The aquaporin 8 (AQP8) membrane channel gene is present in the elasmobranch dogfish (Squalus acanthias) genome and is expressed in brain but not in gill, kidney or intestine

The transport mechanisms for water, ammonia and urea in elasmobranch gill, kidney and gastrointestinal tract remain to be fully elucidated. Aquaporin 8 (AQP8) is a known water, ammonia and urea channel that is expressed in the kidney and respiratory and gastrointestinal tracts of mammals and teleost fish. However, at the initiation of this study in late 2019, there was no copy of an elasmobranch aquaporin 8 gene identified in the genebank even for closely related holocephalon species such as elephant fish (Callorhinchus milii) or for the elasmobranch little skate (Leucoraja erinacea). A transcriptomic study in spiny dogfish (Squalus acanthias) also failed to identify a copy. Hence this study has remedied this and identified the AQP8 cDNA sequence using degenerate PCR. Agarose electrophoresis of degenerate PCR reactions from dogfish tissues showed a strong band from brain cDNA and faint bands of a similar size in gill and liver. 5' and 3' RACE was used to complete the AQP8 cDNA sequence. Primers were then designed for further PCR reactions to determine the distribution of AQP8 mRNA expression in dogfish tissues. This showed that AQP8 is only expressed in dogfish brain and AQP8 therefore clearly can play no role in water, ammonia and urea transport in the gill, kidney or gastrointestinal tract. The role of AQP8 in dogfish brain remains to be determined.

Aquaporin (AQP) channels in the spiny dogfish, Squalus acanthias II: Localization of AQP3, AQP4 and AQP15 in the kidney

Three aquaporin water channel proteins, AQP3, AQP4 and AQP15 were localized to cells within the kidney of the spiny dogfish, Squalus acanthias, using an immunohistochemical approach. Dogfish kidney has two zones, the bundle zone (including five nephron segment bundles) and the sinus zone (with two major loops). In order to discriminate between the two loops, the cilia occurring in the first proximal/intermediate loop were labeled with two antibodies including an anti-acetylated tubulin antibody. The second late distal tubule loop (LDT) was identified, as the nephron in that region has no luminal cilia. Strong staining of the rabbit anti-dogfish AQP3, AQP4 (AQP4/2) or AQP15 polyclonal antibodies localized to LDT tubules. These antibodies were further co-stained with a mouse anti-Na⁺,K⁺-ATPase a5 monoclonal antibody, as Na⁺,K⁺-ATPase has previously been suggested to localize to the early distal tubule (EDT) and LDT and a mouse anti-NKCC T4 antibody, as NKCC2 was previously suggested to be located in the EDT and the second half of the LDT. In the LDT, strong AQP4/2 and AQP15 antibody staining localized together with the strong Na⁺,K⁺-ATPase antibody staining, whereas strong AQP3 antibody staining was largely separate but with an overlapping distribution. Very low levels of AQP4/2 antibody basal membrane staining was also detected in the first proximal /intermediate loop of the sinus zone. There was no mouse anti-NKCC T4 antibody staining apparent in the LDT. In the convoluted part of the bundle zone, the AQP4/2 and Na⁺,K⁺-ATPase but not the AQP3 or AQP15 antibodies stained tubule segments, with both AQP4/2 and Na⁺,K⁺-ATPase staining the EDT, and with low-level AQP4/2 staining of two other tubules of the bundle, which were most likely to be the proximal 1a (PIa) and intermediate II (IS II) tubules. The AQP4/2 antibody also stained the EDT in the straight bundle zone. The mouse anti-NKCC T4 antibody stained the apical region of EDT tubules in the convoluted bundle zone, suggesting that the antibody was binding to the NKCC2 cotransporter. The AQP15 antibody appeared to bind to the peritubular sheath surrounding bundles in the bundle zone. Due to the AQP4/2 antibody staining in the EDT that immediately proceeds and continues into the LDT, this suggested that the strong AQP4/2, AQP15 and Na⁺,K⁺-ATPase antibody staining was located at the beginning of the LDT and therefore the strong AQP3 was located at the end of the LDT. The staining of all three AQP antibodies was blocked by the peptide-antigen used to make each one, suggesting that all the staining is specific to each antibody.

Aquaporin (AQP) channels in the spiny dogfish, Squalus acanthias I: Characterization of AQP3 and AQP15 function and expression, and localization of the proteins in gill and spiral valve intestine

Complementary DNAs (cDNAs) for two aquaporin water channel genes (AQP3 and AQP15) were amplified cloned and sequenced to initiate this study. Northern blot analysis was carried out to confirm the mRNA sizes of these AQP genes with AQP3 mRNA bands exhibiting sizes of 1.2 and 1.6 k bases and AQP15 had a mRNA band of 2.1 k bases. Northern blot analysis was also performed on kidney and esophagus total RNA samples from fish acclimated to 75%, 100% or 120% seawater (SW). The level of AQP15 mRNA expression was shown to significantly decrease following salinity acclimation from 100 to 120% SW. An opposite but non-significantly different trend was observed for AQP3 mRNA levels. Full length cDNAs were then used to generate AQP3 and AQP15 mRNAs for microinjection into Xenopus oocytes. Both AQP3- and AQP15- microinjected oocytes exhibited significantly elevated apparent water permeability compared to control oocytes at neutral pH. The apparent water permeability was mercury-inhibitable, significantly so in the case of AQP3. AQP3 microinjected oocytes showed pH sensitivity in their apparent water permeability, showing a lack of permeability at acidic pH values. The Carboxyl-terminal derived amino acid sequences of AQP3 and AQP15 were used to generate rabbit affinity-purified polyclonal antibodies. Western blots with the antibodies showed a band of 31.3 kDa for AQP3 in the kidney, with minor bands at 26, 24 and 21 kDa. For AQP15 a band of 26 kDa was seen in gill and kidney. Fainter bands at 28 and 24 kDa were also seen in the kidney. There was also some higher molecular weight banding. None of the bands were seen when the antibodies were pre- blocked with their peptide antigens. Immunohistochemical localization studies were also performed in the gill and spiral valve intestine. In the gill, AQP15 antibody staining was seen sporadically in the membranes of surface epithelial cells of the secondary lamellae. Tyramide amplification of signals was employed in the spiral valve intestine. Tyramide-amplified AQP3 antibody staining was observed in the basal membrane of the invaginated epithelial cell layer of secondary intestinal folds in luminal surface of either the side wall of the spiral valve intestine or in internal valve tissue 'flaps'. For the AQP15 antibody, tyramide-amplified staining was instead found on the apical and to a lesser extent the lateral membranes of the same invaginated epithelial cell layer. The localization of AQP3 and AQP15 in the spiral valve intestine suggests that a trans-cellular water absorption pathway may exist in this tissue.