Journal of Mammary Gland Biology and Neoplasia

, Volume 20, Issue 3, pp 159–172

Exome Sequencing of SLC30A2 Identifies Novel Loss- and Gain-of-Function Variants Associated with Breast Cell Dysfunction

  • Samina Alam
  • Stephen R. Hennigar
  • Carla Gallagher
  • David I. Soybel
  • Shannon L. Kelleher
Article

DOI: 10.1007/s10911-015-9338-z

Cite this article as:
Alam, S., Hennigar, S.R., Gallagher, C. et al. J Mammary Gland Biol Neoplasia (2015) 20: 159. doi:10.1007/s10911-015-9338-z

Abstract

The zinc (Zn) transporter ZnT2 (SLC30A2) is expressed in specialized secretory cells including breast, pancreas and prostate, and imports Zn into mitochondria and vesicles. Mutations in SLC30A2 substantially reduce milk Zn concentration ([Zn]) and cause severe Zn deficiency in exclusively breastfed infants. Recent studies show that ZnT2-null mice have low milk [Zn], in addition to profound defects in mammary gland function during lactation. Here, we used breast milk [Zn] to identify novel non-synonymous ZnT2 variants in a population of lactating women. We also asked whether specific variants induce disturbances in intracellular Zn management or cause cellular dysfunction in mammary epithelial cells. Healthy, breastfeeding women were stratified into quartiles by milk [Zn] and exonic sequencing of SLC30A2 was performed. We found that 36 % of women tested carried non-synonymous ZnT2 variants, all of whom had milk Zn levels that were distinctly above or below those in women without variants. We identified 12 novel heterozygous variants. Two variants (D103E and T288S) were identified with high frequency (9 and 16 %, respectively) and expression of T288S was associated with a known hallmark of breast dysfunction (elevated milk sodium/potassium ratio). Select variants (A28D, K66N, Q71H, D103E, A105P, Q137H, T288S and T312K) were characterized in vitro. Compared with wild-type ZnT2, these variants were inappropriately localized, and most resulted in either ‘loss-of-function’ or ‘gain-of-function’, and altered sub-cellular Zn pools, Zn secretion, and cell cycle check-points. Our study indicates that SLC30A2 variants are common in this population, dysregulate Zn management and can lead to breast cell dysfunction. This suggests that genetic variation in ZnT2 could be an important modifier of infant growth/development and reproductive health/disease. Importantly, milk [Zn] level may serve as a bio-reporter of breast function during lactation.

Keywords

SLC30A2 ZnT2 Breast dysfunction Zinc secretion Breast milk zinc Lactation 

Supplementary material

10911_2015_9338_MOESM1_ESM.pptx (613 kb)
Figure S1Frequency distribution of milk [Zn] in breastfeeding women at ~4 months of lactation. (A) Milk [Zn] was determined by atomic absorption spectrophotometry. Milk [Zn] varied ~10-fold. The mean milk [Zn] was 1.32 mg Zn/L and the median was 1.19 mg Zn/L (CI 1.14-1.47). (B) Milk [Zn] was divided into quartiles (Q1<0.85 mg Zn/L “low”), Q4>1.76 mg Zn/L “high”). “Normal” milk [Zn] ranged between 0.90-1.73 mg Zn/L. (PPTX 613 kb)
10911_2015_9338_MOESM2_ESM.pptx (1.4 mb)
Figure S2Determination of ZnT2 non-synonymous variants in regions with divergent base pairs using Sanger sequencing and chromatograms of exons 1-8. Genomic DNA (100 ng) isolated from buccal swab samples was used and amplified by PCR using primers specific for each of the 8 exons of ZnT2. The PCR products were purified and further sequenced using the 5’ primers used to amplify the DNA. Sanger sequencing reactions were then performed using an Applied Biosystems 3730XL Capillary sequencer. Chromatograms (G, black; A, green; T, red; C, blue) were visually analyzed for changes in DNA sequence and compared with the predicted protein sequence (WT) deposited in the National Center for Biotechnology Information database (http://www.ncbi.nlm.nih.gov/protein/ NP_001004434.1. Select variants were confirmed by TA cloning the resultant PCR products and sequencing 30 independent clones. (PPTX 1440 kb)
10911_2015_9338_MOESM3_ESM.pptx (186 kb)
Figure S3Representative immunoblot of proteins isolated from HC11 cells transiently transfected with wild type ZnT2-HA and select variant constructs. (A) Representative immunoblot of wild-type (WT) ZnT2 and select variants. Approximately 10 μg of total membrane protein from each sample was run on 10% SDS-PAGE gels under non-denaturing conditions. Immunoblotting was performed using anti-HA antibody. A ~42 kDa band representing ZnT2 monomer was detected. Actin was used as a loading control. (B) Transfection efficiencies of cells expressing wild-type and mutant ZnT2 constructs were measured by co-transfecting HC11 cells with ZnT2-HA and ptdTomato-N1 (expressing dsRed). For each transfection, the number of fluorescent cells were counted and expressed as a percentage of the total number of cells within the field. Data represent mean % fluorescent cells ± SD, n=5 images/variant. There were no significant differences in transfection efficiency relative to cells expressing WT ZnT2. (PPTX 185 kb)
10911_2015_9338_Fig6_ESM.gif (537 kb)
Figure S4

Localization of ZnT2 variants in MECs in vitro. Representative confocal images of HC11 cells transfected with wild-type ZnT2-HA (WT) or select HA-tagged ZnT2 variants. Proteins were detected and visualized with anti-HA Alexa 488-conjugated anti-mouse IgG (1 μg/mL; green). Nuclei were stained with DAPI (1 μM; blue). (A) Endoplasmic reticulum was visualized using an antibody directed against calnexin (red) and detected with Alexa 568-conjugated anti-rabbit IgG. (B) Late endosomal compartments were visualized with mannose 6-phosphate receptor, (M6PR, red) and detected with Alexa 568-conjugated anti-mouse IgG (1 μg/mL). (C) Mitochondria were visualized by immunofluorescence using an antibody directed against cytochrome-c oxidase subunit IV (COXIV, red) and detected with Alexa 568-conjugated anti-mouse IgG (1 μg/mL). (D) Lysosomes were visualized with Alexa Fluor 647-labelled lysosome membrane associated protein 1 (Lamp1, red). (E) Exocytotic vesicles were visualized using an antibody directed against Rab3a (red), and detected with Alexa 568-conjugated anti-rabbit IgG (1 μg/mL). (GIF 537 kb)

10911_2015_9338_Fig7_ESM.gif (496 kb)
Figure S4

Localization of ZnT2 variants in MECs in vitro. Representative confocal images of HC11 cells transfected with wild-type ZnT2-HA (WT) or select HA-tagged ZnT2 variants. Proteins were detected and visualized with anti-HA Alexa 488-conjugated anti-mouse IgG (1 μg/mL; green). Nuclei were stained with DAPI (1 μM; blue). (A) Endoplasmic reticulum was visualized using an antibody directed against calnexin (red) and detected with Alexa 568-conjugated anti-rabbit IgG. (B) Late endosomal compartments were visualized with mannose 6-phosphate receptor, (M6PR, red) and detected with Alexa 568-conjugated anti-mouse IgG (1 μg/mL). (C) Mitochondria were visualized by immunofluorescence using an antibody directed against cytochrome-c oxidase subunit IV (COXIV, red) and detected with Alexa 568-conjugated anti-mouse IgG (1 μg/mL). (D) Lysosomes were visualized with Alexa Fluor 647-labelled lysosome membrane associated protein 1 (Lamp1, red). (E) Exocytotic vesicles were visualized using an antibody directed against Rab3a (red), and detected with Alexa 568-conjugated anti-rabbit IgG (1 μg/mL). (GIF 537 kb)

10911_2015_9338_Fig8_ESM.gif (405 kb)
Figure S4

Localization of ZnT2 variants in MECs in vitro. Representative confocal images of HC11 cells transfected with wild-type ZnT2-HA (WT) or select HA-tagged ZnT2 variants. Proteins were detected and visualized with anti-HA Alexa 488-conjugated anti-mouse IgG (1 μg/mL; green). Nuclei were stained with DAPI (1 μM; blue). (A) Endoplasmic reticulum was visualized using an antibody directed against calnexin (red) and detected with Alexa 568-conjugated anti-rabbit IgG. (B) Late endosomal compartments were visualized with mannose 6-phosphate receptor, (M6PR, red) and detected with Alexa 568-conjugated anti-mouse IgG (1 μg/mL). (C) Mitochondria were visualized by immunofluorescence using an antibody directed against cytochrome-c oxidase subunit IV (COXIV, red) and detected with Alexa 568-conjugated anti-mouse IgG (1 μg/mL). (D) Lysosomes were visualized with Alexa Fluor 647-labelled lysosome membrane associated protein 1 (Lamp1, red). (E) Exocytotic vesicles were visualized using an antibody directed against Rab3a (red), and detected with Alexa 568-conjugated anti-rabbit IgG (1 μg/mL). (GIF 537 kb)

10911_2015_9338_Fig9_ESM.gif (372 kb)
Figure S4

Localization of ZnT2 variants in MECs in vitro. Representative confocal images of HC11 cells transfected with wild-type ZnT2-HA (WT) or select HA-tagged ZnT2 variants. Proteins were detected and visualized with anti-HA Alexa 488-conjugated anti-mouse IgG (1 μg/mL; green). Nuclei were stained with DAPI (1 μM; blue). (A) Endoplasmic reticulum was visualized using an antibody directed against calnexin (red) and detected with Alexa 568-conjugated anti-rabbit IgG. (B) Late endosomal compartments were visualized with mannose 6-phosphate receptor, (M6PR, red) and detected with Alexa 568-conjugated anti-mouse IgG (1 μg/mL). (C) Mitochondria were visualized by immunofluorescence using an antibody directed against cytochrome-c oxidase subunit IV (COXIV, red) and detected with Alexa 568-conjugated anti-mouse IgG (1 μg/mL). (D) Lysosomes were visualized with Alexa Fluor 647-labelled lysosome membrane associated protein 1 (Lamp1, red). (E) Exocytotic vesicles were visualized using an antibody directed against Rab3a (red), and detected with Alexa 568-conjugated anti-rabbit IgG (1 μg/mL). (GIF 537 kb)

10911_2015_9338_Fig10_ESM.gif (571 kb)
Figure S4

Localization of ZnT2 variants in MECs in vitro. Representative confocal images of HC11 cells transfected with wild-type ZnT2-HA (WT) or select HA-tagged ZnT2 variants. Proteins were detected and visualized with anti-HA Alexa 488-conjugated anti-mouse IgG (1 μg/mL; green). Nuclei were stained with DAPI (1 μM; blue). (A) Endoplasmic reticulum was visualized using an antibody directed against calnexin (red) and detected with Alexa 568-conjugated anti-rabbit IgG. (B) Late endosomal compartments were visualized with mannose 6-phosphate receptor, (M6PR, red) and detected with Alexa 568-conjugated anti-mouse IgG (1 μg/mL). (C) Mitochondria were visualized by immunofluorescence using an antibody directed against cytochrome-c oxidase subunit IV (COXIV, red) and detected with Alexa 568-conjugated anti-mouse IgG (1 μg/mL). (D) Lysosomes were visualized with Alexa Fluor 647-labelled lysosome membrane associated protein 1 (Lamp1, red). (E) Exocytotic vesicles were visualized using an antibody directed against Rab3a (red), and detected with Alexa 568-conjugated anti-rabbit IgG (1 μg/mL). (GIF 537 kb)

10911_2015_9338_MOESM4_ESM.tiff (1.5 mb)
High Resolution Image (TIFF 1521 kb)
10911_2015_9338_MOESM5_ESM.tiff (1.5 mb)
High Resolution Image (TIFF 1521 kb)
10911_2015_9338_MOESM6_ESM.tiff (1.5 mb)
High Resolution Image (TIFF 1521 kb)
10911_2015_9338_MOESM7_ESM.tiff (1.5 mb)
High Resolution Image (TIFF 1521 kb)
10911_2015_9338_MOESM8_ESM.tiff (1.5 mb)
High Resolution Image (TIFF 1521 kb)
10911_2015_9338_MOESM9_ESM.pptx (89 kb)
Figure S5ZnT2 variants Q71H and A105P alter RhodZin3 fluorescence suggesting changes in mitochondrial Zn pools compared with wild-type ZnT2. Data represent mean fluorescence of RhodZin-3 (arbitrary fluorescence units/μg protein) ± SD, n=8 replicates/variant from 2 independent experiments. Cells not treated with RhodZin3 (HC11 only) reflects background fluorescence. Asterisk (*) indicates a significant change in RhodZin3 fluorescence compared with cells expressing wild-type ZnT2 (wtZnT2), p<0.05. (PPTX 89 kb)
10911_2015_9338_MOESM10_ESM.pptx (91 kb)
Figure S6Relative densitometric analysis of immunoblots comparing protein levels of ZnT1 and ZnT4 in HC11 cells expressing wild-type ZnT2 compared with cells expressing the D103E variant. Data represent mean density normalized to ZnT2-HA or D103E-HA expression ± SD. Asterisk (*) represents a significant difference in ZnT1 protein levels in cells expressing D103E relative to cells expressing wild-type ZnT2, p<0.05. (PPTX 90 kb)
10911_2015_9338_MOESM11_ESM.pptx (1.8 mb)
Figure S7Cell cycle analysis of cells expressing select ZnT2 variants. HC11 cells, mock transfected cells (Lipo), cells expressing ZnT2-HA (WT) or HA-tagged ZnT2 variants were permeabilized and stained with propidium iodide. Cells were sorted by FACS. Histogram, cell gating analysis and the percentage of cells in G1, S, and G2/M phases of the cell cycle were determined using Cell Quest. Data were analyzed using ModFit LT. A) Example cell cycle plots, histogram and cell gating of ZnT2-HA (WT) compared with A28D, K66N, Q71H, A105P, Q137H and T288S. (B) Example cell cycle plots, histogram and cell gating of ZnT2-HA (WT) compared with D103E, T312K. Statistical analysis is presented in Table 3. (PPTX 1792 kb)
10911_2015_9338_MOESM12_ESM.pptx (899 kb)
(PPTX 898 kb)
10911_2015_9338_MOESM13_ESM.pptx (271 kb)
Figure S8Chromatogram of PRLR gene exon 6 mutation and the I146L missense substitution. Top: Exon 6 sequence obtained from one homozygous study participant (2 WT alleles). Bottom: one heterozygous participant harboring both the normal and A-to-C mutated alleles. Genomic DNA (100 ng) isolated from buccal swab samples was used and amplified by PCR using primers specific for exon 6 of PRLR. The PCR products were purified and further sequenced using the 5’ primers used to amplify the DNA. Sanger sequencing was performed using an Applied Biosystems 3730XL Capillary sequencer. Chromatograms (G, black; A, green; T, red; C, blue) were visually analyzed for changes in DNA sequence and compared with the predicted protein sequence deposited in the National Center for Biotechnology Information database (http://www.ncbi.nlm.nih.gov/protein/NP_000940.1. (PPTX 270 kb)
10911_2015_9338_MOESM14_ESM.doc (50 kb)
Table S1(DOC 50 kb)
10911_2015_9338_MOESM15_ESM.docx (70 kb)
Table S2(DOCX 69 kb)

Copyright information

© Springer Science+Business Media New York 2015

Authors and Affiliations

  • Samina Alam
    • 1
    • 3
  • Stephen R. Hennigar
    • 4
  • Carla Gallagher
    • 5
  • David I. Soybel
    • 1
    • 3
  • Shannon L. Kelleher
    • 1
    • 2
    • 3
    • 4
  1. 1.Department of Cellular and Molecular PhysiologyThe Pennsylvania State University Hershey College of MedicineHersheyUSA
  2. 2.Department of PharmacologyThe Pennsylvania State University Hershey College of MedicineHersheyUSA
  3. 3.Department of SurgeryThe Pennsylvania State University Hershey College of MedicineHersheyUSA
  4. 4.Department of Nutritional SciencesThe Pennsylvania State UniversityUniversity ParkUSA
  5. 5.Department of Public Health SciencesThe Pennsylvania State University Hershey College of MedicineHersheyUSA

Personalised recommendations