD19 and anti-CD25 antibodies, was also unaffected. Spleen cell suspensions showed no abnormalities in marginal zone B cells (CD21^highCD23^–) and follicular B cells (CD21^med CD23^high) in Prnd^neo/neo mice (data not shown).

Fig. 1. Generation of mice lacking Dpl. (A) Schematic representation of the targeting strategy to generate ES cells harbouring a homozygous mutation of Prnd. Exons of the Prnd gene are represented by rectangles, thin lines represent intronic regions of the Prnd locus. The Dpl ORF is represented as a dark gray rectangle. The neomycin resistance gene (PGK-Neo) and the diphtheria toxin α gene (DTα) are represented by light gray rectangles, while loxP sites are shown as black triangles. (B) Northern analysis of total testes RNA from wild-type, Prnd^neo/+ and Prnd^neo/neo mice, hybridized with a probe encompassing the Prnd and Prnp ORF. (C) Prnd^+/+, Prnd^neo/+ and Prnd^neo/neo males were mated to wild-type females and copulation plugs were recorded in the morning after mating (Prnd^+/+, n = 1; Prnd^neo/+, n = 1; Prnd^neo/neo, n = 2). (D) Epididymal sperm number of Prnd^+/+and Prnd^neo/neo males (Prnd^+/+, n = 2; Prnd^neo/neo, n = 2). (E) Motility of epididymal sperm isolated from Prnd^+/+and Prnd^neo/neo males (Prnd^+/+, n = 2; Prnd^neo/neo, n = 2). Bars in (D) and (E) represent the standard error of the mean. (F) Similar number of copulation plugs in Doppel expressing and deficient mice, indicative of sexual activity similar to that of controls.

Absence of Dpl causes male sterility

Female Prnd^neo/neo mice were fertile and yielded litter sizes similar to those of wild-type. In contrast, we noted that male Prnd^neo/neo mice were infertile. Breeding of 13 Prnd^neo/neo mice with wild-type females resulted in only a single litter, whereas all tested Prnd^+/+ and Prnd^+/neo littermate controls were fertile and yielded multiple litters (Table I). Whereas wild-type males produced an average litter size of 5.7 offspring, corresponding to a fertility rate of 100%, Dpl-deficient males were dramatically less fertile (0.7%) (Table I). To determine whether the absence of Dpl caused a defect in breeding and copulation behaviour, Prnd^neo/neo, Prnd^+/neo and Prnd^+/+ males were mated to hormone-primed females. Prnd^neo/neo males showed normal mating behavior: their sexual activity was similar to that of controls as revealed by a normal number of copulation plugs (Figure 1D). To clarify the basis of the infertility, we studied the mature sperm of Prnd^neo/neo males. The number of spermatozoa in the cauda epididymis of Prnd^neo/neo males was reduced to 50% of wild-type controls (Figure 1E). Importantly, also the motility of mutant sperm was significantly decreased (Figure 1F). We concluded that the sterility of Dpl-mutant males is not due to behavioural abnormalities, but may be due to a spermatogenesis defect.

Table I. Breeding performance and fertility of Prnd^neo/neo and control mice.

Genotype of males	Prnd+/+	Prndneo/+	Prndneo/neo
Number of males mated	5	4	13
Number of females mated	8	8	21
Number of litters	8	19	1
Average litter size	5.7	5.6	0.04
Fertility rate (%)	100	98	0.7

Normal testes development without Dpl

There was no reproducible difference in the average testicular weight between Prnd^neo/neo mice (185.9 ± 27.1 mg; n = 4) and Prnd^+/neo littermate controls (175.9 ± 27.8 mg; n = 4). The overall histological appearance of mutant testes was normal (Figure 2A), as were brain, colon, heart, kidney, liver, lung, spleen and thymus (data not shown). Histological examination revealed normal distribution and numbers of spermatogonia and spermatocytes, and vimentin staining confirmed the presence of Sertoli cells (data not shown).

Fig. 2. Normal histology and gene expression in Prnd^neo/neo testes. (A) Histological analysis of Prnd^+/+ and Prnd^neo/neo testes. (B) RT–PCR analysis of total testes RNA from wild-type, Prnd^neo/+ and Prnd^neo/neo mice. Reverse transcription reaction was performed with (+RT) and without (–RT) the addition of the reverse transcriptase to control for DNA contamination of the RNA preparation.

To detect more subtle defects in male gametogenesis, the expression of genes that characterize various stages of sperm development was investigated by RT–PCR. The mRNA levels of BMP-8, Hox1.4, protamine 1 and 2 and several other marker genes analyzed was similar in Prnd^+/+, Prnd^+/neo and Prnd^neo/neo mice (Figure 2B). Therefore, expression of maturation stage-specific genes in testes appears to be normal despite the absence of Dpl. Moreover, TUNEL staining did not reveal differences in the number of apoptotic cells between mutant and control testes (data not shown).

Abnormal spermiogenesis in Dpl-deficient testes

We next investigated the expression pattern of Dpl in testes using immunohistochemistry. Dpl protein was detected in the luminal aspect of wild-type seminiferous tubules (Figure 3A and C; indicated by arrows), but not in testes of Dpl mutant mice (Figure 3B and D). Both round and elongated spermatids were strongly immunoreactive for Dpl (Figure 3C). To analyze spermiogenesis in the absence of Dpl in greater detail, we examined Toluidine blue-stained semithin sections by light microscopy. The initial Golgi phases of acrosome formation did not require Dpl, since the development of round spermatids was normal (Figure 4A and B). However, the transformation of round spermatids into testicular spermatozoa was abnormal in Dpl-deficient testes. Whereas wild-type spermatids were undergoing synchronous differentiation, the number of spermatozoa was reduced in Prnd^neo/neo testes, and the regional separation of spermiogeneic differentiation stages was disturbed. Mutant spermatids and more differentiated spermatozoa were detectable in close proximity, indicating a partial blockade of spermiogenesis in the absence of Dpl (Figure 4C and D). Differentiated spermatozoa at late stages of spermiogenesis were present at reduced numbers in mutant seminiferous tubules, in agreement with the reduced numbers of mature sperm found in the cauda epididymis. These findings suggested a role for Dpl in spermiogenesis.

Fig. 3. Dpl is expressed in spermatids. Histological analysis of wild-type (A and C) and Prnd^neo/neo testes (B and D). Expression varies among individual seminiferous tubuli, a finding suggestive of stage-specific expression. Arrowheads in (A) and (C) indicate Dpl immunoreactivity.

Fig. 4. Dpl is required for late stages of spermatogenesis. Semi-thin sections of Prnd^+/+ (A and C) and Prnd^neo/neo testes (B and D) were stained with Toluidine blue. Arrows in (A) and (B) indicate round spermatids. Arrows in (C) and (D) indicate elongating spermatids, arrowheads in (D) show early stages of the elongation phase.

The absence of Dpl causes sperm malformations

We next investigated whether the absence of Dpl had any effect on mature sperm. The spermatozoa from mutant males showed several structural abnormalities. In Prnd^neo/neo spermatozoa, the sperm head was disoriented with respect to the flagellum, and often the flagellum appeared to fold back towards the sperm head instead of being straight and elongated (Figure 5A and B). However, the composition of the mitochondrial sheet was normal in spermatozoa from Prnd^neo/neo males as shown by mitochondrial staining using mitotracker (Figure 5C and D). Moreover, Acridine orange staining did not reveal impaired chromosomal condensation (data not shown). Prnd^neo/neo sperm heads were also severely malformed and lacked a discernible well-developed acrosome (Figure 5E–H). As the acrosome is essential for sperm–egg interaction, this defect could explain the sterility of Prnd-deficient males.

Fig. 5. Spermatozoa from Dpl-deficient mice are malformed. (A and B) Photographs of bright field images of spermatozoa isolated from Prnd^+/+ (A) and Prnd^neo/neo mice (B). (C and D) Spermatozoa isolated from Prnd^+/+ (C) and Prnd^neo/neo mice (D) were stained with mitotracker to detect mitochondria (green) and with the DNA stain Hoechst (blue). (E–H) Photographs of bright field images of sperm heads from Prnd^+/+ (E) and Prnd^neo/neo mice (F–H).

Dpl is required for acrosome function

To gain further insight into the defect caused by the absence of Dpl and to validate the role of Dpl in acrosome function, in vitro fertilization (IVF) experiments were performed. After IVF, the percentage of 2-cell embryos recovered in the control group was 77%, but spermatozoa isolated from Prnd^neo/neo males were unable to fertilize wild-type oocytes (Table II). Spermatozoa from Prnd^neo/neo males occasionally stuck to, but never penetrated, the zona pellucida. In a further set of experiments, we performed a partial zona dissection (PDZ) to test whether the acrosomal reaction could be triggered in spermatozoa from homozygous mutants. The zona pellucida was partially dissected and IVF was performed with sperm suspension from Prnd^neo/neo males. Out of 180 PDZ oocytes fertilized with Prnd^neo/neo spermatozoa, after 24 h, 39 zygotes had progressed to 2-cell stage embryos and 15 zygotes displayed a prominent second polar body, indicative of a fertilization rate of 30% (Table II). Spermatozoa isolated from wild-type males achieved a significantly higher percentage (80%) of fertilization. These data indicated that Prnd^neo/neo spermatozoa are capable of oocyte fertilization, albeit at a lower frequency than controls, but that they cannot overcome the barrier imposed by the zona pellucida.

Table II. In IVF experiments, Dpl-deficient sperms could only fertilize eggs upon mechanical removal of the zona pellucida (ZP).

Genotype of sperm	Prnd+/+	Prndneo/+	Prndneo/neo
2-cell stage (cells with intact ZP)	51/66 (77%)	Not performed	0/118 (0%)
2-cell stage (cells with dissected ZP)	31/39 (80%)	74/60 (78%)	54/180 (30%)

Discussion

In this study, we have investigated the biological function of the prion protein homologue Dpl by targeted disruption of its coding gene Prnd. Whereas females lacking Dpl are normal, Dpl-deficient male mice revealed an essential function of Dpl in spermiogenesis and acrosome biogenesis. A number of gene inactivation studies have identified gene products involved in male fertility, but in most cases female reproduction was unaffected, presumably because spermatogenesis and oogenesis are controlled by largely non-redundant genetic programmes (Grootegoed et al., 2000; Wassarman and Litscher, 2001). It has previously been observed that a significant amount of PrP^C is expressed in mature spermatozoa. The PrP protein found in testes was truncated in its C-terminus in the vicinity of residue 200 (Shaked et al., 1999). A protective role for PrP against copper toxicity has been proposed since sperm cells originating from Prnp- deficient mice were significantly more susceptible to high copper concentrations than sperm from wild-type mice. However, male Prnp^0/0 knock-out mice are not sterile and produce normal litter sizes. As we have shown here, PrP expressed in testes is not capable of compensating for the absence of Dpl, suggesting at least partially non-redundant functions of these two proteins. Also, PrP^C protein levels are comparable in testes and a number of other organs from Prnd^neo/neo and control mice (data not shown). However, it cannot be excluded that Dpl masks a minor function of PrP^C in testes development or spermiogenesis, and that males lacking both PrP^C and Dpl function would display a more severe defect than Prnd^neo/neo single mutant mice.

Spermatozoa are formed from spermatogonial stem cells by a tightly orchestrated process that can be divided into three sequential phases of cell proliferation and differentiation (Grootegoed et al., 2000). First, there is an extensive multiplication and proliferation of spermatogonial stem cells to produce an optimal number of spermatogonia, which give rise to primary spermatocytes, and also to maintain a pool of stem cells. Secondly, the primary spermatocytes undergo a lengthy meiotic prophase, followed by the first meiotic division resulting in the formation of two secondary spermatocytes, each undergoing the second meiotic phase to produce four haploid round spermatids. Finally, there is a gradual remodelling of the nuclear and cellular components of round spermatids during transformation into sperm cells by a process referred to as spermiogenesis (Abou-Haila and Tulsiani, 2000; Grootegoed et al., 2000).

The initial phases of spermatogenesis appear to proceed normally in the absence of Dpl, since the production of round spermatids is unaffected in Prnd^neo/neo mice. The elongation of the spermatid characterizes the late stages of spermiogenesis when the sperm cell flattens as the chromatin condenses and the acrosome becomes compact. The spermatids are in close association with the Sertoli cells until spermatozoa are fully differentiated. Spermatozoa then detach from Sertoli cells, and at this point, the spermatozoon becomes an isolated cell characterized on the basis of its electron-dense nucleus and fully developed acrosome (Abou-Haila and Tulsiani, 2000). Dpl protein expression is confined to round and elongated spermatids, and the presence of Dpl appears to be required for the correct development of the sperm head and the acrosome.

It is worth noting that mature spermatozoa from Prnd^neo/neo males also display reduced motility, although the flagella of Prnd^neo/neo sperm developed normally as judged by transmission electron microscopy, and the mitochondrial sheath was correctly configured. At the onset of sperm elongation, the acrosomic vesicle undergoes a series of highly complex morphological changes that shape the sperm cell. Failure of acrosome biogenesis can result in sperm malformations (Abou-Haila and Tulsiani, 2000; Kang-Decker et al., 2001). At present, the molecular mechanism of Dpl-regulated acrosome development is unclear. Dpl may be present on the acrosomic vesicles through its GPI-anchor, and possibly directly participate in acrosome morphogenesis. Alternatively, Dpl may regulate acrosome formation in a more indirect way. We have also observed that sperm isolated from Prnd^neo/neo mice is greatly impaired in the sperm–egg interaction, and that Prnd^neo/neo spermatozoa fail to trigger the acrosomal reaction. Oligosaccharides have been strongly implicated as being responsible for both sperm binding and signalling for the acrosome reaction, but the composition and structure of the essential carbohydrate moieties remain controversial (Wassarman and Litscher, 2001). Dpl is a highly glycosylated protein that has been shown to be mainly located outside of the plasma membrane (Moore et al., 1999; Silverman et al., 2000). It is plausible, but not yet proven, that Dpl present on the sperm plasma membrane is directly involved in the sperm–egg interaction.

A better characterization of the genetic factors of male infertility in humans is required to improve the precision of diagnosis and counselling for couples, and to arrive at new methods of contraception targeting the male (Grootegoed et al., 2000). Therefore, the identification of genetic causes of male infertility has practical clinical relevance. It will be interesting to determine whether Dpl mutations are present in human patients suffering from infertility. Finally, with the need to develop new methods of contraception, in particular methods targeting the male, inhibition of Dpl function may provide a novel target for contraceptive intervention.

Materials and methods

Targeting the Prnd locus in ES cells

The genomic Prnd locus was cloned from genomic DNA of TC-1 ES cells by PCR. In the targeting vector, the Prnd ORF was replaced with a floxed neomycin resistance selection cassette. A diphteria toxin (DTα) gene was inserted for selection against random integrants. Gene targeting was performed as described (Behrens et al., 2001).

Histological analysis and TUNEL assay

Tissues were fixed in 4% formaldehyde, dehydrated, embedded in paraffin and sectioned (5 μm). Sections were stained with Harris haematoxylin and eosin (Sigma). TUNEL stains were performed as described (Behrens et al., 1999). For immunohistochemistry, tissue sections were fixated with acetone, air-dried and blocked in 2% normal donkey serum (Jackson Immunoresearch Laboratories). Anti-Dpl antibody (G-20; sc-16863, Santa Cruz, CA) diluted 1:20 in 2% normal donkey serum was used as primary antibody. Donkey-anti-goat AP (Jackson) diluted 1:80 in 5% normal donkey serum was used as a secondary antibody. Sections were counterstained with new fuchsin.

Northern and western blot analysis

Mice were killed and RNA isolation, protein isolation and northern and western blots were performed as described (Behrens et al., 1999). The Prnd and Prnp ORFs were used as probes for northern blot analysis. For western blot analysis, we used a polyclonal anti-Dpl antiserum (kindly provided by Dr Hermann M.Schätzl).

FACS analysis

Two- and three-colour fluorescence-activated cell sorting (FACS) analyses were performed on a FACS Calibur (Becton Dickinson, CA) as described recently (Heppner et al., 2001). Bone marrow cells were prepared by flushing the femur with 5 ml ice-cold FACS buffer (phosphate-buffered saline, 2% fetal calf serum, 10 mM EDTA pH 8, 15.3 mM NaN₃). Spleen cell suspensions were obtained by meshing total spleen through a 40 µm nylon cell strainer (Falcon). Mouse blood was obtained retro-orbitally. The following anti-mouse antibodies were used (all PharMingen, Europe): peridinin chlorophyll-a protein (PerCP)-labelled anti-B220, fluorescein isothiocyanate (FITC)-labelled anti-CD21, phycoerythrin (PE)-labelled anti-CD19, FITC-labelled anti-CD25, PE-labelled anti-CD23, PE-labelled anti-CD8 and FITC-labelled anti-CD4. Lysis of red blood cells was performed with FACS lysing solution (Becton Dickinson) according to the manufacturer’s protocol. Living cells were gated using a combination of forward scatter and side scatter. Data were acquired with CellQuest software (Becton Dickinson); analyses were performed with WinMDI (Version 2.8, available at http://facs.scripps.edu).

RT–PCR

Total testes RNA from wild-type, Prnd^neo/+ and Prnd^neo/neo mice was reverse transcribed using the Geneamp kit (Roche) according to the manufacturer’s instructions. The primer sequences used for PCR amplification are available upon request.

Semen processing and examination

Testis and epididymis were dissected from both sides of each mouse and the weight was measured. The cauda epididymis was excised and immediately punctured in 500 µl human tubal fluid (HTF) medium at 37°C. After 5 min, a 6 µl drop of medium with dispersed sperm cells was placed on a warmed glass slide and covered with a coverslip (24 × 24 mm). Progressive motility was estimated by phase-contrast microscopy at a magnification of ×200 (Olympus BX50, UplanFl 20x/0.50 Ph1). The sperm concentration was measured 30 min after puncturing the epididymis, using an improved Neubauer hemocytometer counting chamber.

Sperm viability was determined 5 min after puncturing the epididymis by dual DNA staining (LIVE/DEAD^® Sperm Viability Kit, Molecular Probes Europe, Leiden, NL) using SYBR-14 in combination with propidium iodide (PI). The SYBR-14 (component A) was diluted with anhydrous dimethyl sulfoxide (DMSO) 1:50 (SYBR-14 working solution), while PI (component B) was used in the original concentration. One microlitre of the SYBR-14 working solution was added to 100 µl of semen. After incubation of 5 min at room temperature, 1 µl of component B was added. Five minutes later, 5 µl of stained semen were placed on a slide, covered with a coverslip (24 × 24 mm) and examined using fluorescence microscopy (Olympus BX50, UPlanApo 40×/0.85, FITC filter, high pressure Hg-lamp). Different sequences of fluorescence-stained spermatozoa were monitored by connecting a video camera (SANYO VCC-2972) to the microscope. At least 200 sperm cells were counted on the screen and the percentage of green fluorescing spermatozoa was defined as viability.

IVF and PZD

Cumulus masses of super ovulated B6D2F1 females were collected in HTF medium with 300 iU hyaluronidase/ml. The cumulus-free oocytes were introduced into a drop of 0.3 M sucrose solution (to increase the perivitelline space) but without bovine serum albumin (BSA; to attach the oocytes on the dish surface). The zona pellucida was partially dissected by a single downward motion with a 30G needle. The attached oocytes were then removed from the dish surface by adding medium with 4% BSA. PDZ oocytes were washed gently in HTF medium and transferred into a 200 µl fertilization drop. For IVF, 60 µl of sperm suspension isolated from the epididymis (after 1 h culture for capacitation) were added.