Main

MNK is an X-linked neurodegenerative disorder characterized by copper sequestration in various organs and cells, including intestine, kidney, and cultured skin fibroblasts(1, 2). Excessive intestinal copper accumulation reflects decreased copper absorption, which results in relative copper deficiency in multiple tissues, including the brain and liver. The MNK gene (MNK) has been identified and is predicted to encode a copper-transporting P-type ATPase(35). The MNK gene, and the murine homolog, mottled, are expressed in many tissues, with no or very low expression in liver(6, 7). Diverse mutations have been reported in patients with MNK, and these variously result in absent, reduced, or altered expression of the MNK gene(810). It is apparent that the MNK gene product plays a role in copper transport and homeostasis in normal tissues. However, little is known about the cellular patterns of MNK gene expression in individual organs and tissues.

Hemizygosity for mutations at the murine mottled locus results in phenotypic and biochemical manifestations which resemble those of MNK, and mutations such as blotchy, brindled, dappled, and macular have therefore been proposed as animal models for MNK(2). A specific mutation for blotchy has been identified, and altered mottled gene expression has been reported for dappled(6, 7, 11).

The macular mouse is also similar to human MNK in clinical phenotype and biochemical abnormalities(1214). In the present study, we report the identification of a missense mutation in the mottled gene of the macular mouse, validating macular as a murine model of MNK. We then compare mottled gene expression in the intestine, kidney, and brain of macular mice and normal mice, using Northern analysis and in situ hybridization, in an effort to relate gene expression to the pathogenesis of disease manifestations.

METHODS

Materials. Male hemizygote macular mice were treated with a s.c. injection of cupric chloride solution (50 μg of CuCl2) on postnatal d 7(15). They were kept on a normal diet under standard conditions for 4 wk and then prepared for study. Age-matched male C3H strain mice (Charles River, Tokyo, Japan) were used as controls. All animal studies were approved by the Teikyo University School of Medicine Committee on Use and Care of Animals in research.

cDNA sequencing. Poly(A)+ RNA was isolated from kidney tissue and skin fibroblast cell cultures derived from a macular mouse using a Pharmacia QuickPrep Micro RNA kit. Random primed cDNA was synthesized from 500 ng of Poly(A)+ RNA as described in Das et al.(11). PCR was performed on 30-50 ng of cDNA using mottled cDNA-specific primers designed to amplify the coding region into four overlapping segments. Segment 1, from nucleotide positions 56-1326, was amplified using primers 5′-CAGAGCTCGAACCCCAGCCCTG and 5′-CAATAGTCCCTGTGCTGTTTGCG followed by nested PCR with 5′-GAAACCCAGGAATGTAAAGAC and 5′-CTGTTTGCGAGGGACACGTGG. Segment 2, from nucleotide positions 1026-2450, was amplified with primers 5′-CCAGTATGTAAGCAGTATAG and 5′-GTTCCAGCCATCGTCCTAGTGCG and hemi-nested with 5′-GGTCAGCCATTGTAAAGTAC. Segment 3, from nucleotide positions 2241-3570, was amplified using primers 5′-GTGTCTACCTGTACAGTTTTG and 5′-CGGTTACCAATGAGGACTTTG and nested with 5′-GGTACTTCTACATTCAGGCTTAC and 5′-GCATTTGAGAGATGAGCATCAATG. Segment 4, from nucleotide positions 3308-4661, was amplified using primers 5′-CAATAAGATCCTGGCCATTGTGGGG and 5′-CCTTGCACGTAAGAGCATGAC and heminested with 5′-GAACATCCTTTAGGAGCAGCTG. The PCR products were sequenced, on both strands, using an ABI prism T™ 377 DNA sequencer.

RNA preparation and Northern blot analysis. The intestine, kidney, and brain were removed from a macular mouse and a control mouse after decapitation. Total RNA was isolated from each tissue using Stratagene's RNA isolation kit (Stratagene; La Jolla, CA). Approximately 30 μg of RNA were fractionated on a 0.8% agarose-formaldehyde gel and then transferred to a GeneScreen Plus nylon membrane (Dupont NEN, Boston, MA). This membrane was hybridized for 20 h at 42 °C with a 32P- labeled partial mottled cDNA probe corresponding to bp 1819-4163(6). The membrane was washed twice in 2 × SSPE, 0.1% SDS for 10 min at room temperature, twice in 1 × SSPE, 0.1% SDS for 10 min at 50 °C, one in 0.1 × SSPE, 0.1% SDS for 10 min at room temperature, and then subjected to autoradiography. To quantify the relative level of the transcript, the same Northern blot was hybridized with a β-actin cDNA probe. The level of the transcript was measured using a BAS-1500 Mac, Bio Imaging Analyzed (Fuji Photo Film, Tokyo, Japan).

Preparation of RNA probes. A 2.3-kb fragment of mottled cDNA(bp 1819-4163)(6) was subcloned into SKII Bluescript vector. The plasmid (1 μg of DNA) was linearized by cutting with Xho I (for the antisense strand) and BamHI (for the sense strand). In vitro transcription was performed using the appropriate RNA polymerases (T3 RNA polymerase for the antisense strand; T7 RNA polymerase for the sense strand) and digoxigenin-labeled UTP (Boehringer Mannheim, Germany), as described previously(16). Alkaline-treatment was used to reduce the probes to an average size of 150 nucleotides.

In situ hybridization. Ether-anesthetized mice were transcardially perfused with 4% paraformaldehyde, PBS, pH 7.4. The intestine, kidney and brain were excised, embedded in OCT compound (Miles, Elkhart, IN), and frozen. Cryostat sections (5 μm for the intestine and kidney, 8 μm for the brain) were cut and mounted on glass slides coated with poly-L-lysine(Matsunami, Osaka, Japan). In situ hybridization was performed as described previously(17), with a few modifications. Prehybridization was performed with 2 × SSC in 50% formamide at 50°C for 60 min. The sections were immersed in 50 μL of a hybridization buffer consisting of 50% formamide, 10% dextran sulfate, 20 mM Tris-HCl, 0.3 M NaCl, 5 mM EDTA, 10 mM vanadyl ribonucleoside complex, 1 × Denhardt's solution, 1 mg/mL tRNA, 0.25% SDS, and digoxigenin-labeled RNA probe(50× a probe solution for the intestine and kidney sections, 100× for the brain). The sections were covered with Parafilm and incubated in a moisturized chamber at 50 °C for 19 h. Then, the sections were immersed in 2 × SSC, 50% formamide for 1 h at 50 °C for the intestine and kidney and at 55 °C for the brain. After being washed in 2 × SSC for 1 h, and in 0.1 × SSC for 10 min at room temperature, the sections were treated with blocking reagents (Boehringer Mannheim, Germany) and incubated with an alkaline phosphatase-labeled anti-digoxigenin antibody solution(Boehringer Mannheim, Germany). The samples were developed using 4-nitroblue tetrazolium chloride and 5-bromo-4-chloro-3-indolyl-phosphate. After development, the sections were lightly counterstained with methyl green.

RESULTS

Sequence analysis of macular cDNA. RT-PCR amplification was performed using macular mottled mRNA isolated from tissue and cultured fibroblasts, and the resulting products were sequenced as described in“Methods.” The sequence of the mottled coding region (4473 bp) revealed a single base change T → C, at nucleotide position 4223 resulting in an amino acid change from serine 1382 to proline. This mutation is located in the eighth transmembrane domain, a region highly conserved in many P-type ATPases, including the human MNK gene.

To test the possibility that this may represent a polymorphism, DNA from a series of inbred mouse strains and other mottled alleles, including C57BL/6J, CBA, PT, brindled, dappled, and the macular parental strain C3H, were amplified and tested for the presence of the mutation. Because the mutation results in a loss of a BamHI site, the PCR products were digested with BamHI, and only the macular PCR sample remained uncleaved.

The phenotype of the macular mouse is remarkably similar to that of brindled, another allele at the mottled locus(18). Both mutations cause severe neurologic disease, which can be ameliorated by copper supplementation during the first 10 postnatal days(1, 2, 1214, 19). The brindled mutation is not the same as macular, as demonstrated by documentation of the presence of the BamHI site at the 4223 nucleotide position in brindled mutation. We have found a 6-bp deletion at the position of nucleotides 2479-2484 of the brindled cDNA. This mutation would be predicted to result in the loss of two amino acids, leucine and alanine. This same deletion was also found by J. F. B. Mercer B. Levinson, J. Gitschier, and S. Packman (personal communication), Bankier(20), and Reed et al.(21).

Northern blot analysis of mottled mRNA from normal and macular mice. Northern blot analysis revealed an 8.3-kb transcript in the intestine, kidney, and brain of normal mice, and a transcript of the same size in the corresponding tissues of macular mice. The level of the mottled transcript in the intestine, kidney, and brain of macular mice was approximately 80% that of normal mice in each tissue (Fig. 1).

Figure 1
figure 1

Northern blot analysis of RNA from control and macular mice using mottled cDNA as a probe. Total RNA samples were prepared from the brain (lanes 1 and 2), intestine (lanes 3 and 4) and kidney (lanes 5 and 6) of a control mouse(lanes 1, 3, and 5) and of a macular mouse (lanes 2, 4, and 6). Hybridization with β-actin probe on the same filter is also shown. When the ratios of each transcript to the level of the respective β-actin transcript were compared, the level of the transcript in the intestine, kidney, and brain of the macular mouse was all about 80% of the respective tissues in the normal mouse.

Detection of mottled mRNA in the liver, intestine, and kidney by in situ hybridization. Normal mouse liver was not labeled by in situ hybridization, indicating that mottled mRNA is not detectable in normal liver by this analytic method (Fig. 2). In contrast, intense signals were observed in the absorptive epithelial and Paneth cells of the normal mouse intestine (Fig. 3B). In the normal kidney sections (Fig. 4B), some of the tubular cells were intensely labeled, whereas glomeruli showed a low level of expression. Periodic acid-Schiff staining indicated that the intensely labeled cells were proximal tubular cells (data not shown). The intensity of labeling in mutant and control kidney and intestine were comparable, but slightly weaker in the macular.

Figure 2
figure 2

In situ hybridization of mottled mRNA in the liver tissue from a control mouse (A, with the sense probe;B, with the antisense probe). Labeling was hardly observed in the normal liver tissue. Bar = 100 μm.

Figure 3
figure 3

In situ hybridization of mottled mRNA in the intestine sections from a control mouse (A, with the sense probe;B, with the antisense probe) and from a macular mouse (C, with the antisense). Bar = 50 μm.

Figure 4
figure 4

In situ hybridization of mottled mRNA in the kidney sections from a control mouse (A, with the sense probe;B, with the antisense) and from a macular mouse (C, with the antisense probe). Proximal tubulus (pt), distal tubulus(dt), bar = 25 μm.

Detection of mottled mRNA in the brain by in situ hybridization. In sections of normal brain, intense signals were observed in the choroid plexus (Fig. 5E), in Ammon's horn and the dentate gyrus in the hippocampus (Fig. 5C), and in Purkinje cells and the granular layer of the cerebellum (Fig. 5G). The ependyma, piriform region, amygdaloid body, and medial habenular nucleus were also intensely labeled (Fig. 5A). Neurons were clearly labeled, whereas glia, including astrocytes, were faintly labeled. Blood vessels were also faintly labeled. In the brain of macular mice, the intensity and localization of signals were similar to those of control mice (Fig. 5, B, D, F, and H).

Figure 5
figure 5

The distribution of the mottled mRNA in the brain sections of a control mouse (A, C, E, and G with the antisense probe) and of a macular mouse (B, D, F, and H with the antisense probe). Intense signals were observed in the choroid plexus(cp) (E and F), the Ammon's horn and dentate gyrus in the hippocampus (C and D), and the Purkinje cells(arrows) and granular layer (gl) (G and H) in the cerebellum of both control and macular mice. Bars:A and B = 800 μm; C and D = 300μm; E, F, G, and H = 50 μm.

DISCUSSION

The MNK gene shows an overall identity of 57% with that for Wilson disease, another genetic disorder of copper transport(2224). Although both genes encode copper-transporting P-type ATPases, the MNK gene is widely expressed, except in liver(3, 6, 7), whereas the Wilson disease gene (WND) is expressed most highly in liver (and in kidney)(22, 24). The probe used in the present in situ hybridization studies resulted in no detectable label in the normal liver sections, indicating specificity of the probe for MNK mRNA, without cross-reactivity with WND mRNA.

Histochemical studies have shown that copper accumulates excessively in the absorptive epithelial cells, Paneth cells, and proximal tubular cells of the macular mouse and of the brindled mouse, another animal model of MNK(25, 26). Moreover, renal proximal tubular damage caused by copper accumulation has been reported in the patients with MNK(27) and the brindled mouse(28). Accordingly, it was of interest to examine expression of the mottled gene in sections of normal and mutant intestine and kidney.

Northern analyses showed that mottled mRNA was present in macular intestine and kidney, at a level approximately 80% of that in the respective tissues of normal mice. By in situ hybridization, intense signals were observed in the absorptive epithelial cells and Paneth cells of normal intestine, and in the proximal tubular cells of normal kidney. The corresponding cells in macular mice were also labeled. The localization of the mottled message was in identical cell types in both normal and mutant sections, and these cell types correspond to those that show either copper accumulation or disease manifestations in MNK, and in the macular and brindled mice. The aggregate data are consistent with the notion that the MNK/mottled gene product plays a role in copper transport in intestinal epithelial cells and renal proximal tubular cells.

In both MNK(1, 2) and the macular mouse(29, 30), neuronal cell death and neuronal degeneration have been reported. Copper accumulates in the capillary epithelium of macular and brindled mice(25, 31), and in cultured astrocytes of the macular mouse brain(32). P-type ATPase activity associated with copper efflux has been reported in rat glioma cells(33). These findings suggest that the MNK/mottled gene product plays a role in copper transport in the capillary epithelium and glia of brain, and that abnormal function of the MNK/mottled gene in those cell types leads to the neuropathology of mutants such as the macular mouse. We therefore undertook a direct examination of mottled gene expression in normal and mutant mouse brain.

Northern analysis of brain mRNA showed that the mutant transcript was identical in size to that of normal brain, and was present at about 80% of the level in normal brain. The results of in situ hybridization studies of brain tissue revealed that the localization and intensity of gene expression in macular mice are almost the same as in normal mice. In contrast to expectations derived from the studies of copper accumulation noted above, the in situ hybridization results showed that mottled gene expression was more intense in neurons than in glia. Intense gene expression was observed in both normal and mutant Purkinje cells and granular layer, in Ammon's horn and the dentate gyrus in the hippocampus, and in the choroid plexus. It is possible that expression of the mottled gene in capillary epithelium and glia may be too low to be detected by in situ hybridization, even in normal tissues. We also note that, in this work, we are not distinguishing between expression of full-length, functional mRNA,versus mRNA which may represent alternatively spliced message such as that reported for the WND gene in human brain(24, 34) and for the human MNK gene in some tissues(35). Nevertheless, as a first approximation, the expression of mottled gene in neuronal cells suggests that the MNK/mottled gene product plays a direct role in copper transport in specific neuronal cells, and that copper homeostasis in neurons is not entirely governed by transport mechanisms in nonneuronal cells such as those of the capillary epithelium and glia. Certainly, the data we report are consistent with neuropathologic studies documenting severe damage to cerebellar Purkinje cells and the hippocampal region in MNK patients and in mice with mutations at the mottled locus(28, 30, 36).

Iwase et al.(29) have also reported that the mottled gene is expressed at normal levels in the brain of macular mice, but the distribution of expression reported by that group differed some-what from ours. They noted low expression in Purkinje cells and the CA3 region of the hippocampus. Although they used a human cDNA probe, the probe used in our study is 95% homologous to the same region of the human cDNA, and so the explanation for the discrepancy does not reside in the use of different probes. We note that the mice used in our study were more mature than those used by Iwase et al., and so the discrepancy may be a function of the age differences in the mice of the two studies. This would represent a very interesting developmental progression, and bears further investigation.

The finding of a serine-to-proline substitution in the mottled cDNA of the macular mouse confirms that the macular mouse is a valid animal model for MNK. Evidence that the T4233C change is the causative mutation is based on four points: 1) no other base change was found in the cDNA; 2) the change is unlikely to be a polymorphism; 3) the change results in a nonconservative amino acid replacement: serine is a hydrophilic amino acid, whereas proline is very rigid and creates a fixed kink in a polypeptide chain; and 4) the predicted consequence of this mutation is to alter the highly conserved eighth transmembrane domain. Interestingly, Theophilus et al.(37) have shown a methionine-to-valine amino acid substitution in the eighth transmembrane domain of the mouse WND gene in the toxic milk mouse, a mutant that had been put forth as a possible mouse homolog of Wilson disease. Theophilus et al.(37) pointed out that the methionine in that position was conserved in all eukaryotic and prokaryotic copper-transporting ATPases. It appears that the region of the eighth transmembrane domain may be critical to the proper functioning of copper-transporting ATPases.

The deletion found in brindled is an in-frame deletion, and occurs in a region not yet confirmed as critical for ATPase function. Nevertheless, the deletion was not found in other tested mouse strains (B, Levinson, A. Grimes, C. J. Hearn, P. Lockhart, D. F. Newgreen, J. F. B. Mercer, unpublished results) and is a reasonable candidate for the causative mutation in brindled. As more mutations become associated with mottled alleles of specific and distinct phenotype(18), we expect to gain a better understanding of the structure-function relationships of the MNK/mottled protein, as well as the molecular basis for individual components of the complex phenotypic manifestations of mutations at the MNK and mottled loci.