The Journal of Nutrition Vol. 128 No. 8 August 1998, pp. 1311-1314

The Acrodermatitis Enteropathica Mutation Affects Protein Expression in Human Fibroblasts: Analysis by Two-Dimensional Gel Electrophoresis^1,2,3

Arthur Grider⁴ and Michael F. Mouat

Department of Foods and Nutrition, The University of Georgia, Athens, GA 30602

	ABSTRACT

Abstract Introduction Methods Results Discussion References

The acrodermatitis enteropathica (AE) mutation affects zinc uptake in human fibroblasts. However, the specific biochemical lesion has not been identified. We have used the technique of two-dimensional gel electrophoresis to identify protein differences in total cell lysate isolated from normal and AE fibroblasts. Two proteins with estimated molecular weights of 49.6 and 49.9 kDa and an isoelectric point of 5.1 were identified in normal fibroblasts but absent from AE fibroblasts. The proteins were purified, subjected to in-gel trypsin digest and the resulting peptides separated by HPLC. Sequences from three peptide fragments (8, 15 and 18 amino acids) were obtained after Edman degradation. None of the fragments exhibited homology to any amino acid sequences in the nonredundant Genbank database. The 15 and 18 amino acid fragments each exhibited 100% homology to a 136 amino acid expressed sequence tag that was homologous (43%) to adipophilin. However, the 15 and 18 amino acid fragments were only 30 and 44% homologous, respectively, to corresponding regions within the expressed sequence tag. Therefore, the 49.6/49.9 kDa protein absent from AE fibroblasts was not related to adipophilin. The 8 amino acid fragment did not exhibit homology to any expressed sequence tag. Therefore, the 49.6/49.9 kDa proteins are novel and may be the cause of the reduced zinc uptake and abnormal zinc metabolism characteristic of fibroblasts carrying the AE mutation.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

The acrodermatitis enteropathica (AE)⁵ mutation causes abnormal zinc metabolism in humans. Investigators have characterized the clinical symptoms of the disease, which include low plasma zinc concentrations, skin lesions, hair loss, growth retardation and diarrhea (Dillaha et al. 1953, Van Wouwe 1989). The physiologic lesion associated with AE was first described in intestinal biopsies (Atherton et al. 1979). Although AE patients were treated with zinc supplementation, their intestinal biopsies exhibited a 77% reduction in the accumulation of radioactive zinc. This was the first report indicating that the AE mutation affected cellular zinc uptake, albeit into cells of the intestinal mucosa.

Human fibroblasts isolated from a patient with AE exhibit altered zinc metabolism compared with normal fibroblasts. Zinc transport is significantly reduced in the AE fibroblasts (Chang et al. 1998, Grider and Vazquez 1996, Grider and Young 1996). AE affects the cell zinc content and the activity of the zinc-dependent enzyme, 5'-nucleotidase (Grider et al. 1998, Grider and Young 1996). The specific cause of the defects in zinc metabolism is unknown; however, it is possible that a zinc transport protein, or protein associated with a zinc transporter, is involved. Our working hypothesis is that the AE mutation affects the expression of such a protein. We have begun to test this hypothesis using two-dimensional polyacrylamide gel electrophoresis (2DPAGE) to identify differential protein expression between normal and AE fibroblasts.

	MATERIALS AND METHODS

Abstract Introduction Methods Results Discussion References

Cell Culture.  Fibroblasts were purchased from the Coriell Institute for Medical Research Genetic Mutant Cell Repository (Camden, NJ). The fibroblasts had been isolated from Caucasian males, ages 14 mo (GM5659B; normal) and 6 mo (GM2814; AE). The cells were grown in minimal essential medium (Eagle's) (GIBCO/BRL, Grand Island, NY) containing 0.06 g/L penicillin, 0.1 g/L streptomycin, 20% uninactivated fetal calf serum and double the normal concentrations of essential and nonessential amino acids and vitamins. The fibroblasts were subcultured in 75 cm² cell culture flasks at a density of 1.3 × 10⁴ cells/cm² and grown for 4 d. The 4-d time period was chosen because previous studies (Grider and Young 1996) have shown that the cellular zinc content between normal and AE fibroblasts is similar at this time. Thus, any differences in protein expression will not be due to differences in the cellular zinc content.

Preparation of total cell lysate.  The flasks were rinsed three times (1 min each time) with 10 mL ice-cold wash buffer (10 mmol/L Tris-HCl, pH 7.4, 150 mmol/L NaCl). After removal of the final volume of wash buffer, 240 µL of boiling sample buffer 1 (50 mmol/L Tris-HCl, 200 mmol/L dithiothreitol, 10.4 mmol/L sodium dodecyl sulfate, pH 8.0) was added to each flask, the cells scraped together into the buffer and the cell lysate transferred into 1.5-mL microfuge tubes. The cell lysate was heated for 5 min at 100°C, chilled on ice for 5 min, and 24 µL of sample buffer 2 (500 mmol/L Tris-HCl, 50 mmol/L MgCl₂, 1 g/L DNAse 1, 0.25 g/L RNAse A, pH 8.0) was added. After incubation on ice for 8 min, the cellular proteins were precipitated by the addition of acetone to 13.8 mol/L and incubation on ice for 20 min. The microfuge tubes were then centrifuged at 12,000 × g for 10 min at 4°C, the supernatant discarded and the pellet dried at room temperature for 5 min. The pellet was resuspended in 240 µL of sample buffer mix (22.4 mmol/L Tris-HCl, 17.6 mmol/L Tris, 7.92 mol/L urea, 2.1 mmol/L sodium dodecyl sulfate, 17.6 g/L ampholytes (pH 3-10), 120 mmol/L dithiothreitol, 51 mmol/L Triton X-100), and an aliquot was removed for assaying the protein concentration. The remainder was stored at 80°C. The protein concentration was determined using the Bradford method (Bradford 1976, Bio-Rad, Hercules, CA).

2DPAGE-first dimension isoelectric focusing (IEF).  The proteins from normal and AE fibroblasts were separated by 2DPAGE by using a dedicated system (ESA, Chelmsford, MA). The analytical gels were cast in 180 mm × 1.2 mm (i.d.) tubes and contained a matrix made of 9.5 mol/L urea, 32 mmol/L Triton X-100, 0.58 mol/L acrylamide, 5 mmol/L CHAPS and 0.58 g/L ampholytes (pH 3-10). Sample overlay buffer (0.5 mol/L urea, 3.2 mmol/L Triton X-100, 1 g/L ampholytes (pH 3-10) 50 mmol/L dithiothreitol) was first applied to the tubes, then the cell lysate (50-100 µg protein) was applied under the overlay buffer. The cathode buffer was 100 mmol/L NaOH and the anode buffer was 10 mmol/L phosphoric acid. The gels were run, without prefocusing, at 100 µA per gel for 17.5 h.

2DPAGE-second dimension.  After IEF, the gels were extruded into gel equilibration buffer (0.375 mol/L Tris, 104 mmol/L sodium dodecyl sulfate, 50 mmol/L dithiothreitol, 1.4 mmol/L bromophenol blue) for 2 min and loaded onto large format (22 cm × 22 cm × 1 mm) 10% acrylamide slab gels. The gels were run at 25 W/gel at 4°C until the dye front reached within 1 cm of the bottom of the gel. The cathode buffer contained 50 mmol/L Tris base, 384 mmol/L glycine and 6.9 mmol/L sodium dodecyl sulfate. The anode buffer contained 25 mmol/L Tris base, 192 mmol/L glycine and 3.5 mmol/L sodium dodecyl sulfate. The separated proteins were visualized by silver staining. A cultured rat fibroblast cell lysate (50 µg protein) was used as a molecular weight and isoelectric point standard (ESA, Chelmsford, MA).

Silver staining.  A modified Rabilloud staining method was used (Rabilloud 1992). Gels were fixed with 6.9 mol/L ethanol and 1.7 mol/L acetic acid for 1 h, then overnight with 50 mmol/L glutaraldehyde, 5.2 mol/L ethanol, 8.3 mmol/L potassium tetrathionate and 500 mmol/L sodium acetate. The gels were then rinsed with 18 M water (4 × 15 min) and incubated with 5.9 mmol/L silver nitrate and 8.3 mmol/L formaldehyde for 30 min. Development of the gels (for 30 min) followed a brief rinse in 18 M water. The development solution contained 0.2 mol/L potassium carbonate, 5 mmol/L formaldehyde and 0.05 mmol/L sodium thiosulfate. Development was stopped when the gels were transferred to a solution containing 0.41 mol/L Tris and 0.35 mol/L acetic acid for 10 min. The gels were stored in 0.27 mol/L glycerol until they were dried between sheets of cellophane.

Preparative 2DPAGE.  The preparative IEF gels were cast in 180 mm × 3 mm (i.d.) tubes and contained 8 mol/L urea, 68.4 mmol/L octyl--D-glucopyranoside, 0.6 mol/L acrylamide and 55 g/L ampholytes (pH 3-10). Sample loading was similar to the analytical IEF procedure except that 1.5 mg protein was loaded. The gels were run at 600 µA per gel for 17.5 h. The preparative procedure for the second dimension was similar to that for the analytical gels, except that the slab gels contained a 0.56 mol/L acrylamide stacking gel and the top of the slab gel sandwich was fitted with an adaptor. The preparative IEF gel was placed in this adaptor, secured with 10 g/L agarose and run as described for the analytical gel procedure.

After the running of the second dimension, the gels were incubated for 20 min in 0.6 mmol/L Coomassie blue staining solution containing 12.4 mol/L methanol and 1.7 mol/L acetic acid. The gels were destained in 1.2 mol/L methanol and 1.2 mol/L acetic acid. Although two spots were observed after the highly sensitive silver staining, only a single spot was observed after Coomassie blue staining. The spot, present in normal fibroblast cell lysate and absent from AE fibroblast cell lysate, was excised from nine preparative gels. The gel pieces were pooled and sent to the Microchemical Facility (Emory University, Atlanta, GA) for in-gel digestion, HPLC and amino acid sequencing. The in-gel digestion procedure was as follows. The gel pieces were incubated with 150 µL of 50 mmol/L NH₄HCO₃ containing 12.2 mol/L acetonitrile to remove the Coomassie blue stain. The gels were then dried and rehydrated with 5 µL of 50 mmol/L NH₄HCO₃ containing 0.5 µg trypsin. This solution was added in 5-µL aliquots until the gel pieces regained their original size. The gel pieces were then cut into 1-mm cubes, covered with 50 mmol/L NH₄HCO₃ and incubated overnight at 30°C. After the addition of 1.5 µL of 0.88 mol/L trifluoroacetic acid to stop the reaction, the peptides were extracted with 50 mmol/L NH₄HCO₃ containing 14.6 mol/L acetonitrile. The supernatants were dried to ~10 µL and resuspended in 30 µL of 0.88 mol/L trifluoroacetic acid. Before injection on the HPLC, 0.88 mol/L trifluoroacetic acid and dithiothreitol (added to 50 mmol/L) were added to the sample. After HPLC, the separated peptide fragments were sequenced using Edman degradation.

	RESULTS

Abstract Introduction Methods Results Discussion References

A typical analytical 2DPAGE gel from normal fibroblast total cell lysate is shown in Figure 1. Over 100 proteins were visible after silver staining. The proteins indicated by the arrow were 0.26% of the total gel stain and exhibited a molecular weight of 49.6-49.9 kDa and an isoelectric point (pI) of 5.1. 

View larger version (142K): [in this window] [in a new window]   Fig 1. Two-dimensional polyacrylamide gel electrophoresis (2DPAGE) pattern from normal fibroblasts. The proteins from total cell lysate prepared from normal fibroblasts were separated by 2DPAGE and visualized by silver staining. The migration of the standard rat fibroblast proteins is indicated by the isoelectric point (pI) at the top of the figure and by the molecular weight along the left side of the figure. The arrow indicates the proteins with molecular weights of 49.6-49.9 kDa and a pI of 5.1. These proteins were absent from the total cell lysate prepared from acrodermatitis enteropathica fibroblasts. This gel is representative of at least five different total cell lysate preparations.

A typical 2DPAGE gel from AE fibroblast total cell lysate is shown in Figure 2. Although several proteins exhibited decreased expression in the AE cell lysate in this particular gel, the differences were inconsistent among different lysate preparations. The only difference between the AE and normal cell lysates is indicated by the arrow. Consistently, the 49.6/49.9 kDa proteins were absent from all AE, and present in all normal, total cell lysate preparations. The proteins could not be induced in AE fibroblasts grown in cell culture medium in which 100 µmol/L Zn was added. The proteins were also absent from any AE preparations grown for 2 d before preparing the cell lysate.

View larger version (128K): [in this window] [in a new window]   Fig 2. Two-dimensional polyacrylamide gel electrophoresis (2DPAGE) pattern from acrodermatitis enteropathica fibroblasts. The proteins from total cell lysate prepared from acrodermatitis enteropathica fibroblasts were separated by 2DPAGE and visualized by silver staining. The migration of the standard rat fibroblast proteins is indicated by the isoelectric point (pI) at the top of the figure and by the molecular weight along the left side of the figure. The arrow indicates the area where the proteins with molecular weights of 49.6-49.9 kDa and a pI of 5.1 should migrate. These proteins were absent from the total cell lysate prepared from acrodermatitis enteropathica fibroblasts, but were present in the cell lysate prepared from normal fibroblasts. This gel is representative of at least five different total cell lysate preparations.

Three peptide sequences, containing 8, 15 and 18 amino acids, were obtained from these proteins (Table 1). The Genbank nonredundant database was searched using the blastp and the tblastn programs. None of the sequences exhibited homology to known proteins or nucleotide sequences. Using the tblastn program, the expressed sequence tag (EST) database (dbest) was searched. No homology was found between the octapeptide and any EST. However, a 136 amino acid EST from Jurkat T-cells (EST176941) was homologous to both the 15 amino acid and 18 amino acid peptide fragments (Fig. 3). Both the 15 amino acid and 18 amino acid peptide fragments were 100% homologous to segments of EST176941. The EST176941 exhibited only 43% homology to adipophilin.

View larger version (19K): [in this window] [in a new window]   Fig 3. Homology between the 15 amino acid and 18 amino acid peptide fragments, EST176941 and adipophilin. The 15 amino acid peptide fragment and the 18 amino acid peptide fragment exhibited 100% homology to segments of EST176941 (emboldened and underlined). The EST exhibited 43% homology to adipophilin. The segment of the adipophilin amino acid sequence homologous to the EST is lined up below the EST. Homologous amino acids are emboldened and have an asterisk underneath them.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

The biochemical lesion associated with the AE mutation has not been identified, nor have the affected proteins been characterized. The 2DPAGE, due to its separation of proteins on the basis of isoelectric point and molecular weight, is the most sensitive technique for the separation of proteins from complex protein mixtures. Therefore, this was used to compare the proteins within total cell lysate prepared from normal and AE fibroblasts. The results indicate that two proteins with molecular weights between 49.6 and 49.9 kDa and pI of 5.1 are absent from AE fibroblasts. This is the first report of any protein differences between normal and AE cells. The difference in the abundance of the 49.6/49.9 kDa proteins is not due to a difference in the cellular Zn content because the lysate was prepared from cells grown for 4 d, the time when AE fibroblasts have a normal cell Zn content (Grider and Young 1996). Furthermore, growing the cells in 100 µmol/L Zn did not induce the expression of this protein in AE cells. The slight difference in the molecular weights of the two proteins may indicate a difference in phosphorylation, glycosylation or a difference in two or three amino acids. The observation that they exhibit the same pI suggests that they are related, or are the same protein.

The 49.6/49.9 kDa proteins are related to the expressed sequence tag isolated from a Jurkat T-cell leukemia cDNA library. Currently, the identity of the protein encoded by this cDNA is unknown; however, the expressed sequence tag exhibited 43% homology to adipophilin (Heid et al. 1996, Jiang and Serrero 1992). Adipophilin has a molecular weight of ~50 kDa (Heid et al. 1996, Jiang and Serrero 1992) and is associated with the membrane fraction of cells (Heid et al. 1996). At this time, we do not believe that the 49.6/49.9 kDa proteins are related to adipophilin for the following reasons. The amino acid structure of adipophilins is conserved among the mouse, cow and human (Heid et al. 1996). In comparison, the expressed sequence tag exhibited only 43% homology to adipophilin. Furthermore, the 15 amino acid and 18 amino acid peptide sequences exhibited 100% homology to portions of the expressed sequence tag, and only 30 and 44% homology, respectively, to corresponding regions of adipophilin. Although adipophilin exhibited a molecular weight of ~50 kDa, two variants with pI between 7.5 and 7.8 were observed (Heid et al. 1996). The 49.6/49.9 kDa proteins exhibited a pI of 5.1 and thus were more acidic than adipophilin. Because the nonredundant database did not contain homologous protein sequences for the 49.6/49.9 kDa protein, we believe that it is a novel, previously undescribed, protein.

Several eukaryotic zinc transport proteins have been identified. These include a wheat root-cell plasma membrane Zn uptake protein (Rengel and Hawkesford 1997) and a family of Zn uptake (Zhao and Eide 1996a and 1996b) and efflux proteins (Kaminzono et al. 1989) in yeast. The mammalian zinc transport proteins that have been identified to date are all involved in either cellular Zn efflux (Palmiter and Findley 1995) or the transport of Zn from cytosol into vesicles (Huang and Gitschier 1997, Palmiter et al. 1996a and 1996b). The yeast and mammalian proteins are related in that a moderately conserved histidine-rich region is present in a cytoplasmic loop between two of the transmembrane regions of these proteins (except in the protein ZnT3). It has been hypothesized that this region represents a putative Zn binding site.

Our current data do not indicate any sequence homology between the 49.6/49.9 kDa proteins and this family of zinc transporters. Nevertheless, their absence from AE fibroblasts supports the hypothesis that these proteins are related to the phenotypic expression of the AE mutation in these cells. Future experiments will be designed to characterize the structure of the 49.6/49.9 kDa proteins to determine their relationship, if any, to known families of zinc transport proteins and to cellular Zn uptake.

	FOOTNOTES

Manuscript received 26 February 1998. Initial reviews completed 13 April 1998. Revision accepted 19 May 1998.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Atherton D. J., Muller D.P.R., Aggett P. J., Harries J. T. A defect in zinc uptake by jejunal biopsies in acrodermatitis enteropathica. Clin. Sci. 1979; 56:505-507[Medline]
• Bradford M. M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 1976; 72:248-254[Medline]
• Chang C., Muga S. J., Grider A. Zinc uptake into fibroblasts is inhibited by probenecid. Biochim. Biophys. Acta 1998; 1368:1-6[Medline]
• Dillaha C. J., Lorincz A. L., Aavik O. R. Acrodermatitis enteropathica. Review of the literature and report of a case successfully treated with diodoquin. J. Am. Med. Assoc. 1953; 152:509-512
• Grider A., Lin Y., Muga S. J. Differences in the cellular zinc content and 5'-nucleotidase activity of normal and acrodermatitis enteropathica (AE) fibroblasts. Biol. Trace Elem. Res. 1998; 61:1-8[Medline]
• Grider A., Vasquez F. Nystatin affects zinc uptake in human fibroblasts. Biol. Trace Elem. Res. 1996; 54:97-104[Medline]
• Grider A., Young E. M. The acrodermatitis enteropathica mutation transiently affects zinc metabolism in human fibroblasts. J. Nutr. 1996; 126:219-224
• Heid H. W., Schnolzer M., Keenan T. W. Adipocyte differentiation-related protein is secreted into milk as a constituent of milk lipid globule membrane. Biochem. J. 1996; 320:1025-1030
• Huang L., Gitschier J. A novel gene involved in zinc transport is deficient in the lethal milk mouse. Nature Genet. 1997; 17:292-297[Medline]
• Jiang H. P., Serrero G. Isolation and characterization of a full-length cDNA coding for an adipose differentiation-related protein. Proc. Natl. Acad. Sci. U.S.A. 1992; 89:7856-7860[Abstract/Free Full Text]
• Kamizono A., Nishizawa M., Teranishi Y., Murata K., Kimura A. Identification of a gene conferring resistance to zinc and cadmium ions in the yeast Saccharomyces cerevisiae. Mol. Gen. Genet. 1989; 219:161-167[Medline]
• Palmiter R. D., Cole T. B., Findley S. D. ZnT2, a mammalian protein that confers resistance to zinc by facilitating vesicular sequestration. EMBO J. 1996a; 15:1784-1791[Medline]
• Palmiter R. D., Cole T. B., Quaife C. J., Findley S. D. ZnT3, a putative transporter of zinc into synaptic vesicles. Proc. Natl. Acad. Sci. U.S.A. 1996b; 93:14934-14939[Abstract/Free Full Text]
• Palmiter R. D., Findley S. D. Cloning and functional characterization of a mammalian zinc transporter that confers resistance to zinc. EMBO J. 1995; 14:639-649[Medline]
• Rabilloud T. A comparison between low background silver diammine and silver nitrate protein stains. Electrophoresis 1992; 13:429-439[Medline]
• Rengel Z., Hawkesford M. J. Biosynthesis of a 34-kDa polypeptide in the root-cell plasma membrane of a Zn-efficient wheat genotype increases upon Zn deficiency. Aust. J. Plant Physiol. 1997; 24:307-315
• Van Wouwe J. P. Clinical and laboratory diagnosis of acrodermatitis enteropathica. Eur. J. Pediatr. 1989; 149:2-8[Medline]
• Zhao H., Eide D. The yeast ZRT1 gene encodes the zinc transporter protein of a high-affinity uptake system induced by zinc limitation. Proc. Natl. Acad. Sci. U.S.A. 1996a; 93:2454-2458[Abstract/Free Full Text]
• Zhao H., Eide D. The ZRT2 gene encodes the low affinity zinc transporter in Saccharomyces cervisiae. J. Biol. Chem. 1996b; 271:23202-23210

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences