Functional significance and physiological regulation of essential trace metals in fish

Essential trace metals such as iron, copper, zinc and manganese are vital components for a multitude of biochemical reactions and act as cofactors for various enzymes. They can be found in proteins such as the oxygen-transporting hemoproteins (iron as a cofactor), iron–sulfur clusters involved in DNA synthesis, cytochrome c oxidase of the electron transport chain (copper) and superoxide dismutase (SOD; copper, zinc, manganese) of the antioxidant defense system (Dalziel et al., 2006; de Souza and Bonilla-Rodriguez, 2007; Harris, 1992; Puig et al., 2017). In fish, trace metal deficiency results in impairment to health and physiological performance (Clearwater et al., 2002; Makwinja and Geremew, 2020; Song et al., 2017). On the contrary, these essential metals can become detrimental when in excess, often as a result of increased oxidative stress and induced pathophysiological conditions (Wood et al., 2011). Hence, it is important to tightly regulate and maintain trace metal levels within physiologically safe concentrations.

In adult fish, the acquisition of trace metals occurs via the gill and the gastrointestine (Bury and Glover, 2003). The precise molecular mechanisms regulating the epithelial transport and metabolism of essential metals in fish have remained an active research question, as most of our current understanding is primarily based on mammalian studies. Despite this knowledge gap, recent research has provided new information on the biological significance, physiological regulation and functional involvement of various metal regulatory proteins in fish. These research efforts have demonstrated the complex nature of the regulatory and compensatory mechanisms involved in the maintenance of trace metal balance in fish. Additionally, several essential trace metals have been shown to interact at multiple biological levels (e.g. metal transporters, metal-binding proteins, metal-regulatory signalling pathways). Therefore, an alteration in the homeostasis of one trace metal may have the potential to affect the regulation of several other metals. Moreover, a variety of factors have been shown to influence the way fish regulate metal metabolism and respond to changes in metal availability. These factors include, for example, developmental stage, dietary composition and environmental perturbations (Bury and Glover, 2003; Chandrapalan and Kwong, 2020; Glover et al., 2016; Long et al., 2015).

The uptake and metabolism of essential trace metals in fish have been previously reviewed in several articles and book chapters (Bury and Glover, 2003; Wood et al., 2011; Zhao et al., 2014). In the present Review, we provide an overview of the current state of knowledge on the physiological function of essential trace metals in fish, with a focus on iron, copper, zinc and manganese. We also discuss the interactions between metals (e.g. epithelial transport, transporter expression, cellular and systemic handling), the mechanisms underlying their regulation and recent progress in these topics. Finally, we identify knowledge gaps and discuss potential future research directions.

Iron, copper, zinc and manganese are found in an array of metalloproteins and are cofactors for enzymes essential in various physiological functions, including oxygen transport, antioxidant activity, metal storage, and metabolism and redox balance (Table 1). Before the onset of exogenous feeding, fish embryos obtain nutrients, including essential metals, primarily from the maternally derived yolk stores via the yolk syncytial layer (Carvalho and Heisenberg, 2010; Fraenkel et al., 2005; Riggio et al., 2003; Thomason et al., 2017). Parental deficiency in essential metals such as zinc can have adverse effects on the offspring, including elevated mortality, decreased activity and possible alterations in epigenetic regulators (DNA methyltransferases) (Beaver et al., 2017). In vertebrates, iron is known to be essential for the development of the brain, including neurocognitive development and neurotransmitter metabolism (McCann et al., 2020; Wang et al., 2019). In zebrafish (Danio rerio) embryos, both high levels of iron and copper are deposited in the yolk syncytial layer (Bourassa et al., 2014). High levels of iron are also present in the brain and tail of developing zebrafish (Bourassa et al., 2014). The head also appeared to be the major region of iron deposition in larval zebrafish exposed to iron, leading to a disruption in neurobehavioral performance (Hassan and Kwong, 2020). Similarly, dysregulation of manganese balance in the brain is found to adversely affect neurosensory function in larval zebrafish (Bakthavatsalam et al., 2014; Xia et al., 2017). These results suggest that the maintenance of trace metal balance in the brain is of critical importance during periods of neurodevelopment. Likewise, a few previous studies have shown that zinc is involved in embryogenesis in fish. For example, morpholino-mediated gene knockdown of the zinc transporter ZIP7 resulted in morphological defects and reduced zinc concentrations in the brain, eyes and gills of larval zebrafish (Yan et al., 2012). Additionally, zinc was found to be highly deposited in the hatching gland cells of larval zebrafish, and knockdown of ZIP10 resulted in an impairment in the development of the hatching gland, leading to failure in hatching (Muraina et al., 2020). These results indicate the essential role of zinc and specific zinc transporters in the early development of fish.

Recent studies have shown that iron is involved in bone formation signalling pathways and that feeding with an iron-enriched diet can increase the standard body length of developing zebrafish (Bo et al., 2016; Chandrapalan and Kwong, 2020). Similarly, copper is found to be essential for notochord development in zebrafish, primarily owing to its role in the enzyme lysyl oxidase, which promotes elastin–collagen crosslinking in the notochord (Gansner et al., 2007; Mendelsohn et al., 2006). Copper is also important in the formation of pigmentation in zebrafish through its involvement in the copper-dependent melanogenic enzyme tyrosinase (Ishizaki et al., 2010). Additionally, supplementation of diets with manganese and zinc has been shown to improve survival, bone mineralization and skeletal formation in the post-larvae of Senegalese sole (Solea senegalensis) (Viegas et al., 2021). In another study with post-smolt Atlantic salmon (Salmo salar), however, a change in dietary manganese level (5–65 mg kg⁻¹) was found to have no effect on specific growth rate over an 8-week experimental period (Antony Jesu Prabhu et al., 2019). In addition to possible species-specific differences in dietary metal requirements, differences in the developmental stages and dietary composition used across studies may also account for some of these mixed results.

Current understanding of the nutritional requirement for most essential metals is limited to a few fishes, particularly commercially important taxa such as carps, yellow catfish, groupers and salmonids (Chanda et al., 2015; Davis and Gatlin, 1996; Watanabe et al., 1997). For example, a range of 30–170 mg iron, 15–40 mg zinc, 2–20 mg manganese and 1–5 mg copper (per kg of diet) is proposed to be the minimal requirement for the maintenance of fish health in aquaculture practice (Watanabe et al., 1997). However, it should be noted that most of these studies were conducted with metal salts added in formulated diets, whereas biologically incorporated metals such as those found in prey and natural fish feeds are generally more bioavailable (Deforest and Meyer, 2015; Mebane et al., 2020). Therefore, the actual nutritional requirements for dietary metals in feral fish are likely to be lower than those estimated using formulated diets. Assessing the dietary requirements of metals with natural live diets would be invaluable to better understand how fish regulate metal absorption and balance during periods of metal deficiency or overload.

In vertebrates, iron is essential in the synthesis of oxygen transport proteins (e.g. hemoglobin and myoglobin) and enzymes involved in electron transfer and oxidation–reduction (Ganz, 2013). Manganese and copper are also known to be involved in glucose metabolism and mitochondrial functioning (Baker et al., 2017; Li and Yang, 2018). Therefore, these metals play an essential role in energy metabolism. A few studies have demonstrated that iron and copper status are also linked to the energetic parameters of fish. For example, lower critical swim speeds (U_crit) and reduced maximal oxygen uptake (V̇_O2,max) values were observed in rainbow trout (Oncorhynchus mykiss) experiencing anemia (i.e. through blood removal) (Gallaugher et al., 1995). In yellow catfish (Pelteobagrus fulvidraco), changes in dietary copper levels were found to alter lipogenic enzyme activities and lipid metabolism (Chen et al., 2015). Additionally, dietary supplementation of iron or copper was found to improve hematological parameters in Siberian sturgeon (Acipenser baerii) and stinging catfish (Heteropneustes fossilis), including increases in hemoglobin content, erythrocyte count and hematocrit value (Moazenzadeh et al., 2020; Zafar and Khan, 2020). These enhancements are thought to increase oxygen transport capabilities in fish.

Hepcidin (previously termed as liver-expressed antimicrobial peptide; LEAP-1) is a peptide hormone that regulates intestinal iron uptake and is associated with innate immune signalling (Drakesmith and Prentice, 2012). The first study that demonstrated the potential role of hepcidin in the immune response of fish was conducted in white bass (Morone chrysops), which showed that an infection with a pathogen in this fish species resulted in a 4500-fold induction of hepcidin mRNA expression (Shike et al., 2002). Since this pioneering study, increasing evidence has indicated a close association between hepcidin and iron status in immune function across many fish species (Cuesta et al., 2008; Shi and Camus, 2006; Tarifeño-Saldivia et al., 2018). Nevertheless, how hepcidin orchestrates its dual functional roles in iron regulation and immune response in fish has remained unclear. In zebrafish, it was found that intraperitoneal injection of bacterial DNA can stimulate hepatic hepcidin expression and reduce serum iron levels (Jiang et al., 2017). The shift from circulating iron to intracellular storage was suggested to reduce the availability of iron to invading microorganisms, thereby enhancing antibacterial immunity (Drakesmith and Prentice, 2012). Several studies have also shown that the immune response of fish can be influenced by dietary metals. For example, iron deficiency has been shown to impair the immune function of grass carp (Ctenopharyngodon idella) by reducing the content of immunoglobulin and the transcript abundance of various anti-inflammatory cytokines (Guo et al., 2018a). Juvenile Siberian sturgeon fed on a copper-enriched diet also exhibit an increase and a decrease in the levels of neutrophil and lymphocyte, respectively (Moazenzadeh et al., 2020). In mammals, zinc is involved in various aspects of the immune function, including activation of immune cell signalling and stimulation of the development of immune cells (Prasad, 2008; Wessels et al., 2017). A few previous studies have also shown that zinc promotes immune functioning in fish. For example, feeding of iridescent shark (Pangasianodon hypophthalmus) with a zinc-enriched diet has been found to enhance the immune-hematological status and improve their survival rate after pathogen infection (Kumar et al., 2017). Likewise, grass carp (Ctenopharyngodon idella) fed on a zinc-deficient diet have been shown to exhibit an impairement in enteristis resistance (Song et al., 2017). The role of manganese in immune response in fish is less understood, but exposure to high levels of waterborne manganese has been shown to alter the number of immune cells and the level of antimicrobial enzymes (Aliko et al., 2018; Do et al., 2019; Zafar and Khan, 2019).

Increasing evidence has also suggested that the iron-binding proteins ferritin and transferrin are responsive to bacterial infection in fish. For example, in rock cod (Eleginops maclovinus), bacterial infection increases the mRNA abundance of the ferritin-H subunit (Martínez et al., 2017). In Japanese flounder (Paralichthys olivaceus) and rock bream (Oplegnathus fasciatus), bacterial infection increases the mRNA expression of the ferritin-M subunit (Elvitigala et al., 2013; Wang and Sun, 2015). Additionally, both ferritin and transferrin expression are modulated in sea bass (Dicentrarchus labrax) during bacterial infection, with the levels of ferritin increased in both the liver and the brain (Neves et al., 2009). An increase in transferrin expression was also observed in common carp (Cyprinus carpio) during blood parasite infection (Kamińska-Gibas et al., 2020). The induction of iron-binding proteins during infection is proposed to be an acute inflammatory response to sequester iron and thereby reduce its availability for microbial or pathogen proliferation in the host (Ong et al., 2006). Alternatively, when excess iron is accumulated in macrophages, it can result in the production of free radicals via the Fenton reaction and subsequently aid in the elimination of the intruding pathogen (Tarifeño-Saldivia et al., 2018).

During early developmental stages, before the gills and the gastrointestinal tract (GIT) are fully developed, the uptake of essential trace metals from the environment can occur through the skin. Several studies have shown that larval fish can acquire and accumulate essential metals from the water (Guo et al., 2016; Hassan and Kwong, 2020; Puar et al., 2020). However, the precise pathways for their uptake remain unclear. A few previous studies with the zebrafish model have shown that several specific metal transporters, including those mediating the uptake of iron or zinc, do not appear to be expressed in the skin during early development (Donovan et al., 2002; Haller et al., 2018; Puar et al., 2020). Nevertheless, the epithelial calcium (Ca²⁺) channel (ECaC), which is expressed in the skin of larval fish (Kwong et al., 2014; Lin and Hwang, 2016; Pan et al., 2005), can transport not only calcium but also iron and zinc (Qiu and Hogstrand, 2004). In contrast, the uptake of copper has been found to be sodium sensitive and is proposed to occur via sodium transporting channels (Grosell and Wood, 2002; Handy et al., 2002). Although the epithelial sodium channels (ENaCs) were considered a potential candidate in previous studies, current genomic data suggest that ENaCs are lost in the actinopterygian lineage and thus absent in teleost fish (Wichmann and Althaus, 2020). Instead, it has been proposed that acid-sensing ion channels (ASICs), which have been documented in other vertebrates for copper transport, are responsible for the sodium-dependent copper transport in fish. Indeed, ASICs are expressed in several fish species including rainbow trout and zebrafish (Dymowska et al., 2015; Zimmer et al., 2018); however, whether ASICs can in fact transport copper awaits investigation. Overall, it seems possible that the waterborne uptake of trace metals during the early developmental stages is mediated by ion transporters (i.e. via ionocytes) expressed in the skin. Interestingly, the skin is also suggested to be an important site for the absorption of iron in adult Pacific hagfish (Eptatretus stoutii) (Glover et al., 2016), although the molecular pathways that mediate iron uptake in this species are yet to be determined.

In juvenile and adult fish, the acquisition of essential trace metals occurs via the gill and the GIT. Despite the capability of the gill to absorb trace metals from water, the diet is thought to be the major source of essential metals in fish (Bury and Glover, 2003; Watanabe et al., 1997). This is because in clean environments, the concentration of essential metals is generally quite low and at circumneutral pH, some of these metals such as iron may precipitate and form insoluble metal oxides and hydroxides. A previous study with the marine teleost Psetta maxima and elasmobranch Scyliorhinus canicular has shown that over 60% of the uptake of essential metals, including manganese, cobalt and zinc, occurs via the dietary route (Mathews and Fisher, 2009). In the GIT, the predominant region for the absorption of essential metals varies for different metals and among species. For example, iron absorption occurs primarily in the posterior region of the GIT in marine flounder (Platichthys flesus) (Bury et al., 2001) as opposed to the anterior region in freshwater rainbow trout (Kwong and Niyogi, 2009). Interestingly, the stomach is the major region for the absorption of copper in rainbow trout (Nadella et al., 2011). The mechanisms for the absorption of copper also appear to be different between the stomach and intestine, with the latter partially mediated by sodium- or iron-sensitive pathways (Nadella et al., 2007, 2011). In African walking catfish (Clarias gariepinus), copper uptake occurs predominantly in the distal intestine and is dependent on a chloride gradient (Handy et al., 2000). Furthermore, the levels of dietary trace metals and major ions (e.g. sodium, calcium, magnesium and potassium) have been shown to influence the acquisition of metals from water. For example, waterborne iron uptake was found to increase in zebrafish fed on an iron-deficient diet (Cooper et al., 2006a). Marine medaka (Oryzias melastigma) fed on a low-iron diet also exhibited an increase in waterborne iron uptake (Wang and Wang, 2016). Interestingly, in rainbow trout, it has been shown that increasing the levels of dietary sodium reduced copper acquisition via the gills (Pyle et al., 2003). Similarly, elevated dietary calcium levels have been found to reduce the branchial absorption of zinc in rainbow trout (Niyogi and Wood, 2006). These findings suggest that the levels of trace metals and major ions in the diet can modulate trace metal uptake at the gill. How fish coordinate metal uptake between the gill and the GIT requires further investigation.

Iron in the diet primarily exists in the ferric form (Fe³⁺) and is first converted to the ferrous form (Fe²⁺) by ferric reductase (also called duodenal cytochrome b; Dcytb) before its absorption from the intestine via the divalent metal transporter-1 (DMT1). According to the NCBI database, the gene encoding for ferric reductase has been identified in various fish species, such as rainbow trout, medaka, Indian glassy fish (Parambassis ranga) and zebrafish. Additionally, the activity of ferric reductase has been detected in rainbow trout and the air-breathing fish Anabas testudineus (Carriquiriborde et al., 2003; Rejitha and Peter, 2013). In mammals, ferric reductase is also proposed to mediate the reduction of cupric ion (Cu²⁺) to cuprous ion (Cu⁺) before its absorption by the copper transporter CTR1 (Wyman et al., 2008). In fish, DMT1 is thought to be the major transporter for the acquisition of iron from the diet and water (Bury, 2003). Notably, zebrafish mutants that produced truncated dmt1 mRNA developed hypochromic and microcytic anemia (Donovan et al., 2002), suggesting its essential role in maintaining iron homeostasis. The gene encoding for DMT1 has been detected in early life stages and various tissues in adults (Chandrapalan and Kwong, 2020; Craig et al., 2008; Donovan et al., 2002; Hassan and Kwong, 2020; Wang and Wang, 2016). Interestingly, both an increase and a decrease in dmt1 expression levels following iron exposure have been reported (Chandrapalan and Kwong, 2020; Cooper et al., 2006a; Craig et al., 2008; Hassan and Kwong, 2020; Kwong et al., 2013; Wang and Wang, 2016). The changes in dmt1 expression could be indirect effects induced by iron because the iron transport pathways are also involved in the absorption and homeostatic regulation of several other essential metals (e.g. compensation for other metals; discussed below).

In mammals, DMT1 is capable of transporting not only iron but also other divalent metals such as manganese and copper (Arredondo et al., 2003; Garrick et al., 2003). However, a few recent studies have argued that DMT1 is not physiologically important in the intestinal absorption of copper or manganese (Illing et al., 2012; Shawki et al., 2015). In fish, direct evidence that shows the substrate selectivity of piscine DMT1 is limited to iron and the nonessential metal cadmium (Cooper et al., 2007). Nevertheless, many studies have demonstrated the interactions between iron and other essential metals, such as copper, manganese and zinc (Chandrapalan and Kwong, 2020; Craig et al., 2008; Hassan and Kwong, 2020; Kwong and Niyogi, 2009; Nadella et al., 2007). These interactions were observed either at the uptake surfaces or in regulating internal metal balance. In mammals, copper and zinc were also found to affect DMT1 expression (Iyengar et al., 2009; Tennant et al., 2002; Yamaji et al., 2001), whereas calcium appeared to be a non-competitive inhibitor for iron absorption in the enterocytes (Shawki and Mackenzie, 2010; Thompson et al., 2010). In addition to DMT1, the apical uptake of manganese and copper can also be mediated by other transporters, such as the zinc transporters ZIP14 and CTR1 (Mackenzie et al., 2004; Tuschl et al., 2016; Zhao et al., 2014). In the stomach of rainbow trout, CTR1 is proposed to be the major transporter responsible for the absorption of copper (Nadella et al., 2011).

The basolateral extrusion of iron and copper from absorptive cells is known to be mediated by ferroportin and copper transporting ATPases (e.g. ATP7a), respectively (Donovan et al., 2000; Mendelsohn et al., 2006). Ferroportin is also thought to mediate the extrusion of manganese (Madejczyk and Ballatori, 2017; Seo and Wessling-Resnick, 2015). In a human colorectal adenocarcinoma cell line, Caco-2, copper has been shown to induce the expression of ferroportin (Tennant et al., 2002). These findings indicate that in addition to the apical surfaces (e.g. via DMT1), metal–metal interactions may also occur at the basolateral membrane, via either competition for export or modulation in the expression of metal export proteins. Advances in gene editing also provide a useful tool to reveal the functional significance of metal transporters and related proteins (Table 2). A recent study with zebrafish larvae has demonstrated that CRISPR/Cas9-mediated knockout of the zinc transporter ZnT10, a putative manganese exporter, results in manganese overload and hypermanganesaemia-related phenotypes, which can be partially rescued by iron treatment (Xia et al., 2017). This finding suggests that the interaction between iron and manganese is likely to be competitive in nature. Moreover, feeding with different levels of dietary iron can modulate the whole-body content of zinc and mRNA expression levels of the zinc transporters ZIP8 and ZIP14 in developing zebrafish (Chandrapalan and Kwong, 2020). Interestingly, zebrafish experiencing a knockout of ZIP14, which is a major transporter for zinc, but also capable of transporting manganese and iron, were found to increase manganese accumulation without affecting zinc and iron balance in their body (Tuschl et al., 2016). This result suggests that ZIP14 is important in manganese clearance in the kidney and that there are additional transport pathways to regulate and maintain the homeostasis of zinc and iron. To conclude, trace metal interactions may occur during the uptake process, potentially via direct metal–metal competition for absorption and/or modulation in the expression of metal transport proteins (Fig. 1).

In vertebrates, including fish, ferritin and transferrin regulate intracellular iron storage and iron delivery to various tissues, respectively (Bury et al., 2011). Their levels and binding capacities are known to be induced by iron exposure in fish (Carriquiriborde et al., 2003; Kwong et al., 2013). Recent advances with size or subcellular fractionation have provided some new information on the distribution of trace metals among cytosolic proteins within a cell. For example, in the liver of European chub (Squalius cephalus), it was suggested that copper was predominantly associated with metallothionein and, to a lesser extent, carbonic anhydrase and SOD (Krasnići et al., 2013). Iron appeared to be primarily bound to ferritin, whereas manganese was associated with albumin and transferrin (Krasnići et al., 2013). These results suggest that in addition to their interactions during absorption (see above), trace metals may also compete for binding sites within a cell or in circulation.

The delivery of copper in the blood is primarily mediated by ceruloplasmin (Das and Sahoo, 2018). In mammals, ceruloplasmin is also known to play an important role in systemic iron handling by oxidizing ferrous iron into ferric iron to facilitate its binding to transferrin (Hellman and Gitlin, 2002). Ceruloplasmin has been identified in various fish species, including icefish (Chionodraco rastrospinosus), Indian carp (Aeromonas hydrophila) and channel catfish (Ictalurus punctatus) (Liu et al., 2011; Sahoo et al., 2013; Scudiero et al., 2007). The mRNA expression level of ceruloplasmin has also been found to increase in channel catfish experiencing bacterial infection and iron overload (Liu et al., 2011). Copper is transported into cells by copper transporters (e.g. CTR1 and DMT1), and within the cell, virtually all free copper is sequestered by metal binding proteins including the copper chaperones (Kaplan and Maryon, 2016; Rae et al., 1999). The copper chaperones perform the dual functions of intracellular copper storage and trafficking. These proteins include the antioxidant 1 copper chaperone (ATOX1) for transporting copper to copper-ATPase in the Golgi network, the cytoplasmic copper chaperone (also called copper chaperone for superoxide dismutase; CCS) for delivering copper to SOD, and cytochrome c oxidase copper chaperone (i.e. COX17) for delivering copper to cytochrome c oxidase in the mitochrondria (Cheng et al., 2017; Kwok and Chan, 2019; Leung et al., 2014).

Our current understanding of the signalling pathways (e.g. initiation of transcription, post-translational modifications) involved in trace metal regulation is primarily derived from studies from mammalian counterparts. Several iron transport- and metabolism-related proteins, including DMT1, ferroportin, ferritin and transferrin receptors, can be regulated post-transcriptionally. These proteins have an iron-responsive element (IRE) located either on the 3′- or 5′-untranslated region of the mRNA (Eisenstein, 2000). This region allows for the binding of iron regulatory proteins (IRPs) to the IRE, which can alter mRNA stability and translation of these proteins. IRPs can act as either a translational activator (e.g. binding to the 3′-end of an mRNA to protect against endonuclease cleavage) or a translational inhibitor (e.g. binding to the 5′-end of an mRNA to block the translation process) (Pantopoulos, 2004) (Fig. 2). Another mechanism to regulate the expression of metal storage proteins is the activation of the metal-responsive element (i.e. DNA-binding motif) by the zinc finger metal-responsive transcription factor (i.e. MTF1). Such induction enhances the transcription of the metal-binding protein metallothionein (MT) to store excess zinc and copper (Chen et al., 2020; Tapia et al., 2004). The induction of MT by excess zinc and copper and its role in sequestering free metals are well documented in fish (Bervoets et al., 2013; De Boeck et al., 2003b; Hashemi et al., 2008; McDonald et al., 2021; Wang et al., 2014). However, the mechanism underpinning MTF1–MT activiation by metals is not completely understood. Mulitple models for MTF1–MT activiation have been proposed, including: (i) direct activation of MTF1 by excess zinc in the cytosol, (ii) displacement of zinc from MT by other metals, (iii) stimulation of MTF1 phosphorylation and thereby its transactivation activity, and (iv) promotion of MTF1 translocation to the nucleus through conformational changes and uncovering of the nuclear localization motif (LaRochelle et al., 2001; Smirnova et al., 2000; Wang et al., 2014).

Certain metals can also be regulated by hormones. For example, hepcidin is a peptide hormone that acts as a central regulator of iron homeostasis. It functions by promoting the internalization and degradation of ferroportin at the basolateral membrane during iron overload (Nemeth et al., 2004). In both mammals and fish, hepcidin is primarily synthesized in the liver (Chandrapalan and Kwong, 2020; Nemeth et al., 2004). Several studies have suggested that the mRNA expression of hepcidin is positively regulated by iron in fish (e.g. Neves et al., 2017; Wang and Wang, 2016). Likewise, microinjection of hepcidin cRNA decreases whole-body iron content in zebrafish experiencing iron overload (Jiang et al., 2019). Moreover, intraperitoneal injection of adrenaline or triiodothyronine (T₃) has been found to reduce the activity of ferric reductase in various organs of climbing perch (Anabas testudineus), and the effects appeared to be dependent on the feeding status of fish (Rejitha and Peter, 2013). Several previous studies have also shown that trace metal exposure can affect the production of the stress hormone cortisol in fish (Gagnon et al., 2006; Tellis et al., 2012); however, whether cortisol can directly influence metal transport has not been fully characterized. Interestingly, cortisol is reported to reduce copper absorption by common carp (De Boeck et al., 2003a), whereas it stimulates zinc uptake in rainbow trout (Bury et al., 2008). These results suggest that cortisol may have different effects on the uptake of different metals. Alternatively, cortisol is known to promote major ion uptake by increasing the expression of ion transporters and the density of ion-transporting cells, including those that facilitate the uptake of calcium and sodium (Guh et al., 2015; Kwong et al., 2016a). These pathways may also mediate the uptake of essential trace metals (see above). Additionally, there is emerging evidence from mammalian studies that the status of essential metals in the body may affect various endocrine systems, such as iron in gastrin activity and copper in oxytocin signalling (Stevenson et al., 2019). The mechanism underlying the interplay between essential trace metals and hormonal status in fish is a relatively poorly studied area and requires further characterization.

The regulation of essential trace metals can be influenced by many environmental/abiotic factors. The effects of water chemistry (e.g. water pH, hardness, dissolved organic matter) on the branchial uptake of trace metals are primarily due to its influence on metal bioavailability and are often discussed in the context of toxicology. The toxicology of trace metals is beyond the scope of the present Review, and we refer the readers to other articles (e.g. Wood et al., 2011). In this section, we provide an overview of our understanding of the influence of temperature, dissolved oxygen content and GIT chemistry/diet composition on the transport and metabolism of essential metals in fish.

The temperature in aquatic environments fluctuates throughout the day according to the day/night cycle and seasonally, and could also be affected by climate change. Many studies have demonstrated that trace metal absorption is positively related to water temperature in fish. For example, zebrafish conditioned to elevated water temperature exhibited an increase in copper accumulation (Pilehvar et al., 2019). Similarly, zinc uptake was found to increase in rainbow trout acclimated to increased water temperature (Glover et al., 2003). In common carp and turbot (Scophthalmus maximus), the assimilation efficiency of dietary zinc was shown to increase with increasing water temperature (Pouil et al., 2018; Van Campenhout et al., 2007). One possible reason for the increased metal absorption at higher temperatures could be associated with the general increase in metabolic activities and ventilation rates. For instance, when metabolic parameters were compared in common carp at 10°C and 20°C, standard metabolic rate, maximum metabolic rate and aerobic scope were elevated at the higher temperature (Pillet et al., 2021). In female zebrafish, increases in water temperature induce the hepatic expression of metallothioneins, but not ctr1 nor zip8 (Guo et al., 2018b). In contrast, a transient increase in the branchial expression of ctr1 was observed in common carp during acclimation to increased water temperature (Castaldo et al., 2021). In another study with common carp, however, changes in water temperature were found to have no effect on copper and zinc accumulation in the gill (Pillet et al., 2021). Therefore, it seems likely that changes in water temperature may affect the expression of metal transporters in a tissue-specific manner and that different species may have different capacities to regulate and maintain trace metal balance during acclimation to elevated water temperature. Additionally, the different responses to temperature stress can also be attributed to differences in acclimation duration and the magnitude of temperature changes. Interestingly, cold stress was also found to increase the mRNA abundance of ceruloplasmin, transferrin and metalloreductase (i.e. steap3; reduces both Fe³⁺ and Cu²⁺ ions) in zebrafish larvae (Long et al., 2015). These proteins facilitate metal transport in the circulatory system and endosomal entry of metals into cells; however, the physiological significance of these increases by cold stress requires further study. Finally, the temperature can have indirect effects on the absorption of trace metals from the GIT. For example, an increase in temperature is known to affect GIT physiology by altering phospholipid membrane compositions and intestinal barrier function in fish (Fadhlaoui and Couture, 2016; Sundh et al., 2010).

In mammals, exposure to hypoxia can induce the expression of a variety of iron transport-related proteins through the activation of the hypoxia-inducible factors (HIF) (Renassia and Peyssonnaux, 2019). The increases in iron transport-related proteins (e.g. DMT1, transferrin receptor) by hypoxia are probably due to the increases in iron demand for enhancing oxygen transport capacity (Renassia and Peyssonnaux, 2019). Direct evidence that links the HIF-signalling pathway to micronutrient balance in fish is currently limited to ionic regulation (e.g. calcium) (Kwong et al., 2016b); nevertheless, because HIF signalling is highly conserved among vertebrates, this signalling pathway is also thought to control the expression of iron transport proteins in fish (Pelster and Egg, 2018). In agreement with this hypothesis, a study with larval zebrafish has shown that hypoxia exposure increases mRNA expression of transferrin, metalloreductase and ceruloplasmin (Long et al., 2015). Genes that are involved in oxygen transport and the heme biosynthesis process are also induced by hypoxic treatment (Long et al., 2015). Interestingly, reduction in dissolved oxygen does not affect iron uptake across the skin and GIT of Pacific hagfish, likely because of their tolerance to hypoxic conditions (Glover et al., 2016).

The transport and metabolism of metals are influenced by the chemistry in the GIT. For example, in rainbow trout, the intestinal absorption of copper was found to decrease when luminal pH was increased from 7.4 to 8.0 (Nadella et al., 2007). Similarly, a reduction in intestinal iron absorption was observed when luminal pH was raised to a more alkaline condition (Kwong et al., 2010). In contrast, the presence of the reducing agent ascorbic acid, which promotes the conversion from ferric iron to ferrous iron, was shown to enhance the intestinal uptake of iron in various fish species including rainbow trout, European flounder and gulf toadfish (Opsanus beta) (Bury et al., 2001; Cooper et al., 2006b; Kwong and Niyogi, 2008). These observations are likely a result of the transport properties of DMT1 (a Fe²⁺/H⁺ symporter), which preferentially transports divalent metals and is driven by a proton gradient (Gunshin et al., 1997). In Atlantic salmon, dietary supplementation with long-chain polyunsaturated fatty acids of the n-3 family [eicosapentaenoic acid/docosahexaenoic acid (EPA/DHA)] has been found to reduce hepatic iron content, which appears to be associated with a decrease in the mRNA abundance of transferrin and ferritin in the liver (Rørvik et al., 2003). In rainbow trout, it has also been suggested that copper–histidine and zinc–histidine complexes can be directly absorbed through specific amino acid transporters (Glover and Wood, 2008; Glover et al., 2003). These results indicated that diet composition has a significant impact on trace metal uptake and metabolism in fish.

Iron, copper, zinc and manganese are constituents of an array of metalloproteins and act as cofactors to facilitate various enzymatic reactions. They are also involved in a variety of biological processes, including development, oxygen transport and the immune response. In fish, their uptake may occur through specific metal transporters and via major ion transporters. These trace metals can interact at multiple biological levels from transepithelial transport to intracellular and systemic handling. Their interactions may also influence the function and expression of various metal transport- and metabolism-related proteins. All of this crosstalk may subsequently influence the whole-body homeostasis of trace metals in fish. Recent research efforts have revealed the complex processes involved in the regulation of trace metal homeostasis in fish. However, it is still unclear how the various metal transport pathways in the gill and GIT coordinate to maintain trace metal balance in fish. To date, the molecular machinery in the transport and metabolism of essential trace metals in fish is not completely understood. For example, the functional involvement of specific metal transporters and their cellular/subcellular localization in various organs have not been fully characterized. Additionally, most of our understanding of metal–metal interactions in fish relies primarily on measuring changes in mRNA expression levels following exposure or kinetic analysis through competition study. Direct in vivo evidence showing the functional significance of specific metal transporters in metal metabolism and interactions in fish has remained limited. Among the metals discussed, the regulation of manganese homeostasis is much less understood (e.g. uptake mechanism and cellular/systemic handling) and should be addressed in future studies. Recent advances in functional genetics may prove beneficial to enhancing our knowledge of the molecular pathways involved in trace metal homeostasis (Zimmer et al., 2019). This strategy may also have the potential for the discovery of novel compensatory mechanisms in response to trace metal dysregulation. Additionally, most previous studies on examining the influence of abiotic factors on trace metal homeostasis were primarily focused on the toxicology context, thus key knowledge gaps exist in how environmental stressors such as hypoxia and temperature affect the physiological regulation of metals. Moreover, our understanding of essential trace metal homeostasis is mostly limited to a few fish species, including some commercially important species and traditional model organisms such as zebrafish and rainbow trout. Fish in different habitats likely employ different physiological strategies to maintain trace metal balance, and therefore future research should address how the regulation of metal transport function may differ across species.

We thank Dr Chun Chih Chen for his contributions to the illustrations. We also thank Stefan Galander and the reviewers for their insightful suggestions.