Living with a leaky skin: upregulation of ion transport proteins during sloughing

The skin, the largest vertebrate organ, has evolved into a complex, pluri-stratified, multi-membrane system (Lillywhite and Maderson, 1988) with vital, often contrasting, functions (reviewed in Chuong et al., 2002). It must act as an impermeable barrier to protect the organism from environmental stressors (abiotic and biotic) (Proksch et al., 2008) while simultaneously being able to sense, interact and exchange some solutes with the surrounding environment (Hillyard et al., 2007). For amphibians, this is especially important as their skin regulates the exchange of electrolytes, water and respiratory gases with the environment to maintain internal homeostasis (Boutilier et al., 1992; Shoemaker et al., 1992).

Histologically, the skin of amphibians consists of an outer multilayered epidermis and an inner dermal layer (dermis). The most basal layer of the epidermis is the stratum germinativum, which is attached to the dermis by a basement membrane. The columnar cells of the s. germinativum continually regenerate (via mitotic division), to replenish and maintain the rest of the epidermal cell population (Heatwole et al., 1994). Above the s. germinativum is the s. spinosum (‘spiny layer’), where cellular keratinisation begins. Above the s. spinosum is the s. granulosum (granular layer) where the keratinocytes become known as granular cells. These cells are connected by tight junctions (zonulae occludens) that separate the apical plasma membrane from the basolateral membrane lining the lateral intercellular space (Farquhar and Palade, 1964). The apical plasma membrane of granular cells therefore forms the limiting barrier for transcellular solute and water transport across the epidermis, with connections between this layer and the underlying layers (s. spinosum and s. germanitivum) forming an electrical syncytium. The most superficial layer of the epidermis is the s. corneum, which consists of 1–2 thin layers of dead cells (reviewed in Heatwole et al., 1994).

Because of constant interactions with the external environment, the skin must be renewed in order to maintain functionality (Ling, 1972). All metamorphosed amphibians go through a physiological and behavioural phase known as ‘ecdysis’ or ‘sloughing’ (Ling, 1972; Larsen, 1976), where the old s. corneum is removed and replaced periodically with a new layer (derived from the underlying s. granulosum). In the early shedding stage, the old s. corneum becomes separated from the underlying cell layer via a gradual dissolution of the desmosomes between these two layers (Budtz, 1977). The new s. corneum (derived from the underlying s. granulosum) swells, flattens (Ernst, 1973) and becomes cornified (Elias and Shapiro, 1957; Whitear, 1975). Mucus also appears beneath the separated s. corneum (known as a ‘slough’) to facilitate removal (Larsen, 1976). During the post-shedding phase, the new s. corneum flattens and becomes denser in appearance (Budtz and Larsen, 1973). Fusion of the s. granulosum layer leads to the formation of tight junctions (Budtz and Larsen, 1975), which serve to limit the paracellular movement of salts and water (Tsukita et al., 2001; Niessen, 2007). While sloughing maintains the integrity of the skin, physiological changes have been observed to coincide with sloughing that may disturb homeostatic balance. Jørgensen (1949) first observed an increase in cutaneous permeability in sloughing Bufo bufo, Rana temporaria and Rana esculenta which increased the rate of both water gain and sodium loss. These changes in skin permeability can start 12 h ahead of sloughing (Ewer, 1951), and continue 3–4 h post-sloughing (Jørgensen, 1949).

The temporary disruption of cutaneous integrity that occurs during sloughing corresponds with changes to the skin’s electrophysiological properties. During sloughing there is a decrease in the short-circuit current (indicative of active ion transport activity), transepithelial potential (indicative of the electrical potential difference between the two sides of the cell membrane) and skin resistance (Larsen, 1970, 1971a; Nielsen and Tomilson, 1970). These changes seem to result from the physical shedding of the slough, as they are initiated by the separation of the slough from the underlying skin layer (reviewed in Erlij and Ussing, 1978). However, these studies artificially induced sloughing through the administration of hormones such as aldosterone (Nielsen, 1969; Larsen, 1971a), which can potentially create unrealistic electrophysiological data (relative to natural or spontaneous sloughing), because aldosterone also acts to regulate ion and water reabsorption in the kidneys (Garty, 1986; Eaton et al., 2001) and skin (Bentley, 2002), and increases protein synthesis in the bladder and skin (Crabbé and de Weer, 1964). Thus, many of these earlier sloughing studies are difficult to interpret, as the observed electrophysiological changes of the skin may be an artefact of aldosterone administration instead of the natural physiological changes that occur during sloughing.

While earlier studies have suggested that the physical alterations that occur to amphibian skin during sloughing (e.g. removal of tight junctions) cause the increase skin permeability (Smith, 1975; Larsen, 1976), Katz (1978) suggested disruption of tight junctions was not solely responsible for the observed changes in specific permeability of the skin to Na⁺ and K⁺ during sloughing. A reduction in the abundance or activity of transepithelial ion transport proteins such as epithelial sodium channels (ENaC) or the sodium–potassium pump (Na⁺/K⁺-ATPase; NKA) during sloughing may also cause disruption to the transportation of ions. ENaC is a membrane-bound protein channel found in the apical membrane of epithelial cells that is responsible for the uptake of Na⁺ from the environment, while NKA is an active membrane protein pump found in the basolateral layer of epithelial cells that establishes an electrochemical gradient for ion channels like ENaC to transport Na⁺ against their concentration gradients (Hillyard et al., 2008). Thus, examining the relative expression of these proteins in sloughing amphibians relative to non-sloughing amphibians may clarify the mechanism through which this disruption of ion transport during sloughing occurs.

To date, there is no direct link with the activities of the epithelial transport proteins and the changes in ion and osmotic transport during sloughing. As there are physical and physiological changes in the skin associated with sloughing (Larsen, 1976; Erlij and Ussing, 1978), does the change in skin function cause temporary electrolyte disruption in the animal’s homeostasis (e.g. blood biochemistry), or does the skin actively upregulate expression of epithelial ion transporters (e.g. ENaC and NKA) to maintain homeostasis? Taking an integrative approach, this study aimed to investigate the effects of sloughing on epidermal ion transport in vivo and in vitro, the distribution, abundance and expression of associated regulatory transport proteins, and blood plasma biochemistry in cane toads, Rhinella marina (Linnaeus 1758). This study also avoided the influence of hormone-induced sloughing by investigating epithelial transport in spontaneously sloughing anurans.

We hypothesised that as a result of physical changes to the skin during spontaneous sloughing there would be disruption of cutaneous ion and water transport (Jørgensen, 1949), with a reduction in the abundance and expression of ion transport proteins (e.g. ENaC and NKA), resulting in a transient disruption to internal ion and water homeostasis.

Rhinella marina (n=24) were collected from The University of Queensland campus, Brisbane, Australia, between November 2014 and January 2015, and housed in large black bins (n=5 per bin) containing a thin layer of pine-bark mulch (Richgro, WA, Australia). Toads were checked daily, fed weekly on vitamin-dusted (Aristopet Pty Ltd, QLD, Australia) crickets (Acheta domesticus), and enclosures were cleaned weekly. After approximately 2 weeks, toads were housed individually in ventilated polypropylene plastic containers (130×140×325 mm) with a thin layer of pine-bark mulch saturated with tap water, which were tilted to allow a wet and dry gradient for the toads to move between. All toads were kept at 25±0.3°C on a 12 h:12 h light:dark cycle. All procedures in this study were carried out with the approval of The University of Queensland's Animal Ethics Unit (approval no. SBS/316/14/URG).

To determine sloughing events, R. marina (n=24) were marked with a small amount of non-toxic waterproof zinc cream (Key-Sun Laboratories, NSW, Australia) on the dorsal surface following Meyer et al. (2012). Animals were checked twice daily to record the disappearance of the marks as an indication of sloughing. Once a mark disappeared, it was reapplied. The time (in days) between mark application and loss was recorded as the intermoult interval (IMI). The sloughing behaviour for R. marina was divided into four stages: (1) intermoult, the period in between each slough; (2) pre-sloughing, when animals started to extend their limbs to lift the abdomen off the ground; (3) sloughing, which begins with mouth gaping, followed by abdominal contractions, upper body wiping and removal of the old skin; and (4) post-sloughing, up to 1 h after sloughing when normal behaviour resumes.

To measure changes in in vivo ion efflux during sloughing, each animal was placed in an open-ended ventilated clear chamber (75×110×160 mm) containing 300 ml of reverse osmosis water with a magnetic stirrer to circulate the solution. The rate of ion efflux was measured as the change in conductivity (in µS h⁻¹) for toads in each of the following moult stages: (1) intermoult, (2) pre-sloughing, (3) sloughing and (4) post-sloughing (Fig. 1) for 9 animals. A conductivity electrode connected to a conductivity pod (ML307, ADInstruments, NSW, Australia) was placed at each end of the chamber. The output was digitised with a PowerLab 4/35 interface (ADInstruments) and recorded on Labchart software (ADInstruments). The baseline of the bathing solution was measured (average from the two electrodes) for roughly an hour before the animal was placed into the chamber (Fig. 1). A change in the conductivity of the bathing solution associated with sloughing was determined via video surveillance, and duration of the sloughing event (min) was recorded. The experiment was conducted at room temperature (22±0.5°C). Animals were fed weekly, but because of the difficulty in determining when animals would slough (7 day variation), animals were not fasted prior to measurements and may have been in different digestive states. Both fasted and recently fed animals were included in the analysis as there was no significant difference in the rate of conductivity (t₅₀=−0.76, P=0.45) between fasted and recently fed animals.

The net loss of Na⁺ and K⁺ into the bathing solution was measured using flame photometry (BWB-XP flame photometer, BWB Technologies Ltd, UK). The net loss of Cl⁻ was determined spectrophotometrically (DTX 880 Multimode Detector, Beckman Coulter, IA, USA) using a commercially available chloride kit according to the manufacturer's instructions (MAK023, Sigma-Aldrich, MO, USA). The conductivity readings (µS h⁻¹) were converted to NaCl efflux (mmol l⁻¹) using known concentrations of NaCl solutions (0–1 mmol l⁻¹). The conductivity readings of the solution and known NaCl concentrations showed a predictable linear relationship (r²=0.99; Fig. S1). To estimate the overall effect of sloughing on the total sodium content of the extracellular fluid (ECF), conductivity measures were converted to a rate of sodium loss (mmol h⁻¹) based on [Na⁺] of bath water samples collected after sloughing and assuming that sodium was the primary ion contributing to solution conductivity and that there was an equal proportion of sodium lost relative to other ions. The net amount of Na⁺ lost during sloughing was calculated assuming an ECF volume of approximately ∼25% of the body mass of the animal (Thorson, 1964). The actual Na⁺ concentration of the ECF was measured from plasma samples (see below). The proportion of the total ECF Na⁺ (as a percentage of ECF Na⁺ h⁻¹) lost during sloughing was calculated by dividing the rate of ion loss (mmol l⁻¹ h⁻¹) by the total amount of ECF Na⁺ (mmol) multiplied by 100.

Before and after each conductivity trial, animals were voided of urine by gently applying pressure on the ventral side. Body mass (M_b; g) was then measured (Adam Equipment Co. Ltd, CT, USA) to calculate any change over the sloughing period as percentage change in M_b per hour. The change in M_b over the tested interval represents osmotic water flux. Animals that urinated or defected during the experiment were taken out and re-weighed before being returned to the chamber, with fresh bathing solution.

Toads (n=14 intermoult, n=10 post-sloughing) were killed via double pithing at one of two time points: (1) post-sloughing (no more than 1 h after sloughing had occurred), or (2) intermoult (at a point halfway through the IMI). Snout–vent length (SVL) and M_b were measured immediately. Blood was collected via cardiac puncture into a lithium heparinised syringe. Samples were then centrifuged at 5000 g for 5 min and the plasma collected and stored at −20°C for subsequent electrolyte analysis. Two heparinised capillary tubes of whole blood were also centrifuged at 5000 g for 3 min to determine haematocrit (Hct). Plasma Na⁺ and K⁺ levels (mmol l⁻¹) were measured using flame photometry, and plasma osmolality (mosmol l⁻¹) was measured using a Vapro 5520 vapour pressure osmometer (Wescor, Logan, UT, USA). Plasma Cl⁻ levels (mmol l⁻¹) were determined spectrophotometrically (DTX 880 Multimode Detector, Beckman Coulter, IA, USA) using a chloride kit according to the manufacturer's instructions (MAK023, Sigma-Aldrich).

The transepithelial potential and resistance were also measured under more ecologically realistic conditions consistent with ‘freshwater’ on the apical skin side (26 mmol l⁻¹ NaCl in distilled water), where the electrochemical gradient across the skin would favour a net efflux of ions during sloughing. The traditional approach of using Ringer solution on both surfaces of the membrane could potentially negate this condition.

Epidermal thickness was compared between intermoult and post-sloughing toads. Ventral skin samples (<1 cm²) from the lower abdominal pelvic patch region were collected and placed into aqueous buffered zinc formalin fixative (Z-fix, Anatech, MI, USA) for 24 h, then transferred to 70% ethanol and stored at 4°C. Tissue samples (intermoult n=6, post-sloughing n=6) were then dehydrated through an ascending ethanol series, cleared in xylene and embedded in paraffin wax (Histoplast Paraffin, ThermoFisher Scientific, Sydney, Australia). Tissue samples were then transversely sectioned into approximately 6 µm-thick sections (Leica RM2245, Leica Microsystems, NSW, Australia), stained with Mayer's haematoxylin and 1% eosin in 70% ethanol, and photographed with NIS-Elements software (v. 4.10, Nikon Instruments Inc., Tokyo, Japan) under bright-field illumination (Nikon Eclipse E200 MV, Nikon Instruments Inc.).

Distribution of the NKA α-subunit in the ventral skin tissues was examined via immuno-fluorescence staining. Distribution of the ENaC α-subunit in skin sections was not examined as the antibody (Anti-SCNN1A antibody, HPA012939, Sigma-Aldrich) did not function well in our immunohistochemistry protocol. Sections were de-paraffinised then washed in washing buffer (0.01 mol l⁻¹ PBS, 0.05% Tween-20, pH 7.4; 2×3 min washes) and blocked at room temperature in normal goat serum (2% goat serum, 1% BSA, 0.1% cold fish skin gelatine, 0.1% Triton X-100, 0.05% Tween-20 and 0.05% thiomerosal in 0.01 mol l^–1 PBS, pH 7.4). Sections were then incubated in a humidified chamber overnight at 4°C with the NKA primary antibody α5 (developed by the Developmental Studies Hybridoma Bank, created by the NICHD of the NIH, and maintained at The University of Iowa, Department of Biology), diluted 1:500 in 1% BSA, 0.1% cold fish skin gelatine and 0.05% thiomerosal in 0.01 mol l⁻¹ PBS. Subsequently, sections were incubated with a fluorophore-conjugated secondary antibody (goat anti-mouse IgG Dylight, ab96879, Abcam Inc., Cambridge, UK) diluted 1:500 in 0.01 mol l⁻¹ PBS and 0.05% Tween-20 for 1 h in the dark at room temperature. They were then mounted with Fluoroshield DAPI mounting medium (Sigma-Aldrich), and viewed using a Nikon Eclipse E200 MV series epi fluorescence microscope and photographed using NIS-Elements software.

Semi-quantitative Western blotting was performed to estimate the relative abundance of NKA α-subunit and ENaC α-subunit in the ventral skin tissues of intermoult (n=4) and post-sloughing (n=4) toads. Ventral skin samples (<1 cm²) from the lower abdominal pelvic patch region were collected and stored at −80°C. Frozen tissues were then homogenised in ice-cold membrane extraction homogenisation buffer [250 mmol l⁻¹ sucrose, 30 mmol l⁻¹ Tris, 1 mmol l⁻¹ EDTA, 100 µg ml⁻¹ phenylmethylsulfonyl fluoride (PMSF) and 5 mg ml⁻¹ protease inhibitor cocktail], and centrifuged at 1200 g for 15 min at 4°C. The supernatant was then removed and centrifuged at 23,000 g for 25 min at 4°C. The resulting pellet was re-suspended in 50 µl of homogenisation buffer. An aliquot was used for Bradford protein quantification (Sigma-Aldrich).

A 5 µg sample of total membrane protein in NuPAGE LDS sample buffer (Invitrogen) was loaded in triplicate into Bolt 4–12% Bis-Tris Plus gels (ThermoFisher Scientific) and electrophoresed at 140 V for 1 h. Gels were subsequently transferred onto 0.45 µm Westran PVDF blotting membranes (Sigma-Aldrich) at 20 V for 75 min. Membranes were then blocked in blocking buffer [5% skim milk in TBST (20 mmol l⁻¹ Tris, 150 mmol l⁻¹ NaCl and 1% Tween-20, pH 7.6)] for 1 h at room temperature before incubation overnight at 4°C with their respective primary antibody [0.2 µg ml⁻¹ NKA primary antibody α5 and 0.15 µg ml⁻¹ ENaC α-subunit primary antibody (Anti-SCNN1A antibody, HPA012939, Sigma-Aldrich)] diluted in blocking buffer. Membranes were then incubated in secondary goat anti-mouse IgG horseradish peroxide (HRP)-conjugated antibody (1.2 µg ml⁻¹; C22P20, Antibodies Australia, VIC, Australia) and goat anti-rabbit IgG HRP-conjugated antibody (0.6 µg ml⁻¹; C24P03, Antibodies Australia), respectively, for 1 h at room temperature, and stained with 1-Step Ultra TMB-Blotting solution (ThermoFisher Scientific). The membranes were dried and digitised for densitometry analysis with Image J (https://imagej.nih.gov/ij/). The pixel densities were used to estimate protein abundance in the post-sloughing animals, expressed relative to the intermoult group.

Ventral skin samples (<1 cm²) from the lower abdominal pelvic patch region were collected and stored in RNA-later (Ambion Inc., TX, USA) at −20°C. The skin samples were homogenised with stainless steel beads using a TissueLyser II (Qiagen). Total RNA was isolated using an RNeasy Mini kit with on-column DNAse treatment as per the manufacturer's guidelines (RNeasy Mini kit, Qiagen, Hilden, Germany). RNA purity was assayed by spectrometry, and yield measured using a Qubit fluorometer (ThermoFisher Scientific). RNA was then reverse transcribed into cDNA (SensiFAST™ cDNA synthesis kit, Bioline, NSW, Australia), with appropriate controls (no reverse transcriptase), followed by RNase H treatment.

Quantitative PCR (qPCR) primers against the target gene ENaC (scnn1a, sodium channel epithelial 1 alpha subunit) and the housekeeping gene glyceraldehyde 3-phosphate dehydrogenase (GAPDH) were manufactured using previously published species-specific primer sequences (Konno et al., 2007). Species-specific primers against the target gene NKA (atp1a1, ATPase Na⁺/K⁺ transporting subunit alpha 1) were designed from published R. marina NKA-α1 subunit sequence (GenBank: Z11798.2; Jaisser et al., 1992) using OligoPerfect™ Designer (ThermoFisher) with acceptance of the default parameters. qPCR was performed using the SensiFAST™ SYBR No-ROX kit (Bioline, NSW, Australia) in a thermal cycler (MiniOpticon™, Bio-Rad) using the cycling parameters recommended in the qPCR kit. Each assay (in triplicate) included a no-template control and a no-reverse transcriptase control. All PCR efficiencies were >90% and all the assays produced unique dissociation curves. Bio-Rad CXF Manager software (version 3.1, Bio-Rad) results were exported as tab-delimited text files and imported into Microsoft Excel for further analysis. The expression of each gene was quantified relative to the expression of GAPDH using the Pfaffl (2001) method.

All analyses were performed in R.3.0.1 (https://CRAN.R-project.org/package=nlme). Data were presented either as means±s.e.m. or as individual data points depending on the nature of the data (Cumming et al., 2007; Krzywinski and Altman, 2013), and α was set at 0.05 for all statistical tests.

Differences in the rate of conductivity between each group were analysed using linear mixed effects (lme) models in the R ‘nlme’ package (https://CRAN.R-project.org/package=nlme) with groups (intermoult, pre-sloughing, sloughing, post-sloughing) as fixed effects, M_b (g) as covariate, and individual identity as a random effect to account for repeated measurements within individuals. Changes in M_b (M_b before – M_b after) following the in vivo ion loss experiment between intermoult and sloughing groups were analysed using linear mixed effects (lme) models with group (intermoult, post-sloughing) as fixed variable, duration in the experiment as covariate, and individual identity as a random effect to account for repeated measurements within individuals.

Differences between the intermoult and post-sloughing group in plasma biochemistry (Hct, Na⁺, K⁺, Cl⁻ and osmolality), and the electrophysiological properties of the skin (transepithelial potential, resistance and active transport) were analysed by analysis of covariance (ANCOVA) with time (h) of post-mortem blood and skin tissue collection as covariates.

Differences in staining intensity of antibody bands representing NKA α-subunit and ENaC α-subunit relative abundance and relative expression of NKA and ENaC between intermoult and post-sloughing groups were compared using ANCOVA with group (intermoult and post-sloughing) as fixed effect and time (h) of post-mortem skin tissue sample collection as covariate.

The IMI of Rhinella marina held at 25°C was variable within and between individuals and ranged from 5 to 16 days with a mean of 10.5±1.8 days (n=24). There was no diurnal effect on sloughing, with 53% of toads sloughing at night and 47% during the day. The IMI tended to be longer in larger toads (t₁₉=−2, P=0.051, n=24).

The average rate of change in conductivity in the bathing water, representing ion efflux rate from toads, increased 160 times (t₄₈=27.7, P<0.0001; Fig. 2A) from 0.5±0.1 µS h⁻¹ in intermoult animals to 90±6.9 µS h⁻¹ during sloughing. An increase in conductivity was also recorded just prior to sloughing (11.5±1.0 µS h⁻¹; t₄₈=3.2, P=0.002). The rate of efflux of ions returned to baseline levels rapidly (within 30 s; Fig. 1) after sloughing (t₄₈=0.67, P=0.5). The majority of the ions lost by the toads were sodium and chloride. The estimated net loss of Na⁺ during the hour incorporating the sloughing event represented ∼3.3±0.5% of the animal's total ECF Na⁺ pool.

There was no significant difference in the net change in mass between those that sloughed (0.2±0.3% h⁻¹) and the intermoult group (−0.4±0.6% h⁻¹) during the conductivity measurements (t₂₂=−0.01, P=0.99; Fig. 2B), and the duration of sloughing also had no effect on the change in M_b between treatments (t₃₀=1.16, P=0.25).

There was no significant difference in the transepithelial potential (mV) and resistance (Ω cm²) between the intermoult and post-sloughing groups when the toad skin was perfused with Ringer solution on both sides (Table 1). However, the post-sloughing group showed a significant increase in I_sc (15.4±2.7 µA cm⁻¹) compared with the intermoult group (8.0±0.9 µA cm⁻¹; Table 1). This represents a rate of active Na⁺ transport (influx) of 1.6×10⁻¹⁰±3.2×10⁻¹¹ mol s⁻¹ after sloughing.

When the skin was perfused with freshwater on the apical side of the chamber, the transepithelial potential did not differ significantly between the intermoult and post-sloughing group (Table 1); however, there was a significant decrease in transepithelial resistance in the post-sloughing group compared with the intermoult group (Table 1).

Hct, plasma osmolality, and sodium, chloride and potassium concentrations were not significantly different between the intermoult and post-sloughing animals (Table 2).

There were no significant differences in skin thickness between the intermoult and post-sloughing groups (t₈=−0.99, P=0.34; Fig. 3). No morphological differences were observed between the groups.

The NKA α-subunit was mostly distributed in the epidermis, with little or no immunofluorescence staining in the dermis layer (Fig. 4). The submucosal glands within the skin also showed strong NKA α-subunit immunofluorescence in both intermoult and post-sloughing animals. Within the epidermal layer, the NKA α-subunit was concentrated around the basolateral membranes in the principal cell layer (excluding the s. corneum). NKA α-subunit distribution in the epidermis of all intermoult animals remained relatively uniform (Fig. 4A). In the post-sloughing group, however, NKA α-subunit distribution within the epidermis was less well defined. In some animals, NKA α-subunit distribution was concentrated in the most superficial layers of the epidermis, while in other animals the NKA α-subunit appeared to be restricted to cells in the basal layers of the epidermis (Fig. 4B).

Semi-quantitative western blotting for the NKA α-subunit (α5 antibody) identified a single band of approximately 100 kDa from membrane-enriched skin cell homogenates. NKA α-subunit abundance was significantly higher (F_1,12=9.5, P=0.009) in the post-sloughing group relative to the intermoult group (Fig. 5). The ENaC α-subunit antibody identified a single band of approximately 75 kDa. The ENaC α-subunit abundance was also significantly greater (F_1,12=8.8, P=0.01) in the post-sloughing group compared with the intermoult group (Fig. 5), with up to a 2.5-fold increase in abundance occurring in some animals (mean=1.62±0.16).

The relative expression of NKA (ATP1A1) and ENaC (SCNN1A) mRNA transcripts in the epidermis was not significantly different between the intermoult and post-sloughing group (Pfaffl ratio: 1.137, F_1,5=0.06, P=0.8 and Pfaffl ratio: 2.14, F_1,5=2.5, P=0.1, respectively; Fig. 6).

Amphibian skin is a complex multifunctional organ, with the ability to actively regulate the transcutaneous movement of ions and water necessary for maintaining physiological homeostasis (Feder and Burggren, 1992). Understanding the impact of sloughing on skin function is important to comprehend how amphibians maintain homeostatic balance in the face of frequent and significant perturbations. The present study demonstrates that during sloughing, amphibians in freshwater environments are able to compensate for an increase in the rate of transcutaneous sodium loss by increasing the abundance of sodium transporters in the skin, which in turn allows amphibians to maintain their internal electrolyte homeostasis. These findings suggest that although sloughing causes an acute change in skin osmotic function, the effects are relatively transient and are offset by compensatory pathways. How sloughing may impact species with more frequent sloughing regimens or those with disease-related sloughing dysfunction, however, remains to be determined.

Sloughing-associated skin disturbances began ahead of the physical act of sloughing; indeed, there was a 10-fold increase in the rate of ion efflux in live toads prior to the initiation of sloughing. This finding is consistent with previous studies showing that changes in the electrophysiological properties of amphibian skin often precede the physical sloughing event. Specifically, a decrease in I_sc has been observed during the initiation of sloughing (in vitro) in B. bufo (Larsen, 1970, 1971a,b). In isolated amphibian skin, I_sc is predominately linked to active sodium uptake (Koefoed-Johnson and Ussing, 1958). A reduction in I_sc across the skin during sloughing therefore probably reflects a reduction in active transcutaneous sodium movement. The mechanistic basis for this pre-sloughing change in ion transport remains unclear; however, preparatory shifts in cellular differentiation within the epidermis have been hypothesised to contribute to the observed changes in electrical readings in pre-sloughing amphibians (Larsen, 1971b). Potentially, cellular apoptosis in the s. granulosum as a consequence of its transition to the s. corneum (Elias and Shapiro, 1957) may degrade the ion regulatory transporters responsible for maintaining the potential difference across the skin. This reduction in transporter abundance or activity would lower active sodium uptake rates and contribute to the net increase in sodium loss across the skin. Coincidently, an increase in cutaneous granular secretions to form the mucus layer during sloughing may also contribute to ion loss (Larsen and Ramløv, 2013).

Although some physiological changes preceded sloughing, by far the greatest disruption to skin function in R. marina occurred during sloughing itself. Sloughing substantially increased the rate of cutaneous ion efflux in R. marina, indicating that skin function (specifically active ion transport) was temporarily disrupted during sloughing. In addition, contributing to this net increase in Na⁺ efflux, the electrical resistance across the skin was significantly lower in post-sloughed toads, indicating that paracellular resistance to passive ion movements was substantially lower across newly sloughed skin. This reduction in cutaneous resistance following sloughing may be attributed to incompletely formed tight junctions in the new s. granulosum layer (Budtz and Larsen, 1975) that permits the passive paracellular movement of ions out of the animal along their concentration gradient (Larsen, 1970, 1971b, 1972). In B. bufo, the reduction in skin resistance can be observed a few hours before the initiation of sloughing (Larsen, 1970, 1971b), suggesting the process of epidermal renewal and formation of the new s. corneum increases the skin's permeability before sloughing (Budtz and Larsen, 1975).

Despite the increased loss of Na⁺ preceding and during sloughing, toads were able to maintain osmoregulatory homeostasis with no changes in plasma solute concentration or osmolality detected post-sloughing. This indicates that either the transient nature of the sloughing-induced changes are not substantial enough to disrupt whole-animal ionic homeostasis or active mechanisms are employed during or shortly after sloughing to counteract or compensate for ionic and osmotic fluxes. Jørgensen (1949), using in vitro measures of NaCl loss, estimated that the overall amount of Na⁺ lost during a sloughing cycle probably reflects a relatively small component of a healthy, normal animal's extracellular sodium pool, consistent with the small net sodium loss measured in this study. Compensatory changes to ion transport may include a change in the activity or abundance or ion transporters in the skin that allow sodium to be actively taken up across the skin from the environment. Consistent with this hypothesis, we observed a 2-fold increase in the I_sc across the skin of recently sloughed R. marina, suggestive of an increase in active sodium transport. Moreover, we observed a greater abundance of cutaneous ion transporter proteins (NKA α-subunit and ENaC α-subunit) in recently sloughed toads, suggesting that the increase in active sodium transport may, at least in part, be ascribed to an increase in the number of functional sodium transport proteins in the skin. However, the I_sc (the likelihood of the number of channels opened at a particular time) alone cannot determine the functional activity of ENaC, as the open probability is highly variable depending on external factors (Anantharam et al., 2006; Kleyman et al., 2009); thus, measuring amiloride-sensitive I_sc can quantify the functional activity of ENaC (Sariban-Sohraby and Benos, 1986). In addition, an increase in subunit abundance does not always indicate an increase in functional activity of the whole protein (Sardella and Kültz, 2009; Reilly et al., 2011). For example, the ENaC subunits (α-, β- and γ-subunits) alone cannot induce amiloride-sensitive currents (Canessa et al., 1994), and in bullfrog (Rana catesbeiana) tadpoles, ENaC is expressed in the skin but not in a functional state (no amiloride-blockable Na⁺ transport present) until metamorphosis (Takada et al., 2006).

Combined with the increase in I_sc consistent with an increase in active ion transport, R. marina appears to increase NKA and ENaC protein abundance to compensate for Na⁺ losses incurred during sloughing. The changes in I_sc with spontaneous sloughing are also consistent in magnitude with those observed in toads following aldosterone-induced sloughing events (Nielsen, 1969; Larsen, 1971a; Denèfle et al., 1983), suggesting that the suite of physiological changes that accompany sloughing in amphibians is preserved regardless of whether the sloughing event is initiated naturally or via exogenous hormonal stimulation.

While the sloughing-induced disruption to skin ionoregulatory function in R. marina was relatively brief, amphibians display a wide range of sloughing frequencies (Ohmer, 2016) and the net impact of sloughing on physiological homeostasis may be greater in species that slough more frequently. For example, toads in this study sloughed about every 12 days, but some species of amphibians slough as often as every 24 h (Bouwer et al., 1953; Castanho and de Luca, 2001; Ohmer, 2016). In these frequently sloughing species, disruptions to ion and water exchange during sloughing are likely to require a greater energetic investment to maintain ionic and osmotic homeostasis, which cumulatively may represent a significant cost to the animal. In addition, exogenous factors, both biotic and abiotic, that alter sloughing frequency may increase or decrease the overall impact of sloughing on physiological homeostasis. For example, increased sloughing frequency has been observed in amphibians suffering from the devastating disease chytridiomycosis (Ohmer et al., 2015). Chytridiomycosis, a novel and often fatal cutaneous disease of amphibians caused by the fungus Batrachochytrium dendrobatidis (Bd), appears to disrupt electrolyte transport across the skin of infected frogs (Voyles et al., 2007, 2012). Green tree frogs (Litoria caerulea), infected with Bd slough as much as 25% more frequently than uninfected frogs (Ohmer et al., 2015) and many other amphibian species are reported to slough more frequently when infected with Bd, though the evidence for this is largely anecdotal (Berger et al., 1998; Davidson et al., 2003; Meyer et al., 2012). While an increased sloughing rate may assist with removing pathogens from the skin of infected animals (Meyer et al., 2012; Cramp et al., 2014), sloughing also causes a temporary disruption of ionic and osmotic movements, which, if sloughing frequency is increased, may contribute to the fatal loss of internal ionic homeostasis in clinically infected frogs (Voyles et al., 2009). Thus, future studies that integrate the physiology of sloughing with Bd infection are required to more fully understand the mechanistic basis for the morbidity associated with cutaneous Bd infection.

We would like to thank Prof. M. Grosell (University of Miami) and Dr C. R. Campbell (University of Sydney) for advice on the Ussing chamber protocol. We would also like to thank the UQ School of Biomedical Sciences histology staff for advice on histological procedures, and all volunteers who assisted with animal care and maintenance.