In ureotelic mammals, plasma levels of ammonia are extremely low, so ammonia enters the urine almost entirely by secretion. However, the molecular mechanisms of ammonia transport remain controversial and incompletely understood (Weiner and Verlander, 2017). In the proximal tubule, apical NHE3 seems to play a predominant role, with NH₄ ⁺ substituting on the H⁺ site, such that NH₄ ⁺ secretion is coupled to Na⁺ reabsorption. NH₄ ⁺ movement through apical K⁺ channels, as well as the classic “diffusion trapping” mechanism, with NH₃ diffusion coupled to apical H⁺ secretion by NHE3 and/or v-type H⁺ATPase, may also contribute, again facilitating coupling to Na⁺ reabsorption. The sodium, potassium, two chloride cotransporter type 2 (NKCC2) seems to be the principal reabsorptive transporter contributing to ammonia recycling in the loop of Henle, with NH₄ ⁺ substituting on the K⁺ site. The final net addition of ammonia in the collecting duct seems to be largely mediated by two Rhesus (Rh) glycoproteins, which are channels that facilitate the diffusion of ammonia. Rhbg, which is exclusively basolateral, serves for basolateral uptake from the interstitium into the tubular cells, while Rhcg, which is both apical and basolateral, serves for apical secretion while supplementing basolateral uptake. Controversy continues as to whether the Rh proteins facilitate NH₃ movement (perhaps true for Rhcg), NH₄ ⁺ movement, or both (perhaps true for Rhbg). NH₄ ⁺ substitution on the K⁺ site of basolateral NKA also contributes to basolateral uptake in the collecting duct. HCN2, a hyperpolarization-activated cyclic nucleotide gated channel (subtype 2) has also been implicated in basolateral NH₄ ⁺ uptake. On the apical membrane, the “diffusion trapping mechanism” applies, with v-type H⁺ATPase as well as H⁺, K⁺ ATPase providing the apical H⁺ secretion (and perhaps also direct NH₄ ⁺ transport by the H⁺ site), while Rhcg facilitates the apical diffusion of NH_3.

In ammoniotelic fish, plasma ammonia concentrations are higher than those in ureotelic mammals, but still very low (micromolar) relative to urinary levels (millimolar). Therefore, as in mammals the great majority of ammonia enters the urine by tubular secretion rather than glomerular filtration, probably produced in the tubule cells by deamination and deamidation of glutamine and other amino acids. In a range of freshwater species, urine/plasma ratios of ammonia are high, and increase further during compensation of acidosis, indicating increased secretion (Cameron and Wood, 1978; McDonald and Wood, 1981; Wood et al., 1999; Wright et al., 2014; Lawrence et al., 2015; Thin et al., 2019). Renal activities of potential ammoniagenic enzymes (phosphate-dependent glutaminase, glutamate dehydrogenase, ɑ-ketoglutarate dehydrogenase, alanine aminotransferase, aspartase aminotransferase, phosphoenolpyruvate carboxykinase) are substantial, and some increase during experimental acidoses (Walton and Cowey, 1977; 1982; King and Goldstein, 1983; Wood et al., 1999; Lawrence et al., 2015), as well as mRNA for glutamate dehydrogenase (Hirata et al., 2003; Fehsenfeld and Wood, 2018). At least in the goldfish, glutamate dehydrogenase mRNA is expressed in all sections of the nephron, increasing during acidosis, and with highest levels in the proximal tubule, again as in mammals (Fehsenfeld and Wood, 2018). Given the lack of a countercurrent morphology, it seems likely that the ammonia transport mechanisms in fish will not necessarily parallel those in mammals. Nevertheless, the same basic elements appear to be present, with many similarities.

As noted earlier, one parallel is the presence of NHE3 and v-type H⁺ATPase in the proximal tubule of trout (Ivanis et al., 2008) and goldfish (Fehsenfeld and Wood., 2018). Furthermore, in the kidneys of the freshwater Nile tilapia Oreochromis niloticus, H⁺, K⁺ ATPase mRNA is expressed, with apical protein localization only in NKA-rich cells of the collecting tubule, another clear parallel to the mammalian model (Barnawi et al., 2020). Rhbg is expressed at the mRNA level in the kidneys of freshwater rainbow trout (Nawata et al., 2007) and the mangrove rivulus, K. marmoratus (Cooper et al., 2013); in the latter, there was apical co-localization of the protein with NHE3 in distal tubule cells. However, neither Rhbg nor Rhcg transcripts were detected in the kidneys of the pufferfish (Nakada et al., 2007a). Conversely, in both goldfish (Lawrence et al., 2015) and common carp (Wright et al., 2014), Rhcg (two isoforms) and Rhbg are present. Rhcg1a mRNA and protein were upregulated in response to the metabolic acidosis of low environmental pH exposure in the carp, whereas Rhcg1b mRNA responded in the same way in the goldfish. In the carp, Rhbg mRNA decreased during acidosis, though protein expression did not change (Wright et al., 2014). More detailed profiling in the goldfish revealed that Rhbg protein was basolaterally located, Rhcg1b was apical, whereas Rhcg1a was expressed in both membranes, in both proximal and distal tubules (Fehsenfeld and Wood, 2018). Rhcg1b mRNA and protein were both upregulated in the distal tubule, connecting tubule and collecting duct in response to the metabolic acidosis associated with feeding, whereas Rhcg1a mRNA and protein showed a more modest response in the distal tubule only. Rhbg mRNA was upregulated in the connecting tubule and collecting duct, but downregulated in the distal tubule, accompanied by decreased expression of the protein (Fehsenfeld and Wood, 2018). There was also strong expression of Rhcg1 in the kidneys of the zebrafish at both mRNA and protein levels, with localization to the apical membrane in the collecting duct and distal tubule (Nakada et al., 2007a). Please see Figure 3 for a detailed overview of current ideas on renal transport mechanisms in ammonia regulation.

With respect to other putative ammonia transport mechanisms seen in mammals, various HCN genes, including HCN2b are expressed at the mRNA level in the goldfish kidneys, and several (though not HCN2b) respond to feeding and/or HEA exposure (Fehsenfeld and Wood, 2020). NKCC2 (the principal NH₄ ⁺ transporter of the loop of Henle in mammals) is detectable at the mRNA level in the kidneys of seawater acclimated European sea bass, but is downregulated after transfer to freshwater (L'Honoré et al., 2020), so it is unclear whether it is involved in NH₄ ⁺ transport. NH₄ ⁺ substitution on the K⁺ site of renal NKA probably occurs, as in kidneys of the Amazonia pirarucu, NH₄ ⁺ was more effective than K⁺ in activating the activity of this enzyme (Hochachka et al., 1978). Interestingly, in this obligate air-breather, unusually high urinary ammonia excretion was apparently coupled to high urinary HCO₃ ⁻ excretion (Wood et al., 2020). Studies of the specific mechanisms of ammonia regulation in the teleost kidneys are limited to a few number of species to date. Hence, both the localization and specific role of Rh proteins are still under debate and more research is required.

A comprehensive review of phosphate handing by teleosts, with a particular focus on the kidneys, has been presented by Verri and Werner (2019). Unlike ammonia, phosphate is originally obtained mainly from the diet and stored in the bones (Vielma and Lall, 1998). Plasma phosphate levels (millimolar) are much higher than ammonia levels (micromolar). Under control conditions, phosphate filtered at the glomeruli is normally strongly reabsorbed by the tubules in freshwater fish (Vielma and Lall, 1998), but during experimental acidosis, this reabsorption slows and may change over to tubular secretion (Wheatly et al., 1984; Perry et al., 1987; Wood et al., 1999; Lawrence et al., 2015). Tubular secretion may normally dominate in marine teleosts. In the marine winter flounder, Pseudopleuronectes americanus, secretion is driven by a type-II sodium-phosphate cotransporter (NaPi-II), now termed Slc34a2b, that is sorted to the basolateral membrane of the second part of the proximal tubule, whereas in the collecting tubule and collecting duct, the same transporter is located in the apical membranes and appears to drive reabsorption (Gupta and Renfro, 1989; Elger et al., 1998; Verri and Werner, 2019). Low peritubular pH strongly stimulated tubular phosphate secretion in an in vitro culture system (Gupta and Renfro, 1991). Both mRNA and protein for a reabsorptive NaPi-II cotransporter (now slc34a1) were identified in the kidneys of the freshwater rainbow trout, only in the first part of the proximal tubule, with localization to the apical membrane (Sugiura et al., 2003). This transporter is upregulated in response to low dietary phosphate in trout (Coloso et al., 2003; Sugiura and Ferraris, 2004), but we are aware of no information on its response to acid-base disturbance.

Since the mid-2000, investments in larger and technologically more sophisticated recirculating aquaculture systems (RAS) have increased significantly, with the emergence of new production related challenges and risks (Daalsgard et al., 2013; Fossmark et al., 2021; Davidson et al., 2023). Recycling water permits reduced consumption of intake water per unit biomass produced, constant high temperature, use of continuous light and increased salinity which has resulted in increased average smolt size, from 50 to 100 g in 2008 to approximately 300–500 g in 2022, with several companies aiming to produce smolts up to 600–800 g and future prospects of a 1.000 g fish (Ytresøyl et al., 2023). Despite RAS facilities providing greater control of rearing conditions, the industry has experienced increased incidents of disease-related production disorders such as mineral precipitation in the kidneys (Fivelstad et al., 2018; Klykken et al., 2022a; Klykken et al., 2022b) and hemorrhagic smolt syndrome (Nylund et al., 2003; Krasnov et al., 2019), which have negative implications for growth, health and general performance during the land-based phase (Sommerset et al., 2022). These production disorders are likely a consequence of the rapid changes in technology from flow-through systems (FTS) to RAS, and associated with this, altered rearing protocols for more intensive production (changes in photoperiod, respiratory gas levels, temperature, salinity, stocking density, feed composition and use of buffers and alkalizing reagents). These may ultimately change the overall chemistry of the culture water to sub-optimal conditions which in turn may result in physiological perturbations. However, nephrocalcinosis appear to an issue also in FTS systems (Klykken et al., 2022a). Oxygenation of rearing water without removal of accumulated CO₂ combined with high fish density and potentially low water renewal rate may result in increased blood PCO₂ and respiratory acidosis associated with both hypercapnia and hyperoxia (Skov, 2019; McArley et al., 2021). Complex interactions occur between pH, temperature and salinity and the formation of carbonate compounds, often referred to as the carbonate system. This is normally affected by the partial pressure of dissolved carbon dioxide (PCO₂), and the concentrations of bicarbonate ion (HCO₃ ⁻), carbonate (CO₃ ^2-) ion, hydrogen ion (H⁺) (usually expressed indirectly as the negative logarithm, pH), dissolved inorganic carbon (DIC) and the total alkalinity (TA).

Moving juvenile fish from FTS to RAS gives a larger degree of control, yet more demanding technology and complex interactions in pH, temperature, alkalinity, oxygen saturation and suspended solids, which in turn have direct consequences on dissolved total CO₂ (DIC), nitrogen, ammonia and nitrate (Daalsgard et al., 2013; Becke et al., 2020; Fossmark et al., 2021; Davidson et al., 2023). In addition, the use of salinity (in feed or water) has implications on kidney function (Hickman and Trump, 1969; Kodzhahinchev et al., 2017) and the cardiovascular system (Olson and Hoagland, 2008; Brijs et al., 2017; Sundell et al., 2018). In humans these systems are intrinsically connected, and the same is probably true in fish. For example, chronic kidney disease (CKD) impacts the cardiovascular system and increases the risk of heart failure (Said and Hernandez, 2014; Mullens et al., 2020). We do not yet know whether chronic long-term nephrocalcinosis has similar impacts on the cardiovascular system in fish.

Nephrocalcinosis, a disease of aquaculture that has been known for almost 50 years (Landhoilt, 1975), is a renal pathology where a mechanistic understanding of the kidney`s function in acid-base regulation will be critical. Nephrocalcinosis is a widespread disorder found in salmonids (Gillespie and Evans, 1979; Harrison and Richards, 1979; Smart et al., 1979; Fivelstad et al., 1999; 2003; 2018; Klykken et al., 2022a; Minarova et al., 2023), Nile tilapia (Oreochromis niloticus; Chen et al., 2001; 2003), sea bass (Dicentrarchus labrax; Arciuli et al., 2015), Atlantic cod (Gadus morhua; Damsgard et al., 2011), spotted wolffish (Anarhichas minor; Foss et al., 2003; Béland et al., 2020), clownfish (Amphiprion ocellaris; Blazer and Wolke, 1983), cobia (Rachycentron canadum; Klosterhoff et al., 2015), southern flounder (Paralichthys lethostigma; Appelgate et al., 2016), Atlantic wolffish (Anarhichas lupus; Béland et al., 2020), longsnout seahorse (Hippocampus reidi Ginsburg; Lewisch, et al., 2013), lumpfish (Cyclopterus lumpus) and ballan wrasse (Labrus bergylta) (Erkinharju et al., 2021), and is listed as one of the most prevalent perturbations in Norwegian salmonid culture, particularly with respect to land-based productions of large smolt and post-smolts (Sommerset et al., 2022). The physiological perturbations underlying precipitation of minerals in the kidneys are probably related to elevated concentrations in the diet or in the water of the constituents involved in forming the precipitate, principally calcium, magnesium, and phosphate (Smart et al., 1979), and/or because changes in the water chemistry cause changes in internal physiology which promote the formation of these “kidney stones”. Bacterial and viral pathogens are probably not the primary cause (Landholt, 1975), but there is a need to characterize the microbiome associated with the precipitates, as well as their precise mineral composition in fish raised in culture waters of different composition and fed diets of different composition. According to Klykken et al. (2022a), there are at least 5 different chemical compositions for kidney stones in Atlantic salmon, involving precipitates of calcium, magnesium, phosphate, and carbonate. The most prevalent was amorphous carbonate apatite (Ca₁₀(PO₄)₆CO₃).

In mammals the carbonate apatite (Ca₁₀(PO₄)₆CO₃) forms and precipitates at a pH ≥ 6.8 indicating that there is a specific pH threshold in the pre-urine/urine that needs to be sustained in order for stones to form or not form (Olszynski et al., 2015). Similarly, struvite (MgNH₄PO₄ 6H₂O) is commonly found in both mammals and salmonids alike (Olzinsky et al., 2015; Klykken et al., 2022a), and in mammals this appears to form and precipitate at pH ≥ 7.2 (Olzinsky et al., 2015). As noted earlier, urine pHs are generally higher in FW fish (similar to blood plasma pH, 7.0—8.0) while the urine pH is more acidic in marine fish, 1.0–2.0 pH units below that of the blood plasma (Hickman and Trump, 1969). This may be one reason why nephrocalcinosis generally has been observed to be reduced when salmonids have been transported to SW (Klykken et al., 2022a). However, little is known about the specific pH threshold for stone formation in the urine of salmonids. More studies are required to better understand changes in urine pH, but the interpretation of such results should be done with caution as pH changes in the final urine may not reflect pH changes locally in the different segments of the nephron where stones likely starts to form. In vitro experiments recreating a similar environment (artificial urine), similar to those performed by Olzinsky et al. (2015), are needed in fish. Environmental changes that increase the urinary pH of salmonids may be a significant risk factor, as higher pH in mammals definitely increased precipitation events of both carbonate apatite and struvite.

The association between nephrocalcinosis and elevated water PCO₂ (hypercapnia) was originally documented by Smart et al. (1979) and has now been confirmed by numerous studies (e.g.,; Fivelstad et al., 1999; 2003; Foss et al., 2003; Hosfeld et al., 2008; Petochi et al., 2011). A recent review (Skov, 2019) outlined effects of environmental hypercapnia in fish culture. Elevated PCO₂ in the production water will initially affect oxygen transport and acid-base regulation, which in turn may result in increased prevalence of nephrocalcinosis. When hypercapnia is combined with hyperoxia, the incidence of nephrocalcinosis increases, presumably due to greater PCO₂ build-up in the fish (Wood and Jackson, 1980; Hobe et al., 1984; Wood, 1991b; Skov, 2019), which causes respiratory acidosis. In turn, the fish compensates by increasing the concentrations of inorganic phosphate, total ammonia, and acidic equivalents in the finally excreted urine, thereby retaining and raising blood plasma HCO₃ ⁻ concentrations to compensate the respiratory acidosis as outlined earlier (Wood and Jackson, 1980; Wheatly et al., 1984; Perry et al., 1987; Wood, 1991b; Wood et al., 1999). Urine pH falls only slightly due to the buffering action of ammonia and phosphate, and urinary calcium and magnesium concentrations increase only slightly (Wheatly et al., 1984; Perry et al., 1987). A recent diagnostic case study reported 5-fold greater PCO₂ and 10-fold greater HCO₃ ⁻ concentration in the blood plasma of rainbow trout that were afflicted with a high incidence of nephrocalcinosis in an aquaculture facility with water hypercapnia and hyperoxia (Minarova et al., 2023). Hence, while HCO₃ ⁻ concentrations may not be high in the finally excreted urine, they may be excessively high in the primary urine formed by glomerular filtration, as well as in the surrounding peritubular blood plasma. This, combined with greatly elevated PCO₂, ammonia and phosphate, and slightly elevated calcium and magnesium concentrations in the urine may be important for mineral precipitates to form and grow. A model, incorporating this idea, is offered in the next section and Figure 5.

Although environmental hypercapnia causes acid-base perturbations and is a well-established risk factor for development of nephrocalcinosis in farmed salmon (Skov, 2019), it is clearly not the only culprit. Atlantic salmon post-smolts exposed to high total CO₂ levels (5–40 mg L^-1; see Skov et al. , 2019, for conversion to PCO₂ units) and brackish water (12 ppt) for 12 weeks did not exhibit any signs of nephrocalcinosis until after 6 weeks in full strength seawater (<5 mg L^-1 CO₂; 34 ppt) (Mota et al., 2019; 2020). Similarly, Good and co-authors (2018) found no evidence of nephrocalcinosis in Atlantic salmon exposed to 20 mg L^-1 CO₂ for 384 days. Salmon in the latter study experienced higher water alkalinity (233–241 mg L^-1) and freshwater conditions, while seawater alkalinity ranged between 116 and 165 mg L^-1 in the study by Mota and co-authors (2019). Both studies used NaHCO₃ buffer to maintain alkalinity. Interestingly, Nile tilapia reared in water buffered with NaHCO₃ exhibited higher plasma HCO₃ ⁻ concentrations and only modest prevalence of nephrocalcinosis whilst tilapia reared in FW buffered by CaCO₃ displayed the opposite response, low plasma HCO₃ ⁻ concentrations and high prevalence of nephrocalcinosis (Chen et al., 2001). Hence, both salinities and the specific composition of the ions in the water may affect the prevalence of nephrocalcinosis differently. Divalent cations (Ca²⁺, Mg²⁺) and the divalent anion (HPO₄ ^2-) in the water and/or feed are probably important risk factors. Elevated Ca²⁺ and Mg²⁺ are often a result of adding seawater to increase rearing salinity (Table 2). Additionally, the use of calcium-containing buffers and/or increased salt load via the feed may have osmoregulatory impacts, including kidney dysfunction. How much of an impact depends on type of ions and concentrations and the manner in which the fish are exposed (i.e., via the feed or via the water).

In mammals, low GFR has been identified as one of several causes underlying the formation of kidney stones (Gillen et al., 2005; Worcester et al., 2006) but also high dietary salt and protein intake (Ferraro et al., 2020). Interestingly, GFR in freshwater acclimated fish is high, producing copious amounts of dilute urine (i.e., high UFR), while GFR in seawater acclimated fish is lower, resulting in low UFR of a urine rich in divalent ions (Hickman, 1968a; 1968b; Takvam et al., 2021a). Based on the mammalian model (Gillen et al., 2005; Worcester et al., 2006; Ferraro et al., 2020), the expectations would be that more salinity (through feed and water) would reduce GFR and be a risk factor for developing nephrocalcinosis. Commercial diets often contain high salt levels (Na⁺, Cl⁻, Ca²⁺, Mg²⁺ and inorganic phosphate). Furthermore, commercial feed given to fish in the freshwater stage usually contains ingredients from marine sources which increases the salt load (Shaw et al., 2011; Phillip et al., 2022). From a biological standpoint, this is unnatural for fish in freshwater. In freshwater, fish usually drink very little. Increased salinity of the water will increase drinking rate, though the threshold for this effect is unclear (Damsgaard et al., 2020), and increased salt content of the food will have the same effect (Pyle et al., 2003). Interestingly in mammals, higher salt intake appears to be a big risk factor while drinking adequate levels of water appears to somewhat offset the risks of developing kidney stones (Ferraro et al., 2020). How many of these mechanisms in the mammalian model that can be translated to the fish model is unclear, but it is noteworthy that nephrocalcinosis in euryhaline salmonids, has been observed in both freshwater and in brackish water. Although brackish water (12–15 ppt) has been pointed out as a potential risk factor (Klykken et al., 2022a), other triggers are in play earlier in the production as farmers report incidences of nephrocalcinosis down to 7 g fish which may point to components in the feed. Future research should address these interactions between salt load (water and feed), urine production rates (GFR and UFR) and link this to water chemistry levels discussed here at various specific life-cycle stages during production. Yet, the proximate mechanism by which nephrocalcinosis develops in teleost fishes remains unknown almost 50 years after its first discovery (Landholt, 1975).

In light of our understanding of both ion and water homeostasis (Takvam et al., 2021a) and acid-base balance (current review), we suggest a new model for this proximate mechanism (Figure 5). Increased risk of nephrocalcinosis may be coming from the increased salt load in feed (commercial diets; high in Mg²⁺, Ca²⁺ and PO₄ ^2-) and the addition of buffers in the water such as calcium hydroxide (Ca(OH)₂) and sodium bicarbonate (NaHCO₃) (most commonly used buffers in RAS) in the early freshwater phase. However, we suggest that the key condition for kidney stone formation may start when fish are experiencing elevated PCO₂ and PO₂ levels in the water of their holding tanks. These fish will be in the condition of a chronic respiratory acidosis where the pH has been compensated by HCO₃ ⁻ retention in the body fluids, which is achieved by excretion of acidic equivalents in the urine (see Figures 1–3 for a recap of the mechanisms). Aquaculture facilities report that nephrocalcinosis often occurs when salmonids held in these conditions are moved to new tanks for grow-out (Sommerset et al., 2022). If water PCO₂ and PO₂ are lower in the new tanks, then PCO₂ in the blood plasma and primary urine will quickly fall while [HCO₃ ⁻] in the plasma and primary urine will stay high for some time (e.g., Hobe et al., 1984; Wheatly et al., 1984; Perry et al., 1987; Perry et al., 2010). This may be the proximate trigger. The pH will quickly rise in the blood plasma and primary urine during this period, creating optimal conditions for nucleation, precipitation, and aggregation events to occur in urine that still has elevated Ca²⁺, Mg²⁺, NH₄ ⁺, HPO₄ ^2- and HCO₃ ⁻ concentrations. As discussed earlier, the kidney stones will form at pHs above 6.8 (carbonate apatite) and 7.2 (struvite), and this process can occur in less than 2 h for the former, somewhat longer for the latter (Olyniszy et al., 2015):

To make matters worse, once stones start to form, this will further increase the pH of the pre-urine, thereby accelerating the rate of aggregation (Olzinsky et al., 2015). Once the stones aggregate locally into larger stones in the nephron tubules, inflammation and cell damage will disrupt multiple transport mechanisms, and create mechanical blockages to flow. Reduced urine flow itself will further accelerate stone formation. Keep in mind that this is still very early in the process, and it may take weeks or months for the damages to be visible through the use of X-ray or macro-scoring (Klykken et al., 2022b). The issues may further continue as producers use more salinity in the water (brackish water; 12–15 ppt), salted feed (even higher in salts than commercial feed) and/or calcium and bicarbonate buffers (only in RAS) which may further increase Ca²⁺, Mg²⁺ and HCO₃ ⁻ levels, having a cumulative effect. Long-term pathophysiological disturbances in ion and water regulation, acid-base regulation, hormone balance, immune response and metabolism will occur, impacting overall fish health and welfare. In the future, better understanding of the urine pH in different environmental settings and pH thresholds for aggregation of kidney stones of different composition will be useful. As these aggregation processes likely start very early, disrupting transport mechanisms in the nephron, an important objective will be to identify key biomarkers involved in the different transport pathways that may help us detect disturbances very early, such that adjustments in O₂ and CO₂ levels, water chemistry, and feed composition can be done early in development in modern aquaculture.

Section-specific profiling of gene and protein expression in all sections of the fish nephron is a promising advance. The detailed localization of the relevant transporters in acid-base, ammonia, phosphate, calcium, and magnesium regulation to the specific segments (proximal I-, proximal II-, distal-, collecting tubule and collecting duct) provides important information. This can be done through either manual separation (Fehsenfeld and Wood, 2018) or laser capture micro-dissection (Madsen et al., 2020). Combining such methods with the application of scanning ion-selective electrode techniques (SIET; Fehsenfeld et al., 2018) and/or electrophysiology (Kurita et al., 2008; Weiner and Verlander, 2011) using various experimental pharmacological inhibitors/activators and pre-treatments should be pursued with a range of teleost species to better understand the function of these different transport pathways. Additionally, single-cell technologies such as RNAseq and transcriptomic profiling of specific nephron regions would be beneficial to better understand transport mechanisms in physiology and during the development of kidney disease (Potter, 2018). These technologies are already being applied in clinical nephropathology in humans where the main drawbacks at this point are high costs per sample and analytical complexity (Stewart and Clatworthy, 2020).

Advances in immunochemical and molecular approaches in the past decade have been accompanied by a decline in the use of more traditional physiological methods, approaches that give a more holistic understanding of kidney function in fish at the whole organism level. Only by utilizing catheterization techniques can you collect urine, measure urine flow rate (UFR) and glomerular filtration rate (GFR), whereby the researcher can determine the rates of excretion, secretion, reabsorption, and plasma clearance of different solutes (e.g., acid-base equivalents, ammonia, phosphate, etc.) in fish (Wood and Patrick, 1994). This becomes a very powerful approach when combined with either section-specific mRNA localization and expression (in situ hybridization) and/or protein localization (immunohistochemistry and Western blot), and/or SIET analysis of transport rates and/or single-cell technologies, all of which are useful when finding the location of the target solute carrier or channel. These can often also identify whether the transporter is located on the apical or basolateral membrane of the tubule cell. Several of the transporters addressed in the current review are likely located in the same cell(s). Thus, co-localization of transporters/channels is another approach that would deepen our understanding of the co-operative actions these have in dealing with acid-base, ammonia and phosphate regulation. We now know, for example, that Rh proteins, Na⁺ and H⁺ transport pathways, and carbonic anhydrase are organized into a functional metabolon in the freshwater gill (Wright and Wood, 2009; Ito et al., 2013). The same may be true for these and other transport systems in the fish kidneys (Takvam et al., 2021a). In addtion, a better understanding of paracellular transport (tight junctions proteins; claudins) in conjuction to the trancellular transport (adressed in this review) and their role in acid-base balance, which convincingly have been demonstrated to play an important role in the goldfish kidneys (Figure 3; Fehsenfeld et al., 2019). Future studies should include the individual and collective roles of paracellular and trancellular transport pathways in the fish kidneys. In the future these integrative approaches should be combined in order to better understand complex issues in fish. For example, formation of kidney stones (nephrocalcinosis) is a major issue in intensive aquaculture but so far approaches have usually focused on diagnostic tools (histopathology, blood chemistry, X-ray analyses e.g., Harrison and Richards, 1979; Klykken et al., 2022a; Klykken et al., 2022b; Minarova et al., 2023) rather than on understanding the underlying molecular and physiological mechanisms or the consequences of the disease for normal renal function. While such diagnostic approaches can provide very useful tools to detect the disease, they bring us no closer to finding the root cause of the problem, which requires more integrative understanding on how these transport systems work in the kidneys. The collection of extensive water chemistry and feed composition data in combination with this integrative understanding would be valuable towards solving the problem.

Research on mammalian models is more extensive and often better funded due to the many underlying pathological conditions of renal physiology in humans. Fish researchers can learn a great deal from following the mammalian kidney literature as highlighted in several previous sections in this review. However, benefits may also go in the opposite direction: the ability of euryhaline fishes to rapidly adapt to very different environments, and the plasticity of their kidneys where drastic changes in filtration and flow rates (GFR and UFR) as well as substantial shifts (upregulation, downregulation, and net reversal of direction) in their transport systems may provide a powerful experimental model. For example, human kidney disease is often related to reductions in glomerular filtration rates and changes in transport mechanisms. Normal teleost kidney function in freshwater (high GFRs and UFRs) may be a proxy for normal function of the human kidneys, whereas normal teleost kidney function in seawater (low GFRs and UFRs) may be a proxy for abnormal function of the human kidneys. The ability of many teleost fishes to acclimate to high ammonia, hypercapnic, hyperoxic, hypoxic, acidic, and alkaline environments is remarkable, and renal acid-base regulatory function plays a key role in these adjustments. Human diseases ranging from obstructed breathing to liver failure and diabetic coma present some of these same physiological impacts on the human kidneys. Indeed, most of the protein domains related to kidney diseases in humans are present in fish which is why the zebrafish is one of the most popular experimental models for research for human disease (e.g., Bradford et al., 2017). Collectively, increasing our knowledge on the kidneys both at a molecular and physiological levels will be beneficial in addressing these issues in the future as kidney disease is a large and ongoing problem in both mammals and fish.