Marginal Maternal Zinc Deficiency in Lactating Mice Reduces Secretory Capacity and Alters Milk Composition

Abstract

Dietary analysis predicts that marginal Zn deficiency is common in women of reproductive age. The lack of reliable biomarkers limits the capacity to assess Zn status and consequently understand effects of maternal Zn deficiency. We determined effects of marginal maternal Zn deficiency on mammary gland function, milk secretion, and milk composition in mice. Mice (n = 12/diet) were fed marginal (ZD; 15 mg Zn/kg diet) or adequate (ZA; 30 mg Zn/kg diet) Zn diets for 30 d prior to conception through mid-lactation. Mice fed the ZD had a higher plasma Zn concentration (~20%; P < 0.05) but lower milk Zn concentration (~15%; P < 0.05) compared with mice fed the ZA. ZnT2 abundance was higher (P < 0.05) in mice fed the ZD compared with mice fed the ZA; no effect on ZnT4 abundance was detected. The Zn concentration of mammary gland mitochondria tended to be ~40% greater in mice fed ZD (P = 0.07); this was associated with apoptosis and lower milk secretion (~80%; P < 0.01). Total milk protein was ~25% higher (P < 0.05), although the abundance of the major milk proteins (caseins and whey acidic protein) was lower (P < 0.05) in mice fed the ZD. Proteomic analysis of milk proteins revealed an increase (P < 0.05) in four proteins in mice fed the ZD. These findings illustrate that marginal maternal Zn deficiency compromises mammary gland function and milk secretion and alters milk composition. This suggests that lactating women who consume inadequate Zn may not produce and/or secrete an adequate amount of high quality milk to provide optimal nutrition to their developing infant.

Introduction

Zn deficiency is of global concern (1–3) and results from low Zn intake and intake of poorly available Zn (4), predominantly affecting children and women of reproductive age (5). Estimates of low Zn intake, combined with increased Zn requirements during reproduction (6), suggest that ~80% of pregnant and lactating women are at risk for marginal Zn deficiency (5, 7). Efforts to explore the effects of Zn deficiency in women are hampered by the inability to accurately assess Zn status (4). Studies in animal models illustrate that marginal maternal Zn deficiency impairs bone development (8) and immunocompetence (9) and is associated with insulin resistance, obesity (10), and hypertension (11) in rat offspring. We previously found that moderate maternal Zn deficiency in rats reduces milk secretion (12), which may have important implications for nutrient transfer and the overall health and development of the nursing infant. This is consistent with observations noted in Zn-deficient lactating women, suggesting a relationship between maternal Zn deficiency and hypogalactia (7). To our knowledge, only one study has explored effects of maternal Zn deficiency on milk nutrient concentration; that study noted that moderate maternal Zn deficiency increased the β-casein concentration in rat milk (13). This observation may be physiologically relevant to the nutrition of the offspring, because β-casein can chelate Zn (14) and calcium (15) and thus alter Zn bioavailability in the developing infant.

Whereas some reports have not observed a correlation between maternal Zn intake and milk Zn concentration (16, 17), others have found that maternal Zn intake and milk Zn concentration are positively correlated in women (18) and rats (12). These discrepancies may reflect the timing and/or degree of Zn deficiency and perhaps genetic factors (19, 20). The regulation of Zn secretion into milk is not well understood. The Zn transporters ZnT2 and ZnT4 have been implicated as key players in Zn secretion into milk (19–26). Mice with a mutation in SLC30A4, which eliminates ZnT4 protein, have smaller mammary glands, decreased milk secretion, and ~35% less Zn secreted into milk (27). We previously determined that female rats fed a marginal Zn diet had higher ZnT4 abundance in the mammary gland, which was associated with the maintenance of normal Zn concentrations in milk (23). The precise role of ZnT4 in mammary gland Zn metabolism remains to be characterized. There is a greater understanding of the role of ZnT2 as a key modifier of milk Zn secretion and mammary gland metabolism. A mutation in SLC30A2 (the gene encoding ZnT2) in lactating women is associated with an ~75% decrease in milk Zn concentration (19). This defect may result from reduced Zn import into secretory vesicles (24) or alterations in Zn import into mitochondria, which may modulate energetics or apoptosis in the mammary cell (28). No effect of marginal Zn intake on ZnT2 abundance has been described (23). The high prevalence of Zn deficiency in women of reproductive age presses our need to further understand the physiological and molecular consequences of Zn deficiency on mammary gland function and lactation.

Materials and Methods

Animals.

This study was approved by the IACUC Committee at the Pennsylvania State University, which is accredited by the American Association for the Accreditation of Laboratory Animal Care. Female C57BL/6 mice were obtained commercially (Charles River) and individually housed in polycarbonate cages. Mice (n = 12/diet) were fed a commercially available, purified, casein-based diet based on AIN93 (MP Biomedicals) containing 30 mg Zn/kg (ZA⁷) or a diet reduced only in Zn (15 mg/kg; ZD) (29) for 30 d prior to conception and were fed these diets throughout the study. The Zn concentration of the diet was verified by atomic absorption spectrophotometry. Food intake and body weight were monitored every other day to confirm that there were no effects of a marginal Zn diet on food cycling. Mice were maintained on a 12 h-light/-dark cycle under controlled temperature and humidity. Mice were bred and allowed to deliver naturally and examined during mid-lactation (lactation d 7–10). Milk secretion was assessed using weigh-suckle-weigh (described below). The following day, dams were removed from pups for 2 h to control for effects of suckling, and milk was collected as previously described (23). Mice were killed by CO₂ asphyxiation. Blood was drawn by cardiac puncture and collected into heparinized tubes and plasma was separated by centrifugation at 1500 × g for 10 min at 4°C. Mammary glands were removed and snap-frozen. Mammary glands from mice that had not been milked (n = 4 mice/diet) were fixed in phosphate-buffered paraformaldehyde (0.04 wt:v), pH 7.4, overnight.

Milk secretion.

Milk secretion (n = 7 mice/diet) was assessed using weigh-suckle-weigh adapted from McDonald and Nielsen (30). The litters were separated from dams for 1 h, returned to the dams to suckle for 2 h, then again separated from the dams for 3 h to standardize milk production and hormone secretion. Litters were weighed and pups were allowed to suckle for 2 h and were weighed one final time at the end of the suckling period. Milk secretion was calculated as the difference between the final and initial litter weights.

Creamatocrit.

Frozen milk samples (n = 7 mice/diet) were warmed in a 37°C water bath for 10 min to thaw. Samples were mixed by vortexing and milk (20 μL) was drawn into a micro-hematocrit tube (VWR) and sealed with Critoseal (Fischer Scientific). Capillary tubes were centrifuged in a hematocrit microcentrifuge for 10 min at 500 × g. The relative amount of milk fat was calculated by dividing the volume of the milk fat by the total volume of the milk.

Lactose concentration.

Lactose in skimmed milk (n = 7 mice/diet) was measured using the Lactose Assay kit (Abcam) as described in the instruction manual. OD was measured using a microplate spectrophotometer (Epoch BioTek).

Milk protein concentration.

Frozen milk samples (n = 7 mice/diet) were thawed on ice and centrifuged at 2000 × g for 15 min at 4°C to skim. Skimmed milk (50 μL) was diluted in two volumes of buffer [100 μL; sodium phosphate (50 mmol/L), sodium chloride (150 mmol/L), and EDTA (50 mmol/L, pH 7.4)] and centrifuged at 11,600 × g for 15 min at 4°C. The total protein in the supernatant was measured using the Bradford assay.

Proteomic analysis.

Milk samples (50 μL, 100 μg protein/μL; n = 3 mice/diet) were prefractionated into two fractions: MFGM and milk proteins (31). Following resuspension, the protein within the MFGM pellet was concentrated using methanol/chloroform (32). The MFGM contain between 1–2% of the total milk protein (33). For each sample, the entire methanol/chloroform pellet was resuspended in 125 μL of protein solubilization buffer (Bio-Rad) and used to passively rehydrate immobilized pH gradient isoelectric focusing strips (7 cm; pH 3–10; Bio-Rad) for 12–15 h at room temperature. The milk proteins were also concentrated using methanol/chloroform, then resolubilized in 300 μL of protein solubilization buffer. Isoelectric focusing was performed at 20°C as follows: 250 V for 1 h, 500 V for 1 h, 6000 V for 0.5 h, 6000 V for 8000 V-h, and 1000 V for 1 h. Following isoelectric focusing, the proteins were separated according to their molecular weights using a Tris/glycine buffering system and fixed and then destained with water (34).

Histology.

Fixed mammary glands (glands 4 and 5) from mice that had not been previously milked (n = 4 mice/diet) were stained with hematoxylin and eosin and mounted onto slides using DePex Mounting Medium (Electron Microscopy Sciences). The TACS-XL Basic In Situ Apoptosis (TUNEL) Detection kit (Trevigen) was used to identify apoptotic cells in the mammary gland per the manufacturer’s instructions. A DAB Peroxidase Substrate kit (Vector) was used to detect nicked DNA. Sections were counterstained with toluidine blue. The number of TUNEL-positive apoptotic cells was visually quantified using 10× magnification from five separate fields/section. Data represent the mean number of TUNEL-positive cells ± SD from 4 mice/diet. Sections were viewed using a Leica DM IL LED microscope and images were collected using Leica Application Suite (V3.6). Sections were analyzed twice by investigators who were unaware of the treatment.

Isolation and purification of mitochondria and mitoplasts.

Mammary glands (0.2 g) were homogenized in ice-cold buffer [HEPES (10 mmol/L, pH 7.5), mannitol (210 mmol/L), sucrose (70 mmol/L), and EDTA (1 mmol/L)]. Mitochondria and mitoplasts were generated as previously described (28).

Analysis of Zn concentration.

The Zn concentrations of plasma, milk, tissues, and isolated mitoplasts were determined by atomic absorption spectrophotometry as previously described (35). The protein concentration of mitoplasts was measured using the Bradford assay and the Zn concentration was normalized to protein.

Immunoblotting.

Mammary gland (50 μg protein), mitoplast (10 μg protein), or skimmed milk (100 μg protein) was electrophoresed and transferred and immunoblots performed as previously described (22). Primary antibodies used were as follows: rabbit anti-mouse ZnT2 antibody, rabbit anti-mouse ZnT4 antibody, anti-WAP (Santa Cruz Biotech), or anti-β-casein antibody (Santa Cruz Biotech). Proteins were detected using donkey, anti-rabbit IgG conjugated to HRP (Amersham Pharmacia Biotech), visualized with a Super Signal Femto Chemiluminescent Detection System (Pierce), and exposed to autoradiography film. Relative band density was quantified using the Carestream Gel Logic 212 Pro (Carestream). Membranes used for ZnT2 and ZnT4 immunoblots were stripped and reprobed for β-actin (mammary gland) or succinate dehydrogenase (mitoplasts) to control for equal protein loading. Membranes used for WAP or β-casein immunoblots were stained with Ponceau S for 1 h then rinsed in distilled water to control for equal protein loading.

Statistical analysis.

Results are presented as mean ± SD, sample size as indicated. Statistical comparisons were performed using Student’s t test (Prism Graph Pad) and a significant difference was demonstrated at P < 0.05.

Results

Food intake and weight was not affected (data not shown). Consistent with previous observations in a marginal Zn deficiency models (12, 23, 29, 36), the plasma Zn concentration was higher (P < 0.05) in mice fed the ZD compared with mice fed the ZA, illustrating the redistribution of body Zn pools in response to a marginal Zn diet (Table 1). The milk Zn concentration in mice fed the ZD was ~15% lower (P < 0.05) than in the milk from mice fed the ZA. Diet did not affect the total amount of Zn in the mammary gland or liver (data not shown).

TABLE 1.

Plasma and milk Zn concentrations in mice fed a ZA or ZN diet from 30 d prior to pregnancy through mid-lactation¹

	ZA	ZD
Plasma, μmol/L	10.6 ± 1.2	12.8 ± 2.5*
Milk, μmol/L	374 ± 35	320 ± 37*

¹Values are mean ± SD, = 8. *Different from ZA, P < 0.05. ZA, Zn adequate; ZD, Zn deficient.

ZnT2 and ZnT4 abundance.

No effect on ZnT4 was observed (data not shown). The total abundance of ZnT2 was ~90% greater in mice (n = 6 mice/diet) fed the ZD (3.34 ± 0.66 arbitrary units) compared with mice fed the ZA (1.76 ± 0.51 arbitrary units) (P < 0.05).

Milk secretion.

Milk secretion was ~80% lower in mice (n = 7 mice/diet) fed the ZD (0.2 ± 0.1 g) compared with mice fed the ZA (0.9 ± 0.2 g) (P < 0.01).

Mammary gland morphology.

Hematoxylin and eosin staining illustrated that the morphology of mammary glands from mice fed the ZD was severely compromised, because we observed distorted mammary gland architecture and less dense alveoli compared with mice fed the ZA (Fig. 1A). The number of TUNEL-positive apoptotic cells was higher in mammary glands from mice fed the ZD (95 ± 50 TUNEL-positive cells) compared with mammary glands from mice fed the ZA (17 ± 15 TUNEL-positive cells) (P < 0.05). Consistent with previous reports, we noted that the TUNEL-positive cells were detected within the lumen of the alveoli (Fig. 1B), indicative of apoptotic cell extrusion in the lactating mammary gland (37). No significant effect on cell proliferation was noted (data not shown).

FIGURE 1.

Mammary gland morphology in mice fed ZA or ZD diets from 30 d prior to pregnancy through mid-lactation. Representative mammary gland sections were stained with hematoxylin and eosin (5× magnification) (A) or were stained for TUNEL to assess apoptosis (10× magnification) (B). Note the location of apoptotic cells (arrows) within the lumen of the alveoli. ZA, Zn adequate; ZD, Zn deficient.

Mitochondrial Zn concentration.

The Zn concentration of mammary gland mitochondria tended to be ~40% greater in mice fed the ZD (3.39 ± 0.62 nmol Zn/mg protein) compared with mitochondria isolated from the mammary glands of from mice fed the ZA (2.46 ± 0.15 nmol Zn/mg protein; P = 0.07; n = 4 mice/diet). We next tested whether a higher mitochondrial Zn concentration was associated with increased ZnT2 abundance in mitochondria. Although mice fed the ZD had 90% greater total ZnT2 abundance, marginal Zn deficiency did not affect ZnT2 abundance in mitochondria (Fig. 2).

FIGURE 2.

Abundance of ZnT2 in mitochondria isolated from the mammary glands of mice fed ZA or ZD diets from 30 d prior to pregnancy through mid-lactation. A representative immunoblot (n = 3/diet) of ZnT2 in mitoplasts isolated from mammary glands is shown. Membranes were stripped and reprobed for SDHB as a loading control. ZA, Zn adequate; ZD, Zn deficient; SDBH, succinate dehydrogenase.

Milk protein concentration.

Marginal Zn intake did not affect the percentage of fat or lactose concentration of milk (data not shown). We found that the protein concentration was ~30% higher in mice (n = 7 mice/diet) fed the ZD (77.8 ± 10 g/L) compared with mice fed the ZA (60.2 ± 11.8 g/L) (P < 0.05).

Relative distribution of milk proteins.

We separated milk proteins (n = 4 mice/diet) by electrophoresis (50 μg protein/well) and used immunoblotting to determine that the abundance of both β-casein (ZD, 1.27 ± 0.07 arbitrary units; ZA 1.42 ± 0.08) and WAP (ZD, 2.28 ± 0.1 arbitrary units; ZA, 2.59 ± 0.07 arbitrary units) was significantly lower in mice fed the ZD compared with mice fed the ZA (P < 0.05). Next, we used 2-dimensional electrophoresis to fractionate milk proteins and noted that marginal maternal Zn deficiency altered the abundance of numerous proteins in the MFGM; however, these changes were highly variable and these observations will require extensive further investigation (data not shown). In contrast, marginal maternal Zn deficiency resulted in an increase in four milk proteins (Fig. 3) (P < 0.05) in the sucrose fraction isolated from the milk of mice fed the ZD compared with mice fed the ZA.

FIGURE 3.

The milk proteome from mice fed ZA or ZD diets from 30 d prior to pregnancy through mid-lactation. Representative 2DE gels of proteins isolated from milk were stained with colloidal Coomassie Blue. Numbers 1–3 and 5 indicate proteins with spot volumes that differed between groups (P < 0.05, n = 3 mice/group). ZA, Zn adequate; ZD, Zn deficient.

Discussion

Mild-to-moderate maternal Zn deficiency during reproduction is of global concern (5). The lack of reliable biomarkers of Zn status (4) limits the capacity to correctly assess the Zn status of women of reproductive age and consequently to understand the effects of maternal Zn deficiency. We used a marginally Zn deficient mouse model to explore the effects on mammary gland biology, milk secretion, and milk composition. We noted that marginal maternal Zn deficiency significantly impaired milk secretion. Consistent with this observation, we noted numerous, although variable, alterations in MFGM, which is derived from the lipid monolayer of the endoplasmic reticulum membrane surrounded by the lipid bilayer from extruded apical membrane (38). The MFGM contains proteins that are key to vesicle trafficking and cytoskeletal dynamics (31), milk proteins (31, 38–40), immune function, and metabolism (39), and its composition changes throughout the course of lactation (39). Thus, alterations in the complement of MFGM proteins may reflect changes in the secretory capacity of the mammary gland and/or milk composition. Further studies are required to ascertain the identity of these proteins and determine the relevance of these changes to mammary gland function, milk composition, and maternal and infant health. Consistent with our findings, Scheplyagina (7) noted that maternal Zn deficiency was associated with hypogalactia in women; however, it is not understood if this effect resulted from low Zn intake or general malnutrition. However, 35% of women who initiate breast-feeding wean their infants early, or do not exclusively breastfeed because they perceive their milk supply is insufficient (41). The reality of this perception has not been rigorously evaluated, but it would be interesting to determine if Zn intake is inadequate in this subpopulation of lactating women. Importantly, this further presses the need to establish adequate biomarkers to ascertain Zn status, ultimately allowing us to develop targeted Zn supplementation strategies in lactating women to improve infant health and development.

A second key finding noted that mammary gland architecture was dramatically impaired by marginal maternal Zn deficiency, which occurred despite the fact that plasma and mammary gland Zn concentrations were not compromised. One explanation is that marginal Zn deficiency redistributed subcellular Zn pools in the mammary epithelial cell (42). Alterations in subcellular Zn distribution modulate a vast array of key cellular processes and thus may affect mammary gland function (43). For example, Zn accumulation in mitochondria releases cytochrome c and activates caspases, which leads to apoptosis (28). This is consistent with our observations of Zn enrichment in mitochondria isolated from the mammary glands of marginally Zn-deficient mice. We hypothesized that mitochondrial Zn accumulation was facilitated by increased mitochondrial ZnT2 abundance (28); however, this was not observed and reaffirms our previous documentation of a role for other Zn transporters or Zn transporting mechanisms in mitochondrial Zn accumulation (28). Secondly, cell signaling pathways such as MAPK (44) and PI3K/Akt (45) are Zn dependent and regulate mammary gland expansion, differentiation, and secretion [reviewed in (46)]. Finally, numerous transcription factors are Zn dependent [reviewed in (47)]. A specific and potentially interesting example is Lmo4, which is a Zn finger LIM domain protein that is highly expressed in ductal and alveolar luminal cells of the mature mammary gland (48). Sum et al. (48) found that reduced expression of Lmo4 results in impaired lobuloalveolar development in mouse mammary glands and overexpression induces hyperplasia and cell invasion (49). We speculate that marginal Zn deficiency alters subcellular Zn distribution, which modulates the function of key regulatory factors in the mammary gland.

Alternatively, marginal Zn deficiency may alter reproductive hormones that are critical for mammary gland expansion or secretion. Marginal Zn deficiency increases systemic prolactin concentrations (12), which may downregulate the prolactin receptor in the mammary gland (50), impairing mammary gland differentiation or secretion [reviewed in (51)]. Estrogen, luteinizing hormone, and progesterone are key to mammary gland expansion and differentiation during pregnancy and lactation (52, 53). Severe Zn deficiency significantly reduces luteinizing hormone, estrogen (54), and progesterone (55) concentrations in male rats and is associated with reproductive defects such as prolonged diestrous phase, increased atresia, and cessation of oogenesis and ovulation in female mice (56). However, effects of marginal maternal Zn deficiency on reproductive hormones have not been reported. Although mammary gland function during lactation was compromised, it is possible that it was caused by defects in expansion and/or differentiation during pregnancy. Further studies are required to understand the mechanisms responsible for the observed defects to better ascertain the optimal window for maternal Zn supplementation.

The relationship between maternal Zn intake and milk Zn concentration is inconsistent. We and others have determined that ZnT2 and ZnT4 (21–26, 57) play key roles in mammary gland metabolism and Zn secretion into milk. Our previous report in rats noted that ZnT4 abundance was increased and associated with the maintenance of milk Zn concentration (23). However, we noted herein that ZnT4 abundance was not affected and milk Zn concentration was significantly lower. These differences are curious, because the previous study used a diet slightly lower in Zn (10 mg Zn/kg) for a longer period of time (70 d prior to lactation) and milk Zn concentration was refractory. It is possible that differences in species exist or that compensatory mechanisms are activated once a threshold is surpassed. The importance of ZnT2 in milk Zn secretion predicts that ZnT2 abundance would parallel milk Zn concentration; however, milk Zn concentration and ZnT2 abundance were inversely related. This suggests that either ZnT2 function or vesicle trafficking in general (24) may have been affected. The mechanism(s) through which ZnT2 transports Zn have not yet been determined. Together, these observations provide evidence that marginal maternal Zn deficiency impairs mammary gland function and milk secretion and reduces the ability of the mother to provide optimal nutrition for the developing infant.

Compromised mammary gland function will have negative implications for the health and development of the nursing infant. We found that marginally Zn deficient mice had a significantly greater total milk protein concentration despite slight decreases in specific milk proteins (β-casein and WAP). This is inconsistent with previous reports in moderately Zn deficient rats and may reflect the degree of Zn deficiency (13). β-Casein is a highly phosphorylated milk protein and assists in the digestion and absorption of macronutrients and micronutrients from milk. Casein phosphopeptides are formed during casein digestion and keep calcium in a soluble form, contributing to the high bioavailability of calcium from milk (15). Furthermore, casein phosphopeptides have been associated with antithrombotic, antihypertensive, and opioid activities, which may contribute to sleeping behavior and fluid transport in the small intestine of breast-fed infants [reviewed in (58)]. Caseins are also important Zn-binding proteins (14); thus, a reduction in milk caseins may alter the pool of calcium and Zn that is available for absorption during infancy. WAP is a Zn-dependent protease inhibitor that is found in milk. WAP contains two 4-disulfide core motifs (WAP domains) that are composed of eight cysteine residues (59), which may serve as a Zn-binding domain essential for proteolytic function. WAP proteins play a role in the proliferation of mammary cells (60), which may be relevant to the defects we detected in the mammary glands of Zn-deficient mice. In addition, WAP has potent antimicrobial activities and although it is not found in human milk, other WAP domain proteins are found in human milk (61). As a result, WAP family members are thought to play key roles in immunity [reviewed in (62)]. Therefore, understanding of the function of WAP during lactation may provide crucial insight into mammary gland function and infant health.

In summary, our results indicate that marginal Zn deficiency dramatically alters mammary gland morphology and reduces secretory capacity. The prevalence of marginal Zn intake in women of reproductive age suggests that marginal Zn deficiency may be a key modifier of mammary gland structure and function in lactating women, which may compromise the health and development of nursing neonates.