The sperm chemoattractant “allurin” is expressed and secreted from the Xenopus oviduct in a hormone-regulated manner

Developmental Biology 275 (2004) 343 – 355 www.elsevier.com/locate/ydbio

The sperm chemoattractant ballurinQ is expressed and secreted from the Xenopus oviduct in a hormone-regulated manner Xueyu Xiang, Lindsey Burnett, Alan Rawls, Allan Bieber, Douglas Chandler* Molecular and Cellular Biology Program, School of Life Sciences, Arizona State University, Tempe, AZ 85287-4501, United States Received for publication 2 June 2004, revised 3 August 2004, accepted 6 August 2004 Available online 11 September 2004

Abstract Recently, we cloned and sequenced the cDNA of allurin, a sperm chemoattractant isolated from the jelly of Xenopus laevis eggs [Proc. Natl. Acad. Sci. U.S.A. 78 (2001) 11205]. In this report, we demonstrate that allurin mRNA is expressed almost exclusively in the oviduct and that its expression is increased 2.5-fold by human chorionic gonadotropin over a 12-h period. Both dot blots and immunocytochemistry show that allurin is secreted from the upper two thirds of the oviduct that includes the pars recta and the proximal pars convoluta. Allurin appears to be deposited on the ciliated surfaces of luminal epithelial cells that come in direct contact with eggs as they move through the oviduct. Immune staining also demonstrates the presence of allurin in the serosal capsule of the oviduct. In contrast, allurin is not found within the tubular jelly-secreting glands or ducts that constitute a major portion of the oviduct wall. Therefore, we hypothesize that allurin is synthesized by nonciliated secretory cells in the luminal epithelium of the oviduct, is displayed on the ciliary layer and then mechanically mixed with jelly, and applied to eggs as they progress down the oviduct. This hypothesis is consistent with the fact that eggs progressing down the oviduct initially show evidence of allurin being incorporated into the J1 layer. Subsequently, allurin within J1 diffuses outward to J3 and eggs stored in the uterus now demonstrate a J3 localization of this chemoattractant. D 2004 Elsevier Inc. All rights reserved. Keywords: Sperm chemotaxis; hCG; Xenopus laevis; Egg jelly; Fertilization

Introduction A highly coordinated set of interactions between sperm and the egg extracellular matrix is necessary for fertilization in all animal species. These interactions include activation of sperm motility, sperm chemotaxis towards the egg, sperm penetration of the extracellular coats, induction of an acrosome reaction, and finally binding to and fusion of the sperm with the egg plasma membrane (Foltz and Lennarz, 1993; Hedrick and Nishihara, 1991; Katagiri, 1987; Romanoff and Myles, 2002; Ward and Kopf, 1993; Wassarman et al., 2001). An important step in these

* Corresponding author. Fax: +1 480 965 9699. E-mail address: [email protected] (D. Chandler). 0012-1606/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ydbio.2004.08.011

interactions is sperm chemotaxis, that is, the movement of sperm up a gradient of chemoattractant toward the ovulated egg. It is prevalent throughout the Metazoa, from marine species with external fertilization to humans (Cosson, 1990; Eisenbach, 1999a,b; Fuchs et al., 2001; Miller, 1985; Nishigaki et al., 1996; Ward and Kopf, 1993). Our primary understanding of sperm chemotaxis has benefited from studies of invertebrates. Sperm chemoattractant peptides from invertebrate eggs and their receptors have been well characterized (Kaupp et al., 2003; Matsumoto et al., 2003; Miller, 1985; Riffell et al., 2002; Spehr et al., 2003; Suzuki, 1985; Yoshida et al., 2002). For example, resact and speract, isolated from A. punctulata and S. purpuratus, respectively (Cook et al., 1994; Kaupp et al., 2003; Suzuki et al., 1984; Ward et al., 1985), are the most extensively studied peptides and there is a growing

344

X. Xiang et al. / Developmental Biology 275 (2004) 343–355

understanding of their actions at the molecular level (Wood et al., 2003; Dangott et al., 1989; Daszon et al., 2001; Solzin et al., 2004). Resact, for example, binds to a guanylate cyclase receptor on the sperm surface, switches on a cGMP signal transduction pathway, and mediates sperm movement toward the egg (Dangott et al., 1989; Matsumoto et al., 2003). In humans and mammals, it has been demonstrated that follicular fluid contains one or more chemoattractants for sperm that are able to initiate changes in sperm direction, drive sperm accumulation, and enhance sperm motility and velocity in vitro (Cohen-Dayag et al., 1995; Eisenbach, 1999a,b; Falcone et al., 1991; Ralt et al., 1991, 1994; Villanueva-Diaz et al., 1990). However, no active factor in vertebrates had been identified until recently when our laboratory purified and sequenced allurin, a 21-kDa sperm chemoattractant was found in the jelly layers of Xenopus laevis eggs (Al-Anzi and Chandler, 1998; Olson et al., 2001). Purified allurin has been tested for chemotactic activity, molecularly cloned from oviductal mRNA and sequenced. The allurin sequence shows that this protein has homology to members of the mammalian cysteine-rich secretory protein (CRISP) family (Fernandez et al., 1997; Kratzschmar et al., 1996), including testis-specific spermatocyte protein (TPX-1; Foster and Gerton, 1996; Hardy et al., 1988; Maeda et al., 1998; O’Bryan et al., 2001) and acidic epididymal glycoprotein (AEG; Brooks et al., 1986; Cohen et al., 1992, 2000; Hayahi et al., 1996; Rochwerger and Causnicu, 1992; Xu and Hamilton, 1996), both of which have been demonstrated to bind to sperm. Addition of allurin to this family supports the hypothesis that CRISP family members escort or guide sperm on their voyage all the way from the testes to the egg: TPX-1 functions to link spermatocytes to Sertoli cells (Maeda et al., 1998), AEG, then stimulates sperm maturation in the epididymis (Cohen et al., 2000), and finally in the female reproductive tract, allurin binds to sperm and directs them toward the egg by its concentration gradient. In this study, we provide evidence that allurin is produced almost exclusively in the upper oviduct as detected at both the mRNA and protein levels. To determine which region of the oviduct is responsible for the secretion of allurin, we measured allurin expression in three different regions of oviduct: upper (OVU), middle (OVM), and lower (OVL). We find that allurin mRNA and protein are highly expressed in the upper third (OVU) that consists of the pars recta and pars convoluta 1 region as defined by Yoshizaki (1985) and further illustrated by Wake and Dickie (1998). Expression is regulated by human chorionic gonadotropin (hCG) with allurin mRNA levels increasing about 2.5-fold over a 12-h period. Surprisingly, allurin is not secreted together with jelly from the tubular secretory glands, which make up the largest part of the oviduct wall. Although the exact type of cell responsible for allurin secretion is not known, detection of allurin on the extracellular surfaces of the

ciliated luminal epithelium points to the cells of this epithelium being involved.

Materials and methods Animal, tissue, and egg water preparation Male and female X. laevis were obtained commercially (Xenopus Express) and kept on a 12:12-h light/dark cycle at 208C in circulating tap water. Eggs, ovarian, oviductal, and uterine tissues as well as nonreproductive tissues were obtained by dissection of euthanized females primed with 900 units of hCG (Sigma) at 0–14 h prior. Testes were obtained from euthanized male frogs by dissection. For RNA isolation, the entire length of oviduct was cut into three portions: OVU (upper), OVM (middle),, and OVL (lower). As defined by a previous anatomical observation (Yoshizaki, 1985), OVU refers to the uppermost region closest to the ovary including the pars recta (PR1 and PR2) and pars convoluta 1 (PC1) regions; OVM refers to the pars convoluta 2 (PC2) region; and OVL refers to the lowermost region close to the uterus including the pars convoluta 3 and 4 (PC3 and PC4) regions. For immunocytochemistry, oviduct tissue and eggs were taken from the specific regions described by Yoshizaki (1985): PR1, PR2, PC1, PC2, PC3, and PC4. The diffusible components of egg jelly, referred to as begg water,Q were prepared by incubating 1 g of freshly spawned eggs in 4 ml of F-1 buffer (41.25 mM NaCl, 1.25 mM KCl, 0.25 mM CaCl2, 0.06 mM MgCl2, 0.5 mM Na2HPO4, 2.5 mM Hepes, pH 7.8) at 228C with gentle swirling for 3 h (Al-Anzi and Chandler, 1998; Sugiyama et al., 2004). The conditioned medium was removed with a micropipette taking care not to break or transfer eggs and stored at 208C until use. Twelve-hour egg water (12 HEW) was prepared by incubating 1 g of eggs in 4 ml of 1.5 OR2 buffer (124 mM NaCl, 3.75 mM KCl, 1.5 mM CaCl2, 1.5 mM MgCl2, 1.5 mM Na2HPO4, 10 mM Hepes, pH 7.8) for 12–16 h at 48C and the medium removed and stored frozen at 208C. Construction of antisense oligonucleotide probes for allurin expression Antisense probes specific for allurin mRNA were produced from a 600-bp sequence containing the coding and 3Vuntranslated regions of allurin cDNA inserted into a TA cloning vector (Invitrogen). The plasmid DNA was linearized by BamHI (Promega) digestion and transcribed in vitro using T7 RNA polymerase and digoxigeninlabeled nucleotides (Boehringer-Mannheim). Antisense probes specific for Xenopus glyceraldehyde-3-phosphate dehydrogenase (XGAPDH) were produced by an identical protocol except that the plasmid DNA was linearized by E. coli I (Promega) digestion and transcribed by SP6 RNA

X. Xiang et al. / Developmental Biology 275 (2004) 343–355

polymerase (Boehringer-Mannheim). The XGAPDH plasmid was a generous gift from Dr. D. C. Eaton of Emory University. Production of polyclonal antibodies to allurin Allurin was purified from 12-h egg water (Olson et al., 2001) and 110 Ag was used in the immunization of one rabbit for production of polyclonal antibodies (Rockland, Inc.), yielding approximately 217 ml of antiserum. Antiallurin antibodies were characterized by slot blots and Western blots using 12-h egg water as the target (Xiang et al., 2004). Slot blots using antiallurin serum produced a strong signal at antibody titers as low as 1:50,000. Western blots exhibited label at one major band allurin at a relative mobility of 23 kDa. Densitometry of Western blots showed that the allurin band accounted for 97% and 93% of the labeling when egg water protein and oviduct protein, respectively, were used as targets. Although two very weak bands at 110 and 150 kDa were present in Western blots of oviductal tissue, these bands were not hormone dependent as was allurin. In contrast, use of preimmune serum or no serum for the primary antibody step resulted in no signal (Xiang et al., 2004). Reverse transcription PCR (RT-PCR) Portions of the oviduct were frozen in liquid nitrogen, crushed into powder in frozen Trizol Reagent (Invitrogen), and total RNA was isolated according to the manufacturer’s instructions. Complementary DNA (cDNA) was generated by reverse transcription of total RNA using SuperScriptk First-Strand Synthesis System (Gibco). cDNA was amplified using gene-specific primers for allurin (5VTTCGTGGTATAATGAAAGAT3Vand 5VCGCGCGTTTTTTTTTTTTTT3V) and for GAPDH (5VTGACCCCTTCATCGACTTGG3V and 5VGACACGGAAAGCCATTCCG3V). Products were analyzed by 1% agar gel electrophoresis, the gel stained with SYBR Green I (Molecular Probes) and visualized on a Molecular Dynamics Storm PhosphoImager. Northern and dot blotting For Northern blotting, electrophoresis of total RNA was performed using a 1% agarose gel containing 2.2 M formaldehyde with a 1 MOPS running buffer, pH 7.0 (Sigma #M5755). A total RNA sample mixture containing 5 Ag RNA in 11 Al, 5 Al 10 MOPS, pH 7.0, 9 Al 12.3 M formaldehyde, and 25 Al formamide was applied to each well and 50 V applied for 3–4 h. The gel was either stained with SYBR Green II (Molecular Probes) or diffusion transferred to Immobilonk-Ny + membrane (Millipore) in the presence of 20 SSC, and the membrane cross linked with a UV lamp (Stratalinker, autocross link setting, 254 nm; Stratagene). Nonspecific membrane binding sites were blocked with DIG Easy Hyb Buffer (Roche, #1603558) at

345

688C for 2 h, and the membrane incubated with 50 ng/ml DIG-labeled antisense RNA probes (for allurin and GAPDH) in DIG Easy Hyb Buffer. After washing at moderate stringency (0.2 SSC/0.1% SDS, 428C), the membrane was incubated for 1 h in nonfat milk blocking buffer (Roche, #1585762) to minimize nonspecific binding. Anti-DIG-alkaline phosphatase (Roche, #1363514) was then used to probe the DIG-labeled hybrids and labeling was visualized by CSPD chemiluminescent substrate (Roche) and exposure to CL-XPosurek film (Pierce). The X-ray film was preflashed and the optical density of the signals on the film was scanned and analyzed by Scion Image software. For dot blotting, 1 Ag total RNA was dissolved in 2 Al DEPC H2O, denatured at 958C for 5 min, cold shocked on ice, then applied to an Immobilonk-Ny + membrane. Cross-linking, hybridization, and detection were identical to those for Northern blotting. Polyacrylamide gel electrophoresis and Western blotting Protein was isolated from the oviducts using Trizol Reagent according to manufacturer’s instructions (Invitrogen). SDS-PAGE was performed on protein extracts using 4–12% gradient gels (NuPage) with an MES/SDS running buffer (Invitrogen, #46-5626). The gels were either stained with Sypro Orange (Molecular Probes) for protein visualization or electrophoretically transferred to PVDF membrane in the presence of Towbin buffer (Towbin et al., 1992; 25 mM Tris, 193 mM glycine, 0.1% SDS, pH 8.4, 15% methanol v/v) for Western blotting. Nonspecific membrane binding sites were blocked with a 5% solution of nonfat milk powder in Tris-buffered saline (TBS; 12.5 mM Tris, 140 mM NaCl, 4 mM KCl, pH 7.6) and the membrane probed with antiallurin antibodies 1:10000 in TBS with 1% Tween v/v. Goat anti-rabbit IgG antibodies (Sigma) conjugated to alkaline phosphatase and extravidin–alkaline phosphatase were then used to probe the primary antibodies and biotinylated molecular weight markers (Sigma), respectively. A chemifluorescent substrate (ECF, Pierce) was used to stain the membrane. Both staining of gels with Sypro Orange and of membranes with ECF were visualized on a Molecular Dynamics Storm PhosphoImager. Histology and immunocytochemistry For histology, the PC1 region of oviduct from hCGprimed frogs was cut into 2-mm-long tubes, flushed with fixative (2% glutaraldehyde and 1% fresh formaldehyde in 1.5 OR2 buffer), and the tissue further sliced into 1-mmthick rings and fixed at room temperature for 2 h. The tissue was then postfixed in 1% osmium tetroxide at room temperature for 1 h, dehydrated in a graded series of ethanol, and embedded in Spurr’s epoxy resin (Electron Microscopy Sciences) according to the manufacturer’s

346

X. Xiang et al. / Developmental Biology 275 (2004) 343–355

instructions. One-half-micrometer-thick sections were cut on an RMC MT6000 ultramicrotome, stained with 0.1% toluidine blue or with EMS Epoxy Tissue Stain (Cat.# 14950), and photographed on a Nikon E300 inverted microscope using a Quantix digital camera (Photometrics). For immunocytochemistry, specimens from the PR1, PR2, and PC1 regions of the oviduct were cut into 2-mmlong pieces and fixed in 10% neutral-buffered formalin (Richard-Allan Scientific) for 2 h at room temperature. After washing for 15 min in phosphate-buffered saline (PBS) three times, the tissues were dehydrated by washing for 15 min each in 25%, 50%, and 75% ethanol, three times in 100% ethanol, and finally two times in Micro-Clear tissue clearing agent (Thomas Scientific). The tissues were then incubated for 15 min at 608C in Micro-Clear agent: paraffin wax, 1:1, then for 15 min three times in paraffin wax. After transferring the tissues into molds, they were allowed to harden at room temperature overnight. The wax blocks were trimmed, 10-Am-thick sections cut on a microtome, and the sections placed on polylysine-coated slides at 378C overnight for relaxation. For immunocytochemistry, the sections were dewaxed with Micro-Clear (three washes, 5 min each) and rehydrated by washing for 5 min three times in 100% ethanol then one time in 75%, 50%, and 25% ethanol sequentially. After removal of the remaining ethanol with PBT (PBS, 0.1% TritonX-100), the sections were incubated for 1 h in nonfat milk blocking buffer and exposed to antiallurin serum at 1:100 dilution in blocking buffer for 16 h at 48C. After washing four times with PBT, the sections were doublestained sequentially with goat anti-rabbit IgG conjugated to rhodamine (Pierce) at 1:100 dilution in blocking buffer for 1 h and with DAPI (10 mg/ml; Sigma) at 1:1000 dilution in blocking buffer for 5 min. Eggs for whole mount immunocytochemistry were dissected from appropriate regions of the oviduct after fixation of 5-mm-long sections of the oviduct in cold methanol. The methanol-fixed eggs were then transferred into ethanol, rehydrated, and stained with antibodies as described above for sections. Both eggs and sections were imaged using an inverted Leica NTS Laser Scanning Confocal Microscope with

10 and 40 objective lenses. Control specimens were treated by an identical protocol except that the primary antibody serum was either omitted or preimmune serum from the same rabbit used instead. Laser power and photomultiplier voltage were kept constant and electronic images files were manipulated in a similar manner for experimental and control specimens by Photoshop 6.0 software.

Results Expression of allurin mRNA in Xenopus reproductive and nonreproductive tissues Since several studies have pointed to the possibility that plasma proteins such as albumin cross the oviduct wall and are incorporated into the oviductal fluids without having originated in the oviduct proper (Gerena and Killian, 1990; Buhi et al., 2000), we first sought to establish that allurin is a product of the oviduct. To investigate the distribution of allurin expression, total RNA was extracted from 11 different Xenopus tissues and allurin and GAPDH transcripts detected by specific antisense oligonucleotides. Both oligonucleotide probes were first tested by Northern blot to ascertain that they labeled only one major RNA species (see Fig. 1A). Dot blots (Fig. 1B, top row) show that substantial production of allurin mRNA was observed in only one X. laevis tissue—the oviduct. With the exception of a low level signal in the ovary and the testis, all other tissues tested, both reproductive (uterus) and nonreproductive (brain, heart, lung, liver, kidney, muscle, and skin), showed no evidence of allurin mRNA expression. Indeed, allurin mRNA levels, measured quantitatively, were undetectable in all tissues except for oviduct, ovary, and testis where the allurin/GAPDH expression ratios were 1.1, 0.26, and 0.20, respectively. Localization of allurin expression in the Xenopus oviduct To determine the distribution of allurin in the oviduct, the oviduct was cut into thirds (labeled OVU, OVM, and

Fig. 1. (A) Digoxigenin-labeled antisense oligonucleotide probes for XGAPDH and allurin each detect a single mRNA species in Northern blot analysis. Each lane was loaded with 10 Ag of total RNA isolated from oviduct. (B) Dot blot analysis of the distribution of allurin mRNA expression in 11 different Xenopus tissues harvested from an hCG-primed frog. Each dot contained 1 Ag total RNA from each of the following organs: ovary (OY), oviduct (OV), uterus (U), testis (T), brain (B), heart (H), lung (LG), liver (LV), kidney (K), muscle (M), and skin (S). The ratio of allurin dot density/XGAPDH dot density was obtained by densitometry as described in Materials and methods. The results are representative of three similar experiments.

X. Xiang et al. / Developmental Biology 275 (2004) 343–355

OVL from anterior to posterior) and dot blots of the total RNA isolated from each portion probed by allurin- and GAPDH-specific antisense oligonucleotides (see Fig. 2A). High levels of allurin transcripts were detected in the OVU and OVM regions, whereas low levels of allurin transcripts were detected in the OVL region. Liver, used as a negative control, showed no expression. These results were confirmed by semiquantitative RT-PCR. cDNA products from OVU and OVL RNA templates were amplified by allurin- and XGAPDH-specific primers (see methods). As shown in Fig. 2B, the upper third of the oviduct (U12) expressed considerably higher amounts of allurin mRNA template compared with the lower third (L12) while they contained the same amount of XGAPDH mRNA indicating equal loading of lanes. A similar pattern was present in allurin protein expression as shown by Western blot (see Fig. 2C). High levels of allurin were found in OVU and OVM, but not in OVL, ovary, or uterus.

347

Hormonal regulation of allurin expression in the Xenopus oviduct Injection of hCG into female X. laevis appeared to increase the expression level of allurin mRNA. This was first ascertained by RT-PCR using total RNA from the upper oviduct of animals primed with 900 units of hCG 12 h earlier (U12; Fig. 2B). The allurin cDNA product was prominent compared with that from nonprimed oviduct (U0; Fig. 2B). There was no allurin expression in the lower oviduct either before (L0) or after (L12) hCG stimulation. Amplification of XGAPDH was used to control for the amount of RNA used in the RT-PCR reaction. To study the kinetics of hCG-induced allurin mRNA and protein expression, we analyzed total RNA and total protein of upper oviducts obtained from X. laevis euthanized at 0 (non-hCG-primed), 3, 6, 9, and 12 h after hCG injection by dot blot and Western blot, respectively (see Fig. 3). In dot blots (Fig. 3A), low levels of allurin mRNA expression were detected at 0 h, a level that increased over the 12-h period after hCG injection. To estimate accurately the increase in expression, we used densitometry to quantitate dot blot density, then calculated the ratio between allurin and GAPDH signals for each time point. As early as 3 h after hCG injection, allurin transcripts were increased. The expression level remained steady at 6 and 9 h but increased again at 12 h post-hCG to 2.5-fold over that at 0 h (Fig. 3B). Allurin protein expression also increased over the same time period as shown by Western blotting. Allurin accumulated steadily following hCG administration reaching sevenfold higher levels at 12 h post-hCG (Fig. 3C). Immunohistochemical localization of allurin in the Xenopus oviduct

Fig. 2. Allurin is expressed mainly in the upper region of the X. laevis oviduct. (A) Localization of allurin mRNA expression in different regions of the Xenopus oviduct by dot blot analysis. LV: liver (negative control); OV: the entire oviduct (positive control); OVU: the uppermost portion of the oviduct including PR and PC1; OVM: the middle portion of the oviduct including PC2; OVL: the lowermost portion of the oviduct including PC3 and PC4. (B) Comparison of allurin mRNA expression in non-hCG primed and hCG-primed Xenopus oviducts by semiquantitative RT-PCR as described in Materials and methods. Both XGAPDH and allurin PCR products are approximately 600 bp. In unprimed oviduct, the lower third (L0) showed no detectable allurin expression while the upper third (U0) showed weak expression. Twelve hours after priming with hCG, the lower third of the oviduct (L12) still showed no allurin expression while the upper third (U12) showed increased expression. (C) Western blot analysis of allurin protein expression in different regions of the oviduct. The allurin band (Mr = 23 kDa) is present in the upper and middle thirds of the oviduct but not in the lower third. Egg water (EW), known to contain 10% allurin (Xiang et al., 2004), was used as a positive control. These results are representative of three similar experiments.

The X. laevis oviduct was divided into subregions PR and PC1 to PC4 by Yoshizaki (1985). In our hands, the histology of all regions showed a number of consistent features documented here in 0.5-Am-thick sections of resin-embedded specimens from the PC1 subregion. In Fig. 4A, a sector of the oviduct wall in this region exhibits simple tubular glands, each gland (dashed outline) having a single secretory duct (SD) extending up through the mucosal epithelium to reach the lumen of the oviduct. These glands are composed solely of jelly secreting cells packed with large basophilic granules and having a single nucleus at their basal end (Fig. 4C). Capillaries (C; Figs. 4A and C) are frequently found between the tubular glands. The mucosal epithelium surrounding the oviduct lumen above is thrown into folds or domes that appear as arches in cross section (Fig. 4B). Each arch consists of two types of epithelial cells radiating from a central capillary: ciliated epithelial cells (CE) that are more numerous interspersed with nonciliated secretory cells having clusters of small granules (asterisks; Fig. 4B). The outer (serosal)

348

X. Xiang et al. / Developmental Biology 275 (2004) 343–355

optics (red arrows, Fig. 5B), also coincide with a networklike pattern of staining (white arrows, Fig. 5A). Of particular interest were funnel-shaped regions of staining that extend from the ciliary border into the interior of the arch (light blue arrows, Fig. 5A). These may represent either granule-containing regions of secretory cells or points where the ciliary layer invaginates into the arch. Although the tubular jelly-secreting glands below were generally absent of staining, occasional small patches of immunostaining were seen on the lateral surfaces of these glands (data not shown). Finally, on the serosa, the connective tissues were strongly stained (Fig. 5C). Staining formed convex caps around the bases of the tubular glands that coincide with the cup-shaped layers of connective tissue that form the capsule of the oviduct (Fig. 4D). DAPI staining points out nuclei of fibroblasts residing in this layer (blue fluorescence, Fig. 5C). Kinetics of allurin expression and application to the egg

Fig. 3. Time course of allurin expression in the upper region of the hCGprimed Xenopus oviduct. (A) Typical dot blot analysis of allurin mRNA expression. (B) Graph of densiometric data (mean F SEM) from three experiments. Allurin dot density was first expressed as a ratio to XGAPDH dot density, then all ratios normalized by setting the ratio at zero time to a value of one. (C) Western blot analysis of allurin protein expression in the oviduct between 0 and 12 h after hCG injection. The relative density of the allurin band (Mr = 23 kDa, arrow), indicated below each lane, increases steadily with time. Egg water (EW) was used as a positive control and its density set to one. All time course studies were performed by extracting total RNA and protein from the Xenopus oviduct at the indicated times after hCG stimulation. Results are representative of three similar experiments.

surface of the oviduct is bounded by a capsule containing collagen and fibroblasts (Fig. 4D). In order to localize allurin, paraffin sections of the PC1 region of the oviduct were processed for immunocytochemistry using antiallurin primary antibodies and anti-rabbit IgG-rhodamine second antibodies. At low magnification, strong immunostaining was found at the ciliated border of the mucosal epithelium and along the serosal boundary of the oviduct wall (Fig. 4E). In contrast, virtually no staining was seen over the tubular jelly glands (Fig. 4E, dashed line). The signal observed was specific immune staining as indicated by the fact that no signal was observed when preimmune serum was used (Fig. 6E) or when the primary antibody was omitted (Fig. 6F). Comparison of immune staining patterns (Fig. 5A) with the cellular architecture (Fig. 5B) in the arches of the mucosal epithelium indicated that the majority of the signal comes from the cilia layer proper and possibly from the apical regions of these cells. The lateral boundaries between epithelial cells, seen as dark lines in phase contrast

The time course of allurin protein expression was studied in the PC1 region of the oviduct using immunocytochemistry (see Fig. 6). Prior to hCG injection of the frog (0 h; Fig. 6A) a sparse distribution of allurin was detected, mostly located on the cilia of the arches. By 6 h post-hCG, heavy staining indicated that large amounts of allurin had accumulated not only on the cilia but also in the cell–cell extracellular matrix, cell–capillary extracellular matrix, and the serosal extracellular matrix (Figs. 5A and 6B). A similar pattern of allurin distribution was detected at 12 h post-hCG (Fig. 6C), except that less staining was observed, especially between cells in the epithelial arches. Again, no staining was detected within cells of the tubular jelly secreting glands at any time point. No eggs were present in the oviducts of 0 and 6 h specimens but at 12 h post-hCG, the oviduct was filled with eggs moving in single file toward the uterus. Immunostaining of sections showing the oviduct engorged with eggs exhibited an epithelial ciliary layer rich in allurin as well as allurin being incorporated into the egg coats (Fig. 6D). Allurin was never seen within the ducts or cells of the jelly glands, but instead large aggregates of allurin appeared to be sloughing off the ciliary surface and mixing into the jelly strands being applied to the eggs (arrows, inset, Fig. 6D). Allurin in egg jelly relocates during passage of the egg down the oviduct Previously, we reported that the allurin of freshly spawned eggs is found primarily in the J2 and J3 jelly layers and not in the J1 layer (Xiang et al., 2004). Since our studies above suggest that allurin is applied primarily in the PC1 and PC2 regions of the oviduct along with the J1 jelly layer, we carried out immunocytochemical studies to verify the relocation of allurin as eggs progress from the ovary to

X. Xiang et al. / Developmental Biology 275 (2004) 343–355

349

Fig. 4. Histology of the proximal pars convoluta (PC1) region of the X. laevis oviduct. (A) The oviduct wall contains a radial array of simple tubular glands extending from the mucosal epithelium to the connective tissue capsule at the serosa. One gland is outlined by white dashes. At the mucosal surface these glands are capped by domes of ciliated epithelial cells (CE). (B) The epithelial domes at the mucosal surface consist largely of two cell types, ciliated cells, and secretory cells packed with small granules (asterisks). Both cell types extend to a centrally located capillary (C) that underlies each dome. Secretory ducts (SD) from the tubular glands below exit to the oviduct lumen between adjacent domes. (C) Longitudinal cross section of a tubular jelly-secreting gland showing a duct (SD) surrounded by a single layer of secretory cells packed with large basophilic granules. Just outside the glands are capillaries (C) and extracellular matrix (ECM). (D) On its serosal surface, the oviduct is protected by a capsule of connective tissue. (E) Survey of immunostaining for allurin in the wall of the oviduct. No staining is seen within the tubular jelly glands (outlined) while strong staining is seen at the ciliated epithelium of the mucosa and the connective tissue capsule of the serosa. Scale bars are 10 Am (B–D) or 50 Am (A and E). The results are representative of three to five experiments.

the oviduct to the uterus. Twelve hours after hCG injection (a time at which the oviduct is filled with eggs moving single file to the uterus), eggs were removed from specific regions of the female tract. Eggs were either examined without fixation by phase contrast optics to verify jelly layer application (Figs. 7B, D, and G) or fixed in cold methanol, processed for immunocytochemistry, and either embedded and sectioned or examined as a whole mount (Figs. 7A, C,

E, F, H, and I). Fig. 7A demonstrates that in the ovary, mature oocytes were surrounded by a single layer of follicle cells recognized by their DAPI-stained nuclei (blue; white arrow). These cells were embedded in a thin extracellular matrix that stained strongly with antiallurin antibody (red; black arrow). This result, combined with the fact that dot blots demonstrated weak expression of allurin in the ovary (Fig. 1B), led to the unexpected finding that allurin is found

350

X. Xiang et al. / Developmental Biology 275 (2004) 343–355

in the ovary, albeit in a highly specific location. In the PR2 region of the oviduct, the egg is coated with a bprefertilizationQ (PF) layer seen in Fig. 7B; allurin is clearly associated with the outer PF as shown in Fig. 7C and in other specimens not shown allurin staining is actually seen throughout the PF. From these observations, we conclude that the PR2 region of the oviduct begins application of allurin to the egg. Although we cannot rule out secretion of allurin from the PR1 region as well, we are unable to provide evidence for this since all of our specimens exhibited a PF layer that is thought to be added in the PR2. Application of the J1 layer and the thinner, refractile J2 layer of jelly in the PC1 and PC2 regions of the pars convoluta (see Fig. 7D) was accompanied by heavy deposits of allurin in both layers (see Fig. 7E). As the J3 layer began to be deposited in the PC3, allurin continued to be present due either to continued application or to migration of the allurin originally applied in J1 or J2 (see Fig. 7F). Near the end of PC3 and in PC4, all three jelly layers were present (Fig. 7G) and allurin was now localized exclusively in J3 (see Fig. 7H). Allurin continued to be localized in J3 in eggs that had reached the uterus (Fig. 7I), and an identical localization is seen in freshly spawned eggs (Xiang et al., 2004).

Discussion

Fig. 5. Immunocytochemical localization of allurin in the X. laevis oviduct. (A) A single epithelial arch showing very strong staining at the ciliated border that lines the lumen of the oviduct. This staining exhibits funnelshaped extensions (light blue arrows) into the cellular areas of the arch, possibly representing allurin-secreting cells. Weaker staining is seen between epithelial cells of the arch (white arrows) and in association with the endothelium that surrounds the central capillary (C). (B) Transmission micrograph of the same arch shown in A. Red arrows indicate apparent boundaries between cells that are weakly stained in A. Blue arrows indicate funnel-shaped regions of allurin staining. (C) Strong immunostaining is also seen in the serosal capsule of the oviduct that covers the bases of the tubular glands (TG). Blue DAPI-stained nuclei are likely those of fibroblasts. Scale bars are 10 Am (A and B) or 25 Am (C). The results are representative of five similar experiments.

Allurin, isolated from Xenopus egg jelly, was the first vertebrate chemoattractant to be purified and sequenced. This protein shares homology with other cysteine-rich secretory protein (CRISP) family members some of which are mammalian sperm binding proteins expressed in the male reproductive tract, for example, TPX-1 and AEG (Olson et al., 2001; Fernandez et al., 1997; Kratzschmar et al., 1996; Cohen et al., 1992; Foster and Gerton, 1996; Hardy et al., 1988). Although the mechanisms that couple allurin activation to behavioral changes in sperm swimming are not yet known, it is clear that allurin constitutes a major protein of Xenopus egg jelly (3% of total protein) that diffuses rapidly into the medium as eggs are spawned into pond water (Xiang et al., 2004). The jelly layers of amphibians are synthesized in the oviduct and deposited sequentially onto the egg during its travel through this organ (Bakos et al., 1990). The high molecular weight bstructuralQ glycoconjugates that compose most of the jelly are manufactured by tubular gland cells that store these products in large basophilic secretory granules (Bonnell et al., 1996; Yoshizaki, 1985; and Fig. 4C). In addition, lower molecular weight proteins such as the acrosome reaction-inducing substance of jelly (Ueda et al., 2003) and coelomic envelope processing proteins such as oviductin (Hiyoshi et al., 2002; Lindsay et al., 1999) are also secreted by the oviduct in a hormone-dependent manner. Thus, we anticipated that allurin might be expressed and secreted by the oviduct in a hormone-

X. Xiang et al. / Developmental Biology 275 (2004) 343–355

351

Fig. 6. Immunocytochemistry demonstrates that allurin stored at the luminal epithelium (CE) reaches a peak at 6 h post-hCG and is subsequently applied to the egg as it passes down the oviduct. (A–C) Time course series of micrographs demonstrating that the small amount of allurin found at the arches before hCG injection (0 h) is dramatically increased at 6 h post-hCG in preparation for egg arrival. By 12 h post-hCG, egg passage is well underway and the arch epithelium exhibits substantial albeit reduced amounts of allurin. (D) As eggs are moved down the oviduct, allurin is transferred from the ciliated epithelium to the forming jelly layers in what appears to be a mechanically driven process. Composite transmission–fluorescence images show allurin-stained strands of jelly being added to the egg (arrows, inset, panel D). (E) No primary antibody control. (F) Preimmune serum control. The oviducts were obtained from the frogs 0, 6, and 12 h after hCG stimulation, and the sections double stained by rhodamine and DAPI for allurin and nuclei, respectively. CE: ciliated epithelium; TG: tubular gland; ECS: extracellular space; SD: secretory duct; PF, prefertilization layer. These results are representative of three similar experiments. Scale bars are 50 Am (A–C) or 25 Am (D–F).

dependent manner as are other jelly components. The current report demonstrates that this hypothesis is correct. Dot blots using an allurin-specific oligonucleotide probe show that allurin mRNA is expressed almost exclusively in the oviduct. The fact that allurin is not expressed in amphibian tissues outside of the reproductive system suggests that allurin is a reproduction-specific protein having a reproductive function. Localizing the region of allurin expression in the oviduct has allowed us to determine the stage at which allurin is

incorporated into the jelly and to determine the richest source of allurin mRNA. The oviduct is composed of a pars recta region (PR1 and PR2) about 2 cm in length that is followed by the pars convoluta (PC1-4) about 40 cm in length. It is known that electron-dense particles characteristic of the prefertilization layer are produced in the PR2 region, and that jelly layers J1 to J3 are added in the PC1 to PC4 regions sequentially (Yoshizaki, 1985). Since allurin in spawned eggs is found in the outermost jelly layer (Xiang et al., 2004), we expected that allurin would be expressed near

352

X. Xiang et al. / Developmental Biology 275 (2004) 343–355

Fig. 7. Immunolocalization of allurin associated with Xenopus eggs during progression through the female reproductive tract. (A) Ovarian oocytes are surrounded by a thin, immunoreactive extracellular matrix (red) in which follicle cells are embedded. The nuclei of these cells are stained with DAPI (blue). (B) Phase contrast micrograph showing the prefertilization (PF) layer of eggs isolated in the PR2 region of the oviduct. (C) Section of an egg isolated from the PR2 region of oviduct showing strong antiallurin immunoreactivity at the transition between PF and JI deposition. Weak staining is also seen at the vitelline envelope in some eggs (arrow). (D) Phase contrast micrograph showing the J1 and J2 jelly layers of eggs isolated from the pars convoluta 2 region of the oviduct. (E) Strong antiallurin immunoreactivity is seen in the J1 and J2 of eggs isolated from the pars convoluta 2 region. (F) Eggs from the pars convoluta 3 region have now started to exhibit a J3 layer; allurin is present in J1, J2, and the new J3. (G) Phase contrast micrograph showing that all three jelly layers are present in eggs isolated from the late pars convoluta 3 region. (H and I) Eggs in the pars convoluta 3 region and in the uterus both show allurin located exclusively in the J3 layer suggesting that allurin has relocated from the J1 layer to the J3 layer during passage of the jellied egg down the oviduct. These results are representative of three similar experiments. Scale bars are 10 Am (A), 100 Am (B and C), or 50 Am (D–I).

the end of the oviduct, for example, PC3 and PC4. However, in contrast to our prediction, dot blots and RTPCR detection of allurin mRNA (Fig. 2) show that allurin is highly expressed in the upper half of the oviduct including

the pars recta, PC1, and PC2 regions. The basis for this discrepancy is now clear. Immunolocalization of allurin in eggs harvested from specific points in the oviduct (Fig. 7) demonstrates that allurin is secreted and incorporated into

X. Xiang et al. / Developmental Biology 275 (2004) 343–355

the inner layers of the jelly in the PR2, PC1, and PC2 regions of the oviduct but then diffuses from inner to outer layers as the egg travels through the remainder of the oviduct. It is well known that most vertebrates release their eggs periodically or seasonally under the control of sexual hormones. In amphibians, ovulation in the ovary and secretion of jelly constituents in the oviduct is initiated by a surge of pituitary gonadotropin (Evennett and Thornton, 1971; Kelly, 1982; Kim et al., 1998; King and Millar, 1979; Licht, 1979, 1990; McCreery and Licht, 1983). Our data clearly demonstrate that human chorionic gonadotropin (hCG), an LH-like hormone, induces the production of allurin in addition to triggering ovulation and jelly glycoconjugate secretion. In dot blots, the relatively low level of allurin mRNA expression observed in non-hCGprimed frogs was increased 2.5-fold after hormonal stimulation for 12 h. At the protein level, Western blots showed that allurin is increased sevenfold over basal levels at 12 h after hCG. As expected, allurin mRNA levels rose early (within 3 h) while the protein levels rose gradually over the remaining 9 h. Although there appears to be a second increase in allurin mRNA level at 12 h post-hCG, the underlying basis for this is not known. One interesting possibility is that given the impressive number of eggs released from the ovary at 9–12 h post-hCG, the secretion of allurin may be elicited by mechanical stimuli rather than in response to hormones. The contact between eggs and oviductal wall and the peristalsis of the oviduct might act as potential inducers of allurin secretion during this second increase. The low level of allurin expression in non-hCGstimulated frogs may be due to constitutive expression or due to a long-acting trophic effect of previous hCG injections. Since we inject our frogs with hCG about once every 2 months, these effects must last at least this amount of time. A more complete study of long-term hCG effects on allurin expression is needed. However, our preliminary studies do show that sexually immature females and females, which have not been injected with hCG for 6 months, show no detectable levels of allurin mRNA or protein. An unexpected finding was that allurin is not produced in the same cells as the jelly glycoconjugates. Immunocytochemistry showed that the simple tubular glands of the oviduct contain no staining for allurin, a result found for the PC2 to PC4 regions as well as the PC1 region of the oviduct (data not shown). Likewise, the secretory ducts that lead from these glands contained no staining for allurin. Thus, our focus became the epithelial folds or barchesQ that line the lumen of this passage. As shown in Fig. 4B, the arches contain ciliated cells and, in addition, another type of secretory cell whose granules appear ovoid and much smaller than the large basophilic secretory granules of the jellysecreting glands below. Both cell types extend their bases to the endothelium of a single capillary that runs though each

353

arch. These observations are consistent with Yoshizaki’s (1985) at both the light and electron microscope level. Allurin appeared to be located at extracellular surfaces, most heavily on the ciliary surface and to a lesser degree in cell–cell interstices within the arches. Also present were inverted cones of patchy staining that could represent the granule-containing regions of arch secretory cells, which when triggered to undergo exocytosis would release allurin to the ciliary bforestQ above. Alternatively, these regions could represent deeper recesses of the ciliary border. If so, this might mean that allurin is secreted by an alternative pathway. Unlike the classic regulated pathway involving secretory granules, allurin protein might be synthesized continuously, carried to the apical plasma membrane by small transport vesicles, and deposited onto the surface of microvilli. In either case, allurin would be stored temporarily while anchored to the ciliated cell surface. Then, during the passage of eggs, the jelly secreted from the tubular glands would be ejected into the lumen of the oviduct, mixed mechanically with allurin, and the mixture applied to the egg surface by the force of oviductal peristalsis and ciliary beating. Since transport vesicles containing allurin would be small, they are unlikely to be observed at the light microscope level and study of this possibility will need to be carried out by electron microscopy.

Acknowledgments We would like to thank the W. M. Keck Bioimaging Laboratory for use of their laser scanning confocal microscopy instrumentation and to Dr. Dennis McDaniel who provided technical training in the microscopy work. This study was supported by NSF grant IBN-0130001. References Al-Anzi, B., Chandler, D., 1998. Xenopus laevis egg jelly releases a sperm chemoattractant during spawning. Dev. Biol. 198, 366 – 375. Bakos, M.A., Kurosky, A., Hedrick, J.L., 1990. Physicochemical characterization of progressive changes in the Xenopus laevis egg envelope following oviductal transport and fertilization. Biochemistry 23, 609 – 615. Bonnell, B.S., Reinhart, D., Chandler, D.E., 1996. Xenopus laevis egg jelly coats consist of small diffusible proteins bound to a complex system of structurally stable networks composed of high-molecular-weight glycoconjugates. Dev. Biol. 174, 32 – 42. Brooks, D.E., Means, A.R., Wright, E.J., Singh, S.P., Tiver, K.K., 1986. Molecular cloning of the cDNA for androgen-dependent sperm coating glycoproteins secreted by the rat epididymis. Eur. J. Biochem. 161, 13 – 18. Buhi, W.C., Alvarez, I.M., Kouba, A.J., 2000. Secreted proteins of the oviduct. Cells Tissues Organs 166, 165 – 179. Cohen, D.J., Munuce, M.J., Causnicu, P.S., 1992. Mammalian sperm-egg fusion: the rat egg has complementary sites for a sperm protein that mediates gamete fusion. Dev. Biol. 153, 83 – 90. Cohen, D.J., Ellerman, D.A., Causnicu, P.S., 2000. Mammalian sperm-egg fusion: evidence that epididymal protein DE plays a role in mouse gamete fusion. Biol. Reprod. 63, 462 – 468.

354

X. Xiang et al. / Developmental Biology 275 (2004) 343–355

Cohen-Dayag, A., Tur-Caspa, I., Dor, J., Mashiach, S., Eisenbach, M., 1995. Sperm capacitation in humans is transient and correlates with chemotactic responsiveness to follicular factors. Proc. Natl. Acad. Sci. U. S. A. 92, 11039 – 11043. Cook, S.P., Brokaw, C.J., Muller, C.H., Babcock, D.F., 1994. Sperm chemotaxis: egg peptides control cytosolic calcium to regulate flagellar responses. Dev. Biol. 165, 10 – 19. Cosson, M.P., 1990. Sperm chemotaxis. In: Gagnon, C. (Ed.), Controls of Sperm Motility: Biological and Clinical Aspects. CRC Press, Boca Raton, FL, pp. 103 – 135. Dangott, L.J., Jordan, J.E., Bellet, R.A., Garbers, D.L., 1989. Cloning of the mRNA for the protein that crosslinks to the egg peptide speract. Proc. Natl. Acad. Sci. U. S. A. 86, 2128 – 2132. Daszon, A., Beltran, C., Felix, R., Nishigaki, T., Trevino, C.L., 2001. Ion transport in sperm signaling. Dev. Biol. 240, 1 – 14. Eisenbach, M., 1999a. Mammalian sperm chemotaxis and its association with capacitation. Dev. Genet. 25, 87 – 94. Eisenbach, M., 1999b. Sperm chemotaxis. Rev. Reprod. 4, 56 – 66. Evennett, P.J., Thornton, V.F., 1971. The distribution and development of gonadotropic activity in the pituitary of Xenopus laevis. Gen. Comp. Endocrinol. 16, 606 – 607. Falcone, L., Gianni, S., Piffaretti-Yanez, A., Marchini, M., Eppenberger, U., Balerna, M., 1991. Follicular fluid enhances sperm motility and velocity in vitro. Fertil. Steril. 55, 619 – 623. Fernandez, C., Szyperski, T., Bruyere, T., Ramage, P., Mosinger, E., Wuthrich, K., 1997. NMR solution structure of the pathogenesis-related protein P14a. J. Mol. Biol. 266, 576 – 593. Foltz, K.R., Lennarz, W.J., 1993. The molecular basis of sea urchin gamete interactions at the egg plasma membrane. Dev. Biol. 158, 46 – 61. Foster, J.A., Gerton, G.A., 1996. Autoantigen 1 of the guinea pig sperm acrosome is the homologue of mouse Tpx-1 and human TPX1 and is a member of the cysteine-rich secretory protein (CRISP) family. Mol. Reprod. Dev. 44, 221 – 229. Fuchs, T., Glusman, G., Horn-Saban, S., Lancet, D., Pilpel, Y., 2001. The human olfactory subgenome: from sequence to structure and evolution. Hum. Genet. 108, 1 – 13. Gerena, R.L., Killian, G.J., 1990. Electrophoretic characterization of proteins in oviduct fluid of cows during the estrous cycle. J. Exp. Zool. 256, 113 – 120. Hardy, D.M., Huang, T.T., Driscoll, W.J., Tung, K.K., Wild, G.C., 1988. Purification and characterization of the primary acrosomal autoantigen of guinea pig epididymal spermatozoa. Biol. Reprod. 38, 423 – 437. Hayahi, M., Fujimoto, S., Takano, H., Ushiki, T., Abe, K., Ishikura, A., Yoshida, M., Kirchoff, C., Ishibashi, T., Kasahara, M., 1996. Characterization of a human glycoprotein with potential role in sperm-egg fusion: cDNA cloning, immunohistochemical localization, and chromosomal assignment of the gene (AEGL1). Genomics 32, 367 – 374. Hedrick, J.L., Nishihara, T., 1991. Structure and function of the extracellular matrix of anuran eggs. J. Electron Microsc. Tech. 17, 319 – 335. Hiyoshi, M., Takamune, K., Mita, K., Kubo, H., Sugimoto, Y., Katagiri, C., 2002. Oviductin, the oviductal protease that mediates gamete interaction by affecting the vitelline coat in Bufo japonicus: its molecular cloning and analyses of expression and posttranslational activation. Dev. Biol. 243, 176 – 184. Katagiri, Ch., 1987. Role of oviductal secretions in mediating gamete fusion in anuran amphibians. Zool. Sci. 4, 1 – 14. Kaupp, U.B., Solzin, J., Hildebrand, E., Brown, J.E., Helbig, A., Hagen, V., Beyermann, M., Pampaloni, F., Weyand, I., 2003. The signal flow and motor response controlling chemotaxis of sea urchin sperm. Nat. Cell Biol. 5, 109 – 117. Kelly, D.B., 1982. Female sex behaviors in the South African clawed frog, Xenopus laevis: gonadotropin-releasing, gonadotropic, and steroid hormones. Horm. Behav. 16, 158 – 174. King, J.A., Millar, R.P., 1979. Hypothalamic luteinizing hormone-releasing hormone content in relation to the seasonal reproductive cycle of Xenopus laevis. Gen. Comp. Endocrinol. 39, 309 – 312. Kim, J.W., Im, W.B., Choi, H.H., Kwon, H.B., 1998. Seasonal fluctuations

in pituitary gland and plasma levels of gonadotropic hormones in Rana. Gen. Comp. Endocrinol. 109, 13 – 23. Kratzschmar, J., Haendler, B., Eberspaecher, U., Roosterman, D., Donner, P., Schleuning, W.D., 1996. The human cysteine-rich secretory protein (CRISP) family. Primary structure and tissue distribution of CRISP-1, CRISP-2 and CRISP-3. Eur. J. Biochem. 236, 836 – 872. Licht, P., 1979. Reproductive endocrinology of reptiles and amphibians: gonadotropins. Annu. Rev. Physiol. 41, 337 – 351. Licht, P., 1990. Hypothalamo-pituitary axes and feedback control in amphibians and reptiles. J. Exp. Zool., Suppl. 4, 159 – 160. Lindsay, L., Wieduwilt, M., Hedrick, J., 1999. Oviductin, the Xenopus laevis oviductal protease that processes egg envelope glycoprotein gp43, increases sperm binding to envelopes, and is translated as part of an unusual mosaic protein composed of two protease and several CUB domains. Biol. Reprod. 60, 989 – 995. Maeda, T., Sakashita, M., Ohba, Y., Nakanishi, Y., 1998. Molecular cloning of the rat Tpx-1 responsible for the interaction between spermatogenic and Sertoli cells. Biochem. Biophys. Res. Comm. 248, 140 – 146. Matsumoto, M., Solzin, C., Helbig, A., Hagen, V., Ueno, S., Kawase, O., Maruyama, Y., Ogiso, M., Godde, M., Minakata, H., Kaupp, B., Hoshi, M., Weyand, I., 2003. A sperm-activating peptide controls a cGMPsignaling pathway in starfish sperm. Dev. Biol. 260, 314 – 324. McCreery, B.R., Licht, P., 1983. Induced ovulation and changes in pituitary responsiveness to continuous infusion of gonadotropin-releasing hormone during the ovarian cycle in the bullfrog, Rana catesbeiana. Biol. Reprod. 29, 863 – 871. Miller, R.L., 1985. Sperm chemo-orientation in the Metazoa. In: Metz, C.B., Monroy, A. (Eds.), Biology of Fertilization, Biology of the Sperm, vol. 2. Academic Press, New York, pp. 275 – 337. Nishigaki, T., Chiba, K., Miki, W., Hoshi, M., 1996. Structure and function of asterosaps, sperm-activating peptides from the jelly coat of starfish eggs. Zygote 4, 237 – 245. O’Bryan, M.K., Sebire, K., Meinhardt, A., Edgar, K., Keah, H.H., Hearn, M.T., De Kretser, D.M., 2001. Tpx-1 is a component of the outer dense fibers and acrosome of rat spermatozoa. Mol. Reprod. Dev. 58, 116 – 125. Olson, J.H., Xiang, X., Ziegert, T., Kittelson, A., Rawls, A., Bieber, A.L., Chandler, D.E., 2001. Allurin, a 21-kDa sperm chemoattractant from Xenopus egg jelly, is related to mammalian sperm-binding proteins. Proc. Natl. Acad. Sci. U. S. A. 98, 11205 – 11210. Ralt, D., Goldenberg, M., Fetterolf, P., Thompson, D., Dor, J., Mashiach, S., Garbers, D.L., Eisenbach, M., 1991. Sperm attraction to a follicular factor(s) correlates with human egg fertilizability. Proc. Natl. Acad. Sci. U. S. A. 88, 2840 – 2844. Ralt, D., Manor, M., Cohen-Dayag, A., Tur-Kaspa, I., Ben-Shlomo, I., Makler, A., Yuli, I., Dor, J., Blumberg, S., Mashiach, S., Eisenbach, M., 1994. Chemotaxis and chemokinesis of human spermatozoa to follicular factors. Biol. Reprod. 50, 774 – 785. Suzuki, N., 1985. Structure, function and biosynthesis of sperm-activating peptides and fucose sulfate glycoconjugate in the extracellular coat of sea urchin eggs. Zool. Sci. 12, 13 – 27. Suzuki, N., Shimomura, H., Radany, E.W., Ramarao, C.S., Ward, G.E., Bentley, J.K., Garbers, D.L., 1984. A peptide associated with eggs causes a mobility shift in a major plasma membrane protein of spermatozoa. J. Biol. Chem. 259, 14874 – 14879. Riffell, J.A., Krug, P.J., Zimmer, R.K., 2002. Fertilization in the sea: the chemical identity of an abalone sperm attractant. J. Exp. Biol. 205, 1439 – 1450. Rochwerger, L., Causnicu, P.S., 1992. Redistribution of a rat sperm epididymal glycoprotein after in vivo and in vitro capacitation. Mol. Reprod. Dev. 31, 34 – 41. Romanoff, P., Myles, D.G., 2002. Penetration, adhesion, and fusion in mammalian sperm-egg interaction. Science 296, 2183 – 2185. Spehr, M., Gisselmann, G., Poplawski, A., Riffell, J.A., Wetzel, C.H., Zimmer, R.K., Hatt, H., 2003. Identification of a testicular odorant receptor mediating human sperm chemotaxis. Science 299, 2054 – 2058.

X. Xiang et al. / Developmental Biology 275 (2004) 343–355 Solzin, J., Helbig, A., Van, Q., Brown, J.E., Hildebrand, E., Weyland, I., Kaupp, U.B., 2004. Revisiting the role of H+ in chemotactic signaling of sperm. J. Gen. Physiol. 124, 115 – 124. Sugiyama, H., Al-Anzi, B., McGaughey, R., Chandler, D.E., 2004. Assays for vertebrate sperm chemotaxis. Methods Mol. Biol. 253, 27 – 48. Towbin, H., Staehelin, T., Gordon, J., 1992. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Biotechnology 24, 145 – 149. Ueda, Y., Kubo, H., Iwao, Y., 2003. Characterization of the acrosome reaction-inducing substance in Xenopus (ARISX) secreted from the oviductal pars recta onto the vitelline envelope. Dev. Biol. 264, 289 – 298. Villanueva-Diaz, C., Vadillo-Ortega, F., Kably-Ambe, A., Diaz-Perez, M.A., Krivitzky, S.K., 1990. Evidence that human follicular fluid contains a chemoattractant for spermatozoa. Fertil. Steril. 54, 1180 – 1182. Wake, M.H., Dickie, R., 1998. Oviduct structure and function and reproductive modes in amphibians. J. Exp. Zool. 282, 477 – 506. Ward, C.R., Kopf, G.S., 1993. Molecular events mediating sperm activation. Dev. Biol. 158, 9 – 34. Ward, G.E., Brokaw, C.J., Garbers, D.L., Vacquier, V.D., 1985. Chemotaxis

355

of Arbacia punctulata spermatozoa to resact, a peptide from the egg jelly layer. J. Cell Biol. 101, 2324 – 2329. Wassarman, P.M., Jovine, L., Litscher, E.S., 2001. A profile of fertilization in mammals. Nat. Cell Biol. 3, E59 – E64. Wood, C., Daszon, A., Whitaker, M., 2003. Speract induces calcium oscillations in the sperm tail. J. Cell Biol. 161, 89 – 101. Xiang, X., Kittelson, A., Olson, J., Rawls, A., Bieber, A., 2004. Allurin, a 21 kD sperm chemoattractant, is rapidly released from the outermost jelly layer of the Xenopus egg by diffusion and convection processes. Mol. Reprod. Dev. In press. Xu, W., Hamilton, D.W., 1996. Identification of the rat epididymis-secreted 4E9 antigen as protein E: further biochemical characterization of the highly homologous epididymal secretory proteins D and E. Mol. Reprod. Dev. 43, 347 – 357. Yoshida, M., Murata, M., Inaba, K., Morisawa, M., 2002. A chemoattractant for ascidian spermatozoa is a sulfated steroid. Proc. Natl. Acad. Sci. U. S. A. 99, 14831 – 14836. Yoshizaki, N., 1985. Fine structure of oviductal epithelium of Xenopus laevis in relation to its role in secreting egg envelopes. J. Morphol. 184, 155 – 169.