Structure, Function, and Regulation of a Subfamily of Mouse Zinc Transporter Genes*

Zinc is an essential metal for all eukaryotes, and cells have evolved a complex system of proteins to maintain the precise balance of zinc uptake, intracellular storage, and efflux. In mammals, zinc uptake appears to be mediated by members of the Zrt/Irt-like protein (ZIP) superfamily of metal ion transporters. Herein, we have studied a subfamily of zip genes (zip1, zip2, and zip3) that is conserved in mice and humans. These eight-transmembrane domain proteins contain a conserved 12-amino acid signature sequence within the fourth transmembrane domain. All three of these mouse ZIP proteins function to specifically increase the uptake of zinc in transfected cultured cells, similar to the previously demonstrated functions of human ZIP1 and ZIP2 (Gaither, L. A., and Eide, D. J. (2000) J. Biol. Chem. 275, 5560–5564; Gaither, L. A., and Eide, D. J. (2001) J. Biol. Chem. 276, 22258–22264). No ZIP3 orthologs have been previously studied. Furthermore, this first systematic comparative study of the in vivo expression and dietary zinc regulation of this subfamily of zip genes revealed that 1) zip1 mRNA is abundant in many mouse tissues, whereas zip2 and zip3 mRNAs are very rare or moderately rare, respectively, and tissue-restricted in their accumulation; and 2) unlike mouse metallothionein I and zip4 mRNAs (Dufner-Beattie, J., Wang, F., Kuo, Y.-M., Gitschier, J., Eide, D., and Andrews, G. K. (2003) J. Biol. Chem. 278, 33474–33481), the abundance of zip1, zip2, and zip3 mRNAs is not regulated by dietary zinc in the intestine and visceral endoderm, tissues involved in nutrient absorption. These studies suggest that all three of these ZIP proteins may play cell-specific roles in zinc homeostasis rather than primary roles in the acquisition of dietary zinc.

Zinc is an essential trace element that is required for the catalytic activity of numerous metalloenzymes (4,5) and can also serve a purely structural role by stabilizing the conformation of certain zinc-dependent protein domains, such as zinc fingers, zinc clusters, and RING fingers, that are commonly found in transcriptional regulatory proteins (5,6). Deficiency of this essential metal can have devastating effects. In mammals, inadequate levels of zinc in the diet lead to dermatologic lesions, growth retardation, mental disorders, and compromised function of the immune and reproductive systems (7)(8)(9). Likewise, high levels of zinc can be cytotoxic. Thus, cells must maintain tight control over intracellular zinc levels. This control is achieved through a balance of zinc efflux, cellular zinc storage, and zinc uptake. Each of these activities is mediated by a distinct family of proteins (8).
In mammals, the zinc transporter (ZnT) 1 family of proteins function in a tissue-, cell-, and organelle-specific manner to regulate intracellular zinc homeostasis. These proteins contain six predicted transmembrane domains and are thought to function as multimers. Seven members of the Znt family have been identified to date in mammals (Znt1- 7), and genetic studies have confirmed the importance of many of these genes in mammalian zinc metabolism (10 -17). Recent studies demonstrated that zinc can regulate expression of the Znt1 gene (18), and dietary zinc modulates Znt1 and Znt2 mRNA levels in the kidney, intestine, and liver (19). Zinc can also regulate the intracellular localization of the ZnT4 and ZnT6 proteins (16). Intracellular zinc is bound by small cysteine-rich proteins called metallothioneins (MTs) (20). In the mouse, the MT family consists of four members. Exposure to high levels of zinc causes MT accumulation through increased gene expression, whereas dietary zinc deficiency leads to decreased MT abundance through decreased gene expression and protein destabilization (21). These proteins are thought to sequester zinc when present at high levels, protecting against heavy metal toxicity, and to provide a labile pool of zinc under limiting conditions that can be released for use by other proteins.
In eukaryotes, uptake of several essential metals is mediated by members of the Zrt/Irt-like protein (ZIP) superfamily of metal ion transporters (22,23). The first member of this family to be identified was iron-regulated transporter-1 (Irt1) from Arabidopsis. Irt1 was found to preferentially import iron, but can also transport manganese, cadmium, and zinc (24 -27). Subsequently, two other members of this superfamily (zincregulated transporter-1 (Zrt1) and Zrt2) were identified that function in zinc uptake in yeast (28,29). The acronym for the superfamily (i.e. ZIP) reflects Zrt-and Irt-like proteins. ZIP proteins contain eight putative transmembrane domains with conserved histidine, serine, and glycine residues within the fourth transmembrane domain, which may play a role in metal binding during transport. These proteins also often contain a histidine-rich loop between the third and fourth transmembrane domains.
Computer searches of the human and mouse genome sequences reveal multiple members of the ZIP superfamily (12 in human) (30). However, little is known about the regulation and function of most of these mammalian ZIP genes. A conserved subfamily of three zip genes (zip1, zip2, and zip3) has been identified in mice and humans. Human ZIP1 (SLC39A1) and ZIP2 (SLC39A2) have been demonstrated to function as zinc transporters, and ZIP1 is the major zinc uptake protein in K562 cells (1,2). No studies on the function of ZIP3 have been reported. The recent findings that the human ZIP4 gene (SLC39A4) is mutated in the genetic zinc metabolism disorder acrodermatitis enteropathica (31) and that this conserved gene encodes a zinc-regulated zinc transporter in mice (3) suggest that mammalian ZIP genes are integral components of the zinc homeostatic mechanism.
Although several recent studies have examined the expression and regulation of members of this conserved subfamily of three zip genes, there have been no systematic comparative studies of their expression or regulation in vivo. zip1 mRNA has been detected in most adult rat and many embryonic mouse tissues (32), and ZIP1 mRNA is also found in most human tissues (2) and cultured human cell lines. In contrast, ZIP2 mRNA was not detectable by Northern blot analysis of human tissue RNAs (1). Expression of ZIP3 has not been examined in any system, and expression of zip1 and zip2 has not been examined in adult mouse tissues. A modest hormonal regulation of ZIP1 mRNA in cultured prostate cell lines has been reported (33); and in human monocytes and THP-1 cells treated with a zinc chelator, ZIP2 mRNA abundance increased significantly, whereas ZIP1, ZIP3, and ZIP4 mRNA abundance did not change (34,35). The objective of the studies presented herein was to characterize in detail the structure and function the mouse zip1, zip2, and zip3 genes and to compare and contrast their regulation in vivo in nutrient-absorptive tissues (intestine and embryonic visceral endoderm) during periods of dietary zinc deficiency.

EXPERIMENTAL PROCEDURES
Computer Analyses of Sequence Data-Multiple sequence alignments were performed using the AlignX application within the Vector NTI Suite program (Informax, Bethesda, MD).
Plasmid Construction-The expressed sequence tags (ESTs) for zip1 (GenBank TM /EBI accession number BE572790) and zip3 (accession number BF100710) were obtained from Incyte Genomics (St. Louis, MO). Sequencing of these ESTs revealed that they contain the expected open reading frames. The mouse zip2 cDNA was obtained by reverse transcription (RT)-PCR from poly(A) ϩ RNA isolated from the placentas of day 12 zinc-deficient pregnant mice (described below). Reverse transcription was carried out using Improm-II reverse transcriptase (Promega, Madison, WI), and PCR of the open reading frame was carried out using Pfu DNA Polymerase (Stratagene, La Jolla, CA). For zinc uptake studies, selectable mammalian pcDNA3.1Puro(ϩ) expression plasmids (36) encoding each of the three mouse ZIP proteins with C-terminal hemagglutinin (HA) tags were constructed through a multistep cloning strategy. The resulting plasmids contained the open reading frame, including at least 6 bp preceding the initiator methionine for optimal translation fused in-frame at the 3Ј-end with an oligomer encoding an HA tag. In the case of zip3, the potential upstream initiator methionine that does not adhere to the Kozak consensus sequence (37) (see "Results") was included in the plasmid in case it is the actual translation start point. Each plasmid was sequenced in its entirety to confirm the absence of mutations introduced during the subcloning process.
Animal Care and Use-All experiments involving mice were performed in accordance with National Institutes of Health guidelines for the care and use of animals and were approved by the Institutional Animal Care Committee. CD-1 mice (48 -60 days old) were purchased from Charles River Breeding Laboratories (Raleigh, NC). Mouse diets were purchased from Harlan Teklad (Madison, WI) and have been described in detail previously (38). Zinc levels in the diets were as follows: zinc-deficient, 1 ppm zinc; and zinc-adequate, 50 ppm zinc. These diets are otherwise identical.
To examine tissue-specific expression, female or male CD-1 mice (six per group) maintained on zinc-adequate feed were killed, and the indicated tissues were harvested and snap-frozen in liquid nitrogen for subsequent extraction of RNA for Northern analysis. Pancreatic RNA was extracted immediately from fresh tissue.
To examine the effect of dietary zinc deficiency during pregnancy on mouse zip mRNA levels, pregnant mice were subjected to moderate dietary zinc deficiency as described previously (39 -41). CD-1 females were mated with CD-1 males; the day a vaginal plug was found was considered day 1 of pregnancy. On day 1, mice were placed in pairs in cages and provided free access to zinc-adequate feed and deionized distilled water. On day 8, mice were placed in pairs in cages with stainless steel false bottoms to reduce the recycling of zinc (42). Water bottles were washed with 4 M HCl and rinsed with deionized water to remove zinc. The diet was changed to the zinc-deficient diet, or where indicated, the zinc-adequate diet was maintained. The visceral yolk sacs, placentas, and maternal small intestines were harvested on days 11-15 (six mice per group) and snap-frozen in liquid nitrogen for RNA extraction and subsequent Northern analysis or RT-PCR.
RNA Extraction and Northern Analysis-Total RNA was isolated either using TRIzol reagent (Invitrogen) according to the manufacturer's instructions or as described previously (18,40). Polyadenylated RNA was purified from total RNA using the Oligotex mRNA isolation kit (QIAGEN Inc., Valencia, CA) according to the manufacturer's instructions.
Total RNA (3 g/lane) was fractionated on formaldehyde-containing 1% agarose gels, transferred to nylon membranes, and immobilized via UV cross-linking (43). Duplicate gels were stained with acridine orange to monitor RNA integrity and to normalize for loading.
DNA fragments containing full-length zip1, zip2, or zip3 cDNA were excised from plasmids and used as probes for Northern blot hybridization. Probes were labeled with [ 32 P]dCTP using the Random Primers DNA labeling system (Invitrogen) according to the manufacturer's instructions. The mouse MT-I probe was as described previously (38). Membranes were hybridized and washed under stringent conditions as described previously (18,38,43). Hybrids were detected by autoradiography with intensifying screens at Ϫ70°C.
Cell Culture and Transient Transfections-HEK293 cells were maintained in high glucose Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 100 units/ml penicillin, 100 g/ml streptomycin, and 2 mM L-glutamine at 37°C in a humidified 5% CO 2 incubator. For transient transfections, cells were seeded onto poly-Llysine-treated 24-well plates at a density of 2 ϫ 10 5 cells/well and were transfected the following day. Transfections were carried out using LipofectAMINE 2000 reagent (Invitrogen) according to the manufacturer's instructions. Zinc uptake assays were performed 36 -48 h later.
Zinc Uptake Assays-Zinc uptake assays were performed as described previously (1,2). After transfection, cells were washed once with cold uptake buffer (15 mM HEPES, 100 mM glucose, and 150 mM KCl, pH 7.0), followed by a 10-min incubation with prewarmed uptake buffer. Cells were then incubated in prewarmed uptake buffer containing the specified concentration of 65 ZnCl 2 (PerkinElmer Life Sciences) in a 37°C shaking incubator for 15 min unless otherwise indicated. Assays were stopped by addition of an equal volume of cold uptake buffer supplemented with 1 mM EDTA (stop buffer). Cells were collected by filtration on glass fiber filters (Type A/E, Gelman Sciences) and washed three times with stop buffer (total of 10 ml). Cell-associated radioactivity was measured with a Packard Auto-Gamma 5650 ␥-counter. Metal salts were obtained from Sigma. Stock solutions of the chloride salts of various metals (CuCl 2 , CdCl 2 , MnCl 2 , CoCl 2 , MgCl 2 , and NiCl 2 ) and AgNO 3 were prepared at 100 mM in distilled water. A ZnCl 2 stock was prepared at 100 mM in 0.02 N HCl, and an FeCl 3 stock was prepared at 50 mM in 0.1 N HCl. Sodium ascorbate (1 mM) was used to reduce Fe 3ϩ to Fe 2ϩ . Ascorbate treatment alone did not alter zinc uptake activity (data not shown). Cells grown in parallel to those used for uptake experiments were washed three times with ice-cold uptake buffer, resuspended in phosphate-buffered saline containing 0.1% SDS and 1% Triton X-100 for cell lysis, and then assayed for protein content using a Bradford assay kit. Zinc accumulation and uptake rates were normalized to protein concentrations of these cell lysates. Michaelis-Menten constants were determined by nonlinear interpolation of the data using Prism (Version 3.0a for Macintosh, GraphPAD Software, San Diego, CA).
RT-PCR Detection of Mouse zip2 mRNA-Tissue-specific expression of zip2 was examined by RT-PCR. Total RNA from various mouse tissues (1 g/reaction) was DNase I-treated according to the manufac-turer's instructions (Invitrogen). DNase I was inactivated by addition of EDTA to 2.5 mM, followed by a 10-min incubation at 65°C. Reverse transcription was subsequently carried out using Improm-II reverse transcriptase. Samples were then amplified using Platinum Taq DNA polymerase (Stratagene) for 30 cycles. Mouse zip2 was amplified using primers 2TXTS (5Ј-CTTCTTGGGAGCAGGGTTGATGC-3Ј) and 210073AS (5Ј-CGCCACTGTGGCCTGTAGTCC-3Ј) to give a 406-bp product. Mouse zip1 was used as a control for RT reactions because it is widely expressed and was amplified using primers mZIP1KOZS (5Ј-CGGGATCCACAGCCACCATGGGGCCCT-3Ј) and mZIP1GFPAS (5Ј-CGGAATTCTTAGATTTGGACAAAGAGAAGGCCAGTGAGC-3Ј) to give a 1000-bp product.
RT-PCR was used to examine the effect of dietary zinc deficiency during pregnancy on mouse zip2 mRNA abundance. Total RNA (1 g; isolated from embryonic visceral yolk sacs from the zinc-deficient pregnant mice described above) was DNase I-treated and reverse-transcribed as described above. Samples were then amplified using Platinum Taq DNA polymerase for increasing increments of three cycles, starting at 21 cycles and proceeding to 33 cycles. Mouse zip2 was amplified using primers 2TXTS and 210073AS. Mouse zip1 was used as a negative control for regulation by dietary zinc and was amplified using primers mZIP1KOZS and mZIP1GFPAS. Mouse MT-I was used as a positive control for regulation by dietary zinc and was amplified using primers MT1S3 (5Ј-CACCACGACTTCAACGTCCTG-3Ј) and MT1AS604 (5Ј-TCTTGCAGGCGCAGGAGCTG-3Ј) to give a 138-bp product.

RESULTS
Characterization of the Mouse zip1, zip2, and zip3 Genes and cDNAs-To identify mouse ZIP subfamily members, the conserved 12-amino acid signature sequence (HSVFEGLAVGLQ) from the human ZIP1, ZIP2, and ZIP3 proteins ( Fig. 1) was used to search the mouse translated high throughput genome sequences data base on the NCBI Protein Database server. This search identified three different BAC entries that were derived from the C57/B6 strain of mice (see "Experimental Procedures" for accession numbers). Alignment of the cDNA sequences of the human ZIP genes with the nucleotide sequences flanking the signature sequences within each of these three mouse BACs (Fig. 1) provided the identity of the mouse zip1, zip2, and zip3 genes.
The cDNA corresponding to each gene was identified by alignment of ESTs from the mouse EST data base to generate overlapping contiguous consensus sequences. This approach worked well to identify the cDNAs for zip1 and zip3, which are ϳ2300 and 3750 bp in length, respectively, because there were several ESTs for each of these genes in the data base. Unfortunately, it did not work for zip2 because there were only three zip2 EST entries in the data base. Therefore, the full-length human ZIP2 cDNA sequence was aligned with the genomic sequence in the zip2 BAC. The high degree of homology shared between the human ZIP2 cDNA sequence and the mouse zip2 genomic sequence helped reveal the zip2 cDNA sequence, which was found to be ϳ1350 bp in length. PCR primers were designed using this information to amplify the cDNA containing the coding sequence.
Rapid amplification of cDNA ends was carried out for each of these three mouse zip genes to identify the 5Ј-end of the cDNA. In each case, the 5Ј-end was extended either by only a few base pairs or not at all compared with the known cDNA and EST sequences in the data bases. The 3Ј-end of each cDNA is demarcated by a poly(A) tail. In the case of zip3, comparison of ESTs revealed two different messages that are identical in coding sequence, but one ends prematurely due to the use of a non-canonical polyadenylation signal (Fig. 2C). The number of ESTs corresponding to this shorter transcript suggests that this message is significantly less abundant than the longer message. As this alternative poly(A) tail occurs within the 3Ј-untranslated region, the same protein is predicted to be encoded by both messages. Both zip1 and zip3 cDNAs contain a thymidine-rich region within their long 3Ј-untranslated regions, but the significance of this is not yet known.
Comparison of the zip1 cDNA and gene sequences initially revealed a zip1 pseudogene within this BAC in the data base. The pseudogene has no introns, terminates in a poly(A) tract, and displays mismatches and codon deletions compared with the EST sequences. To identify a BAC containing the actual expressed zip1 gene, primers were designed to differentiate between the functional gene and the pseudogene based on the sequence mismatches. These primers were sent to Incyte Genomics for PCR screening of a mouse genomic BAC library. Two positive clones were identified, and partial sequence analysis revealed that they were devoid of the nucleotide mismatches seen in the pseudogene. These BACs were then sequenced in their entirety at the Mouse Genomic Sequencing Group of the University of Oklahoma.
The exon-intron structure for each gene was determined by aligning the gene with the cDNA and by identifying splice donor and acceptor consensus sequences (Fig. 2). All three genes are relatively small, spanning Ͻ10 kilobase pairs of DNA, and contain four exons. Using accession numbers for the BACs containing each zip gene, the chromosomal localization was assigned using the mouse genome server. 2 Localization of zip1 to chromosome 3 is consistent with a previous report (32). The zip2 gene was localized to chromosome 14, the zip3 gene to chromosome 10, and the zip1 pseudogene to chromosome 2.
Analysis of the cDNA sequence revealed that the predicted mouse ZIP1, ZIP2, and ZIP3 proteins are 324, 309, and 317 amino acids in length, respectively. Alignment of the predicted human and mouse ZIP proteins revealed significant sequence conservation along the entire length of each of the three proteins (Fig. 3). The mouse and human ZIP1, ZIP2, and ZIP3 proteins are 93, 78, and 83% identical, and 96, 85, and 90% similar, respectively. Computer analysis of the zip3 cDNA identified an open reading frame for mouse ZIP3 that contains an additional 31 amino acids N-terminal to the initiator methionine shown in Fig. 3C. This extension is due to the presence of an in-frame methionine codon located 93 nucleotides farther 2 Available at www.ensembl.org.
FIG. 1. Amino acid conservation within the fourth transmembrane domain identifies a subfamily of mammalian ZIP proteins. A portion of the predicted amino acid sequence within the fourth putative transmembrane domain was optimally aligned for several known ZIP proteins from Arabidopsis, yeast, and human. The conserved region in the human sequence was used to identify three mouse ZIP proteins belonging to a subfamily. These sequences contain the conserved histidine, serine, and glycine residues (boldface) that are a hallmark of ZIP family members. Human and mouse ZIP1, ZIP2, and ZIP3 belong to a smaller subfamily that shares an extended 12-amino acid signature sequence (underlined), with the exception of one conserved amino acid substitution present within ZIP3.
upstream. This potential upstream start codon does not adhere to the Kozak consensus sequence for optimal translation (37), whereas the downstream codon does. In addition, this N-terminal extension is not present in the human ortholog. We therefore conclude that the downstream methionine is the initiator codon that is used in vivo, although this remains to be proven.
Putative transmembrane domains for the mouse ZIP proteins were predicted using the on-line application TMPred. 3 Consistent with other members of the ZIP superfamily of metal ion transporters, these three mouse ZIP proteins contain eight putative transmembrane domains (Fig. 3).
Mouse zip1, zip2, and zip3 Each Encode Functional Zinc Transporters-Human ZIP1 and ZIP2 have been shown to 3 Available at www.isrec.isb-sib.ch. FIG. 2. Structures of the mouse zip1, zip2, and zip3 genes and cDNAs. The gene structures for mouse (m) zip1 (A), zip2 (B), and zip3 (C) were deduced by comparing the cDNA sequence for each zip gene with the genomic sequence and by identifying the exon-intron splice donor and acceptor consensus sequences. All genes contain four exons and are relatively small, each spanning Ͻ10 kilobase pairs. The predicted translation start and stop points and length are shown for each zip cDNA, and an alternative poly(A) tail resulting from a non-canonical polyadenylation signal present within zip3 is shown in C.
FIG. 3. Sequence alignments of the predicted mouse and human ZIP1, ZIP2, and ZIP3 proteins. The predicted mouse (m) and human (h) proteins were optimally aligned for ZIP1 (A), ZIP2 (B), and ZIP3 (C). Similar amino acids are shaded, and the eight putative transmembrane domains (TM) are indicated by solid lines above the mouse sequences. Amino acid sequence identities between these human and mouse ZIP proteins ranged from 83 to 93%. The conserved 12-amino acid signature sequence that is characteristic of this subfamily is boxed for each ZIP. function in zinc uptake, but the function of ZIP3 has not been reported. To examine the functions of mouse ZIP1, ZIP2 and ZIP3, the open reading frame for each of these ZIP proteins was cloned into the mammalian expression vector pcDNA3.1Puro, and a linker encoding an HA tag was cloned in-frame immediately preceding the stop codon. HEK293 cells were transiently transfected with either empty vector or a vector encoding one of these three HA-tagged mouse ZIP proteins. Pools of transfectants were assayed for 65 Zn uptake activity (Fig. 4). Consistent with the ability to import zinc, cells expressing mouse ZIP1, ZIP2, and ZIP3 accumulated ϳ4-fold more zinc over a 60-min period compared with the endogenous uptake activity assayed in cells expressing vector only (Fig. 4A). Zinc accumulation by mouse ZIP1, ZIP2, and ZIP3 was virtually identical over the time course of this experiment. Only low levels of zinc accumulation were detected when these assays were carried out at 0°C, indicating that zinc accumulation by each of these three mouse ZIP proteins as well as by the endogenous system is temperature-dependent and therefore is likely to be transporter-mediated rather than due to zinc binding to the cell surface. Consistent with this, uptake by the endogenous system as well as by mouse ZIP1, ZIP2, and ZIP3 was concentration-dependent and saturable. When assayed over a range of zinc concentrations, the uptake activities of mouse ZIP1, ZIP2, and ZIP3 showed surprisingly similar Michaelis-Menten kinetics, with apparent K m values of 1.7, 1.6, and 1.6 M, respectively, and V max values of 9.5-10.5 pmol of zinc/min/mg of protein. Compared with mouse ZIP1, ZIP2, and ZIP3, the endogenous system in HEK293 cells had a similar apparent K m of 2.0 M, but a far lower V max of 4.1 pmol of zinc/min/mg of protein, confirming that its activity is distinguishable from that of mouse ZIP1, ZIP2, and ZIP3.
To assess the substrate specificity of uptake by mouse ZIP1, ZIP2, and ZIP3, we examined the effect of various metals on zinc accumulation by cells expressing each of the three mouse ZIP proteins. HEK293 cells were transiently transfected with either empty vector or a vector encoding one of the three HAtagged mouse ZIP proteins. Pools of transfectants were assayed for 65 Zn uptake activity in the presence of either a 10-or 50-fold molar excess of competitor metal (Fig. 5). Zinc uptake by the endogenous system in HEK293 cells was strongly inhibited by excess zinc, copper, and cadmium and to a lesser extent by other metals (Fig. 5A). Consistent with the hypothesis that these three mouse ZIP proteins function as zinc transporters, zinc uptake by mouse ZIP1, ZIP2, and ZIP3 was strongly inhibited by the addition of excess zinc (Fig. 5, B-D). Zinc uptake by mouse ZIP1 and ZIP2 was only modestly inhibited by other metals. Conversely, zinc uptake by mouse ZIP3 was inhibited by all metals tested, except for iron, suggesting that it may be more promiscuous in its substrate selection. Nonetheless, all three appear to be largely zinc-specific.
Expression Patterns of Mouse zip1, zip2, and zip3 Are Unique, but Are Not Responsive to Dietary Zinc in the Adult Intestine and Embryonic Visceral Endoderm-To examine the tissue-specific patterns of expression of zip1, zip2, and zip3, Northern blot analysis was carried out using total RNA extracted from several organs from male and female CD-1 mice as well as from extraembryonic tissues surrounding the developing mouse embryo. Expression of zip1 was detected in all tissues examined, except for the pancreas, with the highest levels of expression seen in the intestine and ovary (Fig. 6A, upper  panel). The integrity of the pancreatic RNA was confirmed by staining a duplicate gel with acridine orange (Fig. 6A, lower  panel). In contrast to zip1, expression of zip3 was tissue-restricted, with the highest level seen in the testes (Fig. 6A,  middle panel). In addition, the abundance of zip3 mRNA was much lower than that of zip1 mRNA in all tissues, except in the testes.
Despite repeated attempts, including the use of purified polyadenylated RNA, zip2 was not detected by Northern blotting. Therefore, the more sensitive method of RT-PCR was used to identify tissues that express the zip2 gene (Fig. 6B, upper  panel). Similar to zip3, expression of zip2 appeared to be tissuerestricted, with the highest levels detected in the skin, liver, ovary, and visceral yolk sac. As a control for reverse transcription, zip1 was amplified from the same reactions (Fig. 6B, lower  panel). These data support the results from the Northern blot experiments shown in Fig. 6A, demonstrating the nearly ubiq- uitous expression of zip1 mRNA and the very low abundance of zip2 mRNA.
To address whether these three mouse zip genes are zincresponsive in nutrient-absorptive tissues, pregnant mice were subjected to dietary zinc deficiency, and the abundance of zip1, zip2, and zip3 was examined in the intestine and embryonic visceral endoderm. Pregnant mice were chosen for this study because dietary zinc deficiency during pregnancy is a model system with which to examine the effects of zinc on gene regulation and development. The demand for zinc increases during mid-gestation because of the rapid growth of the embryo, thus accentuating the effects of zinc deficiency (39,40). Beginning on day 8 of pregnancy, mice were placed on either a zinc-adequate or zinc-deficient diet. Maternal small intestines and embryonic visceral yolk sacs were collected for RNA isolation on days 11-15. All three of these zip genes were expressed in the visceral yolk sac. In addition, the relative abundance of MT-I mRNA was examined because this gene is known to be down-regulated by dietary zinc deficiency in these tissues (40). Our previous studies also demonstrated that expression of zip4 is significantly up-regulated in these tissues under these experimental conditions (3). As expected, MT-I mRNA abundance was reduced in response to dietary zinc deficiency (Fig. 7A, third panel). zip1 mRNA abundance did not change in response to dietary zinc deficiency in either the maternal small intestine or the embryonic visceral yolk sac (Fig. 7A, first panel). Similarly, zip3 mRNA levels did not change in either of these tissues under these conditions (Fig. 7A, second panel). The two transcripts seen for zip3 in Fig. 7 presumably result from the presence of the two different polyadenylation signals, and the relatively low intensity of the smaller band correlates with the fewer number of ESTs with the alternative poly(A) tail that results in a significantly shorter transcript.
Semiquantitative RT-PCR was used to examine zip2, MT-I, and zip1 mRNA abundance during dietary zinc deficiency. Total RNA was reverse-transcribed and subsequently PCR-amplified for increasing numbers of cycles. As shown in Fig. 7B, when the PCR products were first detectable by agarose gel electrophoresis (21 cycles), the abundance of the product directly reflected the relative abundance of MT-I mRNA as determined by Northern blotting (Fig. 7A, third panel). As expected, these differences in abundance diminished with increasing cycle number. This experiment was then adapted to analyze zip2 mRNA abundance in response to zinc deficiency using zip1 as a control for reverse transcription and as a control for an mRNA whose abundance does not change under these conditions (Fig. 7A, first panel). PCR products for zip1 were detectable after 27 amplification cycles, and their abundance did not change during dietary zinc deficiency, consistent with the Northern blot results (Fig. 7, compare C with A). Amplification products for zip2 were detectable after 30 cycles (Fig.  7C), and their abundance also apparently did not change during dietary zinc deficiency. Modest changes (Ͻ2-fold) in zip2 mRNA abundance may not have been detected under these experimental conditions. DISCUSSION The studies reported herein focused on a subfamily of three mouse zip genes, two of whose human orthologs have been demonstrated to function as zinc transporters (1,2). As in humans, there appears to be only three members of this subfamily in the mouse. Given the high degree of amino acid similarity to their human orthologs, we predicted mouse ZIP1 and ZIP2 to have very similar functional characteristics, as was found to be the case. However, we did note several differences. First, human ZIP1 and ZIP2 function as zinc transporters, but their uptake is strongly inhibited by other metals (1,2). Zinc uptake by human ZIP1 is inhibited by copper, iron, and cadmium, and that by human ZIP2 is inhibited by all other metals tested. In contrast, uptake by mouse ZIP1 and ZIP2 reported here was relatively specific for zinc as a substrate, with only modest inhibition by several other metals. These differences in uptake properties could reflect, in part, differences inherent in the cell types that were assayed. The data presented herein for mouse ZIP3 zinc uptake properties and metal specificity represent novel, albeit predictable findings given the structural similarities between mouse ZIP1 and ZIP2 and mouse ZIP3. It seems likely that the conserved 12-amino acid signature sequence found in members of this subfamily of ZIP proteins is important for their zinc transport activity.
Computer searches predict the presence of 12 ZIP genes in the human genome and at least five in the mouse (30). This finding raises the question of why so many ZIP genes exist and what the functional significance of this is. There are numerous scenarios that could explain this. Our data suggest that this can be explained, in part, by differential tissue distribution. The zip1 gene was actively expressed in nearly all tissues examined, consistent with reports of widespread distribution of ZIP1 transcripts in adult human and rat tissues as well as in embryonic mouse tissues (2,32). In contrast, the abundance of zip2 and zip3 transcripts was much more tissue-dependent. As is the case in human tissues (1), zip2 mRNA was rare in all of these mouse tissues. However, higher transcript levels were detected in the ovary, liver, and skin. In contrast, zip3 mRNA was generally more abundant in these tissues compared with zip2 mRNA, but it was also present at the highest levels in only a few tissues (e.g. testes). However, those tissues that expressed zip2 and/or zip3 also expressed zip1. Interestingly, all three of these transcripts could be detected in the adult intestine and embryonic visceral yolk sac. This could represent a redundancy that is necessary in case one of the zip genes is inactivated by mutation. Alternatively, it is possible that they FIG. 6. Tissue-specific expression of the mouse zip1, zip2, and zip3 genes. Total RNA was isolated from the indicated mouse (m) tissues. A, RNA was resolved by formaldehyde-agarose gel electrophoresis (3 g/lane), transferred to nylon membranes, and probed with cDNA probes against either zip1 (upper panel) or zip3 (middle panel). A duplicate gel was stained with acridine orange (lower panel) to verify RNA integrity and to normalize loading. B, RNA was reverse-transcribed (1 g/reaction) and subsequently PCR-amplified using primers corresponding to zip2 (upper panel) or zip1 (lower panel). Plasmids containing the appropriate cDNAs were used as positive controls, and no input template was used as a negative control. Amplification of zip2 yielded a 406-bp product, whereas amplification of zip1 yielded a 1000-bp product. sm int, small intestine.

FIG. 7.
Effects of dietary zinc deficiency on mouse zip1, zip2, and zip3 mRNA abundance in the maternal small intestine and visceral yolk sac. Pregnant CD-1 mice were fed either a zinc-adequate (ϩ) or zinc-deficient (Ϫ) diet on day 8 of pregnancy. Maternal small intestines and visceral yolk sacs were collected on days (d) 11-15 of pregnancy (i.e. days 3-7 of the zinc-deficient diet). A, RNA was isolated; resolved by formaldehyde-agarose gel electrophoresis (3 g/lane); transferred to a nylon membrane; and hybridized with labeled cDNA probes for zip1, zip3, and MT-I. B, RNA was isolated, reverse-transcribed (1 g/reaction), and subsequently PCR-amplified using primers corresponding to MT-I mRNA to produce a 138-bp product. Amplification was carried out for the indicated number of cycles. No input template (NT) was used as a negative control. C, RNA from the embryonic visceral yolk sac was reverse-transcribed (1 g/reaction) and subsequently PCR-amplified using primers corresponding to zip2 or zip1 to control for efficiency of reverse transcription. Amplification was carried out for 27, 30, or 33 cycles. Plasmids containing the appropriate cDNAs were used as positive controls (PT), whereas no input template (NT) was used as a negative control. Amplification of zip2 yielded a 406-bp product, whereas amplification of zip1 yielded a 1000-bp product. m, mouse.
are expressed in the same tissue, but within different subsets of cells within that tissue. This could be addressed through immunohistochemistry using antibodies specific for each ZIP protein.
Another possibility to explain the large number of zip genes in the mouse genome is that different ZIP proteins may exhibit distinct subcellular localization that represents functional differences. This is seen in the case of yeast ZIP proteins. Zrt1 and Zrt2 function at the plasma membrane to import zinc into cells, whereas Zrt3 functions to transport stored zinc out of the yeast vacuole during the transition from zinc-replete to zinc-limiting conditions (28,29,44). A similar scenario is seen in the ZnT family of zinc transporters. ZnT1 functions to export zinc from cells and is localized to the plasma membrane (10), whereas ZnT2, ZnT3, ZnT6, and ZnT7 are located within intracellular membranes, causing zinc accumulation within vesicles or organelles (11,16,(45)(46)(47).
Although many ZIP proteins are known to transport zinc, yet another possibility for the existence of multiple zip genes is that they may differ in the metals that they transport within the physiological context of the animal. There is precedence for this in the case of the Arabidopsis ZIP protein Irt1, whose preferential substrate is iron, but it can also transport manganese, cadmium, and zinc (24,26,27). As mentioned above, the mouse ZIP1, ZIP2, and ZIP3 transporters were found to be largely zinc-specific.
Contrary to what we expected, zip1, zip2, and zip3 mRNA abundance was not regulated in vivo by dietary zinc in tissues that function in the absorption of dietary or maternal zinc. Expression of genes encoding other zinc transporters, such as members of the zip and Znt families, is regulated, at least in part, at the RNA level by their substrate. The first zip gene to be identified, which encodes the Arabidopsis Irt1 iron transporter, is induced in roots under conditions of iron deficiency (23). Likewise, expression of zip genes encoding the yeast Zrt1 and Zrt2 and Arabidopsis ZIP1, ZIP3, and ZIP4 zinc transporters is induced under conditions of zinc deficiency (28,29,48). In addition, mouse zip4, the ortholog of human ZIP4, which is mutated in the genetic human zinc metabolism disorder acrodermatitis enteropathica, is zinc-regulated in the mouse intestine and visceral yolk sac (3). Zinc deficiency causes a marked increase in expression of zip4 in the intestine, which is consistent with the finding of increased intestinal zinc uptake activity during periods of dietary zinc deficiency in mammals (19, 49 -52). However, zip1 mRNA is also relatively abundant in the mouse intestine and may thus be responsible, at least in part, for the constitutive zinc uptake activity found in the zincreplete intestine. Alternatively, zip1 may be regulated by other signals. For example, human ZIP1 may be hormonally regulated in cultured prostate cells (33).
The lack of zinc responsiveness of the mouse zip2 and zip3 genes and the low abundance of zip2 and zip3 mRNAs in tissues that function in the absorption of dietary zinc are consistent with the concept that these proteins may serve cell-specific roles in zinc homeostasis. A zinc chelator has been shown to increase the abundance of ZIP2 mRNA in cultured monocytes and THP-1 cells (35). Thus, it is possible that zip2 and zip3 may be regulated by zinc in a cell-specific manner. Whether dietary zinc can modulate the cell-specific expression of the zip2 and zip3 genes in vivo in tissues other than the intestine and visceral yolk sac remains to be determined.
The lack of regulation of mouse zip1, zip2, and zip3 mRNA abundance by zinc certainly does not preclude the possibility that they are functionally zinc-responsive. It is possible that the mouse ZIP1, ZIP2, and ZIP3 proteins are regulated by translational and/or post-translational mechanisms in response to zinc. The intracellular trafficking of the human Menkes' copper transporter is regulated by copper (53). Under conditions of low copper, the protein is localized in the trans-Golgi network; but when copper levels are elevated, the protein rapidly relocalizes to the plasma membrane (54 -56). Whether the intracellular trafficking of these mouse ZIP proteins is zinc-responsive is under investigation. It is also possible that the uptake activity of these ZIP proteins may be regulated, possibly through post-translational modifications such as phosphorylation, circumventing the need for regulation at the RNA level. The structural basis for the zinc uptake activity of ZIP proteins is not understood.
The data presented herein describe a subfamily of mouse zip genes that can function as zinc transporters and whose expression is differentially regulated in a tissue-specific manner. These results suggest that these genes may function in mammalian zinc homeostasis, and genetic manipulation of these genes in the mouse model system will allow us to directly test this hypothesis.