Abstract
Aquaporin 1 is a membrane water channel protein, which was studied here in spiny dogfish (Squalus acanthias) osmoregulatory tissues using a variety of techniques. The cloning of aquaporin 1 (AQP1) in the spiny dogfish identified a splice variant version of the mRNA/protein (AQP1SV1/AQP1SV1). Polymerase chain reaction (PCR) in a range of tissues showed AQP1 to be expressed at very high levels in the rectal gland with ubiquitous mRNA expression at lower levels in other tissues. Northern blotting showed that AQP1 had a mRNA size of 5.3 kb in kidney total RNA. The level of AQP1 mRNA was significantly lower in the rectal glands of fish acclimated to 120% seawater (SW; vs. 75% SW (p = 0.0007) and 100% SW (p = 0.0025)) but was significantly higher in those fish in the kidney (vs. 100% SW (p = 0.0178)) and intestine (vs. 75% SW (p= 0.0355) and 100% SW (p = 0.0285)). Quantitative PCR determined that AQP1SV1 mRNA levels were also significantly lower in the rectal glands of both 120% (p = 0.0134) and 100% SW (p = 0.0343) fish in comparison to 75% SW-acclimated dogfish. Functional expression in Xenopus oocytes showed that AQP1 exhibited significant apparent membrane water permeability (p = 0.000008-0.0158) across a range of pH values, whereas AQP1SV1 showed no similar permeability. Polyclonal antibodies produced against AQP1 (AQP1 and AQP1/2 antibodies) and AQP1SV1 had bands at the expected sizes of 28 kDa and 24 kDa, respectively, as well as some other banding. The weak AQP1 antibody and the stronger AQP1/2 antibody exhibited staining in the apical membranes of rectal gland secretory tubules, particularly towards the periphery of the gland. In the gill, the AQP1/2 antibody in particular showed staining in secondary-lamellar pavement-cell basal membranes, and in blood vessels and connective tissue in the gill arch. In the spiral valve intestine side wall and valve flap, the AQP1/2 antibody stained muscle tissue and blood vessel walls and, after tyramide signal amplification, showed some staining in the apical membranes of epithelial cells at the ends of the luminal surface of epithelial folds. In the rectum/colon, there was also some muscle and blood vessel staining, but the AQP1 and AQP1/2 antibodies both stained a layer of cells at the base of the surface epithelium. In the kidney convoluted bundle zone, all three antibodies stained bundle sheath membranes to variable extents, and the AQP1/2 antibody also showed staining in the straight bundle zone bundle sheath. In the kidney sinus zone, the AQP1/2 antibody stained the apical membranes of late distal tubule (LDT) nephron loop cells most strongly, with the strongest staining in the middle of the LDT loop and in patches towards the start of the LDT loop. There was also a somewhat less strong staining of segments of the first sinus zone nephron loop, particularly in the intermediate I (IS-I) tubule segment. Some tubules appeared to show no or only low levels of staining. The results suggest that AQP1 plays a role in rectal gland fluid secretion, kidney fluid reabsorption and gill pavement-cell volume regulation and probably a minor role in intestinal/rectal/colon fluid absorption.