The biological significance of zinc can be appreciated given that proteins with zinc-binding domains are estimated at 3–10% of the human proteome (1). Zinc is utilized by numerous enzymes and other proteins as a catalytic or structural component and accordingly contributes to a variety of fundamental biological processes (2, 3).
Zinc homeostasis is tightly regulated at both the cellular and organismal levels (4, 5). Mechanisms to maintain body zinc balance are largely efficient, but chronic low zinc intakes, physiological, and pathological stimuli or genetic defects could disturb this balance and cause abnormalities of fast-turnover tissues and compromised immune and neuronal functions (2, 6).
The gastrointestinal system is central to systemic zinc homeostasis because it serves as the interface of zinc exchange between the organism and environment. Dietary zinc is primarily absorbed in the small intestine and discharged from the body via pancreatic and intestinal excretions (5, 7). Zinc uptake in the intestine exhibits both unsaturable and saturable kinetics, and the latter suggests a carrier-mediated process (8, 9). Our understanding of mammalian zinc transport at the molecular level was initiated by the identification of ZnT1 and rapidly enriched through the studies of a series of zinc transporters in cell culture, rodent models, and diseases linked to their deficiency (10, 11). Briefly, mammalian zinc transporters largely fall into two conserved families, the ZnT (SLC30) family and the ZIP (Zrt- and Irt-like protein, SLC39) family. Members of the ZnT family mediate zinc efflux or sequestration into organelles/vesicles, whereas those of the ZIP family transport extracellular or organellar/vesicular zinc into cytoplasm (10, 11).
Dietary zinc enters the polarized enterocytes through the apical membrane and is released at the basolateral side into circulation. The presence of a variety of zinc transporters in enterocytes has been manifested at the protein level, which includes ZnT1, ZnT2, ZnT4, ZnT5, ZnT6, ZnT7, Zip4, and Zip5 (12–16). This review summarizes our present knowledge of enterocyte-expressed zinc transporters in terms of their coordinated functions in the aforementioned trans-epithelial movement of zinc.
ZINC UPTAKE ACROSS THE APICAL MEMBRANE INTO ENTEROCYTES
Acrodermatitis enteropathica (AE) is an autosomal recessive human disorder of zinc malabsorption. Patients suffering from AE display classic symptoms of zinc deficiency including dermatitis, diarrhea, growth retardation, immune dysfunctions, and occasionally, neuropsychological disturbances, which could be ameliorated by oral zinc supplementation (17, 18). Human ZIP4 has been identified as the gene responsible for AE (19, 20). Further characterizations found that Zip4 mRNA was abundantly expressed throughout the small intestine tract in both human and mice (15, 19, 20). The expression of mouse Zip4 was highly responsive to dietary zinc, exhibiting up-regulation under zinc deficiency and down-regulation under increased zinc concentration at the mRNA level (15). Immunohistochemistry localized mouse Zip4 to the apical membrane of the enterocytes and, in zinc-replete conditions, Zip4 protein was largely internalized and degraded (15, 19, 21). In cultured cells, Zip4 mediated zinc influx with high specificity and saturable kinetics (15). The aforementioned evidence strongly supports the role of Zip4 in zinc uptake into enterocytes across the apical membrane.
Homozygous Zip4 knockout mice died during early embryogenesis. More than 40% of the heterozygotes were dead or born with morphological abnormalities, which were exacerbated by zinc deficiency and ameliorated by zinc supplement (22). In visceral yolk sac Zip4 was distributed on the apical side of visceral endoderm cells, suggesting a role in maternal zinc transport to the fetus (15).
Recent studies have provided insights into the mechanisms of zinc-responsive ZIP4 regulation. It was reported that zinc status had little effect on transcription of Zip4 but rather regulated mRNA stability both in cultured Hepa cells and mouse intestine (21). In contrast, a recent study found that Krüppel-like factor 4 bound to Zip4 promoter and its knockdown impaired Zip4 induction by zinc depletion in a mouse intestinal epithelial cell line (23). At the protein level, apart from endocytosis stimulated by low micromolar zinc (24), ubiquitination and degradation of ZIP4 were observed under higher zinc levels (25). While during prolonged zinc deficiency, the extracellular amino-terminal domain of Zip4 underwent proteolytic cleavage while the remaining part accumulated at the plasma membrane (26).
ZINC RELEASE FROM ENTEROCYTES INTO CIRCULATION
To reach the blood stream for systemic supply, zinc ions must be released from the basolateral side of the polarized enterocytes. The molecular basis underlying this process was proposed to be ZnT1, which directly mediates zinc efflux across the basolateral membrane. Supporting evidence in mammalian systems largely came from functional analysis at the cellular level and immunolocalization studies. In mammalian cell culture, ZnT1 was localized primarily on the plasma membrane and it mediated zinc export to reduce intracellular zinc levels (27, 28). The anatomical and subcellular distribution of endogenous ZnT1 was also in good correlation with a role in dietary zinc absorption. Rodent ZnT1 was most abundant in the proximal small intestine (12, 14), the primary site for zinc absorption (29, 30). Intensive ZnT1 expression was detected in absorptive epithelial cells in the villus, in contrast to weak or undetectable expression in mucos-secreting goblets cells, cells in the crypts or lamina propria (12, 14). Furthermore, ZnT1 was distributed predominantly on the basolateral membrane of the enterocytes (12, 14). Intracellular punctate localization of ZnT1 was also found in the enterocytes but its physiological significance remains undetermined (13). Dietary zinc supplementation induced both ZnT1 mRNA and protein levels whereas zinc deficiency did not significantly affect ZnT1 expression in rat small intestines (12, 31). It has been shown that MTF-1 binds to the metal-responsive elements in ZnT1 promoter and mediates its regulation in response to zinc (32).
Znt1 knockout in mice resulted in early embryonic lethality at a similar stage to that of Zip4−/− embryos. Heterozygote Znt1 knockout mice developed normally but displayed higher risk of conceiving abnormal embryos under dietary zinc deficiency (33). Tissue-specific inactivation of Znt1 or Zip4 in intestinal epithelium is needed to avoid the early lethality and to further characterize their in vivo functions in dietary zinc absorption.
Inspiringly, a recent study in Drosophila provided the genetic evidence for the participation of ZnT1 in dietary zinc absorption (34), which will be discussed later in this review.
INTRACELLULAR ZINC TRAFFICKING IN ENTEROCYTES
The intracellular trafficking of zinc from the apical cytoplasm to the basolateral cytoplasm of the enterocytes remains largely uncharacterized. Another question of great interest is to unveil other putative pathways utilized by the enterocytes to release zinc into circulation, For example, vesicles-mediated exocytosis, in addition to direct pumping across the basolateral membrane. A variety of intracellular zinc transporters are found to be expressed in enterocytes and our current understanding of their functions in dietary zinc absorption is described in this part (Fig. 1, Table 1). The contribution of metallothionein to intracellular zinc trafficking is also briefly reviewed here. In addition, zinc can be released from intracellular storage in response to stimuli, as evidenced by oxidizing agents-induced zinc liberation from metallothionein (35), zinc release from the perinuclear region including the endoplasmic reticulum induced by IgE receptor crosslinking (36), and zinc mobilization from the mitochondrial pool by depolarization or rapid calcium entry (37).
Table 1. Mammalian zinc transporters with manifested expression in intestinal absorptive cells
Anatomical distribution along the intestinal tract
Restricted to the villus instead of the crypts, abundant in the large intestine
Zinc-deficient milk in lm mice
Golgi apparatus (variant A), apical membrane of the enterocytes
Mainly in duodenum and jejunum
Poor growth, osteopenia, male-specific bradyarrhythmias in knockout mice
TGN, apical membrane of the enterocytes
Jejunum, the large intestine
The entire intestinal tract, most abundant in the small intestine
Poor appetite, reduced body weight gain, low adiposity in knockout mice
Zinc Transporters in the Golgi Apparatus
ZnT7 was reported to be predominantly located in the Golgi apparatus and its overexpression led to the accumulation of chelatable zinc in this organelle (38). ZnT7 is also required for the activity of alkaline phosphate, suggesting zinc in the Golgi lumen is incorporated into zinc-requiring secretory and membrane-bound enzymes (39, 40). The expression pattern of ZnT7 suggests a role in dietary zinc absorption. First, ZnT7 protein was restricted to lung and small intestine despite the ubiquitous expression of its transcripts (38). Second, cytoplasmic immunolocalization of ZnT7 was detected along the entire mouse gastrointestinal tract with the highest abundance in duodenum and jejunum (14). Znt7 null mutant mice exhibited poor appetite and reduced body weight gain due to low adiposity, which intriguingly were not alleviated by zinc supplementation (41). The mice did not show hair and skin abnormality either, which are typical symptoms under zinc deficiency. Znt7−/− mice manifested lower serum and tissue-associated zinc and reduced ability to accumulate 65Zn in the small intestine, liver, and kidney in an isotope feeding assay. The low zinc status in the liver was an indication of impaired systemic zinc supply, as ZnT7 protein was not expressed in this tissue. The author thus concluded that ZnT7-mediated zinc incorporation into Golgi is essential for dietary zinc absorption. However, it is unlikely that ZnT7 is directly involved in zinc efflux from enterocytes into circulation as embryonic fibroblasts from Znt7−/− mice displayed enhanced 65Zn efflux with unaffected 65Zn uptake that together caused reduced 65Zn accumulation (41). Overall, the role of ZnT7 in zinc homeostasis needs to be further clarified.
ZnT5 was expressed in mouse small intestine epithelium, exhibiting a similar immunostaining pattern with that of ZnT7 in duodenum and jejunum (14). The antibody used was against a ZnT5 variant A-specific peptide. Like ZnT7, ZnT5A transported zinc into Golgi and contributed to the activity of zinc-requiring enzymes (39, 42, 43). It is possible that ZnT5A and ZnT7 execute partially overlapping functions in intestinal zinc homeostasis. Unlike typical ZnT members, the smaller variant ZnT5B mediated bidirectional zinc transport as manifested by enhanced zinc influx and efflux in ZnT5B-expressing Xenopus oocytes (44). Immunolocalization of ZnT5 (with an antibody unable to distinguish between variant A and B) to the apical membrane of enterocytes was reported in human small intestine (45); however, its physiological significance is obscure. Znt5-null mice displayed poor growth, osteopenia, and male-specific fatal bradyarrhythmias. Nevertheless, their serum zinc levels were normal (46).
It was reported that ZnT6 formed a heteromeric complex with ZnT5A to activate zinc-requiring enzymes in the secretory pathway (40). Upon elevated zinc concentrations, ZnT6 exhibited translocation from the Golgi apparatus to periphery vesicles (47). ZnT6 was detected on the apical membrane as well as intracellularly in absorptive epithelial cells of mouse jejunum, cecum, and colon (14). ZnT6 may function in zinc excretion from enterocytes into lumen via exocytosis.
Zip7, Zip9, and Zip13
Zip7 (KE4) was localized in the Golgi apparatus and antagonized zinc accumulation in Golgi caused by ZnT7 overexpression (48). It was proposed that Zip7 transports zinc out of the Golgi apparatus to increase cytoplasmic zinc levels (48, 49). Zip7 mRNA was ubiquitously expressed in human and mouse tissues including small intestine, whereas the tissue distribution of this protein was not characterized (48, 49). Recent studies have revealed two other Golgi-residing Zip transporters, Zip9 and Zip13 (50, 51). No information was available regarding their expression in small intestine. Zip13-deficient mice showed defects in bone, teeth, and connective tissues but exhibited normal serum zinc levels (51).
Zinc Transporters in Vesicles
Mutation in Znt4 underlies the inherited zinc deficiency “lethal milk (lm)” in mice (52). Pups fostered by homozygous lm dams died in neonatal life (53). lm dams produced zinc-deficient milk due to impaired zinc sequestration into milk-secreting vesicles, where Znt4 was localized (54–56). In cell types other than the mammary epithelium, ZnT4 has been localized to endosomal compartments or trans-Golgi network and intracellular vesicles (47, 57).
Although the exact mechanism remains a mystery, several pieces of evidence indicate that ZnT4 may participate in dietary zinc absorption. First, lm mice suckled on wild type dams exhibited normal zinc homeostasis until 8 months of age, when they began to exhibit signs of zinc deficiency including alopecia, dermatitis, and premature sterility (58, 59). Second, ZnT4 expression in rodent small intestine was restricted to epithelium of the villus compared to the crypts (14, 57), and the immunolocalization of ZnT4 in enterocytes was primarily basolateral in growing rats (57). Third, lm mice exhibited decreased level of Zip4 transcripts and induction of metallothionein in jejunum, indicating increased cytoplasmic zinc levels (59). Finally, ZnT4 translocated from Golgi to peripheral vesicles upon elevated zinc concentrations in the medium (47), and in cultured fibroblasts, a fraction of ZnT4 was detected on the plasma membrane (60). It was therefore proposed that ZnT4 promotes zinc secretion into the blood stream via exocytosis. The zinc transport activity of ZnT4 has been proved by its virtue to rescue the zinc sensitivity of zrc1 yeast (52), but there is no solid evidence supporting ZnT4-mediated zinc secretion in cell types other than the breast epithelium (13, 52, 56). The contribution of ZnT4 to dietary zinc absorption could be limited considering zinc homeostasis appeared normal in young lm mice (58), and notably, a recent immunohistochemical study in young adult mice pointed out that ZnT4 expression in the proximal intestine is rather weak, compared to that in the colon and rectum (14).
ZnT2 mRNA has been detected in small intestine, which is suppressed to barely detectable levels by zinc deficiency and induced by zinc repletion (31). In rat duodenum, ZnT2 was localized to vesicles on the apical side of the enterocytes and was not detected in crypts or lamina propria (13). In cell culture, ZnT2 sequestered zinc into endosome/lysosome-like vesicles (61). ZnT2 might be involved in a storage or excretion mechanism in enterocytes with high zinc influx. Mutations in ZNT2 have been found in some patients with zinc-deficient milk (62).
Chaperones in Cytoplasm for Zinc Delivery
Intracellular zinc is largely bound to proteins or sequestered in organelles, leaving the cytoplasmic labile zinc at very low levels (63). Although it is well characterized in copper homeostasis (64), how the intracellular process of zinc trafficking, particularly in cytoplasm, happens to remain largely unknown. Is it mediated by free diffusion or a specific set of chaperones? There are some indications suggesting that metallothioneins may play roles in helping zinc delivery to zinc-requiring enzymes. Metallothionein binds zinc in two separate clusters with different affinities and serves a dual role in zinc detoxification and also zinc reserve as a bioavailable pool. The function of metallothionein as a zinc donor has been documented for several zinc metalloproteins (35), however, the viability of metallothionein knockout mice suggests the presence of other pathways for zinc incorporation into essential apoproteins (65). Despite prominent expression in the intestine, the role of metallothionein in dietary zinc absorption remains to be elucidated. It was even shown that after an oral gavage of zinc on metallothionein knockout and transgenic mice, the rise in serum zinc was inversely related to metallothionein protein levels (66).
OTHER ZINC TRANSPORTERS INVOLVED IN INTESTINAL ZINC HOMEOSTASIS
Zip5 was localized on the basolateral membrane of enterocytes and was internalized and degraded during dietary zinc deficiency (16, 21). Combining its manifested activity in zinc influx (67), it was suggested that Zip5 transports zinc from blood stream into enterocytes for ultimate excretion into the intestinal lumen under zinc-replete conditions.
Abundant Zip14 mRNA was found in duodenum and jejunum (68). Zip14 mediated influx of Zn2+ and Fe2+ (68) and was proposed to account for hypozincemia of inflammation (69) and hepatic loading of nontransferrin bound iron (68). However, in another report, Zip14 was reported to mediate Cd2+ uptake, which was inhibited by Zn2+, Mn2+, and Cu2+ but not by Fe2+ or Fe3+ (70). The significance of Zip14 in intestinal zinc homeostasis waits to be determined.
DROSOPHILA AS A MODEL TO STUDY DIETARY ZINC ABSORPTION
We recently reported the use of Drosophila model in elucidating the molecular basis of zinc efflux across the basolateral membrane of enterocytes (34). It was shown that ubiquitous or gut-specific knockdown of Drosophila ZnT1 resulted in growth retardation and developmental arrest under zinc deficiency, which could be complemented by human ZnT1. Endogenous dZnT1 resides on the basolateral membrane of the enterocytes and mediates zinc efflux for systemic supply. Thus the mechanism for zinc release from enterocytes into circulation appeared largely conserved between Drosophila and mammals. In addition, we recently identified a Drosophila zinc transporter of the Zip family, which resides on the apical membrane of the enterocytes and is required for dietary zinc absorption (our unpublished results).
Six members of the ZnT family and eight numbers of the ZIP family are found in the Drosophila proteome, displaying homology to mammalian zinc transporters localized to the plasma membrane or intracellular compartments. Knockout of Zip4 or Znt1 leads to early embryonic lethality in mice, while in Drosophila, it is readily accessible to selectively modulate or inactivate zinc transporters in various tissues including the gut using the RNAi strategy and UAS/GAL4 system (34). Drosophila melanogaster could therefore serve as a tractable model to study how an array of zinc transporters functions in concert to fulfill dietary zinc absorption, and to a wider scope, the systemic zinc homeostasis.
The authors apologize to authors whose works are not cited here due to space limitation. Xiaona Tang provided technical help in the art work. Our research was supported by the National Basic Research Program of China (#2005CB522503). Bing Zhou is a recipient of the National Outstanding Youth Grant (#30688001) from the National Scientific Foundation of China.