zinc (Zn) is required for a multitude of cellular mechanisms including cell growth and division, apoptosis, and maintenance of DNA integrity (38). As a consequence, Zn is critical for many physiological processes such as immune function (11), reproduction (1, 17), and growth (18). Approximately 1–3 mg Zn is transferred from the mammary gland into milk each day to provide optimal Zn nutriture for the developing infant (34). In humans, suboptimal Zn transfer to the developing neonate during lactation results in severe adverse effects including growth retardation, compromised cognitive development, and increased morbidity and mortality (4, 39). Therefore, optimal Zn transfer from the mammary gland into milk is an imperative physiological process to ensure the growth and health of the infant. A role for the Zn transporter ZnT2 has been delineated in Zn transfer into milk as a mutation in the gene that encodes ZnT2 (SLC30A2) decreases milk Zn concentration by ∼75% in lactating women (7) and two single nucleotide polymorphisms in ZnT2 alter Zn secretion from mammary cells (37). In addition, a spontaneous truncation mutation in the gene that encodes ZnT4 (SLC30A4) is responsible for the “lethal milk” (lm/lm) syndrome in mice (15). The lm/lm syndrome is associated with decreased (∼35%) milk Zn concentration and results in the death of suckled offspring. Additionally, lm/lm mice have lower milk volume and mammary gland weight compared with wild-type littermates (35), suggesting a critical role for ZnT4 in mammary gland Zn metabolism well beyond a discrete role for Zn secretion into milk. Deciphering the role of ZnT4 is critical for understanding the complex Zn transporting network in the lactating mammary gland.

Studies to explore the role of ZnT4 in mammary gland Zn metabolism have been limited, and the precise physiological function of ZnT4 remains to be elucidated. After the initial identification of ZnT4 in the mouse mammary epithelium (15), Michalczyk et al. (29) noted the presence of ZnT4 in the luminal mammary cells taken from resting and lactating human breast biopsies where it displays a granular and cytoplasmic distribution. Consistent with this observation, we found that ZnT4 was expressed in the mammary gland of the lactating rat and localized to an intracellular compartment (20). Most recently, we have shown that ZnT4 protein is not detected in mouse mammary gland until lactation and was localized in close approximation to the apical membrane in lactating mammary glands (28). A study by Huang et al. (16) elegantly demonstrated ZnT4 localization to the trans-Golgi network (TGN) in normal rat kidney cells, which was redistributed in response to Zn. This may be of particular importance in the mammary gland given the expansion of Zn pools that occurs in the Golgi apparatus of mammary epithelial cells and the need to secrete tremendous Zn into milk during lactation (28). Because of the vital role of the Golgi apparatus in the secretory process (8, 24, 32) including the production of lactose and the modification of proteins that are secreted into milk, these studies suggest that ZnT4 bears the potential to contribute to mammary epithelial cell functions that are important in the production of milk.

Herein, our data indicated that ZnT4 directly transports Zn from the cytoplasm into the TGN of mammary epithelial cells and can be relocalized to the cell membrane in response to Zn. ZnT4-mediated Zn import into the TGN may provide Zn for key Zn-requiring enzymes such as galactosyltransferase and carbonic anhydrase VI, which are either resident (5) or secreted from the Golgi apparatus into milk (23), respectively. To our knowledge, our study is the first to directly characterize the function of ZnT4 as well as define its contribution to the process of Zn secretion from the mammary gland during lactation.

MATERIALS AND METHODS

Cell culture.

HC11 cells were a gift from Dr. Jeffrey Rosen (Department of Molecular and Cellular Biology, Baylor College of Medicine, Houston, TX) and used with permission of Dr. Bernd Groner (Institute for Biomedical Research, Frankfurt, Germany). HC11 cells are a COMMA-D-derived cell line from the mammary gland of midpregnant BALB/c mice (2). Cells were routinely maintained in a nonsecretory phenotype as previously described (26).

Generation of plasmid DNA constructs and expression of ZnT4-hemagglutinin fusion proteins.

SLC30A4 DNA was obtained from a cDNA clone (accession number, BC117997; Open Biosystems) and COOH-terminally tagged with hemagglutinin (HA) using methods previously described (7) to produce pcDNA3.1-ZnT4HA. The orientation and fidelity of the insert and incorporation of the HA tag were confirmed by directed sequencing (The Nucleic Acid Facility at Pennsylvania State University). Large-scale plasmid purification was carried out using the Plasmid Midi Kit (Qiagen). To generate cells expressing ZnT4-HA, cells were plated in antibiotic-free Opti-MEM (Invitrogen) in 6-well plates (for cell surface biotinylation and luciferase reporter assays), 24-well plates (confocal microscopy), or 96-well plates (for fluorometry assays). Cells were cultured overnight until ∼95% confluent and transiently transfected with 0.2 μg (96-well plates), 0.8 μg (24-well plates), or 4 μg (6-well plates) of pcDNA3.1-ZnT4HA in antibiotic-free Opti-MEM using Lipofectamine 2000 (Invitrogen) as previously described (26).

Immunoblotting.

Cells were washed in PBS, scraped into HEPES-based lysis buffer containing protease inhibitor, and sonicated for 20 s as previously described (21). Cellular debris and nuclei were pelleted by centrifugation at 500 g for 5 min. Supernatant was centrifuged at 100,000 g for 20 min, and total membrane pellet was resuspended in lysis buffer. Protein concentration was determined by Bradford assay. Total membrane protein (50–100 μg) was diluted in Laemmli sample buffer containing 100 mM dithiothrietol and incubated at 95°C for 5 min. Total membrane protein extracts were separated by electrophoresis, transferred to nitrocellulose membrane, and immunoblotted with anti-HA (0.5 μg/ml; Invitrogen), anti-β-actin (1:10,000; Sigma-Aldrich), anti-galactosyltransferase (1:1,000; Aviva), or anti-carbonic anhydrase-VI (1:200; Santa Cruz) antibodies and detected with horseradish peroxidase conjugates as previously described (7). Proteins were visualized by chemiluminescence after exposure to autoradiography film.

Small interfering RNA-mediated gene attenuation.

Cells were plated in antibiotic-free OPTI-MEM in six-well plates and cultured until ∼50% confluent. Cells were transfected with 100 pmol of ZnT4-specific small interfering RNA (sense, 5′-GCUAAUUCCUGGAAGUUCA-3′; antisense, 5′-UGAACUUCCAGGAAUUAGC-3′; Sigma-Aldrich) or mismatched control small interfering RNA (sense, 5′-CCGCGUCCUUCCUUAUGUAGGAAUU-3′; antisense, 5′-AAUUCCUACAUAAGGAAGGACGCGG-3′; Invitrogen) using Lipofectamine 2000 at an oligonucleotide/transfection reagent ratio of 25:1 for 48 h before experiments.

Subcellular localization of endogenous ZnT4 and ZnT4-HA.

To visualize the subcellular localization of endogenous ZnT4 in mammary epithelial cells, cells were seeded onto glass coverslips and cultured overnight until ∼50–80% confluent. Cells were fixed and blocked as described below, followed by incubation with affinity purified ZnT4 antibody (1 μg/ml) (20). Cells were washed with PBS, and ZnT4 antibody was detected with Alexa Fluor 488-conjugated anti-rabbit IgG (1 μg/ml; Invitrogen) for 45 min at room temperature, shielded from light. Cells were washed and mounted, and coverslips were sealed.

To determine the localization of ZnT4 in mammary epithelial cells, cells expressing ZnT4-HA were plated on glass coverslips in 24-well dishes and grown to ∼95% confluency. Cells were fixed in cold methanol (70% in PBS) for 10 min, washed briefly in PBS, and permeabilized with Triton X-100 (0.2% in PBS) for 5 min. Detection of ZnT4-HA was achieved by blocking nonspecific binding sites with 4% BSA in PBS, followed by incubation with Alexa Fluor 488-conjugated HA (1 μg/ml, Invitrogen) for 1 h at room temperature and shielded from light. Detection of p58 (TGN marker) was achieved by blocking nonspecific binding sites with 5% goat serum and 1% bovine serum albumin in PBS for 30 min followed by incubation with p58 antibody (1:100 dilution in blocking buffer with the addition of 0.5% Triton-X100; Sigma-Aldrich) for 1 h at room temperature and subsequent detection with Alexa Fluor 568-conjugated goat anti-mouse secondary antibody. Cells were washed and mounted, and coverslips were sealed.

To visualize the spatial redistribution of ZnT4-HA in response to Zn in some experiments, cells were treated with 200 μM Zn for 2 h and ZnT4-HA localization was detected as described above. Nuclei were detected using TO-PRO-3 (1:1,000 dilution in PBS; Invitrogen). Cells were imaged as previously described using an Olympus BX50WI microscope using a ×100 oil lens, and digital images were captured sequentially (LaserSharp2000, version 4.1; Bio-Rad) to eliminate potential interference between fluorochromes (26).

Cell surface biotinylation.

To determine the presence of ZnT4-HA at the cell surface in response to Zn, HC11 cells were seeded in antibiotic-free media and cultured until ∼95% confluent. Cells were then generated to express ZnT4-HA as described above and subsequently treated with Zn (200 μM) for 2 h, and changes in ZnT4 abundance at the cell membrane were assessed. Briefly, cell membrane proteins were labeled with N-hydroxy-sulfosuccinimide biotin (Pierce; 0.5 mg/ml) at room temperature for 30 min. Cells were then scraped into lysis buffer, and biotinylated cell membrane proteins were isolated using Ultralink-neutravidin beads (Pierce) as previously described (26). Biotinylated proteins were then eluted with the use of Laemmli buffer containing dithiothrietol (100 mM), separated by electrophoresis, and immunoblotted with anti-HA antibody as described above.

Cytosolic Zn pool measurement.

To determine the effect of ZnT4 overexpression on cytosolic Zn pools, HC11 cells were seeded in antibiotic-free growth medium in 96-well optical bottom plates and cultured until 90–95% confluent. HC11 cells were transfected using Lipofectamine 2000 as described above with thymidine kinase promoter-linked Renilla luciferase vector (internal control, 0.05 μg) and either pGL3 empty vector (0.8 μg) plus 4×-metal responsive element (MRE)-pGL3 (a luciferase reporter containing 4 MREs from the mouse metallothionein 1A promoter upstream of the firefly luciferase open reading frame kindly provided by Dr. Colin Duckett, Univ. of Michigan Medical School, Ann Arbor, MI; 0.8 μg) as previously described; (27) or pcDNA3.1ZnT4-HA (0.2 μg) plus 4×-MRE-pGL3 for 24 h before experiments. After 48 h, the cells were rinsed with PBS and harvested in passive lysis buffer (Promega, Madison, WI), following the manufacturer's instructions. The dual-luciferase reporter assay system (Promega) was used to measure luminescence (Turner Biosystems, Sunnyvale, CA) for firefly and Renilla luciferase activity (internal control). Data were expressed as relative light units (ratio of firefly:renilla luciferase activity).

FluoZin3 fluorometry assay.

To determine whether ZnT4-HA facilitates labile Zn pool accumulation, cells were seeded in antibiotic-free growth medium in 96-well optical bottom plates and cultured until 90–95% confluent. Cells were transfected as described above for 24 h before experiments. Cells were rinsed with 1× PBS and loaded with FluoZin3-AM (1 μM, Invitrogen) for 1 h at 37°C. Cells were then rinsed with 1× PBS for 30 min at 25°C with constant shaking. Fluorescence of FluoZin-3 (emission, 495 nm; and excitation, 516 nm) was measured at 25°C using a FLUOstar OPTIMA plate reader (BMG Labtech) spectrofluorimeter with FLUOstar OPTIMA software version 1.32R2. Cellular protein concentration was determined using the Bradford assay, and fluorescence measurements were normalized to total protein concentration.

RNA isolation and relative RT-PCR analysis.

Total RNA was extracted from HC11 cells using TRIzol reagent (Invitrogen, Carlsbad, CA). Complementary DNA was synthesized using the ImPromII reverse transcription system (Promega). Mouse carbonic anhydrase VI and β-actin gene-specific primers were designed using Primer3 and synthesized by the Genomics Facility at the Pennsylvania State University (carbonic anhydrase VI, 5′-GCAGTATCCTTCCTGCGGCGG-3′ and 5′-ACATGGAGGGCGGCAGATCG-3′; and β-actin: 5′-AGCCATGTACGTAGCCATCC-3′ and 5′-CTCTCAGCTGTGGTGGTGAA-3′). Real-time PCR was performed with SYBR Green Supermix (PerkinElmer Applied Biosystems, Foster City, CA) using the MJ Research Opticon 2 system (MJ Research, Waltham, MA). cDNA synthesis, real-time relative PCR and data analysis were performed as previously described (36). Each sample was analyzed in triplicate and normalized to β-actin using the equation previously described (22).

Statistical analysis.

Results are presented as means ± SD. Statistical comparisons were formed using Student's t-test on Prism Graph Pad (version 5.00, La Jolla, CA) and a significant difference was demonstrated at P < 0.05. Pearson's coefficient was determined using Olympus Fluoview (v2.0c, Tokyo, Japan).

RESULTS

Previous studies have detected ZnT4 in the TGN in normal rat kidney cells (16) within distinct perinuclear vesicles in Caco-2 cells (31) and in a perinuclear location in mammary epithelial cells (20). Because ZnT4 is not expressed before lactation in mammary epithelial cells (28) and Zn pools in the Golgi apparatus of mammary epithelial cells significantly increase during lactation (28), we postulated that ZnT4 is responsible for Zn transport into the Golgi apparatus of mammary epithelial cells during lactation, which provides Zn for functions specific to secretion. To determine the subcellular localization of ZnT4 within mammary epithelial cells, we used confocal microscopy to first observe the spatial distribution of endogenous ZnT4 within HC11 cells. Visualization of endogenous ZnT4 (Fig. 1A) clearly indicated a perinuclear distribution of ZnT4, similar to what has been previously observed (16, 20), as well as a distinct appearance on the cell membrane. To determine whether ZnT4 is specifically localized to the Golgi apparatus, we constructed a ZnT4-HA fusion protein. We generated HC11 cells to express ZnT4-HA (Fig. 1B), and using confocal microscopy, we determined that ZnT4 colocalized with the TGN marker p58 (Fig. 1C). The Pearson's coefficient for colocalization between ZnT4-HA and p58 was 0.98, indicating that ZnT4 is almost entirely localized within the TGN.

To confirm that ZnT4 draws on cytoplasmic Zn pools to transport Zn into the TGN, we first used a Zn-responsive luciferase reporter assay to determine changes in cytosolic Zn pool levels following ZnT4 overexpression. As noted in Fig. 2A, ZnT4 overexpression significantly decreased cytoplasmic Zn pool concentration (P < 0.001) relative to mock transfected cells, confirming that ZnT4 draws on cytoplasmic Zn pools for Zn transport into the TGN. To determine whether ZnT4 increases labile Zn pools or provides Zn for metallation of proteins in the TGN (13), we next used FluoZin3, a Zn specific fluorophore that has a K_d of 4–15 nM and fluoresces upon Zn binding. Our data indicated that ZnT4 overexpression significantly increased FluoZin3 fluorescence by ∼1.3-fold (P = 0.013) relative to mock-transfected cells (Fig. 2B), suggesting that labile Zn pools (Zn bound with affinity less than 4–15 nM) were increased.

Based on the subcellular localization of ZnT4, we hypothesized that ZnT4-mediated Zn transport into the TGN provides Zn for critical functions specific to secretion in mammary epithelial cells. To test this hypothesis, we first used the activity of galactosyltransferase, a biochemical marker of changes of Zn metallation of resident proteins in the TGN. Galactosyltransferase catalyzes the transfer of galactose to glucose to yield lactose. Zn is required for galactosyltransferase activity, but it has been demonstrated that excess Zn decreases the K_m and V_max for both glucose and UDP-galactose in the lactose synthase reaction, which decreases galactosyltransferase activity (33). Therefore, we used galactosyltrasferase activity as a biomarker for Zn metallation of resident proteins in the Golgi apparatus and postulated that Zn import into the TGN would result in decreased galactosyltransferase activity. Indeed, as noted in Fig. 3A, ZnT4 overexpression in mammary epithelial cells significantly (P < 0.05) decreased galactosyltransferase activity by ∼1.5-fold relative to mock-transfected cells. ZnT4 overexpression did not change galactosyltransferase protein abundance, confirming that the change in galactosyltransferase activity was not due to decreased protein abundance (Fig. 3B).

In addition, metallation of Zn-dependent enzymes occurs in the TGN and is necessary for their conformational stability and subsequent catalytic activity (40). To determine whether ZnT4 provides Zn for metallation of Zn-dependent proteins that are secreted from the Golgi apparatus, ZnT4 expression was attenuated in mammary epithelial cells and the abundance of the Zn-dependent enzyme carbonic anhydrase VI (23) was measured by immunoblotting. As noted in Fig. 4, ZnT4 attenuation significantly decreased the abundance of carbonic anhydrase VI (P = 0.002), while not altering carbonic anhydrase-VI mRNA expression, suggesting that ZnT4 is vital for providing Zn for metallation of Zn-dependent proteins that are secreted from the TGN. We speculate that the limited effect of ZnT4 overexpression on FluoZin3 fluorescence reflects the dual role of ZnT4-mediated Zn accumulation in the Golgi apparatus for both labile and nonlabile pools.

We have previously determined that ZnT4 is localized both intracellularly and at the periphery of mammary epithelial cells in the mammary gland of lactating mice (22). This suggests that ZnT4 may be relocalized to the apical membrane in response to intra-/extracellular cues. We found that ZnT4 was redistributed in response to exogenous Zn exposure (Fig. 5A), consistent with our previous observations (20). We further used cell surface biotinylation and determined that Zn exposure increased ZnT4 abundance at the cell membrane by 80% (Fig. 5B) in mammary epithelial cells, consistent with observations in MCF7 (malignant human breast cells) and NIH/3T3 (mouse fibroblast cells) (14).

DISCUSSION

The phenotypic conversion of the mammary gland from a resting tissue to a secreting organ incurs an abundance of dynamic changes inherent to the needs of milk synthesis and mammary gland function. One critical change during lactation includes Zn accumulation within mammary epithelial cells, particularly in the Golgi apparatus (28), which reflects the need to secrete a profound amount of Zn (1–3 mg Zn/day) into milk (34). Our recent work indicates that Zn accumulation during secretory transition is associated with increased ZnT4 expression (22). The lack of ZnT4 protein before lactation that increases throughout lactation (20, 25) suggests that ZnT4 function provides Zn for functions specific to secretion in the lactating mammary gland. Consistent with a critical role, a mutation in the gene encoding ZnT4 (SLC30A4) results in a phenotype known as the “lethal milk” syndrome (lm/lm) (15). A C > T substitution at base 934 in SLC30A4 substitutes an arginine residue for a premature stop codon and results in the synthesis of a truncated, nonfunctional ZnT4 protein (15). Because lm/lm mice have ∼35% lower-milk Zn concentration compared with wild-type mice, it has been presumed that ZnT4 directly provides Zn for secretion into milk. However, there are several other key physiological outcomes that have been described in lm/lm mice including decreased mammary gland size and decreased milk volume (35). These observations are consistent with a key role for ZnT4 in providing not only Zn specifically for secretion into milk but also overall mammary gland function and the general process of secretion.

The Golgi apparatus is a subcellular compartment central to the process of secretion in mammary epithelial cells. It serves as both a storage pool and a sorting station for further modification and trafficking of secreted milk components including proteins (6, 41), lactose (9, 19), and minerals (3, 28). We recently identified the Golgi apparatus as the subcellular location of a mobilizable and labile Zn pool that expands during lactation (28). Our data herein suggest that Zn accumulation within the Golgi apparatus is modulated, at least in part, by ZnT4 and that the role of ZnT4-mediated Zn transport may be multifactorial. Our data suggest that ZnT4 may provide Zn for metallation of Zn-dependent milk proteins. One Zn-binding protein secreted into milk is carbonic anhydrase VI (23). Zn is critical for carbonic anhydrase-VI protein stability (40). The decrease in carbonic anhydrase-VI protein abundance, despite maintained carbonic anhydrase-VI mRNA expression observed following ZnT4 attenuation, suggests that ZnT4-mediated Zn transport is necessary for carbonic anhydrase-VI stability. In addition, ZnT4 may provide Zn for more labile Zn pools. We found that ZnT4 overexpression increased FluoZin3 fluorescence. This suggests that ZnT4-mediated Zn transport may also provide Zn to small molecular weight compounds such as citrate for secretion into milk. These observations improve our understanding of the process through which Zn is secreted into milk. We have previously shown that ZnT2 significantly contributes to Zn secretion from the mammary gland as evidenced by the significant decrease in milk Zn concentration (∼75%) that occurs in women with a mutation in ZnT2 (7), as well as our identification of two single nucleotide polymorphisms in ZnT2 that modulate Zn secretion in cultured mammary cells (37). Together this suggests that ZnT4 and ZnT2 share a complementary relationship in terms of Zn secretion from the mammary gland and may functionally contribute to separate subcellular Zn pools. Taken together, these data indicated that ZnT4-mediated Golgi Zn accumulation in mammary epithelial cells during lactation is directed to several different intracellular Zn pools for secretion.

The Golgi apparatus plays a critical role in the process of secretion (8, 24, 32) and numerous resident proteins such as galactosyltransferase and metalloproteases are Zn dependent (30, 33). We found that changes in ZnT4 abundance modulated the activity of galactosyltransferase. This enzyme is vital to the production of lactose, a component of milk that is necessary for regulating milk volume and secretion. This observation helps to understand why milk volume is lower in lm/lm mice. In contrast to ZnT5, ZnT6, and ZnT7, Zn import into the TGN through ZnT4 does not appear to metallate alkaline phosphatase (12), consistent with our observation that ZnT4 is localized to the trans- and not cis-Golgi apparatus. It has previously been determined that Zn transport into the secretory pathway is critical to homeostatic maintenance of secretion (10), thus it is likely that ZnT4 plays a key role in this regard. Our studies indicated that Zn exposure redirects ZnT4 localization from the TGN to a more dispersed, punctate pattern and prompts an increase in ZnT4 abundance at the cell membrane. We postulate that Zn-induced relocalization potentially serves to assist in the efflux of Zn from the mammary epithelial cell into milk that takes place during lactation perhaps in response to dramatic Zn accumulation which occurs during lactation (28).

In summary, our study indicates that ZnT4 is critical to several different aspects of mammary gland Zn metabolism that are important during lactation. Herein, we presented evidence that ZnT4 provides Zn for Zn-requiring enzymes within the TGN that are requisite for regulating milk volume as well as the production of milk. Overall, these results reveal critical roles for ZnT4 in producing milk of optimal quantity and quality.

GRANTS

This study was supported by Intramural funds and National Institute of Child Health and Human Development Grant R01-HD-058614 (to S. L. Kelleher).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

N.H.M. and S.L.K. conception and design of research; N.H.M. performed experiments; N.H.M. and S.L.K. analyzed data; N.H.M. and S.L.K. interpreted results of experiments; N.H.M. and S.L.K. prepared figures; N.H.M. and S.L.K. drafted manuscript; N.H.M. and S.L.K. edited and revised manuscript; N.H.M. and S.L.K. approved final version of manuscript.