If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Zinc is a ubiquitous biological metal in all living organisms. The spatiotemporal zinc dynamics in cells provide crucial cellular signaling opportunities, but also challenges for intracellular zinc homeostasis with broad disease implications. Zinc transporters play a central role in regulating cellular zinc balance and subcellular zinc distributions. The discoveries of two complementary families of mammalian zinc transporters (ZnTs and ZIPs) in the mid-1990s spurred much speculation on their metal selectivity and cellular functions. After two decades of research, we have arrived at a biochemical description of zinc transport. However, in vitro functions are fundamentally different from those in living cells, where mammalian zinc transporters are directed to specific subcellular locations, engaged in dedicated macromolecular machineries, and connected with diverse cellular processes. Hence, the molecular functions of individual zinc transporters are reshaped and deeply integrated in cells to promote the utilization of zinc chemistry to perform enzymatic reactions, tune cellular responsiveness to pathophysiologic signals, and safeguard cellular homeostasis. At present, the underlying mechanisms driving the functional integration of mammalian zinc transporters are largely unknown. This knowledge gap has motivated a shift of the research focus from in vitro studies of purified zinc transporters to in cell studies of mammalian zinc transporters in the context of their subcellular locations and protein interactions. In this review, we will outline how knowledge of zinc transporters has been accumulated from in-test-tube to in-cell studies, highlighting new insights and paradigm shifts in our understanding of the molecular and cellular basis of mammalian zinc transporter functions.
). While zinc is fundamental to eukaryotic cellular biology, it is a later addition to proteomes after the advent of a geochemical shift of an ancient high-sulfide ocean to the modern oxidizing, sulfate-rich one (
). The release of the sulfide-bound zinc provided zinc bioavailability to prompt a burst in the innovation of protein structures such as zinc fingers with consequences for quickening the rise and diversification of eukaryotes in evolution (
). However, zinc utilization in eukaryotes came at the expense of increasing risk for interference with preexisting cellular machineries evolved earlier. Accordingly, the human body harnesses potentially toxic zinc chemistry in a spatially confined manner to avoid cytotoxicity. At the organ level, zinc is highly enriched in the hippocampus and neocortex region of the brain, prostate gland, and islets of Langerhans of the pancreas (
). These tissues accumulate abundant zinc into secretory pathways and release the stored zinc on demand. At the cellular level, the total cellular zinc content is submillimolar with uneven distributions ranging from a picomolar range of free cytosolic zinc to over 20 mM in specialized secretory vesicles (
). Common to all zinc transporters is the ability of selective zinc binding, but bound zinc ions in ZnTs and ZIPs are transported in opposite directions. ZnTs are efflux transporters responsible for removing excess zinc in the cytoplasm, whereas ZIPs are uptake transporters that replenish cytosolic zinc (Fig. 1, A–C). The complementary functions of ZnTs and ZIPs stabilize the cytosolic zinc concentration around a homeostatic setpoint while enriching zinc in the lumen of specific subcellular compartments to support zinc-dependent cellular processes.
The multisite localizations of zinc transporters in mammalian cells increase the number of protein–protein interactions that drive functional cross talks between zinc transporters and an array of cellular machineries performing diverse cellular processes beyond zinc transport. Hence, mammalian zinc transporters are organized in a functional hierarchy across scales—from the fundamental chemistry of protein structures through subcellular localizations and macromolecular complexes up to cellular processes that constitute the molecular physiology of mammalian cells. Accordingly, this review will begin with in-test-tube studies of isolated zinc transporters to illuminate how zinc coordination and geometric arrangements of binding residues confer zinc selectivity in protein structures and how zinc bindings drive protein structural dynamics to move up or down zinc concentration gradients across the membrane. In-cell studies of zinc transporters further showcase how zinc transporters are directed to distinct subcellular locations and engaged in different protein complexes in response to environmental stimuli. The next level of functional integrations takes place in zinc-dependent biological processes in which individual zinc transporters coordinate networks of protein–protein interactions involved in cellular zinc signaling, endoplasmic reticulum (ER) homeostasis and unfolded protein response, and activation of zinc ectoenzymes in the early secretory pathway. The involvement of mammalian zinc transporters in diverse cellular processes provides insights into their global functions in human pathophysiology. Finally, we will summarize existing knowledge of loss-of-function (LOF) mutations and polymorphisms in human ZnTs and ZIPs and their clinical manifestations in major human diseases.
Zinc selectivity and its structural basis
Cells acquire a variety of transition elements including iron, zinc, copper, manganese, cobalt, nickel, molybdenum, tungsten, chromium, and vanadium. These biological metals may exist in tightly bound forms such as metal-bound cofactors and proteins or nucleic-acid-bound species, or loosely bound forms in association with a diverse heterogeneous buffer. Biomolecules in the intracellular milieu have an extraordinary metal chelation capacity, probably containing an excess of high-affinity binding sites relative to the number of transition metal ions in cells (
), zinc and copper would form the most stable complexes, preferentially accumulating in complexes with biomolecules if metal ions are presented in equal amounts in the test tube. However, in living organisms, metal ions are selectively acquired in cells, where metal transporters override the intrinsic thermodynamic propensity by the selection and compartmentalization of metal ions to enable metal-specific cellular processes. As such, the metal selectivity of zinc transporters is an essential aspect of cell biology, regulating the composition of the intracellular metallome and safeguarding the fidelity of zinc delivery to the right subcellular compartments in the right amount. While ZnTs and ZIPs transport zinc ions in opposite directions by different transport mechanisms, common to these zinc transporters is an overall metal selectivity for zinc. This raises the question as to how different zinc transporters exploit zinc coordination chemistry to select zinc against other similar metal ions.
Zinc has an unusual electron configuration of [Ar]3d10. The completely filled d-orbital renders zinc redox-inert, and its ionic form has a fixed valence state of +2. Zinc is considered as a borderline soft-hard metal, being coordinated by both the sulfur atom of cysteine and nitrogen atom of histidine (soft base ligands) or by carboxylate oxygen atoms of aspartate and glutamate (hard base ligands) (
). Often, water molecules participate in zinc coordination and stabilize zinc ligation in particular conformations. In some cases, the binding of a water molecule to a positively charged zinc center reduces the pKa of water from 15.7 to ∼7, generating a hydroxide ion as a catalytically active species (
). The coordination geometry of each zinc binding site is broadly categorized by coordination number (N) and bond angles into tetrahedral (N = 4), trigonal bipyramidal (N = 5), and octahedral geometries (N = 6). Despite the variability in zinc coordination environments, the affinity toward Zn(II) is usually high in the μM to pM range in diverse zinc metalloproteins (
Zinc transporters are metalloproteins specialized for selective capture and vectorial movements of zinc ions across the membrane barrier. In this two-step molecular process, chemical properties of Zn(II) dictate the composition and geometry of metal-binding sites and their immediate surroundings to afford zinc selectivity. ZnTs and ZIPs are ubiquitous metal efflux and uptake transporters found in bacteria, archaea, and eukaryotes (
). At present, bacterial zinc transporters, YiiP and ZIPB, are two representative ZnT and ZIP homologs for which crystal structures have been determined with bound Zn(II) and/or its isomorphous Cd(II) (Fig. 2, A and B). The metal selectivities of YiiP and ZIPB have also been explicitly determined, providing the experimental basis for correlating metal selectivity with binding site composition and coordination geometry.
YiiP is an integral membrane protein found in the cytoplasmic membrane of Escherichia coli (
). However, the presence of redundant metal transport systems and compensatory metal homeostatic controls could lead to misinformation on metal selectivity. A more rigorous determination of metal selectivity was developed using direct measurements of metal uptake into proteoliposomes mediated by purified and reconstituted YiiP proteins. The spectrum of YiiP metal substrates was profiled by inductively coupled plasma–mass spectrometry (ICP-MS), showing that YiiP transported Zn(II) and Cd(II), but rejected all other transition metal ions in the fourth period (
). Thus, ZIPB shares a common metal selectivity with YiiP. Zinc and cadmium are two group-12 d-block metals in the fourth and fifth period, respectively. They have similar outer electron configurations but vary in their ionic radii. It appears that common features in the electron configurations of Zn(II) and Cd(II) are exploited to confer Zn(II)/Cd(II) selectivity against transition metal ions with different preference for coordination number and geometry. On the other hand, the accommodation of distinct ionic sizes from 0.74 Å for Zn(II) to 0.97 Å for Cd(II) demonstrates considerable fluidity in size selection, making a critical distinction from the size-based selectivity mechanism used by ion channels to discriminate s-block metals, such as potassium and sodium ion (
) (Fig. 2C). Each protomer comprises an N-terminal transmembrane domain (TMD) followed by a C-terminal domain (CTD) that protrudes into the cytoplasm. Three distinct zinc-binding sites were found in each protomer. The intramembranous site, also known as the transport site, is localized to the hydrophobic core of TMD and responsible for Zn(II)/Cd(II) selectivity (
). Side chains of four highly conserved residues (D49, D53, H153, and D157) are projected from two antiparallel transmembrane helices (TM2 and TM5) to form a classic tetrahedral Zn(II)/Cd(II)-binding site (Fig. 2A). Metal-binding analysis by isothermal titration calorimetry revealed a sub-μM range of Zn(II)/Cd(II)-binding affinity (
). H153 in the tetrahedral transport site is the sole proton-titratable residue under the physiological pH range. This residue may act as a proton donor or acceptor depending on Cd(II) coordination.
The structure of ZIPB showed a monomeric transport unit consisting of eight transmembrane helices (TMs) forming a single helix bundle where the first four TMs (TM1 to TM4) are approximately twofold related to the last four TMs (TM5–TM8) (
) (Fig. 2B). These TMs intertwine to embrace a central binuclear metal center within an inner four-helix bundle stabilized by four peripheral TMs and lipid molecules that fill the inter-TM gaps. The binuclear metal center is situated in the hydrophobic core, likely responsible for Zn(II)/Cd(II) selectivity. ZIPB was cocrystallized with Cd(II), and its structure was solved with or without Zn(II) back soaking to partially replace bound Cd(II), yielding two conformations of the binuclear metal center occupied by either two Cd(II) or a Cd(II) and a Zn(II) (
). The metal ions trapped in the binuclear metal center are termed M1 and M2, respectively. They are bridged by one or two carboxylate residues (E181 or E181 + E211) that form bidentate coordination (Fig. 2B). M1 is penta-coordinated while M2 is hexa-coordinated. Their coordination spheres are primarily filled by ligands from residues located within two conserved hexapeptides: “177HNhPEG182” and “207QD/NhPEG212” (h refers to a hydrophobic residue) in TM4 and TM5, respectively. These TMs are kinked by P180 and P210 to properly place multiple ligating residues. Thus, the binuclear metal center is largely nested between TM4 and TM5, but residues from neighboring TMs, such as M99 from TM2 and E239 from TM6, fill up the coordination sphere. Of note, two coordination sites of M1 and one coordination site of M2 can be occupied by water molecules (
). These coordinated water ligands may exist as hydroxide ions to stabilize two closely associated metal ions within 4.5-Å in a low-dielectric-constant environment of the inner membrane.
The crystal structures of YiiP and ZIPB show that zinc coordination chemistry is richly exploited in different coordination spheres from a classical tetrahedral site to a more complex binuclear site with coordination numbers ranging from 4 to 6. Carboxylate anions of aspartate and glutamate are predominantly employed in the zinc transport sites of YiiP and ZIPB (Fig. 1C). The binding of a carboxylate residue offers two alternative binding modes between monodentate and bidentate coordination. A carboxylate can also form coordination bonds with two separate metal ions as a bridging residue. The ability of the carboxylate group to rearrange coordinating ligands gives high flexibility of the zinc coordination sphere while maintaining a constant coordination number (
). As such, Zn(II) and Cd(II) may be accommodated by virtue of their shared preference for common coordination numbers and geometries, despite a large difference in their ionic radii. Although histidine is the most common coordinating residue in zinc enzymes (
), it occurs less frequently in transport sites. Nevertheless, histidine plays a critical role in regulating proton-coupled or pH-dependent zinc transport as either a proton donor or an acceptor. Cysteine is another prevalent coordinating residue in structural sites of many zinc metalloproteins (
); however, it is conspicuously missing in the transport sites of YiiP and ZIPB.
Structure-guided discovery of functional mutations in zinc transport sites provides insights into the roles of individual coordinating residues. The tetrahedral transport site of YiiP is highly conserved, and alanine substitutions to one of the residue quartets resulted in loss of both zinc binding and transport activity in YiiP and mammalian ZnT homologs (
). Thus, individual coordinating residues in the tetrahedral transport site are functionally important. Moreover, YiiP has a DDHD quartet for transporting both Zn(II) and Cd(II), whereas mammalian ZnTs have an HDHD quartet (
). This finding demonstrates tuneability of metal selection by altering the residue composition of the tetrahedral transport site. However, the residue composition of the tetrahedral site is still insufficient to predict metal selectivity of the CDF family members due to potential second shell interactions (
). Other noted variations to the tetrahedral site are composed mainly of Asn and Asp with additional Asn and Asp residues in the immediate surroundings. Such modified transport sites in ZnT10 and a few bacterial homologs are responsible for selective Mn(II) transport (
In contrast to essential roles of individual coordinating residues in the tetrahedral transport site of YiiP, individual residues participating in the binuclear metal center in ZIPB are functionally dispensable (
). Comparing multiple ZIPB crystal structures with variations in the binuclear metal center reveals two critical clues. First, M1 coordination is fluidic with alternative coordinating residues and water molecules: H177, E181, Q207, E211, and M99 when M1 is a Cd(II), and E181, Q207, E211, and two water molecules when M1 is a Zn(II) (
). Thus, the loss of a coordinating group by a single alanine substitution could be compensated by the recruitment of an alternative coordinating residue or water molecule. Second, M2 coordination is required for neither zinc transport nor M1 binding (
). The absence of M2 binding did not significantly alter the M1 position and its coordination sphere. Moreover, single M1 binding occurs naturally in some ZIP homologs. For example, M2 in mammalian ZIP2 is occupied by a neighboring lysine residue (
). The asymmetric functions of M1 and M2 sites reflect clustering of conserved residues around the M1 site while large residue variations around the M2 site may diversify binding properties of the binuclear metal center through M1–M2 interactions to influence the primary transport site at M1 (
). The overall ZIP structure is evolutionally conserved from bacteria to humans, as demonstrated by a close alignment between the bacterial ZIPB crystal structure and a human ZIP4 structural model generated by coevolution-based contact prediction (
). Nearly all ZIP homologs have at least one carboxylate residue in the TM4 and/or TM5 signature motif, allowing two metal ions to be bridged by one or two carboxylate groups. Within this conserved structural framework, variable protein sequences, local structural dynamics, and incoming water molecules are expected to create far more fluidic coordination environments in ZIPs than a single tetrahedral transport site in ZnTs. Existing metal transport data on mammalian ZIPs suggest promiscuous metal selectivity. For example, ZIP8 and ZIP14 are broad-spectrum metal transporters that mediate cellular uptake of Zn(II), Fe(II), Mn(II), and Cd(II) (
). A glutamic acid residue was found in place of a highly conserved histidine residue in the predicted binuclear metal center. This E-to-H substitution may contribute to a shift of metal selectivity away from zinc (
). High-resolution structures of metal–ZIP complexes and accurate experimental profiling of metal substrates are needed to unravel further the structural elements of metal selectivity in mammalian ZIPs.
Dynamic mechanisms driving zinc mobility
Zinc complexations that obey the 18-electron rule are typically "exchange inert." However, zinc bindings to zinc transporters are directly coupled to translocation across the membrane barrier. The mobility of zinc ions makes a fundamental distinction between zinc transporters and zinc metalloenzymes, where, in the latter, zinc is generally considered as a permanent constituent of catalytic sites (
). Zinc movement in transporter proteins is an energetic process driven by transmembrane electrochemical gradients. A general form of free energy is stored in a proton gradient as an energy reservoir. It drives many secondary transport systems associated with nutrient uptake and maintenance of ionic homeostasis. YiiP transports zinc against its concentration gradient when the membrane becomes energized with a proton gradient. In contrast, ZIPB transports zinc in an opposite direction, down a zinc concentration gradient (
). Hence, distinct mechanisms of zinc transport are expected to couple selective Zn(II) binding to either a proton motive force or a zinc concentration gradient to dislodge bound zinc ions from their coordination spheres to render mobility.
In-cell characterization of individual zinc transporters is often hampered by the presence of redundant zinc transport systems, fluctuations of the proton motive force and zinc concentration gradient. A more direct kinetic analysis of zinc transport was developed using purified proteins in proteoliposomes encapsulated with a zinc-sensing fluorescent dye. The initial rate of zinc influx can be measured on a stopped-flow apparatus before the buildup of a significant intraproteoliposome zinc concertation. This in-test-tube transport assay allows precise manipulations of proton and zinc gradients while a tight seal of the proteoliposome membrane enables robust measurements of zinc flux with a time resolution down to 5–10 milliseconds. Stopped-flow kinetic analyses of YiiP and a second E. coli CDF protein, ZitB, revealed a substrate saturable transport process that can be fitted by the Michaelis–Menten equation (
). This kinetic behavior indicates a two-step process initiated by a Zn(II)/Cd(II) binding followed by a protein conformational change to move the bound metal ion across the membrane. Zn(II)/Cd(II) binding rapidly reaches a steady state while the ensuing conformational change constitutes a rate-limiting step of the transport reaction (
). In an experimental setting with a Cd(II) concentration in equilibrium across the membrane, the stopped-flow application of a proton concentration jump caused a Cd(II) flux in opposition to the imposed proton gradient, whereas depleting protons in the reaction buffer completely stalled Cd(II) transport despite an imposition of a Cd(II) gradient (
). These results clearly demonstrated an obligatory proton-coupled antiport mechanism. The Cd(II)-for-proton exchange stoichiometry was found to be 1:1 by stopped-flow flux analysis, in agreement with the calorimetric titration result as described above (
). Thus, direct water access to the transport site is expected to act as a proton donor or acceptor to drive an obligatory zinc-for-proton exchange. The zinc-driven proton transport could be probed by microsecond X-ray irradiation to activate water molecules in the zinc translocation pathway where residues in close proximity to the passing water molecules are covalently labeled by hydroxyl radicals and identified by bottom-up mass spectrometry (
). Millisecond time-resolved dynamics revealed that zinc binding to the transport site drove a rapid TM5 motion coupled to the gating of L152, resulting in alternate exposures of the transport site to two solvent-filled cavities on either side of the membrane. The in-cell transmembrane proton gradient of E. coli is about one to two pH units around the expected pKa of H153 in the transport site (
). The flipping of H153 as a part of the transport site to either side of the membrane is expected to change its protonation state (Fig. 2C). A deprotonated H153 facing a relatively alkaline cytosol would promote zinc binding from the intracellular cavity, whereas a protonated H153 facing a relatively acidic periplasm may facilitate zinc release into the extracellular cavity. As such, an inward pH gradient drives a vectorial zinc efflux in a 1:1 exchange stoichiometry. In this process, protonation and deprotonation of H153 are mediated by water access through the adjacent L152 that is directly coupled to zinc binding to the transport site. Consequently, the gated water access to the transport site enables a stationary proton gradient to facilitate the conversion of the proton potential energy to the kinetic power stroke of a vectorial zinc transport (
Contrary to a single tetrahedral transport site and a complete lack of outer shell interactions in YiiP, the crystal structure of ZIPB revealed multiple metal coordination sites characterized by fluidic coordination environments and participations of coordinating water molecules that are stabilized by residues in the second coordination sphere (
). These structural features suggest a transport mechanism distinct from the single-site alternating-access model. Indeed, stopped-flow kinetic analysis of purified ZIPB in proteoliposomes showed that ZIPB-mediated zinc flux is nonsaturable, electrogenic, and voltage-dependent (
). Instead, the zinc equilibrium potential exhibits a Nernst relationship predicted for the divalent zinc ion while the voltage dependence of the zinc flux also follows the Goldman–Hodgkin–Katz current equation at a symmetrical zinc concentration. These data provide strong evidence that the ZIPB-mediated zinc flux is electrodiffusional through a zinc permeant channel (
). Of note, the zinc flux through ZIPB is extremely slow in comparison with the potassium ion flux through potassium channels. The restricted zinc flow implies highly constrained zinc bindings that limit Zn(II) mobility along a transmembrane conduit. This kinetic behavior is consistent with the binuclear metal bindings in the ZIPB crystal structure. All bacterial ZIPs identified thus far promote zinc influx into the cytoplasm where zinc is buffered to extremely low levels around homeostatic set points (
). These bound zinc ions are approximately aligned in a single file and also partially hydrated (Fig. 2D). Since ZIPB constitutes a major uptake route for bacterial zinc acquisition, a high zinc-binding affinity is a prerequisite for effective zinc capture (
), but the high affinity may trap bound zinc ions to impede their transmembrane movement. The dynamic process of water access to the binuclear metal center revealed by X-ray footprinting uncovers an active role of hydration water molecules in releasing the trapped zinc ions (
). Contrary to the expectation that zinc binding would block water access as observed in YiiP, multiple zinc bindings in ZIPB were concomitant with increased water accessibility to selective coordinating residues, indicating that zinc ions and water molecules are cotransported (
). The opposite changes in water accessibility suggest that water entry to subpockets of the binuclear metal center may partially rehydrate the bound zinc ions, switching the mode of coordination residues from zinc binding to release. Mapping water-reactive residues to the ZIPB crystal structure revealed a water translocation pathway that overlapped with a zinc translocation pathway defined by X-ray crystallography (
). Following the intramembranous binuclear metal center, zinc ions navigate through the translocation pathway via an interim zinc-binding site to a peripheral binuclear metal center at the cytoplasmic exit (
). These binding sites are closely spaced, allowing a series of ligand exchanges between consecutive binding sites to relay a bound zinc ion from one binding site to another down a zinc concentration gradient (
). The peripheral binuclear metal center is thought to form a high-affinity sink to hold imported zinc before the bound zinc is accepted by cytosolic zinc-binding proteins such as metallochaperones, although zinc-specific metallochaperones have yet to be identified (
). This diffusional mechanism was similarly ascribed to the copper uptake transport Ctr1 where Cu+ diffusion is mediated by consecutive Cu+-binding sites, leading toward a high-affinity copper sink at the cytoplasmic exit (
Comparative structural analyses provide further insights into how protein conformational changes or local structural dynamics achieve differentiated functions of zinc transport in YiiP and ZIPB. Comparing X-ray and cryo-EM structures of YiiP homologs revealed large conformational changes between an inward- and outward-facing conformation, providing direct evidence for the alternating-access mechanism (
). A major conformational change involves pivoting of a four-helix bundle (TM1, TM2, TM4, and TM5) relative to a TM3–TM6 helix pair to flip over the zinc accessibility to either side of the membrane while a minor conformational change involves twisting of TM2 and TM5 in the four-helix bundle (
), a small TM2–TM5 shift could lead to a large readjustment of the zinc-coordination geometry in favor of either zinc binding or release, thereby allowing for allosteric regulation of zinc coordination through protein conformational changes (
). As a result, the zinc turnover rate of YiiP is several orders of magnitude faster than zinc exchange rates in typical zinc metalloproteins.
By comparison, no significant conformational change was observed among ZIPB crystal structures when bound metal ions were partially released by soaking crystals in a metal-free buffer or when metal binding to the M2 site of the binuclear metal center was ablated by triple point mutations (
). The findings of partially hydrated Zn(II)/Cd(II) in ZIPB crystal structures provide a critical clue as to the dynamic mechanism driving zinc movements. Multiple crystallographic water molecules were found in the hydration shells of bound metal ions and also in close proximities to protein ligands that participated in binuclear metal coordination. Positional shifts of coordinating residues via local conformational dynamics would suffice to switch interactions from a bound metal ion to solvation water molecule, favoring zinc release over binding. This notion is supported by comparative X-ray footprinting analyses of ZIPB with and without zinc binding (
). The reciprocal pattern of water accessibility changes in both binuclear metal centers and their associations with vicinal crystallographic water molecules provide clear structural evidence for direct water access to a sequence of five zinc-binding sites in ZIPB (
). Hence, accumulating data support a transport model in which water dynamics complements metal coordination chemistry to confer mobility to trapped zinc ions in ZIPB, highlighting the functional importance of solvated water in driving zinc transmembrane diffusion.
Bacterial YiiP and ZIPB serve as prototypes for mammalian ZnTs and ZIPs, respectively. Detailed structure–function studies lay the foundation for understanding the inner workings of mammalian ZnTs and ZIPs as to how coordination chemistry is built into different protein structures to confer zinc selectivity, but with distinct mechanisms driving zinc mobility. The tetrahedral transport site in YiiP is highly conserved from bacteria to humans. It is likely that mammalian ZnTs harness the proton motive force to actively pump zinc by conformational changes that alternatively expose a single transport site to either side of the membrane surfaces (
). Indeed, a recent cryo-EM structure of human ZnT8 validated a conserved tetrahedral transport site, which is alternately accessible to zinc binding from either side of the membrane during the zinc transport cycle (
). Interestingly, human ZnT8 in the absence of zinc binding to the transport site exhibited heterogeneous dimeric conformations with one protomer in an inward-facing and the other in an outward-facing conformation. The conformational conversion between these two states involved rocking motions of TM1, TM2, TM4, and TM5 relative to the TM3–TM6 helix pair as described for YiiP. The modes of operation in TMD are similar between bacterial YiiP and human ZnT8, but differences were noted in zinc bindings to cytosolic domains. A mammalian-specific HCH motif at the N terminus of ZnT8 contributes to zinc coordination to the CTD of a neighboring subunit. This HCH motif is also connected to the N terminus of TM1, allowing cytosolic zinc binding to influence the stability and transport function of the TMD (
). On the other hand, the binuclear metal center in ZIPB is less conserved and individual coordinating residues are functionally dispensable. This raises a question as to whether a conserved electrodiffusion mechanism of zinc transport exists from bacteria to humans. In bacteria, ZIPB is responsible for zinc uptake as a micronutrient, facilitating passive diffusion of extracellular zinc into the cytoplasm down an inward zinc concentration gradient (
). Likewise, the zinc level in human sera (>10 μM) is many orders of magnitude higher than the homeostatic set point of cytosolic free zinc concentrations in mammalian cells. This imposing zinc gradient is maintained by a multitude of mammalian zinc transporters and intracellular zinc buffering proteins. Up to now, in-cell measurements of zinc transport by various mammalian ZIPs have shown a consistent substrate saturable process with Michaelis–Menten steady-state kinetics (
). Such kinetic behaviors seem consistent with a single-site alternating-access model, but more rigorous functional studies of purified proteins in a well-defined experimental setting are required to unequivocally define the transport mechanism for mammalian ZIPs.
Surfacing of zinc transporters in response to environmental stimuli
While coordination chemistry and protein dynamics dictate selective binding and transport of zinc ions, compartmentalization of zinc transporters in mammalian cells provides the cellular context for the execution of transport functions and their integration into diverse cellular processes. The in-cell functions of individual mammalian zinc transporters are determined by their specific homes in various subcellular compartments and their specific molecular partners interlinked within the cellular protein network (interactome). Depending on the type of cells and cellular environments, a zinc transporter could be trafficked to different subcellular destinations and engaged in different protein interactions. This raises two fundamental questions: how a specific zinc transporter is placed in the right subcellular location at the right time in the right amount, and how protein interactions couple zinc transporters to macromolecular machineries to influence cellular processes? At present, there is a paucity of biochemical data on localizations of individual zinc transporters, even less on the organization of the zinc transporter interactome and the dynamics of protein interactions in response to pathophysiological stimuli. We will use relatively well-studied examples of mammalian zinc transporters to illustrate the molecular mechanisms underlying in-cell functions. The existing knowledge is far from complete, but the granular data may help define knowledge gaps and take us toward a more complete understanding of the inner workings of mammalian zinc transporters.
We begin with the regulation of zinc transporters on the surface membranes of enterocytes, which are intestinal absorptive cells lining the inner surface of the small intestine. Dietary zinc enters the polarized enterocytes through the apical membrane followed by zinc release into the circulation at the basolateral side. The trans-epithelial zinc movement is orchestrated by coordinated functions of many mammalian zinc transporters, including ZnT1, ZnT2, ZnT4, ZnT5, ZnT6, ZnT7, ZIP4, and ZIP5 (
). Their opposite directions of zinc transport and spatial segregation on different surface membranes allow them to play a major role in driving vectorial zinc movement from the intestinal lumen to the blood. Zinc fluxes across apical and basolateral membranes need to be balanced to maintain intracellular zinc homeostasis while the net trans-epithelial zinc flux is also regulated to maintain systemic zinc homeostasis in the face of dietary zinc fluctuations. As a result, the abundance of ZIP4 and ZnT1 on the respective cell surfaces is tightly regulated according to the zinc availability. Zinc deficiency promotes accumulation of ZIP4 on the surface membrane (
). On the other hand, high extracellular zinc levels not only induce internalization of surfaced ZIP4, but also trigger drastic removal of cellular ZIP4 via proteasomal and lysosomal degradation pathways (
). Hence, zinc uptake by ZIP4 across the apical membrane is principally modulated by a putative zinc-dependent brake of ZIP4 internalization. When the extracellular zinc level reaches a threshold, auxiliary structural components such as zinc sensors may be triggered to activate ZIP4 endocytosis by releasing the internalization brake.
Three potential zinc-sensing elements have been proposed in the human ZIP4 (hZIP4) sequence: a histidine-rich loop (His-rich) in a large N-terminal extracellular domain (ECD), a histidine-containing HxH motif in the extracellular loop between TM2 and TM3, and a His-rich cluster in the cytosolic loop between TM3 and TM4 (Fig. 3A). The His-rich loop in the ECD was shown to bind zinc with a low μM affinity (
). Thus far, zinc-dependent endocytosis has been reported for ZIP1 and ZIP4. Generally, endocytosis needs specific motifs such as [DE]XXXL[LI] and tyrosine-based endocytosis signals (YXXØ), which operate at the binding site for membrane-bound adaptor proteins (APs) (
). In hZIP4, a [DE]XXXL[LI]-like motif is found in the cytosolic loop between TM3 and TM4, but substitutions of the two Leu residues with Ala residues were insufficient to produce an effect on hZIP4 endocytosis (
). This conformation-dependent endocytosis of hZIP4 is proposed to occur constitutively in a ubiquitin-independent manner under normal conditions with adequate but nonexcessive zinc. Interestingly, all plasma membranes located ZIPs in the LIV-1 subfamily have some forms of the LQL motif, which fits the consensus motif [I,L,S]Q[L,N,D,P,A], suggestive of functional importance throughout the family. In addition to the LQL motif, the aforementioned cytosolic His-rich cluster in the TM3–TM4 loop is involved in zinc-stimulated ubiquitination and degradation of hZIP4 under excess zinc conditions (
). Because the cytosolic loops are divergent in the protein sequence, the number of His residues, and the length among ZIP proteins (Fig. 3A), they may provide unique properties for each ZIP protein, including apical or basolateral sorting of ZIPs, in addition to the regulation of endocytosis.
Zinc sensing and zinc-stimulated endocytosis of ZIP4 are slightly different between the human and mouse homologs. In mZIP4, the HxH motif in the extracellular TM2–TM3 loop is required for a high sensitivity of mZIP4 endocytosis in response to low zinc concentrations (
). However, the HxH motif is not involved in zinc sensing in hZIP4 as discussed above. mZIP4 is constitutively degraded following endocytosis in cells cultured in medium containing low μM zinc concentrations (
). Detailed analyses of the mouse–human difference in the kinetics of endocytosis and degradation of ZIP4 are required to gain a more complete understanding of the regulated endocytosis of mammalian ZIP4 and their underlying mechanisms.
ZIP4 surfacing is further regulated posttranslationally. The extracellular N terminus of ZIP4 is proteolytically cleaved during a prolonged zinc deficiency (
). Nevertheless, the N-terminally cleaved ZIP4 is internalized at a faster rate than the full-length ZIP4, suggesting that the extracellular N-terminal portion may be involved in regulation of the sensitivity to zinc-stimulated endocytosis in addition to zinc transport activity (
), which are often involved in trafficking the active form of zinc transporters to the plasma membrane.
In contrast to the surface expression of multiple zinc uptake transporters in the ZIP family, the only zinc efflux transporter predominantly targeted to the plasma membrane is ZnT1 in the mammalian ZnT family (
). The response of zinc-dependent ZnT1 surfacing is opposite to that of ZIP4. Excess zinc exposure would increase ZnT1 surface expression as well as its total cellular expression mediated by transcriptional upregulation under the control of metal response element binding transcription factor 1 (MTF1) (
). When translation of ZnT1 was blocked by a protein synthesis inhibitor cycloheximide, ZnT1 was degraded much faster than tubulin under normal culture conditions. As discussed in ZIP4 endocytosis, it is still unclear which APs are involved in ZnT1 endocytosis under zinc deficiency and how high zinc exposure may suspend ZnT1 endocytosis. Furthermore, posttranslational modifications of ZnT1 regulate ZnT1 stability in a zinc-responsive manner. Loss of N-glycosylation at Asn299 in an extracellular loop between TM5 and TM6 stabilized ZnT1, but it did not affect trafficking of ZnT1 to the cell surface. Interestingly, a similar stability control by N-glycosylation was found in ZIP14 (
While ZnT1 and ZnT10 are respectively zinc and manganese efflux transporters targeted to the cell surface membrane, mammalian ZnTs generally are zinc-sequestrating transporters localized to intracellular membrane compartments (Fig. 1A). Some ZnTs targeted to the secretory vesicles can be transiently trafficked to the surface membrane following exocytotic fusion of secretory vesicles with the surface membrane (
). This type of ZnT surfacing control is operative in secretory cells where a rapid recycling mechanism may be necessary to restore the localization of vesicular ZnTs.
Functional diversification by multilayers of protein interactions
Protein interactions add another dimension of functional integration to mammalian zinc transporters in addition to subcellular trafficking, posttranslational regulation, and tissue-specific expression. Protein interactions display a continuum of binding strength and stability (
). Many interactions are transient, and alternative interactions may occur in specific cellular contexts or at particular times during cell lineage specification. The dynamic organization of zinc transporter interactomes results in segregation of zinc transporters and functionally related proteins into interconnected protein communities corresponding to multiprotein complexes, subcellular protein colocalizations, and intracellular signaling cascades (
). At the molecular level, individual zinc transporters are assembled into functional modules through homo- and hetero-oligomerization of monomeric ZnT and ZIP homologs. These functional modules are further assembled into protein networks dedicated to specific cellular processes. At present, architectures of zinc transporter interactomes are largely unknown with one exception (see below), but well-studied examples of ZnT/ZIP-interacting proteins provide a glimpse of context-specific organization of zinc transporter interactomes that drive the integration of zinc transport activity to diverse in-cell functions.
The first layer of protein interactions takes place among monomeric forms of ZnT and ZIP as the functional unit of zinc transport. Crystal structures of YiiP and ZIPB suggest that extramembranous domains may contain signals directing oligomerization. YiiP is purified as a stable homodimer (
). As such, the CTD of YiiP may act as a dimerization joint as well as a cytosolic zinc sensor. Zinc binding to the CTD interface triggers an inter-CTD motion, which alters the coordination geometry of the transmembrane transport site, thereby allosterically regulating its transport activity (
). The additional zinc coordinates in ZnT8 are provided by the HCH motif of NTD from a neighboring subunit to establish NTD–CTD interactions to allosterically regulate the TMD conformation as discussed above (
). In addition, dimerization of mammalian ZnTs could be mediated by inter-CTD dityrosine bonds. Covalent dimerization represents a novel mechanism regulating subcellular localization and zinc transport activity of mammalian ZnTs (
) to higher-order polymerization as observed by cryo-EM imaging of purified ZIPB in detergent micelles. The crystal structure of ZIPB was determined in a lipid cubic phase with four outer TMs that create large gaps for lipid filling on the monomer surface (
). Accordingly, delipidation of ZIPB by different detergent treatments influences the nonspecific oligomeric association. Mammalian ZIPs are generally divided into four subfamilies, ZIPI, ZIPII, LIV-1, and gufA (Fig. 1B) (
), suggesting that it may play additional roles other than dimerization. All LIV-1 family ZIPs contain multiple His-rich sequences, especially in the ECD and the cytosolic loop between TM 3 and TM4, as demonstrated by ZIP4, ZIP5, ZIP6, ZIP7, and ZIP10, where the total number of histidine residues can exceed 50 (Fig. 3A). Zinc binding to the extracellular His-rich cluster was detected at a low μM affinity in ZIP4 as discussed above, while mutations to zinc-binding sites and other conserved ECD sequences could significantly alter zinc transport (
). Hence, the ECD of ZIPs mirrors the CTD of ZnTs in monitoring zinc availability on either side of the membrane to regulate zinc transport through dimerization.
While most ZnTs and ZIPs are thought to be functional as homodimers, heterodimerization occurs naturally among some members of the ZnT and ZIP protein family. A well-characterized heterodimer is ZnT5–ZnT6 (
). The functional roles of ZnT5 and ZnT6 in a heterodimer are asymmetrical. ZnT5 contains a canonical tetrahedral zinc-transport site, whereas ZnT6 misses two zinc-coordinating residues by hydrophobic residue substitutions. Thus, ZnT5 is thought to be an operative zinc transporter while ZnT6 a putative auxiliary protomer in the heterodimer (
). ZnT5–ZnT6 heterodimerization is likely mediated by CTD–CTD interactions. This was demonstrated using chimeric ZnT7 whose CTD was swapped with that of ZnT5. The resultant chimera recapitulated ZnT5-like features, that is, formed heterodimers with ZnT6 (
). For example, ZnT1 was targeted to the plasma membrane in homodimers, but redirected to intracellular vesicles when heterodimerized with ZnT3. In contrast, ZnT3 dimerization was insufficient to redirect ZnT2, ZnT4, and ZnT10 to synaptic vesicles in pheochromocytoma PC12 cells where ZnT3 homodimers were destined to synaptic vesicles (
). These findings revealed a network of ZnT homo- and heterodimers with distinct subcellular localizations, suggesting a potentially prevalent role of heterodimerization in regulating ZnT subcellular distribution. Further studies are needed to validate native ZnT heterodimers and their subcellular localizations in mammalian cells with endogenous ZnT expressions.
Heterodimerization between endogenous ZIP6 and ZIP10 was identified in mouse mammary gland epithelial NMuMG cells based on in situ proximity ligation analysis and immunoprecipitation proteomics with tandem mass tag labeling and bottom-up relative mass spectrometric quantification (
). Among mammalian ZIPs, ZIP6 and ZIP10 are the closest paralogues of the LIV-1 subfamily containing an N-terminal prion protein (PrP)-like sequence within the ECDs. ZIP6 and ZIP10 can form respective homodimers targeted to the plasma membrane where they mediate cellular zinc uptake, which often triggers cells to undergo epithelial-to-mesenchymal transition (EMT) (
), a process by which epithelial cells lose their cell–cell adhesion and gain migratory properties to become multipotent mesenchymal stem cells during embryonic development and tissue regeneration. Interestingly, ZIP6–ZIP10 heterodimers can also trigger mitosis, a process of cell division that requires cell rounding to be initiated (
). The full differences between these functional heterodimers are still unclear, as contributions of other ZIP transporters, such as ZIP5, have yet to be clarified. However, accumulating evidence suggests that ZIP6–ZIP10 heterodimerization allows for functional moonlighting as a key regulator in cell morphogenetic programs (
). Since ECDs of mammalian ZIPs are not well conserved or even missing in certain ZIP homologs, heterodimerization among ZIPs is thought to be limited to a subset of ZIPs carrying PrP-like sequences within the ECDs (
). These ZIPs form a separate clade of LIV-1 ZIPs comprising ZIP5, 6, and 10. Heterodimerization among members of other clades of the mammalian ZIP family has yet to be established.
Zinc transporters greatly expand their protein interaction networks through a second layer of protein interactions involving higher-order heterocomplexes with proteins other than zinc transporters. In mammary gland epithelial cells, ZnT2 was found to directly interact with vacuolar proton-ATPase (V-ATPase), and such interactions are critical for V-ATPase assembly associated with secretory vesicle biogenesis, acidification, and secretion (
). In epidermal keratinocytes, ZnT1 was shown to form heterocomplexes with EVER1 and EVER2 proteins encoded by genes associated with an inherited skin disorder, epidermodysplasia verruciformis. ZnT1–EVER2 interactions altered the intracellular zinc distribution and inhibited a group of zinc-dependent transcription factors (
). In a Xenopus oocytes heterologous expression system, ZnT1 and its C. elegans homolog CDF1 were found to bind to a regulatory domain of the protein kinase Raf1, facilitating Raf-1 translocation to the plasma membrane where ZnT1–Raf1 interactions activated Raf1, leading to downstream activation of mitogen-activated protein kinase (MAPK) members ERK1 and ERK2 (
). The basal Raf1 activity was inhibited by cytosolic zinc binding, raising the possibilities that ZnT1 upregulated Raf1 either directly via protein interactions or indirectly due to zinc efflux in close proximity to Raf1 in a ZnT1–Raf1 complex (
). Hence, accumulating data indicate that heterocomplex formation enables mammalian zinc transporters to perform moonlight functions beyond zinc transport. Finding interaction partners for a zinc transporter in a whole interactome setting may inform on its unconventional functions based on known cellular roles of the associated protein partners.
A mass-spectrometry-based discovery workflow was applied to profiling the ZIP6 interactome followed by global interrogation of its in-cell functions using gene ontology (GO) enrichment analysis to assign putative cellular functions and associated processes. As described above, both ZIP6 and ZIP10 are critical players in EMT, but it is unknown which of the ZIP6 interactors links ZIP6 to the cell migratory machinery and how. The ZIP6 interactome in mouse NMuMG cells was captured by anti-ZIP6 co-immunoprecipitation (co-IP) (
). ZIP6 was found to interact with ZIP10 but not with other ZIP homologs, indicating an exclusive ZIP6–ZIP10 heterodimeric assembly. Perturbation interactomes and corroborating biochemical analyses further demonstrated reciprocal effects of ZIP6 and ZIP10 on each other's expression and coupled responses of ZIP6 and ZIP10 expression to copper stimulation (
). ZIP6–ZIP10 heterodimerization appeared not to exclude homodimerization in distinct subcellular localizations, because subcellular distributions of ZIP6 and ZIP10 were only partially overlapping in neuroblastoma Neuro2a cells (
). The latter prompted a second set of NCAM1-centric interactome analyses, establishing prominent interactions of NCAM1 with integrins, a well-established hub protein in focal adhesion complexes that links intracellular actin bundles with the extracellular matrix (ECM) in many cell types to mediate cell anchoring, migration, and ECM signaling (Fig. 3B). The overall ZIP6 interactome was much smaller than the NCAM1 interactome and comprised only relatively few interactors not shared by NCAM1 (
). The hypothesis-free co-IP proteomics data in combination with bioinformatics and experimental mappings of protein interactions and phosphorylation sites narrowed down the global interrogation of in-cell ZIP6 functions to a putative protein hetero-complex: GSK3A/B-(ZIP6–ZIP10)-NCAM1-focal adhesion complexes (Fig. 3B). This working model assumes that the (ZIP6–ZIP10)-NCAM1 interactions may be mediated by self-templating heteromerization among PrP-like amino acid sequences within the ECDs of ZIP6, ZIP10, and the ECD of NCAM1, which was shown to bind PrP (
). Hence, the ZIP6–ZIP10 heterodimer may serve as a scaffold for binding and directing GSK3 kinases to NCAM1 phospho-acceptor sites localized to a cytosolic domain in juxtaposition to the inner face of the plasma membrane. Phosphorylation of these sites was shown to be critical for NCAM1-mediated signaling in astrocytes (
), an increase of the local zinc concentration by zinc transport through the ZIP6–ZIP10 heterodimer may inhibit the GSK3 activity in close proximity to a ZIP6 heterocomplex, thereby reducing the phospho-occupancy of NCAM1 sites and allowing its dissociation from focal adhesion complexes and the extracellular matrix (Fig. 3B). Taken together, profiling ZIP6- and NCAM1-centric interactomes revealed intricate multicomponent protein networking as an example to illustrate how mammalian zinc transporters are engaged in cellular protein machineries and localized to distinct subcellular locations. The positionings of specific zinc transporters in specific subcellular locations and protein complexes provide the molecular basis for precise controls of the local zinc concentration, which would regulate diverse zinc-dependent cellular processes manifested in unique cell biology of specialized mammalian cells. In the following sections, we will discuss how some of the key mammalian zinc transporters are functionally integrated at the cellular level to regulate three important zinc-dependent cellular processes: zinc signaling, unfolded protein responses in the ER, and zinc ectoenzyme activation in the early secretory pathway.
Cellular zinc signaling
Zinc as a signaling molecule interacts with a multitude of intracellular or extracellular proteins and modulates their activities. While the extracellular zinc concentrations fluctuate with the dietary or serum zinc supplies, transient changes in zinc extracellular concentrations can occur following release of zinc-containing vesicles from neurons, pancreatic β-cells, and secretory cells of the intestine epithelium, mammary gland, and salivary gland. ZnT3 controls vesicular zinc concentrations in presynaptic cells (
). High-glucose exposure stimulates insulin secretion while the cosecreted zinc ions act upon zinc-responsive membrane channels and receptors on the cell surface in a negative feedback loop to inhibit insulin secretion in both autocrine and paracrine fashions (
). These examples illustrate both short- and long-range signaling roles of extracellular zinc signaling in feedback controls of diverse cellular processes, in which zinc transporters regulate the prestimulation zinc loading in secretory vesicles and activate poststimulation clearance of the localized zinc concentration surge. However, what mediates extracellular zinc signaling is only partially understood. An orphan G-protein coupled receptor belonging to the ghrelin/motilin receptor subfamily, GPR39, has been identified as a metabotropic zinc receptor (ZnR) (
). At present, a functional link between GPR39 and mammalian zinc transporters is still unclear. This may be a key to understanding the mechanism by which zinc transporters regulate cellular responses to extracellular zinc fluctuations.
Cytosolic zinc is thought to be an intracellular signal that relays external stimuli to intracellular responses via phosphorylation–dephosphorylation (
). In this process, phosphate residues are transferred from one signaling molecule to the next in a consecutive cascade leading to gene expression in the nucleus. Phosphorylation and dephosphorylation are mediated by protein kinases and phosphatases, respectively. Their interplays regulate spatial and temporal signal processing of cell surface stimulations into two general categories of signaling outputs: a graded downstream response that provides feedback to restore cellular homeostasis, and a binary all-or-none response that pushes the cell out of homeostasis, driving cue-based decision-making on cell fates such as proliferation, differentiation, and apoptosis (
). Hence, cytosolic zinc fluctuations could effectively switch on/off phosphor-transfer by synergizing the actions of zinc-dependent protein kinases and phosphatases. In living cells, basal levels of protein kinases and phosphatases undergo constant fluctuations due to stochasticity in gene expression (
). The dual controls of protein kinases and phosphatases by the cytosolic zinc concentration make zinc transporters effective regulators in a role that damps out stochastic fluctuations in noisy phosphor-transferring cascades, thereby improving the fidelity of signal transduction.
Mammalian ZIPs are emerging as important regulators of two modes of phosphor signaling by modulating a graded cytosolic zinc level as well as directly participating in positive feedback loops of the binary decision-making process. Among mammalian ZIPs, ZIP7 is unique for its localization to the early secretory pathway including the ER (
). Functionally, ZIP7 and the surface-expressed ZIP6 are mutually compensatory in β cells. Double knockdown of ZIP7 and ZIP6 significantly impaired the glucose-stimulated cytosolic zinc surge and insulin secretion (
), suggesting that both ZIP7 and ZIP6 play an important regulatory role in the phosphor transduction from a common extracellular glucose signal diverging to a cytosolic zinc surge and insulin secretion, which is primed in β-cells through activations of both PKC and PKA (
The molecular mechanism linking cell surface stimuli to cytosolic zinc surge and downstream responses was further examined in tamoxifen-resistant MCF-7 breast cancer cells. Extracellular stimuli such as epidermal growth factor plus calcium ionophore elicited a transient protein association between protein kinase CK2 and ZIP7 within minutes (