The role of the rectum in osmoregulation and the potential effect of renoguanylin on SLC26a6 transport activity in the Gulf toadfish (Opsanus beta)

Abstract

Teleosts living in seawater continually absorb water across the intestine to compensate for branchial water loss to the environment. The present study reveals that the Gulf toadfish (Opsanus beta) rectum plays a comparable role to the posterior intestine in ion and water absorption. However, the posterior intestine appears to rely more on SLC26a6 (a HCO₃⁻/Cl⁻ antiporter) and the rectum appears to rely on NKCC2 (SLC12a1) for the purposes of solute-coupled water absorption. The present study also demonstrates that the rectum responds to renoguanylin (RGN), a member of the guanylin family of peptides that alters the normal osmoregulatory processes of the distal intestine, by inhibited water absorption. RGN decreases rectal water absorption more greatly than in the posterior intestine and leads to net Na⁺ and Cl⁻ secretion, and a reversal of the absorptive short-circuit current (I_SC). It is hypothesized that maintaining a larger fluid volume within the distal segments of intestinal tract facilitates the removal of CaCO₃ precipitates and other solids from the intestine. Indeed, the expression of the components of the Cl⁻-secretory response, apical CFTR, and basolateral NKCC1 (SLC12a2), are upregulated in the rectum of the Gulf toadfish after 96 h in 60 ppt, an exposure that increases CaCO₃ precipitate formation relative to 35 ppt. Moreover, the downstream intracellular effects of RGN appear to directly inhibit ion absorption by NKCC2 and anion exchange by SLC26a6. Overall, the present findings elucidate key electrophysiological differences between the posterior intestine and rectum of Gulf toadfish and the potent regulatory role renoguanylin plays in osmoregulation.

seawater teleost fish live in an environment that is hypertonic to their blood plasma, resulting in continual water loss through the gills (34). To compensate for this, seawater teleosts drink copious volumes of water that are continually desalinated by the esophagus and gastrointestinal tract to produce a fluid that is roughly isotonic to blood plasma (17, 34). The constant absorption of ions, by the intestine, is critical as it lowers luminal osmolality and allows for solute-coupled water absorption, and, thereby, compensating for branchial water loss (45, 46). The decrease in luminal osmolality is achieved by ion transport mechanisms that not only favor salt absorption, but also result in the precipitation of osmolytes in the intestinal lumen that further reduce osmolality. Enterocytes express the apical electrogenic HCO₃⁻/Cl⁻-antiporter (SLC26a6) that secretes multiple molecules of HCO₃⁻ in exchange for a single Cl⁻, although the exact ratio is unknown (20, 31). Secreted HCO₃⁻ binds to luminal Ca²⁺ (and to a lesser extent with Mg²⁺) to form CaCO₃ (and MgCO₃) precipitates, resulting in a 70–100 mosmol/kg H₂O reduction in luminal osmolality (52). In addition, the titration of HCO₃⁻ by H⁺, secreted by apical V-type H⁺-ATPase (VHA), can further decrease luminal osmolality by up to 30 mosmol/kg H₂O (16, 17, 19, 21, 22, 51, 52). Na⁺ is primarily taken up by the apical Na⁺/K⁺/2Cl⁻-cotransporter, NKCC2 (SLC12a1), and the apical Na⁺/Cl⁻ cotransporter, NCC (SLC12a3) (25). Water absorption occurs through aquaporins (53), a process coupled to the ion transport mechanisms described above.

The seawater teleost rectum also plays a role in water absorption. For instance, in the sea bream (Sparus aurata) and Japanese eel (Anguilla japonica), fluid absorption by the rectum also occurs via aquaporins, but the rate of absorption is less than in the intestine (8, 15, 28, 36). The rectum of the Gulf toadfish (Opsanus beta) also expresses intestinal ion transporters and enzymes involved in solute-coupled water absorption, such as SLC26a6, the basolateral electrogenic Na⁺/HCO₃⁻-cotransporter [NBCe1 (SLC4a4)], apical VHA, and cytosolic carbonic anhydrase (CA_C) (20, 22, 42, 49). These cellular components are essential for fluid absorption by the marine teleost intestine, but there is a large gap in knowledge concerning their osmoregulatory role in the rectum.

The hydration mechanisms described above are vital to a seawater teleost's survival. However, recent studies have demonstrated the potential for inhibited fluid absorption by the intestine, initiated by the guanylin family of intestinal peptides [guanylin, uroguanylin, and renoguanylin (RGN)] (3, 4, 40, 41). These short-length hormones are expressed in the intestine and rectum of the Japanese eel, European eel (A. anguilla), and Gulf toadfish (41, 47, 48), and they are upstream regulators of apical guanylyl cyclase-C1 and -C2 (GC-C) receptors in the intestine (56). GC-C activation leads to increases in intracellular cGMP formation, whose downstream effects stimulate Cl⁻ secretion by activation of the apical cystic fibrosis transmembrane conductance regulator (CFTR) and greatly inhibit ion and water absorption (3, 4, 6, 40). In the Gulf toadfish and Japanese eel, the guanylin peptide effect appears restricted to the mid- and posterior intestine, by reversing the absorptive short-circuit current (I_SC) (from mucosa-to-serosa to serosa-to-mucosa), mediated primarily by a switch from net Cl⁻ absorption to net secretion (4, 40, 41, 55). This reversal leads to either fluid secretion or inhibited fluid absorption (3, 4, 40, 41, 55). Moreover, the guanylin peptides, GC-C1, and GC-C2, are transcriptionally upregulated when freshwater-acclimated European and Japanese eels are transferred from freshwater to seawater (SW) (9, 10, 27, 54, 56), which suggests that the guanylin peptide system is important for seawater acclimatization. There is also immunohistochemical and molecular evidence for a putative Cl⁻-secretory pathway comprising an apical CFTR and basolateral NKCC1 (SLC12a2) in the posterior intestine of Gulf toadfish, killifish (Fundulus heteroclitus), and sea bream (15, 35, 40, 44).

It is hypothesized that the guanylin-stimulated Cl⁻ secretory response is essential for removing solids (e.g., CaCO₃ precipitates or undigested food) from the intestine (40, 41). The mechanical activity (e.g., peristalsis) along with presence of fluids in the intestinal lumen facilitates the movement of solids along the intestine (43). The presence of fluid in the intestinal lumen likely facilitates clearance of CaCO₃ precipitates from the intestine. In support of this hypothesis, transcription of CFTR and NKCC1 is upregulated in the posterior intestine of Gulf toadfish (41) and Mozambique tilapia (Oreochromis mossambicus) (33) exposed to hypersalinity. Further support for the above hypothesis is the observation that RGN causes a larger reversal of the absorptive I_SC and inhibits water absorption more greatly in the posterior intestine of Gulf toadfish exposed to hypersalinity (41). Although the above hypothesis remains to be fully validated, an important consideration is the fate of the fluid that is either secreted or not absorbed by the distal intestine of seawater teleosts, due to guanylin peptide stimulation.

The retention of water is vital for the survival of a seawater teleost, yet the inhibition of absorptive mechanisms in the intestine that occurs through guanylin peptide stimulation still exists. Either this water is lost by expulsion into the environment or reabsorbed by the rectum. To our knowledge, the duration of the in vivo guanylin peptide response in seawater teleosts is unknown. It is also unknown whether the duration of the guanylin peptide stimulation is long enough that it prevents the distal intestine and rectum from being able to absorb water secreted by the more proximal segments.

The purpose of this study was twofold. First, we sought to rectify the gap of knowledge with respect to transport physiology of the rectum by characterizing the normal osmoregulatory mechanisms of the Gulf toadfish rectum. Second, we wished to determine whether the rectum responded to RGN and compared this response to that of the posterior intestine. A third objective was to determine the mechanisms that decrease HCO₃⁻ secretion in the intestinal tract by RGN.

MATERIALS AND METHODS

Experimental Animals

Gulf toadfish (Opsanus beta) were caught as bycatch from Biscayne Bay, FL, by shrimp fishermen, and placed in a 3-min freshwater bath and treated with malachite green for ectoparasite removal, upon arrival to the laboratory. Individuals were separated by size and placed into 62-liter aerated tanks (33–36 ppt, 20–26°C), with continuous flow-through of sand-filtered SW from Key Biscayne. Gulf toadfish were fed squid to satiation once per week, but food was withheld for 3 days prior to experimentation. For the hypersalinity (60 ppt) studies, Gulf toadfish were held in experimental tanks in groups of 6–10, with recirculating biofilters containing charcoal. The tanks sat in a temperature-controlled water bath at 26°C. Individuals were acclimated to 35 ppt and 60 ppt (adjusted with Instant Ocean; SpectrumBrands, Madison, WI) for up to 4 days for the real-time, quantitative PCR (qPCR) study and at least 7 days for the physiological studies. Tanks were siphoned regularly and topped up with SW adjusted to the proper salinity. Water chemistry was monitored daily (temperature and salinity) and twice weekly (pH). Fish husbandry and experimental procedures followed an approved University of Miami Animal Care Protocol (Insitutional Animal Care and Use Committee no. 13–225, renewal 02). Gulf toadfish were killed using 0.2 g/l MS-222 (Argent, Redmond, WA) solution buffered with 0.3 g/1 NaHCO₃⁻ (Sigma-Aldrich, St. Louis, MO), followed by severing the spinal cord at the cervical vertebra and pithing of the brain.

Composition of Salines, Hormones, and Inhibitors

Mucosal (pH = 7.8) and serosal (pH = 7.8) salines were prepared as described in Table 1 and were used to bathe intestinal and rectal tissues of Gulf toadfish acclimated to both 35 and 60 ppt. The osmolalities of plasma, posterior intestinal, and rectal fluids from Gulf toadfish at these two salinities are not significantly different (13, 37) and, accordingly, would influence neither cell volume nor osmotic transepithelial water movement. Eel RGN (Peptide Institute, Osaka, Osaka Prefecture, Japan) was used in the present study for two reasons. First, the sequence for eel RGN (ADLCEICAFAACTGCL, accession no.: BAC76010.1) is nearly identical to that of Gulf toadfish (tf) guanylin (MDVCEICAFAACTGC, accession no.: AIA09902.1). Second, because of its close homology with tf guanylin, eel RGN has the greatest physiological effects on Gulf toadfish intestinal tissues (40). Ethoxzolamide (Sigma-Aldrich) was dissolved in DMSO (Sigma-Aldrich) and administered at a concentration of 100 μmol/l in saline containing 0.1% DMSO. Vehicle controls for 0.1% DMSO have been shown not to affect the physiological parameters tested in the present study (40). Bumetanide (Sigma-Aldrich) was dissolved in ethanol (Sigma-Aldrich) and administered at 10 μmol/l in saline containing 0.05% ethanol. Control experiments revealed that at this concentration, ethanol does not affect the physiological parameters tested in the present study (data not shown).

Table 1.

Composition of salines for short-circuit current, pH-stat titration, and sac preparation experiments

Compound	Mucosal*	Serosal†	HCO3−/CO2-free serosal
NaCl, mmol/l	169.0	151.0	151.0
KCl, mmol/l	5.0	3.0	3.0
MgSO4, mmol/l	77.5	0.88	0.88
MgCl2, mmol/l	22.5
Na2HPO4, mmol/l		0.5	0.5
KH2PO4, mmol/l		0.5	0.5
CaCl2, mmol/l	5.0	1.0	1.0
NaHCO3, mmol/l		5.0
HEPES, free acid, mmol/l		11.0	11.0
HEPES, Na+ salt, mmol/l		11.0	11.0
Urea, mmol/l		4.5	4.5
Glucose, mmol/l		5.0	5.0
Osmolality§, mosmol/kg H2O	330	330	330
pH	7.8†	7.8	7.8
Gas‡	100% O2	0.3% CO2 in O2	100% O2

^*Mucosal application of blockers and pharmaceuticals: Renoguanylin was added for a final concentration of 10⁻⁷ and 5 × 10⁻⁷ mol/l in Ussing chamber and sac preparation experiments, respectively. Bumetanide and ethoxzolamide were added to Ussing chambers for final concentrations of 10 and 100 μmol/l, respectively. Saline solutions were adjusted for osmolality with mannitol (§). When measuring HCO₃⁻ secretion, mucosal pH in Ussing chambers was maintained at 7.8 by pH-stat titration (†). Salines were gased for at least 1 h prior to experiments (‡).

RNA Isolation and Molecular Cloning

Tissues were collected from Gulf toadfish kept in 33–35 ppt seawater for the tissue distribution study and from Gulf toadfish acclimated to 0 (control), 6, 12, 24, and 96 h in 60 ppt for the hypersalinity study. Collected tissues were homogenized with an Ultra-Turrax T8 homogenizer (Ika-Werke, Breisgau, Baden-Württemberg, Germany) and 50–100 mg/ml of total RNA was extracted with RNA STAT-60 (Tel-Test, Friendswood, TX). Traces of genomic DNA were removed with DNase I (Turbo DNA-free kit, Ambion, Foster City, CA) from each isolate (10 μg/sample). A spectrophotometer (NanoDrop 1000 Spectrophotometer, Thermo Scientific, Grand Island, NY) was used to quantify total RNA at 260 nm. A subsample of isolates was spot-checked using gel electrophoresis to confirm RNA integrity. Isolated RNA (1 μg) was reversed transcribed into cDNA using SuperScript First-Strand Synthesis System (Invitrogen, Carlsbad, CA), according to the manufacturer's protocol. Genes coding for tf guanylin, tf uroguanylin, tf guanylyl cyclase-C (GC-C), tfNKCC1, and tfCFTR were previously cloned (40, 41) by PCR.

Real-Time, Quantitative PCR

Gene expression levels of tf guanylin, tf uroguanylin, tfGC-C, tfNKCC1, tfNKCC2, and tfCFTR were detected using qPCR. Gulf toadfish tissues were dissected and immediately snap-frozen in liquid nitrogen, after which RNA was isolated and cDNA produced, as described above. Gene expression was measured using gene-specific primers (Table 2) and Power SYBR Green (Applied BioSystems, Foster City, CA) as the reporter dye, for qPCR in an Mx4000 cycler (Stratagene, San Diego, CA). Cycling parameters were as follows: 95°C for 10 min, 40–50 cycles of 95°C for 30 s, 55–60°C for 30 s, and 72°C for 30 s. The specificity for all PCR products was confirmed by observing a distinct, corresponding melting peak for each product following each qPCR run using a melting curve analysis (guanylin: 79–85°C, uroguanylin: 81–86°C, GC-C: 80–86°C, NKCC1: 81–86°C, NKCC2: 77–83°C, and CFTR: 80–85°C), as reported previously (41). Gene expression was normalized to EF-1α and was scaled relative to the lowest expressing tissue (tissue distribution study) and the 0-h control (hypersalinity study), which were given a value of 1.0, as described by Pfaffl (39). Primers for tfNKCC1, tfNKCC2, and tfCFTR were designed from an annotated Gulf toadfish intestinal transcriptome (LeMoine CMR, Corradi N, Walshun PJ, unpublished data).

Table 2.

Primers used for qPCR experiments

Name	Sequence
tfGN-F	5′-AGCAAAGGGCAGCATCTGCA-3′
tfGN-R	5′-TGGCAAGATGTTTGTGGCTTTGC-3′
tfUGN-F	5′-CCGACCCTCTCATGCCGCAGG-3′
tfUGN-R	5′-TGCACGGAGGCATCGAGCTG-3′
tfGC-C-F	5′-CAGAGGCCACCATGCGGCGTCACCTA-3′
tfGC-C-R	5′-TTCACCAAGCGCTGCTCCGACCAA-3′
tfNKCC1-F	5′-TCCTGCAGCAGCTCGTTGAG-3′
tfNKCC1-R	5′-GAGCACGTGGCGGCCTTCGAGGAT-3′
tfNKCC2-F	5′-AAGGCACCATAGACGTGTGG-3′
tfNKCC2-R	5′-CCTCAACCGACAGTCCTTCC-3′
tfCFTR-F	5′-GTTTCATCACCGGCATGAACG-3′
tfCFTR-R	5′-GTGCCTTTCTGTAGATGGCTCCAA-3′
EF-1α-F	5′-AGGTCATCATCCTGAACCAC-3′
EF-1α-R	5′-GTTGTCCTCAAGCTTCTTGC-3′
Universal Primer*	5′-CTAATACGACTCACTATAGGGCAAGCGTGGTATCAACGCAGAGT-3′
	5′-CTAATACGACTCACTATAGGGC-3′

^*Universal primer (Clontech SMARTer RACE cDNA amplification kit) consists of both a long and short sequence. tf, Gulf toadfish; F, forward; R, reverse; GN, guanylin, UGN, uroguanylin; GC-C, (guanylyl cyclase-C), NKCC1 (Na⁺/K⁺/2Cl⁻-cotransporter type 1, SLC12a2), NKCC2 Na⁺/K⁺/2Cl⁻-cotransporter type 2, SLC12a1; CFTR, cystic fibrosis transmembrane conductance regulator; EF-1α, elongation factor 1-α.

General Experimental Protocol: Short-Circuit Current and HCO₃⁻ Secretion

Short-circuit current (I_SC), transepithelial potential (TEP), and transepithelial conductance (G_TE) of Gulf toadfish tissues were measured by Ussing chambers (model 2400; Physiologic Instruments, San Diego, CA). These experiments were paired with the posterior intestine and the rectum of an individual Gulf toadfish (15–30 g) measured in parallel. The posterior intestine and rectum were excised, cut open along the midline, and mounted onto P2413 tissue holders (Physiologic Instruments), which exposed 0.71 cm² of tissue and placed between the two half-chambers of the Ussing apparatus. I_SC and TEP were measured by silver (Ag) current and Ag/AgCl voltage electrodes, respectively, and were connected to an amplifier (model VCC600, Physiologic Instruments).

I_SC studies were performed under symmetrical conditions, in which 2 ml of serosal saline (Table 1) was added to both the mucosal (apical membrane/luminal side) and serosal (basolateral membrane/blood side) half-chambers. The salines were continually mixed by airlift gassing (0.3% CO₂ in O₂) and maintained at 25°C by a recirculating bath (model no. 1160S, VWR, Radnor, PA), as described in Grosell and Genz (18). I_SC was measured by current electrodes under voltage-clamp conditions (0.0 mV), with 3 s of 2-mV pulses (mucosal-to-serosal) every 60 s. TEP studies were performed under asymmetrical conditions, in which 2 ml each of mucosal and serosal saline (Table 1) were added to the mucosal and serosal half-chamber, respectively. The mucosal saline was continually mixed by airlift gassing (100% O₂), and the serosal saline, as described above, was maintained at 25°C. TEP was measured by voltage electrodes under current-clamp conditions (0.0 μA), with 3 s of 30-μA pulses (mucosal-to-serosal) every 60 s. G_TE was calculated by Ohm's law and was determined from the deflections in I_SC and TEP during pulsing. Acqknowledge software (v. 3.8.1, BIOPAC Systems, Goleta, CA) recorded I_SC and TEP onto a computer.

A pH-stat titrator (model TIM 854 or 856; Radiometer, Brea, CA) was set up in tandem with an Ussing chamber to measure HCO₃⁻ secretion along with TEP (described above), respectively, on the isolated tissues, following the protocol in Grosell and Genz (18). A pH electrode (Radiometer, model no. PHC4000.8) and a microburette tip (from which acid is delivered) were immersed in the mucosal saline bath and maintained a pH of 7.8, allowing for symmetrical pH conditions on either side of the epithelium. Titramaster software (Radiometer, v. 5.1.0), installed on a computer, recorded pH and the volume of acid titrant (0.005 mol/l HCl) injected into the mucosal bath. The HCO₃⁻ secretion rate from a tissue preparation was calculated from the rate of titrant secreted and its concentration, as described in Grosell and Genz (18).

Experimental set 1.

To determine baseline I_SC and HCO₃⁻ secretion rates, as well as the effects of RGN, on posterior intestinal and rectal tissues of Gulf toadfish acclimated to 35 and 60 ppt. RGN (10⁻⁷ mol/l) was added to mucosal half-chamber once baseline values stabilized.

Experimental set 2.

To determine the source of decreased HCO₃⁻ secretion rates by RGN, bicarbonate was removed from the serosal saline, leaving CA_C as the sole source of intracellular HCO₃⁻. As before, RGN was added to the mucosal half-chamber once the HCO₃⁻ secretion rates of the posterior intestine and rectum stabilized (of fish acclimated to 35 ppt). A subsequent experiment tested the potential influence of RGN on CA_C with ethoxzolamide (10⁻⁴ mol/l) [a known inhibitor of CA_C (18)]. This experimental set served to elucidate the role of RGN on the transport rates of HCO₃⁻ by apical SLC26a6 and basolateral NBCe1, and on HCO₃⁻ formation by CA_C.

Experimental set 3.

To determine the effects of mucosally applied bumetanide on I_SC and HCO₃⁻ secretion. In the Japanese eel and Gulf toadfish, Na⁺ absorption is inhibited by guanylin peptide stimulation, suggesting altered NKCC2 activity (3, 4, 41, 55). To test for this in the Gulf toadfish, bumetanide (10 μmol/l), which specifically binds to and inhibits NKCC isoforms, was added to the mucosal half-chamber, and either I_SC or HCO₃⁻ and TEP were measured.

Intestinal and Rectal Sac Preparations

Sac preparations were used to determine the effects of RGN on water, Cl⁻, and Na⁺ fluxes. The posterior intestine (1.5–3 cm) and rectum (1.5–2 cm) from Gulf toadfish (50–80 g) were excised as described above. The flared end of a PE-50 catheter was inserted into either the posterior intestine or rectum and tied off with a silk suture. Mucosal saline (Table 1) was then injected into the catheter to rinse the tissue of any debris. The tissue was then tied off at the open end with a silk suture to form either an intestinal or rectal sac. The sac preparations were then injected with mucosal saline (control or containing 5 × 10⁻⁷ mol/l RGN). A subsample of injected mucosal saline was collected from which initial Na⁺ and Cl⁻ concentrations were measured. The catheter was sealed, the sac preparations were blot-dried, and initial masses were weighed. The sac preparations were then placed in a scintillation vial containing serosal saline and continually gassed with 0.3% CO₂ (Table 1) and underwent a 2-h flux period. At the end of the flux period, the sac preparations were removed from the serosal saline, blot-dried once again, and final masses were weighed. The sealed catheter was opened, and the fluid was collected. The silk sutures of an empty sac preparation were cut, the tissue was cut down its midline, thoroughly dried, weighed, and its surface area was outlined onto tracing paper. Water, Na⁺, and Cl⁻ flux rates were calculated by the differences in the product between final and initial volumes and concentrations (volume × concentration), divided by the flux period and surface area. Na⁺ and Cl⁻ concentrations were measured using flame spectrometry (Varian 220FS, Palo Alto, CA) and anion chromatography (DIONEX 120, Sunnyvale, CA), respectively.

Statistical Analyses

The data are presented as means ± SE and are evaluated using parametric and nonparametric statistical tests, as described in the figure legends. The 50% effective concentration (EC₅₀) was calculated with a nonlinear regression using a global curve fitting. All data were analyzed and plotted onto graphs with SigmaPlot 13.0. Means were considered significantly different with P ≤ 0.05. Control measurements for I_SC, TEP, and HCO₃⁻ secretion were calculated from the final 30 min of the control fluxes, and treatment values were calculated from the final 30 min of a 70-min treatment exposure. One-tailed tests were used to analyze the qPCR results; a previous study in the Gulf toadfish (41) revealed that tfGC-C, tfCFTR, and tfNKCC1 levels of expression increased in 60 ppt. One-tailed tests were also used to analyze changes in Na⁺, Cl⁻, and fluid absorption, which have been shown to be inhibited after RGN stimulation in the Gulf toadfish (40, 41).

RESULTS

Tissue Density

Tissue masses and surface areas collected for the sac preparations revealed that the rectum of the Gulf toadfish is significantly denser than the posterior intestine (16.16 ± 1.29 vs. 20.59 ± 2.36 mg/cm²). There were no differences in tissue densities of fish acclimated to 35 and 60 ppt. Significant differences were revealed by Mann-Whitney rank sum tests.

Dose Response of the Rectum to Renoguanylin

Stimulation of rectal tissues with increasing RGN concentrations (10⁻⁹→ 10⁻⁶ mol/l) resulted in pronounced decreases in the absorptive I_SC in a sigmoidal, dose-dependent manner. On average, the absorptive I_SC was reversed (mucosa-to-serosa to serosa-to-mucosa) at a concentration of 2.5 × 10⁻⁸ mol/l RGN, with an EC₅₀ of 6.12 × 10⁻⁹ mol/l RGN (Fig. 1A). G_TE also increased significantly at RGN concentrations ≥5.0 × 10⁻⁸ mol/l (Fig. 1B)

Fig. 1.

Changes in short-circuit current (I_SC) (A) and transepithelial conductance (G_TE) (B) in rectal tissues as a function of renoguanylin (RGN) dose after mucosal application (log scale). The concentration of RGN was increased every 30 min when I_SC values became stable. Values are expressed as means ± SE (n = 6). Significant differences from the control (no dose) were revealed by one-way, repeated-measures ANOVAs, followed by Holm-Sidak tests (*P ≤ 0.05). The 50% effective concentration (EC₅₀) of RGN was calculated using a nonlinear regression, global curve fitting (n = 6; P ≤ 0.05). Positive and negative I_SC values indicate secretory and absorptive currents, respectively.

Quantitative PCR Analysis

Gene expression levels were normalized to elongation factor (EF)-1α expression. NKCC1, NKCC2, and CFTR gene expression was present in all intestinal segments and the rectum (Fig. 2). Increased levels of rectal NKCC1 and CFTR gene expression occurred after a 96-h exposure to 60 ppt; conversely, a decreased level of rectal uroguanylin gene expression occurred after 96 h in 60 ppt (Fig. 3).

Fig. 2.

Relative gene expression levels of NKCC1 (A), NKCC2 (B), and CFTR (C) in various tissues of the Gulf toadfish. Gene expression is normalized to elongation factor (EF)-1α and scaled relative to the tissue with the lowest expression value (NKCC1: anterior intestine, NKCC2: gills, and CFTR: muscle), which was given a value of 1.0. NKCC1 gene expression was not detected in the esophagus and liver. Values are expressed as means ± SE (n = 8).

Fig. 3.

Relative gene expression levels of guanylin (GN), uroguanylin (UGN), guanylyl cyclase-C (GC-C), NKCC1, CFTR, and NKCC2 in the rectum of Gulf toadfish acclimated to 0 (control) and 96 h in 60 ppt seawater. The level of expression for each gene is normalized to elongation factor (EF)-1α and scaled relative to the 0-h control, which was given a value of 1.0. Values are means ± SE (n = 8). Significant differences between the 0 and 96-h groups were revealed by one-tailed Student's t-tests (*P ≤ 0.05).

Baseline Parameters: Posterior Intestine Versus Rectum

The posterior intestine and rectum of the Gulf toadfish both displayed an absorptive I_SC, which was significantly greater in the rectum (Fig. 4). The basal TEP of both the posterior intestine and rectum was negative, and significantly more so in the posterior intestine (Fig. 4).

Fig. 4.

Baseline (control) values of the short-circuit current (I_SC), transepithelial potential (TEP), and transepithelial conductance (G_TE) displayed by the posterior intestine and rectum of Gulf toadfish. Values are expressed as means ± SE and were pooled from various experiments in the present study (I_SC: n = 11 and 13, TEP: n = 13 and 13, and G_TE: n = 25 and 27, posterior intestine and rectum, respectively). Significant differences were revealed by Student's t-tests (*P ≤ 0.05). Positive and negative I_SC values indicate secretory and absorptive currents, respectively. Positive and negative TEP values indicate net anion secretion and absorption, respectively.

Effects of Renoguanylin on Bicarbonate Secretion, Transepithelial Potential, and Transepithelial Conductance

Tissues bathed in a serosal saline containing 5 mmol/l HCO₃⁻.

In the posterior intestine, RGN significantly reduced HCO₃⁻ secretion, reversed the TEP (on average), and significantly increased the G_TE (Fig. 5, A–C). Conversely, RGN had no effect on HCO₃⁻ secretion in the rectum, but did significantly reduce the magnitude of the TEP (Fig. 5B). The control TEP of the posterior intestine was significantly more negative than that of the rectum (Fig. 5B).

Fig. 5.

HCO₃⁻ (Bic) secretion (A and D), transepithelial potential (TEP) (B, E), and transepithelial conductance (G_TE) (C and F) displayed by the posterior intestine and rectum of Gulf toadfish before and after treatment with renoguanylin (RGN). Pretreatment values were taken from the final 30 min of the control flux, and post-treatment values were taken from the final 30 min of a 70-min treatment flux. Values are expressed as means ± SE (n = 7 and 8, 5 and 0 mM Bic groups, respectively). Significant differences: control vs. RGN revealed by paired t-tests (*P ≤ 0.05), posterior intestine vs. rectum revealed by Student's t-tests (†P ≤ 0.05), and 5 mM Bic vs. 0 mM (within tissue and within treatment) revealed by Student's t-tests (§P ≤ 0.05). The presence (5 mM Bic) and absence (0 mM Bic) of serosal HCO₃⁻ is indicated by solid and open bars, respectively. Mucosally administered RGN treatment is indicated by hatched bars.

Tissues bathed in a serosal saline devoid of serosal HCO₃⁻.

In both the posterior intestine and rectum, RGN significantly reduced both HCO₃⁻ secretion and the TEP (Fig. 5, D and E). All control and RGN-treated tissues bathed in the serosal saline devoid of HCO₃⁻ displayed significantly lower HCO₃⁻ secretion rates than their counterparts bathed with 5 mmol/l HCO₃⁻ (Fig. 5, A and D). The TEP of the RGN-treated rectum was significantly more negative than that of the RGN-treated posterior intestine (Fig. 5E).

Effects of Renoguanylin and Bumetanide on Short-Circuit Current and Transepithelial Conductance

On average, RGN reversed the absorptive I_SC (from mucosa-to-serosa to serosa-to-mucosa) and significantly increased the G_TE of both the posterior intestine and rectum of the Gulf toadfish (Fig. 6, A and B); these results are in line with previous reports (40, 41). Bumetanide had similar effects as RGN; in both tissues, it significantly increased G_TE and reduced the magnitude of the absorptive I_SC of the posterior intestine and, on average, reversed it in the rectum (Fig. 6, C and D).

Fig. 6.

Short-circuit current (I_SC) (A, C) and transepithelial conductance (G_TE) (B, D) displayed by the posterior intestine and rectum of Gulf toadfish before and after treatment with renoguanylin (RGN) or bumetanide (Bum). Pretreatment values were taken from the final 30 min of the control flux and post-treatment values were taken from the final 30-min of a 70-min treatment flux. Values are expressed as means ± SE (n = 6 and 6 or 7, RGN and Bum experiments, respectively). Significant differences: control vs. treatment revealed by paired t-tests (*P ≤ 0.05) and posterior intestine vs. rectum revealed by Student's t-tests (†P ≤ 0.05). Control and treatment values are indicated by solid and open bars, respectively. Positive and negative I_SC values indicate secretory and absorptive currents, respectively.

Effects of Bumetanide and Renoguanylin on Bicarbonate Secretion, Transepithelial Potential, and Transepithelial Conductance

In the posterior intestine, HCO₃⁻ secretion was significantly decreased in the presence of combined bumetanide and RGN treatment, but not with bumetanide alone (Fig. 7A). Bumetanide reversed the TEP (on average), and this effect was not enhanced by RGN (Fig. 7B). G_TE significantly increased in the presence of combined bumetanide and RGN treatment, but not with bumetanide alone (Fig. 7C). In the rectum, only the TEP was affected by bumetanide, with a significant decrease, and this effect was not enhanced by RGN (Fig. 7B).

Fig. 7.

HCO₃⁻ secretion (A), transepithelial potential (TEP) (B), and transepithelial conductance (G_TE) (C) displayed by the posterior intestine and rectum of Gulf toadfish. Control values (solid bars) were taken from the final 30 min of the control flux. Bumetanide (Bum) (open bars) was then added, and its effects were measured from the final 30 min of a 70-min treatment flux. Finally, renoguanylin (RGN) (shaded bars) was added in combination with Bum, and its effects were measured from the final 30 min of a 70-min treatment flux (t_total = 170 min). Values are expressed as means ± SE. Significant differences were revealed by one-way, repeated-measures ANOVAs, followed by Holm-Sidak tests (^a,bP ≤ 0.05). Positive and negative TEP values indicate net anion secretion and absorption, respectively.

Effects of Ethoxzolamide on Bicarbonate Secretion in the Posterior Intestine and Rectum, When Bicarbonate/CO₂ Is Absent from the Serosal Saline

Under bicarbonate-free conditions, ethoxzolamide decreased bicarbonate secretion in both the posterior intestine and rectum (Fig. 8, A and D). As expected, RGN decreased bicarbonate secretion in both the posterior intestine and rectum, and further by the administration of ethoxzolamide into the mucosal saline (Fig. 8, B and E). The effects of RGN and ethoxzolamide were equal when applied individually and additive when applied together (Fig. 8, C and F).

Fig. 8.

Absolute (A, B, D, E) and relative (C, F) HCO₃⁻ secretion rates displayed by the posterior intestine and rectum of Gulf toadfish when exposed in the absence (0 mmol/l) of serosal HCO₃⁻ (Bic). A and D: ethoxzolamide (Ethox) was added alone to the mucosal saline (n = 7 and 8, posterior intestine and rectum, respectively; t_total = 100 min). Values are expressed as means ± SE. Significant differences were revealed by paired t-tests (*P ≤ 0.05). B and E: RGN was initially added to the mucosal saline, after which Ethox was also added to the mucosal saline (n = 6 and 7, posterior intestine and rectum, respectively; t_total = 170 min). Significant differences were revealed by one-way, repeated-measures ANOVAs, followed by Holm-Sidak tests (^a,b,cP ≤ 0.05). C and F: combined effects of RGN and Ethox were compared with their individual treatments. Significant differences: RGN vs. RGN + Ethox revealed by paired t-tests and Ethox vs. RGN + Ethox revealed by Student's t-tests (*P ≤ 0.05). Control values were taken from the final 30 min of the control flux. Treatment values were taken from the final 30 min of a 70-min treatment flux.

Effects of Renoguanylin on Water, Cl⁻, and Na⁺ Fluxes in Posterior Intestinal and Rectal Sac Preparations of Gulf Toadfish Exposed to 35 and 60 ppt

In the posterior intestine, RGN treatment led to net water secretion of fish exposed to 35 ppt, and decreased water absorption in fish exposed to 60 ppt (Fig. 9A). RGN caused net Na⁺ secretion in posterior sac preparations of fish exposed to both 35 and 60 ppt (Fig. 9B), and led to net Cl⁻ secretion in posterior sac preparations of fish exposed to 35 ppt, but not from the 60 ppt group (Fig. 9C). In the rectum, RGN led to net water secretion in rectal sac preparations of fish exposed to 35 ppt and decreased water absorption in fish exposed to 60 ppt (Fig. 9A). Moreover, the response to RGN of the rectum was greater than in the posterior intestine of fish exposed to 60 ppt (Fig. 9A). RGN led to net Na⁺ and Cl⁻ secretion in the rectal sac preparations of fish exposed to both 35 and 60 ppt (Fig. 9, B and C).

Fig. 9.

Water (A), Na⁺ (B), and Cl⁻ (C) fluxes displayed by the posterior intestine (PI) and rectum (Rec) of Gulf toadfish acclimated to 35 or 60 ppt and treated with RGN. Values are expressed as means ± SE. Three-way ANOVAs, followed by one-tailed Holm-Sidak tests revealed significant differences in Na⁺, Cl⁻, and water fluxes (P ≤ 0.05; n_total = 55). Control vs. RGN (*), PI: 35 vs. 60 ppt (†), and 35 or 60 ppt: PI vs. Rec (§). Positive and negative values indicate net absorption or net secretion, respectively.

DISCUSSION

The present study reveals that the absorptive I_SC of the rectum is greater than that of the posterior intestine, yet, the TEP of the posterior intestine is more negative than that of the rectum. Furthermore, it is demonstrated that RGN affects the rectum similarly to the posterior intestine. However, in the rectum, RGN inhibits water absorption more greatly than in the posterior intestine and, in contrast to the posterior intestine, does not reduce HCO₃⁻ secretion when bicarbonate is present in the serosal saline. The present study also shows that the downstream effects of RGN likely act on SLC26a6 and lead to the observed reduction of the HCO₃⁻ secretion from endogenous CO₂, in addition to stimulating the Cl⁻-secretory pathway, made up of apical CFTR, and basolateral NKCC1 (Fig. 10).

Fig. 10.

Simplified proposed effects of the guanylin peptides in the posterior intestinal and rectal epithelia of Gulf toadfish (Opsanus beta) acclimated to 35 ppt seawater. Guanylin (GN), uroguanylin (UGN), and renoguanylin (RGN) bind to a guanylyl cyclase-C (GC-C) receptor on the apical membrane of an enterocyte (A). The stimulation of GC-C leads to enhanced formation of cGMP, whose downstream effects increase the phosphorylation of PKA and PKG, both of which can activate CFTR and, perhaps, NKCC1 (B). The downstream effects of cGMP also lead to inhibited ion transport activity by NKCC2 and decreased anion exchange activity by SLC26a6 (C). The combined effects of GC-C stimulation results in the reversal of ion flux, from net ion absorption (mucosa-to-serosa) to net ion secretion (serosa-to-mucosa), which result in either inhibition of water absorption or net water secretion. NKA, Na⁺/K⁺-ATPase (maintains electrochemical gradients that enables ion transport); TJ, tight junction, stimulatory (+) and inhibitory (−) effects. Cell diagram modified from Ruhr et al. (41).

Comparative Physiology of the Posterior Intestine and Rectum

The present study reinforces the current knowledge on the tissue distribution patterns of NKCC1, NKCC2, and CFTR gene expression in seawater teleosts. The high relative mRNA expression of CFTR and NKCC1 in gill tissues is not surprising. In the seawater teleost gill, CFTR and NKCC1 are located on the apical and basolateral membranes, respectively, and constitute a Cl⁻-secretory pathway that excretes excess Cl⁻ for the purposes of maintaining homeostasis of plasma osmolality (for a review, see Ref. 34). Conversely, NKCC1 transcripts were not detected in the esophagus of the Gulf toadfish, a result that is unsurprising given the importance of esophageal desalination, rather than salt secretion, for the purposes of decreasing the osmolality of imbibed seawater (from ∼1,000 to ∼500 mosmol/kg H₂O) (for a review, see Ref. 17). It is interesting to note relatively high CFTR gene expression in the esophagus, which could be expressed in the apical membrane as an absorptive isoform or on the basolateral membrane as a secretory transporter. In any case, the Gulf toadfish likely takes advantage of the large inward electrochemical gradient for Cl⁻ that permits its passive absorption from the esophageal lumen into the interstitial fluid. The high mRNA expression of CFTR and NKCC1 in the stomach is also unsurprising given the role of HCl plays as a component of gastric juices used to break down food. All of the intestinal segments and the rectum displayed relatively lower CFTR and NKCC1 gene expression, although they appear to be expressed at the protein level, as evidenced by their immunofluorescent-like reactivity [CFTR in the posterior intestine and NKCC1 in the anterior and posterior intestine (40)]. The anterior intestine is the site of the greatest ion absorption in the seawater teleost intestine, while the middle and posterior intestine are the sites of high water absorption (17, 34). Accordingly, it is not surprising that the entire intestine displays rather low relative mRNA expression of CFTR and NKCC1, since its primary osmoregulatory role is ion and fluid absorption for the purposes of maintaining whole body hydration. Indeed, the seawater teleost intestine expresses absorptive-type ion transporters important for fluid absorption (17), including NKCC2 shown in the present and previous studies (12, 23, 25, 38). Moreover, in the Gulf toadfish rectum, absorptive-type ion transporters and proteins necessary for ion absorption, including NKCC2, as well as CFTR and NKCC1, are also expressed at the mRNA level. These consist of SLC26a6 (20), VHA (22), NBCe1 (49), and CA_C (42). Consequently, the Gulf toadfish rectum has absorbing and secretory ion transporters that enable it to absorb and secrete water, respectively, much like in the posterior intestine.

Indeed, the present study shows that the posterior intestine and rectum of Gulf toadfish acclimated to 35 ppt absorbed Na⁺ and Cl⁻ at roughly the same rates. Interestingly, the rectum displayed a greater absorptive I_SC (under symmetrical conditions) and absorbed more water (under asymmetrical, in vivo-like conditions) than the posterior intestine, while the magnitude of the TEP of the posterior intestine was roughly twice the magnitude of the rectum. These data suggest that the posterior intestine relies more on an HCO₃⁻/Cl⁻ exchange by SLC26a6 and the rectum on Na⁺/K⁺/2Cl⁻-cotransport by NKCC2. This is due to the electrogenic nature of SLC26a6 (20), which exchanges multiple bicarbonate anions for a single Cl⁻, as opposed to the electroneutral transport of NKCC2. The differences in the electrophysiological and osmoregulatory properties between the posterior intestine and rectum are likely not due to transcriptional differences of proteins involved in ion absorption [i.e., SLC26a6, VHA, NBCe1, CA_C, and NKCC2 (present study)], which are expressed equally (20, 22, 42, 49).

An objective of the present study was to determine whether the rectum would respond to RGN stimulation. The present study confirms the presence of the components of the Cl⁻-secretory pathway, CFTR and NKCC1, by qPCR, which seem necessary for responding to RGN, at least for the parameters measured in previous studies (3, 4, 40, 41). In prior studies on the Gulf toadfish and Japanese eel, inhibited water absorption and, in some instances, water secretion have been observed in the middle and posterior intestine, but not in the anterior intestine, after the mucosal administration of guanylin, uroguanylin, or RGN (3, 4, 55). It should first be noted that the rectum seems to be more sensitive to RGN than the posterior intestine, with the EC₅₀ of the rectum (6.12 × 10⁻⁹ mol/l RGN) being significantly lower than that of the posterior intestine [11.6 × 10⁻⁹ mol/l RGN (40)]. Moreover, RGN causes a significant decrease and a reversal of the absorptive I_SC in the rectum at 10⁻⁸ and 2.5 × 10⁻⁸ mol/l (present study), respectively, and at 2.5 × 10⁻⁸ mol/l (for both measurements) in the posterior intestine (40). These differences might be related to a combination of the greater tissue density of the rectum (which could express greater GC-C on the apical membrane) and a greater affinity for RGN by rectal GC-C than in the posterior intestine.

The present study shows that the posterior intestine and rectum of Gulf toadfish acclimated to 35 ppt respond to RGN by inhibiting water absorption and displaying net secretion of Cl⁻ and Na⁺. However, their electrophysiological responses to RGN are distinct: RGN decreases the absorptive I_SC of the rectum more greatly than that of the posterior intestine; conversely, RGN decreases the TEP of the posterior intestine more greatly than that of the rectum. These differences might be due to the proposed reliance on SLC26a6 by the posterior intestine and NKCC2 by the rectum for ion absorption (see above). The larger decrease in the TEP of the posterior intestine is likely the result of SLC26a6 being inhibited more greatly than in the rectum, in contrast to the more tissue-dense rectum, where the effects of RGN might be more pronounced on the overall ion-absorptive capacity of the tissue. In both tissues, however, G_TE was significantly increased, despite inhibition of ion absorption. It has been suggested that anion secretion by the opening of CFTR (40, 41) and, possibly, increased paracellular cation secretion (3) might be responsible for the increases in conductance when intestinal tissues are treated with the guanylin peptides. Inhibition of water absorption partially occurs as a result of Cl⁻ secretion, likely by CFTR activation, which is fueled, in part, by NKCC1, which takes up Cl⁻ from the blood (3, 4, 40, 41, 55), and is necessary to inhibit water absorption. Although counterintuitive, decreased water absorption is a strategy that would benefit the removal of CaCO₃ precipitates that naturally form along the intestine (as part of the normal osmoregulatory processes). It has been demonstrated that the retention of water in the guinea pig ileum can make digested food more aqueous and allow it to be moved out of the intestinal tract more easily by muscle contractions (43). A similar mechanism would allow the removal of CaCO₃ precipitates from the intestine and reduce the occurrence of intestinal blockages. Consequently, the trade-off between decreasing water absorption in the posterior intestine, and the risk of dehydration associated with this behavior, is made by decreasing the risk of harmful obstructions in the narrow intestinal passage. Nevertheless, the osmoregulatory mechanisms necessary for water absorption are maintained in the anterior intestine of the seawater teleost, as it does not seem to be affected by the guanylin peptides (40, 55).

Effects of Renoguanylin on NKCC2 and SLC26a6 Transport Activity

Another objective of the present study was to determine whether the downstream effects of RGN affected the transport activity of NKCC2 and SLC26a6. To test the effects of RGN on NKCC2, bumetanide (a specific blocker of NKCC isoforms) was used to determine whether its effects on electrophysiological characteristics were similar to that of RGN. Indeed, the reductions to the absorptive I_SC and TEP of the posterior intestine and rectum by bumetanide were similar to those of RGN. Of particular interest, however, is the failure of bumetanide to change tissue conductance in the posterior intestine and rectum under in vivo-like conditions (asymmetrical saline exposures), as opposed to RGN. This suggests that RGN does not inhibit NKCC2 as greatly as bumetanide, and the reductions to the absorptive I_SC and TEP, as well as the increases in G_TE, are the sum effects of NKCC2 inhibition, apical CFTR activation, and possible paracellular Na⁺ secretion, which leads to net Na⁺ and Cl⁻ secretion observed in tissues of fish acclimated to 35 ppt. Evidence from the Japanese eel supports this suggestion, because the magnitude of the decreases to the absorptive I_SC and TEP of the middle and posterior intestine were similar for both guanylin and bumetanide (4).

Bumetanide also had no effect on HCO₃⁻ secretion in the posterior intestine and rectum of fish acclimated to 35 ppt (under in vivo-like, asymmetrical conditions). Upon the administration of RGN, however, the posterior intestine displayed reduced HCO₃⁻ secretion and increased G_TE, without further reductions to the TEP. This suggests that, at least in the Gulf toadfish intestine, Cl⁻ does not recirculate across the apical membrane through NKCC2 and SLC26a6 for the purposes of preserving ion and fluid absorption, which is hypothesized to occur in the Japanese eel (4, 55). Interestingly, the present study shows that RGN does not reduce HCO₃⁻ secretion in the Gulf toadfish rectum as it does in the posterior intestine (when HCO₃⁻ is present in the serosal saline). A previous study demonstrated that the downstream effects of RGN inhibited the exchange activity of SLC26a6, when CO₂ was the only source of HCO₃⁻ (via CO₂ hydration and CA_C) with which to fuel SLC26a6 (41). By removing bicarbonate from the serosal saline, the basolateral HCO₃⁻ transporter, NBCe1, was rendered functionless and, as before, CO₂ was the sole source of HCO₃⁻ fueling SLC26a6. Under these conditions, RGN reduced HCO₃⁻ secretion in both the posterior intestine and rectum. These observations support the previous study's conclusions (41) that the transport activity of SLC26a6 is being directly affected by the downstream effects of RGN.

We also tested a possible mechanism in mammals that is proposed to decrease HCO₃⁻ secretion, via SLC26a6, when PKC phosphorylates SLC26a6 at a site that decreases its ability to bind CA_C, and, thereby, disrupts the SLC26a6-CA_C metabolon (1). It is possible that the effects of RGN could lead to PKG or PKA [which are both phosphorylated downstream of GC-C activation (5)] binding to Gulf toadfish intestinal SLC26a6 at a location on or near the CA_C binding site and limit the exchange capacity of SLC26a6 as well. To test this in the Gulf toadfish intestine, the CA_C-specific inhibitor, ethoxzolamide [tested in Gulf toadfish previously (18)], was used along with RGN in tissues bathed with a serosal saline absent of HCO₃⁻, to determine their individual and combined effects on HCO₃⁻ secretion. With this approach, as with the previous experiment (see above), CO₂ was the only source of HCO₃⁻ to fuel SLC26a6. The results of the present study reveal that in both the posterior intestine and rectum, the individual doses of RGN and ethoxzolamide statistically decreased HCO₃⁻ secretion equally, and the combined doses were additive and different than the individual doses. Consequently, these data do not suggest that the downstream effects of RGN stimulation affects SLC26a6 by interfering with the binding of CA_C and disrupting the SLC26a6-CA_C metabolon. Instead, the data support the idea that the downstream effects of RGN have a direct effect on the transport capacity of SLC26a6 for HCO₃⁻ that leads to decreased exchange activity in the posterior intestine (in both the presence and absence of serosal HCO₃⁻) and in the rectum (in the absence of HCO₃⁻). Moreover, three previous studies (including one on the Gulf toadfish) also demonstrate that CFTR does not seem to transport HCO₃⁻ (30, 40, 50), and that Gulf toadfish SLC26a6 transport activity is limited by HCO₃⁻ and not Cl⁻ (40). With respect to the guanylin peptide system in the Gulf toadfish intestine observed in the present and previous studies (40, 41), it is possible that inhibition of SLC26a6 (and HCO₃⁻ secretion) and activation of CFTR due to RGN stimulation leads to a maximal Cl⁻ current that manifests itself, physiologically, as a reversal of the absorptive I_SC.

Increased Response to Renoguanylin in Hypersalinity

The present study reveals changes in gene expression for NKCC1 and CFTR in the rectum after Gulf toadfish are transferred from 35 to 60 ppt, and these results are comparable to what is seen in the posterior intestine (41). The data from both tissues show that gene expression of NKCC1 and CFTR is upregulated in 60 ppt, but differ concerning the expression of GC-C, which is upregulated in the posterior intestine (40), but not in the rectum. The previous study suggested that the response to RGN in the posterior intestine of Gulf toadfish acclimated to 60 ppt was limited by the level of GC-C, NKCC1, and/or CFTR gene expression. At least in the rectum of the Gulf toadfish, the response to RGN is likely limited to NKCC1 and/or CFTR. Despite acclimation to 60 ppt, the current and previous studies (41) demonstrate that water absorption by the posterior intestine and rectum of Gulf toadfish does not increase as expected, as opposed to the sea bream intestine when acclimated to 55 ppt (15). Moreover, both the posterior intestine and rectum of fish exposed to 35 and 60 ppt displayed decreased water absorption in response to RGN that correlates with net Na⁺ and Cl⁻ secretory fluxes (a switch from mucosa-to-serosa to serosa-to-mucosa), with the exception of the posterior intestine from the 60 ppt. When considering our hypothesis that the guanylin peptide-induced inhibition of water absorption is necessary to facilitate the removal of solids (such as CaCO₃ precipitates from the intestine), the occurrence of water absorption being inhibited more greatly in 60 ppt (in response to RGN) would be more important. In 50 ppt, intestinal CaCO₃ precipitation in Gulf toadfish increases 2.3-fold (14). Thus, by upregulating gene expression of the components of the Cl⁻-secretory pathway (apical CFTR and basolateral NKCC1) and inhibiting NKCC2 transport activity by RGN (which seems to occur at 35 ppt as well), more fluid can remain in the intestine to facilitate the removal of increased CaCO₃ precipitates formed as a result of hypersalinity.

A notable difference in Cl⁻ flux between the posterior intestine and rectum of Gulf toadfish exists, as mentioned above, in response to RGN. In the present and previous study, net Cl⁻ secretion does not occur in the posterior intestine of fish acclimated to 60 ppt when treated with RGN, and this is observed in parallel with increased HCO₃⁻ secretion (41). It has been suggested that this occurs to maintain the drinking rate of the fish (2). In hypersalinity, there is even greater branchial fluid loss relative to normal seawater. Consequently, to stave off dehydration, a seawater teleost must increase its rate of drinking concurrently with the increases in branchial water loss. In 35 ppt, decreases in HCO₃⁻ secretion by the posterior intestine is beneficial to promote net Cl⁻ secretion and inhibit both water absorption and CaCO₃ precipitation (40, 41). Conversely, in 60 ppt, stimulating net Cl⁻ secretion and limiting the posterior intestine's ability to absorb water and secrete HCO₃⁻ could be detrimental. The consequence of limiting HCO₃⁻ secretion in 60 ppt is the reduction of CaCO₃ precipitation that is essential for the absorption of water in seawater teleosts. Likewise, increases in luminal Cl⁻ concentration in fish decreases their drinking rate in parallel (2). Accordingly, the RGN response with respect to HCO₃⁻ secretion (41) and Cl⁻ in the posterior intestine of Gulf toadfish acclimated to 60 ppt is a possible adaptation for maintaining a high rate of drinking to compensate for increases in branchial fluid loss. In this fascinating process, Cl⁻ secretion likely still occurs in response to RGN (as evidenced by inhibited water absorption), but is completely reabsorbed by SLC26a6 in exchange for HCO₃⁻. Yet, these mechanisms are not observed in the rectum of fish acclimated to 60 ppt, with Cl⁻ secretion occurring equally to what is seen in both the posterior intestine and rectum of fish acclimated to 35 ppt. The adaptive benefit of this response might be to ensure that enough water is present in the rectum (via inhibited water absorption/stimulation of water secretion) to removed CaCO₃ precipitates (or other solids) from the intestine.

Significance and Perspectives

The posterior intestine and rectum of Gulf toadfish display comparable osmoregulatory physiologies with respect to Na⁺, and Cl⁻ fluxes, yet they possess key distinct mechanisms by which they regulate ion and fluid movement. The posterior intestine appears to rely more on electrogenic anion exchange by SLC26a6 than the rectum (as evidence by the larger magnitude of its TEP), while the rectum appears to rely more on ion cotransport by NKCC2 than the posterior intestine (as evidenced by the large magnitude of its absorptive I_SC and the greater absolute change in I_SC when bumetanide is applied). In addition, the rectum absorbs more water than the posterior intestine, and this observation correlates with the greater absorptive I_SC and tissue density of the intestine. The rectum responds more strongly to RGN than the posterior intestine in reversing the absorptive I_SC and inhibiting water absorption. In support of the hypothesis that the tissue response of the intestine to the guanylin peptides is necessary to facilitate the removal of solids from the intestine, limiting water absorption would permit more water to remain in the rectum, which would enable solids to be removed from the intestinal tract (at least in response to RGN from the Japanese eel).

The downstream effects of cGMP formation, due to RGN binding to GC-C, might act to inhibit both NKCC2 and SLC26a6, in addition to stimulating the opening of apical CFTR. If these intracellular effects do occur, it elevates the guanylin peptides as potent modulators of intestinal physiology due to their modification of the osmoregulatory processes. The intracellular mechanisms that regulate these changes should be investigated further and be focused on the roles that cGMP/PKG and cAMP/PKA might play.

The present study also reveals that uroguanylin gene expression significantly decreases in the rectum of fish exposed to 60 ppt and compares well to what is seen in the posterior intestine in 60 ppt (41). This suggests that Gulf toadfish guanylin and uroguanylin might have different functions. Indeed, in the rat intestine and in T84 cells expressing rat GC-C, uroguanylin was most effective in a pH of 5.5, while guanylin was most effective in a pH of 7.4 and 8.0 (24, 26). In mammals, guanylin and uroguanylin cause increased HCO₃⁻ secretion into the intestinal lumen; accordingly, it has been suggested that uroguanylin might play a role in defending the pH of the intestine, as acidic chyme enters from the stomach (11). Conversely, guanylin transcription and secretion are increased in rat intestinal tissues when perfused with high salt concentrations and decreased with low salt intake (7, 29, 32). Conceivably, in seawater teleosts, uroguanylin plays a more prominent role in digestion, and guanylin takes a more active part in osmoregulatory processes.

It should be emphasized that the guanylin peptide system in Gulf toadfish has, so far, been studied on fasted fish. The potential interactions between osmoregulation and digestion in the context of guanylin peptide function are an interesting avenue of further study.

GRANTS

M. Grosell is Maytag Professor of Ichthyology and is supported by National Science Foundation (Grant IOS 1146695).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

I.M.R. and M.G. conception and design of research; I.M.R. performed experiments; I.M.R. and M.G. analyzed data; I.M.R., Y.T., and M.G. interpreted results of experiments; I.M.R. prepared figures; I.M.R. drafted manuscript; I.M.R., Y.T., and M.G. edited and revised manuscript; M.G. approved final version of manuscript.