Abstract
Expansion of the equine conceptus can be divided into blastocoel and yolk sac phases. The endodermal layer transforming the blastocoel into the yolk sac is completed around day 8 of pregnancy. From that time, the size of the spherical conceptus increases tremendously due mainly to the accumulation of fluid rather than cell multiplication. In this study, we have investigated the abundance and localisation of Na+/K+-ATPases and aquaporins (AQP) in the equine conceptus on days 8, 10, 12, 14 and 16 by multiplex reverse transcriptase PCR, Western blot and immunohistochemistry. During conceptus expansion, the ectoderm of the yolk sac exhibited basolateral abundance of α1ATPase, apical localisation of AQP5, and membrane and cytoplasmic expression of AQP3. With increasing conceptus size its cells showed an extensive enlargement of the apical membrane surface by microvilli. From day 14 onwards, the yolk sac endoderm forms arc-like structures with attaching sites to the ectodermal layer and shows intensive staining for α1ATPase, AQP5 and AQP3 in the membrane as well as in the cytoplasm. In the yolk sac ectoderm, the arrangement of these proteins is comparable with the collecting ducts of kidney with AQP2 being replaced by the closely related AQP5. The detection of phosphorylation sites for protein kinase A suggests a similar AQP5 traffic and regulation as known for AQP2 in the collecting ducts of the kidney. The arrangement of these proteins in equine embryos indicates at least partially the mechanism of conceptus expansion.
Introduction
The equine morula or early blastocyst enters the uterus around days 5 to 6 after ovulation. Until this stage, the embryo does not increase substantially in size but it expands rapidly between days 11 and 18, which appears to be favourable for the maintenance of pregnancy (Ginther et al. 1985, Ginther 1992). This increase in size is mainly caused by influx of fluid. The formation of an osmotic gradient by α1/β1 Na+/K+-ATPase is probably the driving force of water influx into the blastocoel of equine embryos as in other mammals (DiZio & Tasca 1977, Manejwala et al. 1989, Betts et al.1997, Jones et al. 1997) because the pump is present in horse embryos by day 8, the earliest stage that has been examined (Waelchli et al. 1997). Sodium ions are pumped out of the trophectodermal cell layer (which is sealed by tight junctions) into the basolateral intercellular spaces that subsequently form the blastocoel (Watson et al. 2004). After completion of the endoderm around day 8, the term yolk sac is used instead of blastocoel. Therefore, distinction must be made between blastocoel expansion prior to and yolk sac expansion after completion of the endodermal lining. Much less is known about yolk sac expansion than about blastocyst expansion. From at least as early as day 10 the yolk sac fluid is very markedly hypotonic until about day 18 and then gradually increases in tonicity (Betteridge 2007). Hypotonicity within the yolk sac seems to contradict the hypothesis of a Na+/K+ trans-trophoblast gradient responsible for blastocoel expansion before day 8 (Waelchli & Betteridge 1996). Subtrophoblastic compartments described in equine blastocysts (Enders et al. 1993) seem to undergo a sharp increase in tonicity relative to the interior of the yolk sac, forming a third compartment which might be responsible for maintenance of the ion gradient in the equine conceptus larger than 6 mm in diameter (Crews et al. 2007).
The water permeability of the trophoblast and yolk sac wall might be mediated by aquaporins which are members of the major intrinsic protein (MIP) family. They form transmembranous pores for water and, in case of the aquaglyceroporins, also for small solutes (Hara-Chikuma & Verkman 2005). In epithelial tissues, aquaporins act as molecular water channels in the direction of osmotic gradients (Verkman et al. 1996, Verkman 1999). During blastocyst cavitation, they enhance a rapid near-isosmotic fluid transport across the trophoblast (Barcroft et al. 2003).
Aquaporin 2 (AQP2) is most abundant in the apical membrane of the collecting duct cells of the kidney and is controlled at least partially by neurohormonal pathways. In the collecting ducts of the kidney, vasopressin stimulates vesicular transport and fusion with the apical membrane via cAMP phosphorylation (Fushimi et al. 1997, Matsamura et al. 1997, Dibas et al. 1998). Vasopressin is also present in equine yolk sacs during pre-attachment development (Waelchli et al. 2000) and the endoderm of the equine trophoblast shares many ultrastructural features with the intercalated cells of kidney collecting tubules (Enders et al. 1993). In lung epithelial cells and the mouse lung tissue, AQP5 mRNA transcription and protein translocation to the apical membrane are stimulated by a cAMP/protein kinase A (PKA)-dependent pathway (Yang et al. 2003).
In mouse blastocysts, a variety of aquaporin transcripts are present and an increased expression at the morula–blastocyst transition has been demonstrated for the AQP3 transcript (Edashige et al. 2000, Offenberg et al. 2000). The immunolocalization of AQP3, AQP8 and AQP9 was investigated during murine preimplantation development. Cytoplasmic localisation of immunoreactive AQP3 protein in the blastomeres of murine eight-cell-stage embryos changes to an apolar localisation within the membrane domains of all blastomeres at the compact morula stage. At the blastocyst stage, an apolar membrane-associated distribution of AQP3 immunoreactive protein was maintained within the inner cell mass, while trophoblast expression was restricted to the basolateral cell margins (Barcroft et al. 2003).
So far, knowledge of the mechanism of equine conceptus expansion after the blastocyst stage and the role of aquaporins during this process is limited. Contrary to other domestic animals, the pre-attachment horse conceptus is of the ‘fully expanding’ type (Biggers 1972), maintaining a spherical form within the embryonic capsule which replaces the zona pellucida after hatching. This shape is beneficial during the mobile phase of pre-attachment development necessary for maintenance of pregnancy. The aim of our study was thus to elucidate the expression pattern and potential role of aquaporins and ATPases during the mobile phase of equine conceptus development. Preliminary results of this study have been presented and published in abstract form (Budik et al. 2006).
Results
Characterization of equine sodium potassium ATPase subunits and aquaporin transcripts
Amplification of a RNA preparation from a day 12 equine conceptus by RT-PCR with primers specific for α1- and β1-subunits of Na+/K+-ATPase (Table 1) resulted in the expected product length of 559 bp for α1 and 613 bp for β1 visualised on agarose gel (Fig. 1A). Sequence analysis of the amplicon for the α1ATPase subunit using the amplification primers and comparison by nucleotide NCBI BLAST resulted in 100% consensus with exons 4 and 5 of equine Na+/K+-ATPase α1-subunit submitted previously (accession numbers
Na+/K+ ATPase (A) and aquaporin (B) expression pattern of a day 12 equine conceptus obtained by RT-PCR using the primers listed in Table 1. (M: 100 bp DNA ladder, 1.5% agarose gel, ethidium bromide stained and visualised by u.v).
Citation: REPRODUCTION 135, 4; 10.1530/REP-07-0298
Primers used for reverse transcriptase PCR.
| Protein | Organism | Accession number | Forward primer 5′–3′ | Reverse primer 5′–3′ | Amplicon length (bp) |
|---|---|---|---|---|---|
| Na+/K+ ATPase α1 | Horse | X16773-X16795 | taaatatgaacccgcagc | tcctcctttcacttccac | 559 |
| Na+/K+ ATPase β1 | Rat | J02701 | cacagattcctcagatcc | cacattccgcataccact | 613 |
| AQP1 | Mouse | NM007472 | atcaagaagaagctcttctgg | ctgatgtgacccacactttg | 200 |
| AQP2 | Human | NM000486 | atccattacaccggctgctc | tccagaagacccagtggtcatc | 91 |
| AQP3 | Human | NM004925 | ttcctcaccatcaacctg | tatgaactggtcaaagaagcc | 333 |
| AQP4 | Mouse | NM009700 | ccgtcttctacatcattgc | tccccttcttctcttctcc | 594 |
| AQP5 | Rat | NM012779 | tgaaaaaggaggtgtgctc | acgtagaagacagctcggag | 271 |
| AQP5 | Human | NM001651 | atgaagaaggaggtgtgctc | ctgaaccgcttcatgatcac | 627 |
| Horse | AJ514427 | ||||
| AQP6 | Rat | NM022181 | ggttctctgtgtctttgcttc | tacacgctcacttctgtgtc | 375 |
| AQP7 | Mouse | NM007473 | catttttgccacctatcttc | gctgtaaaattctccagtctc | 408 |
| AQP8 | Rat | NM019158 | aggagaccaacatggctgac | aatgacagagaaaccaatggag | 536 |
| AQP9 | Human | NM020980 | cgttcatcttgattgtcc | ccaccagaagttgtttcc | 638 |
| β-Actin | Horse | AF035774 | atggaatcctgtggcatc | gcgcaatgatcttgatcttc | 191 |
AQP3 and AQP5 amplicons with an expected length of 333 and 271 bp respectively were detected by gel analysis (Fig. 1B). After excision and purification, sequencing of the RT-PCR products was performed using the amplification primers. The sequences obtained were submitted to nucleotide NCBI BLAST analysis. BLAST search of the 333 bp sequence obtained with human AQP3 primers (Table 1) resulted in a 91% homology to human AQP3 mRNA (accession number
(A) Aquaporin 5 amplicons after RT-PCR of day 15 trophoblast tissue: f, forward; r, reverse; hf, human forward; er, equine reverse (1.5% agarose gel, ethidium bromide stained, visualised by u.v). (B) Schematic drawing of primer and amplicon sites on equine AQP5 mRNA. rAQP5f/rAQP5r, forward/reverse primer designed from rat AQP5 sequence; hAQP2f, AQP2 forward primer designed from human AQP2 sequence; hAQP5f, forward primer designed from human AQP5 sequence; eAQP5r, reverse primer designed from equine AQP5 sequence (all primers are listed in Table 1).
Citation: REPRODUCTION 135, 4; 10.1530/REP-07-0298
Expression pattern of ATPase α1- and β1-subunits and AQP3 and AQP5 during conceptus expansion
DNase I-digested total RNA from equine conceptuses (days 8–16) was investigated for the expression of ATPase α1- and β1-subunits in comparison with β-actin by multiplex RT-PCR (Fig. 3A and B; Table 2). Small day 8 equine embryos (0.6–1.0 mm diameter) show low or absent expression of α1ATPase transcripts. An increase close to statistical significance in the expression of α1ATPase subunit was found in day 10 embryos and again in day 16 embryos (P=0.056 versus day 1, Table 3). ATPase β1 transcripts were just detectable in the small day 8 equine embryos, increased on day 10 and reached maximal values on day 14 (P<0.001 versus day 1, Table 3). Between days 14 and 16, the expression decreased again to values detected in day 10 embryos.
Age: day after ovulation (ovulation detection=day 0); Diam (mm), diameter in millimetres; all gels, 1.5% agarose, ethidium bromide stained, visualised by u.v, not saturated (A) expression pattern of Na+/K+ ATPase α1-subunit mRNA in comparison with β-actin by multiplex RT-PCR at days 8, 10, 12, 14 (each n=4, left panel) and 16 (n=3, right panel) of equine embryo development. (B) Expression pattern of Na+/K+ ATPase β1-subunit mRNA in comparison to β-actin by multiplex RT-PCR at days 8, 10, 12, 14 (each n=4, left panel) and 16 (n=3, right panel) of equine embryo development. (C) Expression pattern of AQP3 mRNA in comparison to β-actin by multiplex RT-PCR at days 8, 10, 12, 14 (each n=4, left panel) and 16 (n=3, right panel) of equine embryo development. (D) Expression pattern of AQP5 mRNA in comparison to β-actin by multiplex RT-PCR at days 8, 10, 12, 14 (each n=4) and 16 (n=3) of equine embryo development.
Citation: REPRODUCTION 135, 4; 10.1530/REP-07-0298
Primers and primer ratios used for reverse transcriptase multiplex PCR.
| Protein | Organism | Accession number | Forward primer 5′–3′ | Reverse primer 5′–3′ | Primer ratio to β-actin (μM/μM) |
|---|---|---|---|---|---|
| Na+/K+ ATPase α1 | Horse | X16773-X16795 | taaatatgaacccgcagc | tcctcctttcacttccac | 0,6/0,6 |
| Na+/K+ ATPase β1 | Rat | J02701 | cacagattcctcagatcc | cacattccgcataccact | 0,6/0,6 |
| AQP3 | Human | NM004925 | ttcctcaccatcaacctg | tatgaactggtcaaagaagcc | 0,8/0,2 |
| AQP5 | Human | NM001651 | atgaagaaggaggtgtgctc | ctgaaccgcttcatgatcac | 0,8/0,2 |
| Horse | AJ514427 | ||||
| β-Actin | Horse | NM020980 | cgttcatcttgattgtcc | ccaccagaagttgtttcc | – |
Relative mRNA expression of α1ATPase, β1ATPase, AQP3 and AQP5 in relation to β-actin (results of multiplex RT-PCR) in embryos of different age and size, values are means±s.e.m.
| Day | α1ATPase | β1ATPase | AQP3 | AQP5 | β-Actin | Size (mm) diameter |
|---|---|---|---|---|---|---|
| 8 (n=4) | 21.6±8.3a | 27.0±7.0a | 0.0±0.0a | 0.0±0.0a | 100 | 0.8±0.1a |
| 10 (n=4) | 92.5±9.9a,b | 150.0±11.7b | 12.0±3.7b | 8.9±3.7a | 100 | 4.3±0.9a,b |
| 12 (n=4) | 91.3±5.6a,b | 209.9±19.8c | 11.8±1.8b | 9.0±3.4a | 100 | 8.1±1.2b |
| 14 (n=4) | 85.9±7.6a,b | 199.4±23.8b,c | 24.5±3.8c | 18.4±2.6b | 100 | 16.5±2.1c |
| 16 (n=3) | 133.1±63.6b | 152.0±17.5b | 19.0±6.0b,c | 3.4±2.8a | 100 | 26.3±2.0d |
| P | (0.056) | <0.001 | <0.01 | <0.01 | – | <0.001 |
a,b,cValues with different letters within rows differ significantly, see table for P values, for α1ATPase statistical significance is nearly reached (P=0.056).
Transcripts of neither AQP3 nor AQP5 were detectable by multiplex RT-PCR in conceptuses on day 8 of development (Fig. 3C and D; Table 3). AQP3 expression increased thereafter with highest values on day 14 and a subsequent decrease until day 16. Also, AQP5 expression was higher in day 14 embryos (P<0.01 versus days 8, 10 and 12) than in embryos of all other ages (P<0.001; Fig. 3D Table 3). The embryonic size increased continuously with the age of the embryos (P<0.001, Table 3). The diameter of the embryo was significantly correlated with AQP3 expression (r=0.663, P<0.01). No significant correlations existed between diameter of the embryo and either α1-ATPase, β1-ATPase or AQP5 expression.
Analysis of AQP amino acid sequences deduced from the analysed amplicons
The obtained cDNA sequences were translated into protein sequences using the GCG Wisconsin package (Accelrys, San Diego, CA, USA) in order to detect aquaporin-specific amino acid motifs or regulatory sequences. The partial amino acid sequence of equine AQP5 deduced from the pGEM-T cloned 750 bp fragment obtained by RT-PCR using primers specific for human AQP2 (Table 1) is given in Fig. 4A. The second (c-terminal) asparagine–proline–alanine (NPA) aquaporin-specific motif (amino acid residues 185–187) and a cAMP-dependent PKA phosphorylation consensus site were identical to the mouse target sequence (amino acid residues 256–259; Fig. 4A). The amino acid sequence of equine AQP5 deduced from the pGEM-T-cloned 627 bp cDNA amplicon obtained by RT-PCR using human forward and equine reverse primers (Table 1) is given in Fig. 4B (submitted to EBI, accession number AJ555456). During the cloning procedure, the 5′-part of the amplicon was shortened. The amino acid sequence deduced from this part showed aquaporin-specific NPA motifs (aa 69–71 and 185–187), cysteine at position 182, casein kinase II phosphorylation consensus site (aa 148–151) and a putative cAMP-dependent PKA phosphorylation consensus site (aa 152–156). The partial amino acid sequence of equine AQP3 as deduced from 333 bp cDNA amplicon and obtained by RT-PCR using human primers (Table 1) included an aquaporin-specific NPA motif (aa 83–85; Fig. 4C).
Partial amino acid sequences of equine AQP5 and AQP3 with characteristic and putative regulatory sequences: (A) amino acid sequence deduced from the coding part of the ∼750 bp amplicon (second NPA motif position 185–187, cAMP-dependent protein kinase A phosphorylation consensus site identical to the mouse target sequence (aa 256–259). (B) Amino acid sequence deduced from the equine AQP5 627bp amplicon. (C) Amino acid sequence deduced from the equine AQP3 333bp amplicon.
Citation: REPRODUCTION 135, 4; 10.1530/REP-07-0298
Western blotting
Crude protein samples resulted in maximal amounts of Na+/K+-ATPase α1-subunit, whereas the amounts obtained after TriReagent (Sigma–Aldrich) were low (Fig. 5A). In case of AQP3 and AQP5 proteins the recovery after use of TriReagent was better than in case of Na+/K+-ATPase α1-subunit. Microsome preparations provided higher membrane protein amounts than obtained after TriReagent. Embryonic protein samples as well as positive controls resulted in detection of multiple bands, including those of expected molecular masses (∼100 kDa for α1ATPase (Fig. 5A, all lanes); ∼28 kDa for AQP3 (Fig. 5B, lanes E and Ki) and 25–27 kDa for AQP5 (Fig. 5C, lanes E and Lu). Additional bands or band shift in case of AQP5 perhaps caused by glycosylation or degradation (Hendriks et al. 2004) were also detected. With the monoclonal mouse Na+/K+-ATPase α1-subunit antibody (clone C464.6, #05-369) no reproducible results were obtained; therefore, a polyclonal antibody anti-Na+/K+ ATPase α1 (rabbit polyclonal antibody # 06520; Upstate Biotechnology, Lake Placid, NY, USA) was used.
Western blotting of protein samples from embryos and positive controls: (A) α1ATPase: bands at ∼100 kDa for day 12 equine embryos crude protein sample (Ecps), day 12 protein isolation after TriReagent (Eprot), positive control murine kidney medulla (Ki prot); (B) AQP3, bands at ∼30 kDa for day 12 equine embryo crude protein sample (E) and equine kidney microsome preparation (Ki), no bands of the appropriate size in equine skeletal muscle (Mu); (C) AQP5, bands at ∼27 and ∼25 kDa for day 12 equine embryo crude protein sample and equine lung (Lu); no bands of the appropriate sizes in equine skeletal muscle (Mu).
Citation: REPRODUCTION 135, 4; 10.1530/REP-07-0298
Immunohistochemistry
All conceptuses showed positive immunostaining for α1 Na+/K+-ATPase restricted to the basolateral membrane in the trophectoderm (Fig. 6A–D). A positive immunoreaction was also seen in the endodermal layer of the trophoblast in larger (days 12–14) equine conceptuses (Fig. 6B and D). In contrast to the ectoderm not only the basolateral membranes but the whole cytoplasm of the endodermal cells also exhibited intense staining (Fig. 6D). In day 16 equine conceptuses, the trophectoderm exhibited the same pattern of staining as seen in younger conceptuses, but with less intensity. The endoderm, on the other hand, exhibited more membrane abundance and more intense cytoplasmatic staining of immunoreactive protein (Fig. 6E). In the endodermal layer surrounding the subtrophoblastic compartments immunoreactive protein was more abundant in day 16 than in younger conceptuses (Fig. 6E). The embryonic disc never reacted positively, showing an absence of Na+/K+-ATPase α1-subunit in this tissue (Fig. 6C). In the equine kidney, serving as positive control for Na+/K+-ATPase α1-subunit, cells of the proximal tubules exhibited an intense staining of the basolateral membranes (Fig. 6F).
Immunohistochemistry of equine embryos: (day 8 A, C, day 12 B, day 14 D, day 16 E, positive control equine kidney F); Na+/K+ ATPase α1 antibody; (A) basolateral staining is shown in the trophectodermal cells and a diffuse staining in the endodermal cells. The embryonic disc is devoid of any immunostaining (B). YS yolk sac; Ca, capsule; ED, embryonic disc; TEc, trophectoderm; En, endoderm; SC, subtrophoblastic compartment; G, glomerulus; PT, proximal tubule; Scale bars: (A) 100 μm, (B) 50 μm, (C) 50 μm, (D) 20 μm and (E) 20 μm.
Citation: REPRODUCTION 135, 4; 10.1530/REP-07-0298
The cytoplasm of the cells of the yolk sac wall (endoderm and trophectoderm) stained positively for AQP3, exhibiting a granular appearance (Fig. 7). In younger conceptuses (days 8–10) most immunoreactive protein was detected in the apical membrane of trophoblast cells (Fig. 7A and B), whereas in older ones (days 12–14) the basal and lateral membranes also stained more intensely (Fig. 7C and D). In day 16 conceptuses, this tendency increased and the endodermal layer which encloses the subtrophoblastic compartments contained even more immunoreactive proteins in the cell membranes than did the ectoderm (Fig. 7E). The collecting duct cells of equine kidney serving as positive control exhibited an intense basolateral abundance of immunoreactive protein (Fig. 7F).
Immunohistochemistry of equine embryos (capsules were lost during sectioning): day 8 A, day 10 B, day 12 C, day 14 D, day 16 E, positive control (F): equine kidney, AQP3 antibody YS, yolk sac; TEc, trophectoderm; En, endoderm; SC, subtrophoblastic compartment; CD, collecting duct; scale bars: (A) 10 μm, (B) 10 μm, (C) 10 μm, (D) 10 μm, (E) 10 μm.
Citation: REPRODUCTION 135, 4; 10.1530/REP-07-0298
In equine conceptuses, immunohistochemical staining for AQP5 was positive in the apical membrane of conceptuses from day 10 onwards, but negative in younger ones (8 days; Fig. 8A–D). The staining intensity increased with the size of the conceptus. The apical membrane surface enlarged by microvilli of the cells of the trophectoderm appeared to contain high amounts of AQP5 immunoreactive protein. In addition, the cytoplasm of those cells in larger (days 12 and 14) conceptuses showed a weak spotted positive colouring for AQP5 (Fig. 8C and D). AQP5 protein was also detected in the cytoplasm of endodermal cells on days 12 and 14. In day 16 conceptuses, immunostaining of the apical membrane of the trophectoderm remained prominent and the cytoplasm of both trophectodermal and endodermal cells exhibited more intense staining than in younger embryos (Fig. 8E). Equine lung served as a positive control exhibiting an intense staining of the apical membranes of type I pneumocytes (Fig. 8F).
Immunohistochemistry of equine embryos (capsules were lost during sectioning except in B): day 8 A, day 10 B, day 12 C, day 14 D, day 16 E, positive control F: equine lung; the apical membrane of type I pneumocytes stains positive, AQP5 antibody; YS, yolk sac; Ca, capsule; ED, embryonic disc; TEc, trophectoderm; En, endoderm; SC, subtrophoblastic compartment; AL, alveolus; scale bars: (A) 20 μm, (B) 20 μm, (C) 10 μm, (D) 10 μm and (E) 10 μm.
Citation: REPRODUCTION 135, 4; 10.1530/REP-07-0298
Transmission electron microscopy
Conceptuses at days 8, 10, 12, and 14 were investigated for cellular ultrastructure by transmission electron microscopy. Starting on day 10 after ovulation, trophectodermal cells exhibited densely arranged microvilli at the apical membrane (Fig. 9A–C), whereas younger conceptuses (day 8) showed only isolated microvilli. The Golgi apparatus in these cells was composed of enlarged compartments surrounded by large vesicles (Fig. 9C). A subtrophoblastic compartment developed between days 10 and 12 (Fig. 9A). Membrane-bound vesicles close to the apical membrane were frequently observed in these cells (Fig. 9B). In contrast to the trophectoderm, cells of the embryonic disc and endoderm almost completely lacked microvilli at their apical membranes.
Transmission electron micrograph of a 12-day old equine embryo: (A) border of trophectoderm and embryonic disc. Notice the distribution of microvilli at the apical cell membranes (dense at the trophectoderm, only a few at the embryonic disc). Higher magnification shows the membrane bound vesicles at the apical membrane (arrows) (B) and enlarged Golgi compartments in the trophectoderm (C). GA, Golgi apparatus; ED, embryonic disc; TEc, trophectoderm; En, endoderm, SC, subtrophoblast compartment, VS, vesicle; scale bars: (A) 5 μm, (B) 1 μm and (C) 2 μm.
Citation: REPRODUCTION 135, 4; 10.1530/REP-07-0298
Discussion
This study for the first time demonstrates the existence of aquaporins 3 and 5 in pre-attachment equine conceptuses. The development of aquaporin gene expression over time until day 16 of pregnancy suggests important functions of aquaporins during yolk sac expansion. Furthermore, our data confirm the evidence that equine yolk sac expansion is mediated by α1β1 Na+/K+-ATPase (Waelchli et al. 1997). Amplicons specific for both subunits of the isoenzyme were abundant in all conceptuses examined between days 10 and 16 of pregnancy. During completion of the endodermal lining around day 8 neither aquaporin 3 nor aquaporin 5 was detectable. In contrast, AQP3 transcripts were present in all conceptuses examined between days 10 and 16. Amplicons specific for AQP5 transcripts appeared for the first time on day 10, the most pronounced gene expression was found on day 14 corresponding to the phase of the most rapid conceptus expansion. The shift of immunoreactive protein for AQP3 from an apical to a non-polar membrane distribution between days 10 and 12 coincides with the apical membrane abundance of AQP5 and surface enlargement of the apical trophoblast membrane by the formation of microvilli. These microvilli also increase in length. The intensive surface enlargement of the apical trophectoderm membrane concomitant with AQP5 presence indicates an increased ability to transport water through the yolk sac wall. This interpretation is also supported by the active Golgi apparatus and the cytoplasmic vesicles found by transmission electron microscopy which might be associated with AQP5 traffic. After day 14, AQP5 transcripts decreased markedly. This coincides with the fixation of the embryo thus marking the end of the mobile pre-attachment phase of equine gestation (Ginther 1983). The strictly apical ectodermal expression of AQP5 protein in the yolk sac is in agreement with the location of AQP5 in type I alveolar cells and the abundance of AQP2 in the collecting duct cells of the kidney (Kraine et al. 1999, Yang et al. 2003). With regard to protein homology, AQP2 is most closely related to AQP5 within the aquaporin family. A transcriptional analysis of the coding sequence of equine AQP5, besides detecting AQP-specific amino acid motifs, resulted in the discovery of putative Caseine II and cAMP-dependent PKA consensus sequences. From the two cAMP-dependent PKA phosphorylation sites the first one (amino acid residues 152–156) is identical to one of the human target sequences and the second (amino acid residues 256–259) to one of the mouse target sequences. These conserved cAMP-PKA consensus sequences closely resemble with those of vasopressin-regulated AQP2, suggesting that equine yolk sac AQP5 too may be under neurohormonal control (Lee et al 1996).
The occurrence of AQP5, detection of putative target sites for PKA and the abundance of oxytocin and vasopressin within equine conceptuses (Waelchli et al. 2000) suggest a neurohormonal control of yolk sac expansion mediated by cAMP and PKA phosphorylation. The immunohistochemistry of day 8 equine conceptuses clearly showed that localisation of the α1 Na+/K+-ATPase subunit was restricted to the basolateral cell membrane of ectodermal cells of the yolk sac. This is in agreement with the theory of a trans-trophoblast ion gradient created by α1β1 Na+/K+-ATPase isoform leading to blastocyst formation in mammals (Watson 1992, Watson & Barcroft 2001). The endodermal cell layer that develops around day 8 showed an intense but diffuse staining for the α1 Na+/K+-ATPase subunit which affected the whole cell including the cytoplasm. Only large, i.e. day 16, equine conceptuses showed more abundant membrane α1 Na+/K+-ATPase in this layer at the time of enlargement of the subtrophoblastic compartments. The arc-like spaces between the trophectodermal and the endodermal cell layers serve as an additional compartment and show higher osmolarity than the yolk sac (Enders et al. 1993, Crews et al. 2007). Since α1 Na+/K+-ATPase subunit is present on the whole surface, the endodermal cell layer of the yolk sac wall might pump sodium ions out of the yolk sac into the subtrophoblastic compartment causing the hypotonicity of the yolk sac. With this arrangement, the subtrophoblastic compartments may get sodium ions from both cell layers, namely the trophectoderm and the endoderm, explaining their high osmolarity (Crews et al. 2007). AQP3 and AQP5 immunoreactive protein is present in both the cell membranes and the cytoplasm of endodermal cells of the yolk sac. In the endodermal layer, AQP5 membrane abundance is lower than that in the trophectoderm with its kidney collecting duct-like arrangement of AQP3 and AQP5 indicating that the endodermal layer is more resistant to water movement than the trophectoderm. AQP5 as an efficient water conductor is located strictly at the apical membrane of the ectodermal cell layer of the yolk sac which shows an intensive enlargement of its surface by microvilli. Transmission electron microscopy of embryos at different ages showed successive surface enlargement of the apical trophoblast membrane by microvilli, large Golgi apparatus and multiple vesicles sometimes fusing with the apical membrane. A polarisation of the cells of the trophectoderm is necessary to transport ions, solutes and fluids in a directional manner (Simmons 1992). AQP3 is abundant in all membranes of trophectodermal cell and is a less efficient water conductor than AQP5. Therefore, we conclude that in the ectodermal cells of the yolk sac the apical membrane with its enlarged surface and increased abundance of AQP5 is more permeable to water than are the basolateral membranes. This might explain how hypertonicity of the subtrophoblast compartments can be maintained while the yolk sac itself becomes hypotonic. In conclusion, our findings are consistent with indications of an ion gradient between the subtrophectodermal space and the external environment of the conceptus (Crews et al. 2007) being created by the ectodermal cells of the yolk sac. The arrangement of aquaporins 3 and 5 suggests a mechanism analogous to kidney collecting duct with AQP2 being replaced by the closely related AQP5. The control of equine yolk sac expansion might be mediated by changes in the water permeability of the apical ectodermal membrane through differences in AQP5 abundance. This process seems to be under the control of PKA phosphorylation driven by cAMP. Vasopressin in the yolk sac could participate in the regulation of AQP5 function as it serves for AQP2 in kidney collecting ducts. The abundance of α1 Na+/K+ ATPase in the endoderm of the yolk sac might indicate an active ATP-consuming system to create yolk sac hypotonicity and hypertonicity in the subtrophoblastic compartments.
The present investigation does not explain how water from the subtrophoblast compartments can enter into the yolk sac without high expense of energy. In order to understand this process, further studies on transport systems during yolk sac expansion are needed.
Materials and Methods
Animals and embryo collection
Six fertile Haflinger mares (Equus caballus) aged between 4 and 16 years were available for the study. They were kept in a large paddock with access to a shed. They were fed hay and mineral supplements twice daily and water was available ad libitum. The mares were checked for oestrous behaviour and examined daily during oestrus by rectal palpation and transrectal ultrasound scanning of the uterus and ovaries using a 7.5 MHz linear scanner (Aloka SSD-900; Aloka, Wiener Neudorf, Austria). After detection of a 3.5 cm preovulatory follicle and a decrease in uterine oedema, the mares were inseminated with either native or extended (Equi Pro; Minitüb, Germany) non-cooled semen. One insemination dose contained at least 500 million progressively motile spermatozoa. Semen was collected from fertile stallions by artificial vagina using routine procedures (Aurich et al. 1997). The mares were inseminated at intervals of 48 h and checked every 24 h until ovulation was detected. Embryos were flushed between days 8 and 16 after ovulation (ovulation detection=day 0) without sedation of the mares using four times 1 l PBS with Ca and Mg (137.93 mM NaCl, 2.67 mM KCl, 0.901 mM CaCl2, 0.493 mM MgCl2.6H2O, 8.06 mM Na2HPO4.7H2O, 1.47 mM KH2PO4, pH 7.4) without addition of proteins prewarmed to 38 °C. A collection of young embryos (including day 10) was performed with a balloon catheter (V-PEFC-28-100-4S; Cook, Brisbane, Australia, inner diameter 6 mm, outer diameter 9 mm, length 1000 mm) fixed cranial to the uterine cervix. In case of older embryos (days 12–16), a modified equine endotracheal tube (V-PET-26; Global Veterinary Products, Daphne, AL, USA, inner diameter 26 mm, outer diameter 35 mm, length 1000 mm) was used. The tip of the tube was cut close to the front of the balloon and the edges were smoothed. The fluid was recovered directly into sterile glass bottles and subsequently, if necessary, filtered with an embryo filter system (EmCon embryo filter; Immunosystems, Spring Valley, WI, USA). The content of the filter was transferred into a Petri dish and examined under a binocular microscope at 10–20×magnification for the presence of an embryo. Once an embryo was detected it was washed in prewarmed PBS to remove attaching cells and scored for quality and size (McKinnon & Squires 1988). Large embryos (diameter more than ∼0.5 cm) visible to the naked eye were scored and washed directly in the bottles used for collection. If larger conceptuses ruptured during the flushing procedure they were recovered by filtration of the collection fluid through an embryo filter (EmCon, diameter of the pores 75 μm) and subsequently washed in a Petri dish. For RT-PCR, the embryos were transferred into a small amount of PBS in an Eppendorf or Falcon tube and stored at −80 °C till RNA preparation.
For immunohistochemistry, the embryos were placed in formol (4% buffered). Embryos to be analysed by transmission electron microscopy were fixed in 2.5% glutaraldehyde.
Collection of tissue samples for reverse transcriptase PCR (RT-PCR)
Samples from equine lung, kidney cortex and medulla to be used as positive controls were collected from euthanised horses at necropsy. The samples were put into Eppendorf tubes and stored at −80 °C.
Extraction of RNA and protein
After homogenisation of frozen single embryos or tissue samples with a glass homogeniser, total RNA and total protein were extracted by means of TriReagent (T9424 Sigma–Aldrich) according to the manufacturer's protocol. Quantification of total RNA was performed by spectrophotometry and similar amounts were used for RT-PCR. Proteins from the same samples were extracted according to the manufacturer's protocol.
Microsome preparation
For selective enrichment of membrane proteins (positive controls in Western blotting), microsomes were prepared as follows: the tissue was homogenised in 0.1 M sodium phosphate buffer (pH 7.4), centrifuged 9000 g for 20 min at 4 °C. The supernatant was transferred to a new Eppendorf tube and diluted 2.5-fold in 0.1 M sodium phosphate buffer (pH 7.4), 0.25 M sucrose and again centrifuged 9000 g for 20 min at 4 °C.
Crude protein lysates
Small pieces of tissue were homogenised using a 2 ml glass homogeniser (Dounce; VWR, Vienna, Austria) and sonicated subsequently in 2× SDS sample buffer (100 mM Tris–HCl (pH 6.8), 200 mM dithiothreitol (DTT), 4% SDS, 0.1% bromophenol blue, 20% glycerol) using an ultrasonic disintegrator (Sonoplus HD 70, Bandelin, Berlin, Germany) in order to fracture the genomic DNA.
RT-PCR and cloning of RT-PCR amplicons
PCR primers
Primer sequences for PCR were designed from sequences obtained from PubMed nucleotide search (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi) using the Prime program of the GCG Wisconsin package (Accelrys). Preferably, primer pairs resulting in intron spanning products binding in conserved regions of the sequences were selected in order to optimise fitting for equine sequences and allowing distinction between cDNA-derived and genomic products (Table 1).
Reverse transcriptase PCR
Total embryonic or tissue RNA was reversely transcribed using specific primers designed from human, mouse or rat sequences (Table 1). The reaction was performed using the one-step RT-PCR kit (Qiagen) according to the description of the manufacturer, using 30 pmol of each primer and more than 1 pg total RNA. The cycling profile was 30 min at 50 °C (RT), 10 min at 95 °C (polymerase activation); 40 times cycle: 94 °C for 1 min, 52 °C for 1 min, 72 °C for 1 min for PCR amplification; 72 °C for 10 min for final extension and final storage at 4 °C overnight.
Multiplex reverse transcriptase PCR (multiplex RT-PCR)
In order to evaluate the expression of α1, β1ATPase subunits as well as AQP3 and AQP5 relative to the expression of β-actin, multiplex RT-PCR was performed using RNase free DNase I (#EN0521; Fermentas, Vilnius, Lithuania)-digested total RNA obtained after TriReagent (Sigma–Aldrich) extraction from equine embryos flushed on days 8, 10, 12, 14 (each n=4) and 16 (n=3). Primer concentrations were found empirically. The concentrations and primers used are summarised in Table 2. The cycling profile for multiplex RT-PCR was the same as that applied for RT-PCR, but only 25 amplification cycles were done.
Analysis of agarose gels
Evaluation of agarose gels and semiquantitative analysis of the multiplex RT-PCR was performed using Molecular Imager Gel Doc XR System (Bio-Rad Laboratories).
Sequencing of aquaporin RT-PCR products
Products amplified by aquaporin-specific primers were separated by electrophoresis on 1.5% agarose gel containing ethidium bromide using TAE buffer (40 mM Tris base, 20 mM glacial acetic acid, 1 mM EDTA, pH 8.5) and 6× loading dye (Fermentas). After u.v. detection, bands of interest were excised, cleaned by Qiaex II kit (Qiagen) and sequenced using the primers for amplification (IBL, Vienna, Austria). The sequences were compared with those from the database using standard nucleotide blast search (http://www.ncbi.nlm.nih.gov/blast/). Equine sequences not known so far were submitted to EBI databases (www.ebi.ac.uk).
Cloning of RT-PCR fragments
Purified products were ligated to the A/T cloning vector pGEM-T (Promega) overnight at 4 °C and Escherichia coli DHF2α CaCl2 competent cells were transformed by heat shock, rolled 1 h at 37 °C in 2× TY (medium containing tryptone/yeast extract: 16 g tryptone, 10 g yeast extract, 5 g/l NaCl) without ampicillin. Subsequently, the bacterial suspension was plated on 2× TY agar plates for blue white selection (ampicillin, IPTG (isopropyl-β-d-thiogalactopyranoside), X-gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside), both Fermentas). After overnight incubation, white colonies were picked from the plates, put into 3 ml of 2× TY and shaken overnight at 37 °C for miniprep. The next day, miniprep was performed using the Invisorb miniprep kit (Invitek, Berlin, Germany). After control digestion of the plasmids using the restriction enzymes Apa I and Sac I (Fermentas), the plasmid preparations containing inserts with the expected length were sent to the sequencing service (IBL). Identification of the cloned insert was performed using primers M13-47 and M13-48.
Protein analysis
SDS-PAGE of proteins
The discontinuous standard system described by Laemmli (1970) was utilised in a vertical mini gel system (Bio-Rad) with 12% polyacrylamide. For sample preparation, the protein solutions were mixed with the same volume 2× SDS-sample buffer (100 mM Tris–HCl (pH 6.8), 200 mM DTT, 4% SDS, 0.1% bromophenol blue, 20% glycerol) and denatured for 5 min at 95 °C. The conditions for running were: current, 30 mA and voltage, 120 V.
Western blotting
After electrophoresis, the gels were equilibrated for 10 min in transfer buffer (25 mM Tris–Cl, 192 mM glycine, 20% methanol). Blotting was performed with a semi-dry blotting device (Bio-Rad) for 1.5 h at 15 V and 200 mA. Blocking was performed for 30 min in 1× PBS (137 mM NaCl, 2.7 mM KCl, 100 mM Na2HPO4, 2 mM KH2PO4, pH 7.4), 0.05% Tween-20, 2% gelatine powder from fish skin (G7041, Sigma). Then the membrane was washed three times with 1× PBS, 0.05% Tween-20 and incubated with primary antibody overnight at 4 °C. The primary antibodies were diluted as follows: anti-Na+/K+-α1ATPase (rabbit polyclonal antibody #06520, Upstate Biotechnology), 1:1000; 200 μl rabbit anti-AQP3 affinity-purified polyclonal antibody #AB3276 (Chemicon, Temecula, CA, USA), 1:200; goat polyclonal antibody AQP5 (G-19): sc-9890 (Santa Cruz Biotechnology, Santa Cruz, CA, USA), 1:200. After washing three times with 1× PBS, 0.05% Tween-20 the membrane was incubated with HPR-conjugated secondary antibody of the corresponding species for 1.5 h at room temperature. Subsequently, it was washed again three times with 1× PBS, 0.05% Tween-20. Finally, detection by ECL (ECLplus, Amersham) was carried out according to the manufacturer's protocol. Exposure was carried out in an exposure cassette (Kodak, Biome; VWR).
Stripping of Western blot nitrocellulose membranes
Antibodies bound to Western blot nitrocellulose membranes were removed by incubation for 30 min at 50 °C in 50 ml stripping solution (100 mM β-mercaptoethanol, 2% SDS, 62.5 mM Tris–Cl (pH 6.7)). After this treatment, membranes were blocked again and exposed to the primary antibody of choice.
Immunohistochemistry
Equine embryos from days 8, 10, 12, 14 and 16 (n=2 for each age) were immersion-fixed in 4% buffered formalin. Formalin-fixed samples were embedded in Histocomp (Vogel, Giessen, Germany) and cut at 3 μm. Endogenous peroxidase activity was blocked by incubation of the sections in 0.6% H2O2 for 15 min. Incubation in 1.5% normal goat or rabbit serum respectively was used to minimise non-specific antibody binding. Subsequently, slides were incubated with primary antibody (dilution 1:100) overnight at 4 °C. The primary antibodies used were as follows: a polyclonal goat antibody directed against the C-terminus of human AQP5 (G-19: sc-9890; Santa Cruz Biotechnology), a polyclonal rabbit antibody against rat AQP3 residues 275–292 (#AB3276, Chemicon) and a monoclonal mouse Na+/K+-ATPase α1-subunit antibody (clone C464.6, #05-369; Upstate Biotechnology). As secondary antibody for AQP5, a biotinylated anti-goat antibody (Vector Laboratories Inc., Burlingame, CA, USA) was used. Vectastain ABC kit (Vector Laboratories) was utilised to enhance binding signals and for horseradish peroxidase coupling. For AQP3 detection, a poly-horse radish peroxidase anti-rabbit IgG (Powervision; ImmunoVision Technologies, Daly City, CA, USA) was used as secondary antibody. Subsequently, sections were washed and developed in DAB substrate (3′,3-diaminobenzidine, Sigma; 10 mg/50 ml in 0.1 M Tris buffer (pH 7.4) and 0.03% H2O2) for 10 min at room temperature. After washing, the sections were counterstained with haematoxylin, dehydrated and mounted on DPX (Fluka, Buchs, Switzerland).
Transmission electron microscopy
Specimens were fixed in 2.5% glutaraldehyde and postfixed in 1% phosphate-buffered osmium tetraoxide (Plano W. Plannet GmbH, Wetzlar, Germany). Dehydration was performed in a series of graded ethanol solutions. Infiltration with propylene oxide (Merck) was followed by increasing the ratios of epoxy resin (Serva, Heidelberg, Germany) to propylene oxide (1:1, 3:1) and finally pure resin. After an additional change, the resin was polymerised at 60 °C. Ultrathin sections were cut at 70 nm, stained with alkaline lead citrate and methanolic uranyl acetate, and viewed with an electron microscope (EM 900; Zeiss, Oberkochen, Germany).
Statistical analysis
Statistical comparisons were made with the SPSS statistics package (SPSS, Chicago, IL, USA). ANOVA and subsequent Duncan's test were used to compare mRNA expression of α1ATPase, β1ATPase, AQP3 and AQP5 relative to β-actin and embryonic size between embryos of different age. Correlations between α1ATPase, β1ATPase, AQP3, AQP5 and sizes of the embryos were analysed and Pearson's coefficients of correlation were calculated. For all comparisons, P<0.05 was considered significant.
Acknowledgements
The authors are grateful to the Mehl-Muelhens-Stiftung for financial support of the study. The work of Julia Maderner and Gisi Pittner for help with the graphics is greatly acknowledged. The authors declare that there is no conflict of interest that would prejudice the impartiality of this scientific work.
References
Aurich JE, Schönherr U, Hoppe H & Aurich C 1997 Effects of antioxidants on motility and membrane integrity of chilled-stored stallion semen. Theriogenology 48 185–192.
Barcroft LC, Offenberg H, Thomsen P & Watson AJ 2003 Aquaporin proteins in murine trophectoderm mediate transepithelial water movements during cavitation. Developmental Biology 256 342–354.
Betteridge KJ Equine embryology: an inventory of unanswered questions Theriogenology 68 (Supplement 1)2007 S9–S21.
Betts DH, MacPhee DJ, Kidder GM & Watson AJ 1997 Ouabain sensitivity and expression of the Na/K-ATPase alpha - and beta -subunit isoform genes during bovine early development. Molecular Reproduction and Development 46 114–126.
Biggers JD 1972 Mammalian blastocyst and amnion formation Barnes AC, Seeds AEIn The Water Metabolism of the Fetus Springfield, IL:Thomas:3–31.
Budik S, Walter I, Tschulenk W, Helmreich M, Pittner F & Aurich C 2006 Molecular aspects of equine blastocyst formation in relation to ultrastructure. Reproduction in Domestic Animals 41 336.
Crews LJ, Waelchli RO, Huang CX, Canny MJ, McCully ME & Betteridge KJ 2007 Electrolyte distribution and yolk sac morphology in frozen hydrated equine conceptuses during the second week of pregnancy. Reproduction, Fertility and Development 19 804–814.
Dibas AI, Mia AJ & Yorio T 1998 Aquaporins (water channels): role of vasopressin-activated water transport. Proceedings of the Society for Experimental Biology and Medicine 219 183–199.
DiZio SM & Tasca RJ 1977 Sodium-dependent amino acid transport in preimplantation mouse embryos, III. Na+-K+-ATPase-linked mechanism in blastocysts. Developmental Biology 59 198–205.
Edashige K, Sakamoto M & Kasai M 2000 Expression of mRNAs of the aquaporin family in mouse oocytes and embryos. Cryobiology 40 171–175.
Enders AC, Schlafke S, Lantz KC & Liu IKM 1993 Endoderm cells of the equine yolk sac from day 7 until formation of the definitive yolk sac placenta. Equine Veterinary Journal Supplement 15 3–9.
Fushimi K, Sasaki S & Marumo F 1997 Phosphorylation of Serine 256 is required for cAMP-dependent regulatory exocytosis of the aquaporin-2 water channel. Journal of Biological Chemistry 272 14800–14804.
Ginther OJ 1983 Fixation and orientation of the early equine conceptus. Theriogenology 19 613–623.
Ginther OJ 1992 Embryology and placentation. In Reproductive Biology of the Mare, edn 2, pp 345–418. Cross Plains, WI: Equiservices. .
Ginther OJ, Bergfeld DR, Leith GS & Scraba ST 1985 Early embryonic loss in mares: incidence and ultrasonic morphology. Theriogenology 24 73–86.
Hara-Chikuma M & Verkman AS 2005 Aquaporin-3 functions as a glycerol transporter in mammalian skin. Biology of the Cell 97 479–486.
Hendriks G, Koudijs M, van Balkom BW, Oorschot V, Klumperman J, Deen PM & van der Sluijs P 2004 Glycosylation is important for cell surface expression of the water channel aquaporin-2 but is not essential for tetramerization in the endoplasmic reticulum. Journal of Biological Chemistry 279 2975–2983.
Jones DH, Davies TC & Kidder GM 1997 Embryonic expression of the putative γ subunit of the sodium pump is required for acquisition of fluid transport capacity during mouse blastocyst development. Journal of Cell Biology 139 1545–1552.
Kraine CM, Towne JE & Menon AG 1999 Cloning and characterisation of murine AQP 5: evidence of a conserved aquaporin gene cluster. Mammalian Genome 10 498–505.
Laemmli U 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227 680–685.
Lee MD, Bhakta KY, Raina S, Yonescu R, Griffin CA, Copeland NG, Gilbert DJ, Jenkins NA, Preston GM & Agre P 1996 The human Aquaporin-5 gene (molecular characterisation and chromosomal location). Journal of Biological Chemistry 271 8599–8604.
Manejwala FM, Cragoe EJ Jr & Schultz RM 1989 Blastocoel expansion in the preimplantation mouse embryo: Role of extracellular sodium and chloride and possible apical routes of their entry. Developmental Biology 133 210–220.
Matsamura Y, Uchida S, Rai T, Sasaki S & Marmo F 1997 Transcriptional regulation of aquaporin-2 water channel gene by cAMP. Journal of the American Society of Nephrology 8 861–867.
McKinnon AO & Squires EL 1988 Morphologic assessment of the equine embryo. Journal of the American Veterinary Medical Association 192 401–406.
Offenberg H, Barcroft LC, Caveney A, Viuff D, Thomsen PD & Watson AJ 2000 mRNAs encoding aquaporins are present during murine preimplantation development. Molecular Reproduction and Development 57 323–330.
Simmons C 1992 The development of cellular polarity in transport epithelia. Seminars in Perinatology 16 78–89.
Verkman AS 1999 Lessons on renal physiology from transgenic mice lacking aquaporin water channels. Journal of the American Society of Nephrology 10 1126–1135.
Verkman AS, van Hoek AN, Ma T, Frigeri A, Skach WR, Mitra A, Tamaraooii BK & Farinas J 1996 Water transport across mammalian cell membranes. American Journal of Physiology 270 C12–C30.
Waelchli RO & Betteridge KJ 1996 Osmolality of equine blastocyst fluid from day 11 to day 25 of pregnancy. Reproduction, Fertility and Development 8 981–988.
Waelchli RO, Mac Phee DJ, Kidder GM & Betteridge KJ 1997 Evidence for the presence of sodium – and potassium – dependent adenosine triphosphatase alpha1 and beta1 subunit isoforms and their probable role in blastocyst expansion in the preattachment horse conceptus. Biology of Reproduction 57 630–640.
Waelchli RO, Shand NA, Roud HK, Alexander SL & Betteridge KJ 2000 Oxytocin and Vasopressin accumulation in the equine conceptus during the second to fifth weeks of gestation. Theriogenology 53 287.
Watson AJ 1992 The cell biology of blastocyst development. Molecular Reproduction and Development 3 492–504.
Watson AJ & Barcroft LC 2001 Regulation of blastocyst formation. Frontiers in Bioscience 6 708–730.
Watson AJ, Natale DR & Barcroft LC 2004 Molecular regulation of blastocyst formation. Animal Reproduction Science 82–83 583–592.
Yang F, Kawedia JD & Menon AG 2003 Cyclic AMP regulates Aquaporin 5 expression at both transcriptional and post transcriptional levels through a protein kinase A pathway. Journal of Biological Chemistry 278 32173–32180.
