Characterization of ATP7A missense mutants suggests a correlation between intracellular trafficking and severity of Menkes disease

Abstract

Menkes disease (MD) is caused by mutations in ATP7A, encoding a copper-transporting P-type ATPase which exhibits copper-dependent trafficking. ATP7A is found in the Trans-Golgi Network (TGN) at low copper concentrations, and in the post-Golgi compartments and the plasma membrane at higher concentrations. Here we have analyzed the effect of 36 ATP7A missense mutations identified in phenotypically different MD patients. Nine mutations identified in patients with severe MD, virtually eliminated ATP7A synthesis, in most cases due to aberrant RNA splicing. A group of 21 predominantly severe mutations led to trapping of the protein in TGN and displayed essentially no activity in a yeast-based functional assay. These were predicted to inhibit the catalytic phosphorylation of the protein. Four mutants showed diffuse post-TGN localization, while two displayed copper dependent trafficking. These six variants were identified in patients with mild MD and typically displayed activity in the yeast assay. The four post-TGN located mutants were presumably affected in the catalytic dephosphorylation of the protein. Together these results indicate that the severity of MD correlate with cellular localization of ATP7A and support previous studies indicating that phosphorylation is crucial for the exit of ATP7A from TGN, while dephosphorylation is crucial for recycling back to TGN.

Introduction

Menkes disease (MD; OMIM: 309400), including the milder form Occipital Horn Syndrome (OHS; OMIM: 304150), is a rare (1:300.000), X-linked, multisystemic lethal disorder of copper metabolism linked to mutations in the ATP7A gene (OMIM: 300011). The symptoms of MD derive from a lack of dietary copper absorption and copper transfer across the basolateral membrane of intestinal enterocytes into the portal circulation, and from impaired reabsorption of copper in the kidney. This leads to accumulation of copper in intestinal cells and kidney, and lack of copper in vital organs such as heart, liver and brain, and reduced activity of essential cuproenzymes1,2,3. MD can be divided into three subclasses based on phenotype and progression. The severe classical form (observed in 90–95% of MD patients) is characterized by progressive neurodegeneration, connective tissue abnormalities, distinctive “kinky” hair and death typically before the age of three years. Atypical mild MD is characterized by longer survival and/or milder symptoms of the affected patients. OHS is mainly characterized by connective tissue manifestations and is the mildest form1. Recently also two unique mutations have been shown to cause isolated adult-onset spinal muscular atrophy (OMIM: 300489)4.

ATP7A belongs to the P-type ATPase family of ATP-driven membrane pumps that maintain electrochemical gradients as well as cationic and lipid homeostasis. P-type ATPases share a structural core containing a transmembrane (TM) domain responsible for transport, and three soluble domains - N, P and A - required for nucleotide binding, phosphorylation, and dephosphorylation, respectively (Fig. 1a). Copper transporting members of the family, including ATP7A and bacterial CopA proteins, have eight membrane spanning helical segments (TMA, TMB, TM1-TM6), and typically one or more metal-binding domains (MBD) at the amino-terminus with metal-binding CXXC motifs. The mammalian P-type ATPases, ATP7A and ATP7B, the latter mutated in Wilson’s disease, have six MBDs in sequential order5.

Figure 1
figure1

Topology and reaction cycle of ATP7A. (a) Topology of ATP7A. ATP7A, and other P-type ATPases consist of three cytoplasmic domains, nucleotide binding (N, red), phosphorylation (P, blue) and dephosphorylation/actuation (A, yellow). The transmembrane domain encompasses eight membrane-spanning segments (TM), two class I specific (TMA-TMB, cyan) and six conserved segments (TM1-6, wheat). The N-terminus contains six class-specific metal-binding domains (H1-H6, cyan). The copper-donating chaperone ATOX1 (green) and conserved motifs are shown. The non-cytosolic part of ATP7A is located in the TGN or in the extracellular milieu, due to copper-dependent trafficking. (b) The Albers-Post (E1-E2) reaction cycle of ATP7A and other Cu-transporting P-type ATPases. The domains are colored as described above, and copper ions are shown in green. Phosphorylation events in the intracellular domains drive large conformational changes that permit alternating access to transport sites in the membrane about 50 Å from the ATP-targeted catalytic aspartate. A high-affinity state (E1) binds copper and enters an occluded state, which then undergoes phosphorylation (E1.Pi-ADP). Completion of this event (E1P) triggers release of the ion, establishing an outward-facing, low-affinity state (E2P). Release of inorganic phosphate (E2.Pi) yields the fully dephosphorylated conformation (E2), which is followed by restoration of the inward-facing conformation (E1) that initiates a new reaction cycle. (c) Proposed cellular trafficking of ATP7A as an effect of copper concentration. At low cellular copper concentrations the wild-type ATP7A is located in the Trans-Golgi Network (TGN), whereas at higher intracellular copper levels, the steady state distribution of ATP7A shifts to the plasma membrane (and cytosolic vesicles, not shown).

The catalytic cycle of the ATP7A protein is associated with four principal corner stone reaction intermediates (E1, E1P, E2P and E2, Fig. 1b) with considerably different three-dimensional shapes due to structural rearrangements. ATP7A function requires cytosolic copper delivered to the TM domain6,7,8. Copper is donated from copper chaperones such as ATOX19, 10. The E1 state of the protein binds copper with high affinity within the TM domain11, 12. Copper-binding and ATP-recognition allow for auto-phosphorylation of the invariant D1044 in the P-domain leading to the E1P state and occlusion of copper in the TM domain. Completion of phosphorylation triggers large conformational changes that affect access to the ion-binding residues permitting copper release to the non-cytosolic side, reaching the E2P state13,14,15,16. Auto-dephosphorylation of the protein to the E2 conformation follows, which then shifts to the E1 state to initiate a new reaction cycle17. The 875TGE motif of the A-domain is important for auto-dephosphorylation15.

In mammals, trafficking of ATP7A is essential for proper copper homeostasis. At low cellular copper concentrations wild-type ATP7A is located in the Trans-Golgi Network (TGN), whereas at higher intracellular copper levels, the steady state distribution of ATP7A shifts to cytosolic vesicles and to the plasma membrane (Fig. 1c)18, 19. In the TGN, the ATP7A protein is essential for the delivery of copper to cuproenzyme biogenesis, and in the plasma membrane ATP7A is responsible for extrusion of excess copper from the cell19, 20. The copper-mediated trafficking response is reversible; if the intracellular copper level is reduced ATP7A returns to TGN19.

Although the mechanism behind the copper-dependent trafficking of ATP7A is far from fully understood, it has been demonstrated that several elements of the ATP7A protein play important roles. A 38 amino acid sequence containing transmembrane segment 3 was found to be essential for localization to the Golgi complex21. At least one MBD needs to be present for correct copper-dependent trafficking22. The dileucine motifs, 1487LL and 1459LL within the cytosolic carboxy terminal, have been shown to be essential for TGN localization by playing a role in the retrieval of ATP7A from the plasma membrane23. It has recently been demonstrated that Adaptor Protein complexes 1 and 2 interact physically with ATP7A, probably through the 1478LL motif, and this interaction might mediate the trafficking24. Copper-responsive kinase phosphorylation of S1469 and constitutive phosphorylation of S1432 have also been demonstrated to regulate the trafficking of ATP7A25. Importantly, phosphorylation of the catalytic D1004 has been shown to play an important role in copper-induced re-localization of ATP7A from TGN and the subsequent dephosphorylation important for retrieval13.

Few ATP7A variants in man and mice have previously been investigated for the effect on copper-dependent trafficking18, 26,27,28. By studying a large number of missense mutations, all of which hamper copper homeostasis observed in patients with different phenotypes, it is possible to achieve a greater clarification of the complex copper-dependent trafficking process. As a consequence of the E1-E2 reaction cycle and copper-dependent localization of ATP7A, there are a number of ways by which missense mutations may hamper copper homeostasis. Different steps in the reaction cycle such as copper-binding or -release, phosphorylation or dephosphorylation may be prevented. In addition, copper-dependent trafficking may be impaired. At the same time, missense mutations may also influence biogenesis via mRNA splicing, mRNA stability, protein folding or protein stability.

To shed further light on missense mutations in general and the role of MD mutations in particular, we here investigated the effect of 36 missense mutations the ATP7A gene with regard to i) the splicing of ATP7A pre-mRNA, ii) levels of ATP7A transcript and ATP7A protein product, iii) the intracellular localization of the mutated ATP7A protein and iv) in vivo copper transport capacity.

We find that a significant fraction of the mutations lead to the absence or severely reduced amounts of ATP7A protein, rather than affecting the function of the ATP7A protein per se. With respect to the mutations which allow the production of ATP7A protein, most of the mutations affect residues that are implicated in the ATPase-coupled transport function, and abolish copper-dependent trafficking of ATP7A. Enzymatic activity and trafficking thus appear to be coupled processes for ATP7A.

Our work reveals that severe mutations, predicted to inhibit the catalytic phosphorylation of the protein led to trapping of the protein in TGN, whereas milder mutations, predicted to inhibit the catalytic dephosphorylation of the protein led to diffuse post-TGN localization of the protein. Thus, the results suggest a correlation between the severity of the mutations and the cellular localization of the protein.

Results

A library of cultured fibroblasts from 36 different Menkes/OHS patients (24 with the classical MD phenotype, ten with a typical MD, one with OHS and one with a mixed manifestation), each carrying a missense mutation in the ATP7A gene29, was used for investigating the molecular cause of MD in these patients.

Biochemical diagnosis confirmed Menkes disease

The copper-uptake capacity (64Cu uptake/mg protein/20 h) of fibroblasts from the 36 patients was analyzed and found to be higher, whereas the excreting capacity (% of total 64Cu-uptake retained after 24 h) was lower when compared to control fibroblasts, indicating Menkes disease in all cases (Fig. 2).

Figure 2
figure2

Summary. The color represents the mutational effect on ATP7A localization: Normal trafficking of the protein (blue); retention in post-Golgi compartments (grey); retention in the trans-Golgi network (orange); not certain (green); no detectable protein (purple). “A” denote atypical, “O” OHS, and “C” classical phenotype. Pos: indicates affected region (MBD, TM, A, P). Protein: Indicates whether protein was detectable with IF “Yes” or WB only “YesWB”. Localization: Location of ATP7A always in TGN even in the presence of Cu is designated by “TGN(+Cu)”. Never located in TGN is designated by “not TGN”; Location in the TGN only the absence of copper is designated by “normal”. No detectable protein “n.d”. ATP7A-mRNA: indicates % transcript level compared to transcript level in control fibroblasts. Splicing: indicates normal splicing “Normal” or aberrant splicing “All mal”. Empty boxes, indicate “not investigated”. Compl: indicates ability to complement the ccc2Δ yeast strain. “Some56” and “Some80”, indicate some complementation after 56 hours and 80 hours, respectively. “Yes80” indicates full complementation after 80 hours. If no hours are indicated, it means >56 hours. Uptake/retention: Uptake indicates the amount of radioactive copper accumulated after 20 hours incubation (64Cu/mg protein/20 hours). Retention indicated the amount in % of total accumulated 64Cu retained after subsequent 24 hours in the absence of radioactive copper.

Copper-regulated trafficking of ATP7A variants in MD fibroblasts is compromised

Fibroblasts from the 36 patients were evaluated for the presence and localization of ATP7A protein by immunofluorescence (IF) (Fig. 3). Fibroblasts from a healthy person were included as a positive control, whereas fibroblasts from a MD patient with a deletion of exons 3–23 (c.121−?_8333+?del) in ATP7A were used as a negative control. Fibroblasts were either treated with CuCl2 or the copper chelator Bathocuproine disulphonate (BCS) to create a milieu of excess versus low amounts of bio-available copper.

Figure 3
figure3

Cellular investigation of endogenous ATP7A protein in MD fibroblasts by indirect immunofluorescence (IF). MD fibroblasts with various missense mutations were stained with primary antibodies against ATP7A (green; position 1) and the Golgi specific marker GS28 (red; position 2), respectively. Also a merge picture is shown (position 3). The nuclei were counterstained with DAPI. Copper-dependent trafficking was investigated in the presence of BCS (upper panel) and CuCl2 (lower panel), respectively. The pictures are divided into groups defined by the effect of the mutation on cellular location of the ATP7A protein. P1: Normal control cells (C+). P2: ATP7A negative control cells (C−). P3–P4: Copper dependent trafficking. P5-P17: No ATP7A signal was detected. P18–P38: Hampered copper-induced trafficking from TGN.

In positive control fibroblasts, ATP7A showed a copper-dependent localization. In the presence of BCS, ATP7A is located in TGN, and is partly co-localized with the TGN GS28 SNARE protein, whereas in the presence of copper, distinct localization was replaced by a diffuse, cytoplasmic, and very weak ATP7A signal (Fig. 3, Panel P1; Supplementary Fig. 1) in accordance with previous studies17. The loss of a TGN-signal for ATP7A - when conditions change from low to high copper levels - indicates copper-induced trafficking. No ATP7A signal was detected in the negative control cells (C−) under any conditions (Fig. 3, Panel P2; Supplementary Fig. 2).

A copper dependent trafficking pattern was only observed for ATP7A with the predicted amino acid substitutions S761PA or Q924R° (Fig. 3, Panels P3 and P4, A and ° denote atypical and OHS phenotypes, respectively; Supplementary Figs 3 and 4). Although copper dependent trafficking was observed, the pattern of S761PA and Q924R° was more diffuse in the presence of BCS, compared to the wild type ATP7A (C+), with staining in the cytoplasm in addition to TGN. The degree of co-localization with GS28 of Q924R° was however on par with the wild type ATP7A, whereas the co-localization of S761PA with GS28 was reduced compared to wild type ATP7A, indicating mis-localization to other areas of TGN or other nearby located compartments of the endomembrane systems (Supplementary Figs 3 and 4). The Pearson's correlation R values, (co-localization analysis tool, Coloc 2; Fiji) in the presence of BCS/Cu respectively were 0.41/0.06 for Q924R°, 0.18/−0.06 for S761PA, and 0.51/0.07 for wild type ATP7A. The R values for negative control cells (C−) in the presence of BCS/Cu respectively were −0.02/−0.19 (Supplementary Table 1). An R value of 1 corresponds to total positive correlation, 0 corresponds to no correlation, and −1 corresponds to total negative correlation.

In contrast, when looking at ATP7A with the 21 predicted substitutions: S653YA, G727RC, G728DA, C1000RA, G1005RC, A1007VA, D1044GC, G1255RC, K1282EA, G1300EC, D1301GC, G1302VC, G1302EC, N1304KC, D1305AC, D1305GC, A1308DC, G1315RC, A1362VA, M1393TC, and S1397FC (Fig. 3, Panels P18–P38, Cdenote classical phenotype), copper-dependent trafficking was hampered because ATP7A was localized in the TGN regardless of the copper concentration. The Pearson’s correlation R values for co-localization with GS28 were between 0.30 and 0.73 for all cultures with no copper dependent differences, in agreement with no copper dependent trafficking. The signal for G1255RC was however very faint (Fig. 3, Panel P25; Supplementary Table 1).

The remaining 13 mutants (E628VC, K633RC, G666RA, K802NC, R844HC, G860VC, G876RC, G876EA, G1015DA, K1037NC, A1325VA/ O /C, G1369RC and A1373PC) did not reveal any convincing ATP7A signal regardless of the treatment (Fig. 3, Panels P5 to P17). The results are summarized in Fig. 2.

Western blotting confirmed that ATP7A protein is present in a fraction of the IF negative cultures

The ATP7A protein is difficult to visualize by IF staining unless it accumulates in a physically distinct cellular compartment such as TGN. The inability to detect ATP7A protein by IF may therefore not only be due to amino acid substitutions which compromise ATP7A accumulation, but could also be due to substitutions which abolish distinct TGN-location.

The fibroblast cultures with no IF-detectable ATP7A were therefore examined by Western blot analysis (WB) to assess whether any ATP7A protein was present (Fig. 4). We included the G1255RC substitution, with very faint IF signals, to compare the WB and IF results.

Figure 4
figure4

Western blots of lysates from MD fibroblasts. Western blots of lysates from MD fibroblasts with no detectable ATP7A protein when analysed by Immunofluorescence (IF). The specific ATP7A missense mutations are indicated above the lanes. Control + lysate from normal fibroblasts. Control − lysate from MD fibroblasts containing a big deletion of the ATP7A gene (exon 3–23 deleted). The protein product of the housekeeping gene GAPDH was used as an internal control for correct loading of 25 μg protein in each lane. The staining of ATP7A and GAPDH was performed on one membrane cut in two. Gaps indicate separate gels. Small gaps indicate that the gel was cut in producing the figure.

The band intensity for ATP7A in cells with the substitutions G666RA, G876EA, G1015DA and A1325VA/O/C corresponded to the band intensity of the wild-type positive control, indicating that these four amino acid substitutions did not affect protein accumulation, but hamper their localization to TGN.

The band intensity of accumulated G860VC and G1255RC protein was slightly reduced, making it difficult to determine the cellular localization of G860V by IF. In contrast, no ATP7A protein was observed in cells predicted to express the eight variants E628VC, K633RC, K802NC, R844HC, G876RC, K1037NC, G1369RC or A1373PC. The results are summarized in Fig. 2.

Absence of ATP7A protein is in most cases due to aberrant splicing

Undetectable ATP7A protein may relate to severely reduced levels of correctly spliced mRNA, impaired protein synthesis or reduced protein stability. To distinguish between these possibilities, the level of ATP7A mRNA was determined in fibroblasts from all patients by real time PCR. Of the eight amino acid substitutions that prevented ATP7A accumulation, only four - E628VC, K633RC, R844HC and G876RC led to a reduced accumulation of ATP7A transcript (10–25% remained compared to the healthy control). Fibroblasts with the remaining four variants - K802NC, K1037NC, G1369RC and A1373PC - accumulated ATP7A mRNA on par with the healthy control. All twenty-eight additional ATP7A mutants had normal mRNA levels. The results are summarized in Fig. 2.

To test for erroneous ATP7A RNA splicing, reverse transcription-PCR (RT-PCR) was carried out on mRNA from fibroblasts encoding either i) E628VC, K633RC, R844HC or G876RC with a small amount of ATP7A transcript and no detectable ATP7A protein, ii) K802NC, K1037NC, G1369RC or A1373PC, with a normal amount of ATP7A transcript, and no detectable ATP7A protein iii) G860VC or G1255RC, with a normal amount of ATP7A transcript and a reduced amount of ATP7A protein. The PCR primers were designed to amplify the cDNA sequence covering the mutation-containing exon and several flanking exons.

Electrophoretic analysis of the RT-PCR products showed that the mutations predicting the substitutions G876RC, E628VC and K633RC (Fig. 5AI), K802NC, K1037NC, G1369RC and A1373PC (Fig. 5AII) all produced an RT-PCR pattern which differed from that of the wild-type. Except for A1373PC, sequence analysis of the PCR products did not reveal any normally spliced transcripts, but only mal-spliced transcripts in which different exons had been skipped. None of these exons, except for exon 10, were skipped in the healthy control. This concurs with previous observations that show that 10–15% of ATP7A transcripts in healthy cells skip exon 1030. In contrast, the RT-PCR band pattern obtained from R844HC, G860VC or G1255RC fibroblasts was indistinguishable from the wild-type sample (Fig. 5AI and AIII). Accordingly, sequence analysis confirmed that only normally spliced transcripts were present in these fibroblasts. These results are summarized in Table 1.

Figure 5
figure5

(A) RT-PCR investigation of ATP7A transcript. RNA from MD fibroblasts were subjected to RT-PCR spanning the exons with the missense mutations. Control samples are from a normal control. Size (in bp) of marker DNA fragments are indicated. The PCR products were separated by gel electrophoresis, purified and sequenced (Table 1). Primers are given in Supplementary Table 2. (I) No detectable ATP7A protein and a reduced amount of ATP7A transcript. A: PCR amplification of exons 7–14. E628VC and K633RC (mutations in exon 8); skipping of exon 8, exons 8–9 or exons 8–10. G876RC (mutation in exon 12); skipping of exon 12, exons 11–13 or exons 10–13. R844HC indistinguishable from the wild-type sample. (II) No detectable ATP7A protein and a normal amount of ATP7A transcript. B: PCR amplification of exons 7–14. K802NC (mutation in exon 10); skipping of exon 10. C: PCR amplification of exons 12–23. K1037NC (mutation in exon 15); skipping of exon 15. D: PCR amplification of exons 19–23. G1369RC (mutation in exon 21); skipping of exon 21 or exons 20–21, and A1373PC (mutation in exon 21); normal transcripts or skipping of exon 21 or exons 20–21. (III) Reduced amount of ATP7A protein and normal amount of ATP7A transcript. E: PCR amplification of exons 7–14. G860VC indistinguishable from the wild-type sample F: PCR amplification of exons 17–23. G1255RC indistinguishable from the wild-type sample. (B) Mutations presumed to affect donor splice sites. According to “human-Splicing Finder” (http://www.umd.be/HSF/) splice sites with consensus values (CV)’s higher than 80 are strong splice sites, whereas splice sites with CV’s between 65 and 70 are weak30. A relative change in CV (∆CV) of 10 percentage points relative to wild-type sites is predicted to attenuate splicing. (C) Mutations presumed to affect ESS/ESE sites. Disruption of ESE sequences and/or creation of an ESS sequences may lead to exon skipping. In agreement with the observed splicing pattern, the four base pair substitutions are predicted by the web-tools EX-SKIP (http://ex-skip.ing.cas.cz/)32 to lead to exon skipping as a result of increased ESS/ESE values. Selected motifs are illustrated.

Table 1 Sequence of PCR products obtained by spanning RT-PCR.

The nucleotide alterations leading to amino acid substitutions K802NC, G876RC or K1037NC define the 3′ end of exons 10, 12 and 15, respectively. Consequently, the observed aberrant RNA splicing may be due to poor 3′donor splice site recognition as predicted by the web-tool Human-Splicing Finder (http://www.umd.be/HSF/)31 (Fig. 5B).

In contrast, the erroneous splicing of the E628VC, K633RC, G1369RC and A1373PC mutated transcripts are most likely due to effects on exonic splicing regulatory elements, such as exonic splicing enhancers (ESE) or exonic splicing silencers (ESS)32. The effects are illustrated in Fig. 5C. Splicing defects caused by exonic missense mutations in ATP7A are part of the mechanism of pathogenesis in MD and have not been described previously. Only a single case, in which an exonic silent mutation was shown to lead to exon skipping, has recently been described33.

Temperature affects accumulation of G860V

Neither reduced amounts, nor abnormal ATP7A mRNA splicing apparently account for the small amount of ATP7A protein accumulated in fibroblasts from patients with the G860VC or G1255RC substitution. Therefore, in vivo instability of these variants may explain the observed lack of protein accumulation. Increased protein accumulation at reduced growth temperatures has previously been described for unstable proteins with a compromised folding capacity. The accumulation of e.g. ATP7A34, ATP7B35 and P-type ATPase ATP8B136 variants has previously been shown to increase in cells growing at 30 °C instead of 37 °C. The amounts of ATP7A protein in lysates from G860VC and G1255RC fibroblasts grown at 30 °C or 37 °C, respectively, were therefore compared by WB (Fig. 6a). We included fibroblasts with ATP7A substitutions C1000RC or A1362VA (Fig. 6b), as growth at the reduced temperature has previously been shown to increase accumulation of C1000R and A1362D34. While decreased amounts of wild-type ATP7A were observed in fibroblasts from a healthy control and for the G1255RC variant, an increased amount was observed for G860VC, (and for the positive controls C1000RC and A1362VA) in fibroblasts grown at 30 °C rather than 37 °C (Fig. 6).

Figure 6
figure6

WB of lysates isolated from MD fibroblast cultures cultured at 30 °C and 37 °C, respectively. The MD fibroblasts contain different missense mutations as indicated. Control + is lysate from normal fibroblasts. Control − is lysate from MD fibroblast, containing a big deletion of the ATP7A gene (exon 3–23 deleted). The blots are representative for Western blots of three separate cultures that were harvested and prepared individually for each mutation and control. The Western blots were performed on separate occasions; therefore band intensity cannot be compared between different gels. (a) Mutations that lead to low levels of ATP7A in this study. (b) Mutations that are similar to or resemble the mutations analysed by Vonk et al.34. Gaps indicate that the gel was cut in producing the figure.

Proteasomal degradation affects accumulation of G860V

To test whether proteasomal degradation is either fully or only partly responsible for the reduced accumulation of G860VC and G1255RC, we treated the fibroblasts with the proteasome inhibitor Bortezomib for 20 hours, and determined the ATP7A protein content by WB. The data in Fig. 7 show that proteasomal degradation is involved in the degradation of G860VC, as proteasomal inhibition increased the accumulation of G860VC protein, whereas no effect was observed for G1255RC. The variant R844HC was also included and - as expected - no protein was observed, thereby supporting the notion that for this mutant, the absence of protein is due to unstable ATP7A transcript.

Figure 7
figure7

WB of lysates from MD fibroblasts cultured in the absence or presence of 25 µM Bortezomib, respectively. The MD fibroblasts contain different missense mutations as indicated. Control + is lysate from normal fibroblasts. Control − is lysate from MD fibroblast, containing a big deletion of the ATP7A gene (exon 3–23 deleted). The protein product of the housekeeping gene Alfa-Tubulin was used as an internal control for the correct loading of 25 μg protein in each lane.

Bypassing RNA splicing, revealed that mutations affecting RNA splicing do not affect the ATP7A activity per se

Missense mutations might cause disease because of aberrant ATP7A transcript splicing, reduced stability of the transcript and/or protein, or reduced Cu-ATPase activity as a result of the amino acid substitution per se. To distinguish between these effects, and to test for a possible correlation between cellular localization and activity of the ATP7A variants in fibroblasts, we used a yeast complementation assay based on a strain lacking the single ATP7A orthologue, CCC2, to determine in vivo activity of all 36 missense mutations. The assay is based on the fact that a ccc2Δ yeast host is unable to thrive under iron-limited conditions, while growth can be rescued by the expression of wild-type ATP7A.

Seven of the studied mutations caused mal-splicing of ATP7A RNA in patient fibroblasts (Fig. 5). By introducing each mutation into the full length ATP7A cDNA sequence, the effect on RNA splicing was bypassed, allowing the investigation of the amino acid substitution per se. Six of the seven mutants, E628VC, K633RC, K802NC, K1037NC, G1369RC and A1373PC were able to complement the high iron requirement of the ccc2Δ yeast strain, comparable to the wild-type, indicating that these amino acid substitutions did not compromise the Cu-ATPase activity in itself (Fig. 8). Of the seven mutants, only G876RC failed to show any in vivo activity.

Figure 8
figure8

Complementation of the ccc2Δ iron requiring phenotype on agar plates. All 36 variants were investigated for their ability to complement the high iron requirement of a ccc2Δ yeast strain by plating cells on agar plates containing the iron chelator Ferrozine. WT, ccc2Δ yeast cells producing wild-type ATP7A; EV, ccc2Δ yeast cells expressing no ATP7A (empty vector). Mutants are spotted according to their amino acid substitution. To control for cell viability, each yeast strain was also spotted on iron-containing agar plates.

Also R844HC and G860VC which resulted in a very small amount of ATP7A transcript and ATP7A protein in human fibroblasts, respectively, were able to complement the high iron requirement of the ccc2Δ yeast strain.

Complementation of the ccc2 Δ yeast phenotype revealed poorer complementation of the TGN- trapped variants as compared to post-TGN mutants

Of the two mutants with copper dependent trafficking, the S761PA substitution was able to complement the ccc2Δ yeast phenotype after 56 hours, whereas only faint complementation was observed for the Q924RO substitution even after 80 hours. Investigation of the four mutations leading to post-TGN localization of the protein showed that substitutions G666RA, A1325VA/O/C and G1015DA were able to complement, whereas G876EA was not. None of the 21 TGN-trapped variants were able to complement the ccc2Δ yeast strain as well as wild-type ATP7A. However, after 80 hours, complementation was observed for G1315RC and A1362VA and weak complementation for S653YA, G1005RC, A1007VA, K1282EA and D1301GC. No complementation was observed for the last 14 TGN-trapped variants: G727RC, G728DA, C1000RA, D1044GC, G1255RC, G1300EC, G1302VC, G1302EC, N1304KC, D1305AC, D1305GC, A1308DC, M1393TC and S1397FC.

Despite normal splicing, the mutation predicting the R844HC substitution showed little accumulation of ATP7A transcript in human fibroblasts, while G860VC and G1255RC led to a reduced protein accumulation. To examine the activity of these variants in more detail, we analyzed their ability to complement the high iron requirement of the ccc2 yeast knock-out strain at twenty different iron concentrations for 92 hours. The data in Fig. 9 show that R844HC and G860VC were able to complement to the same extent as the wild-type, whereas G1255RC failed to show any complementation.

Figure 9
figure9

Complementation of the ccc2Δ iron requiring phenotype in liquid culture. The mutants R844H, G860V and G1255R leading to no, or only little protein accumulation in patient fibroblasts and the frequent mutant G727R were investigated for their ability to complement the high iron requirement of a ccc2Δ yeast strain in liquid cultures containing 1 mM Ferrozine and increasing amounts of Fe2+, as indicated. Wild-type, ccc2Δ yeast cells producing wild-type ATP7A; Empty vector, ccc2Δ yeast cells expressing no ATP7A.

In contrast to the effect observed in fibroblasts, all three variants led to a substantial accumulation of protein in the yeast (Fig. 10), indicating that the disease-causing effect of the mutations predicting the R844HC and G860VC substitutions in humans is due to an unstable transcript and protein respectively, whereas the disease-causing effect of G1255RC is primarily due to the lack of copper-transport activity. The frequent variant G727RC found in about 4.7% of MD patients3 was also included. The observed absence of transport activity combined with a substantial amount of accumulated protein, verifies that the disease-causing effect of G727RC indeed is due to the amino substitution per se. This does not concur with results published by Tang et al.37 who found a significantly reduced amount of G727R protein in patient fibroblasts, but also found that the G727R mutant was able to complement the ccc2 yeast knock-out strain. The reason for this discrepancy is unknown. All lanes, including the lane representing the empty vector (EV, negative control) show some unspecific bands, whereas the band representing ATP7A is absent in the negative control.

Figure 10
figure10

WB of crude lysates from yeast transformed with ATP7A expression plasmids. Crude membrane proteins were isolated from ccc2Δ yeast cells, transformed with plasmids encoding G727R, R844H, G860V, G1255R, WT, or no ATP7A (empty vector, EV). 25 μg crude membrane proteins were analyzed by Western blotting as described in Materials and Methods.

Discussion

We analyzed the effect of 36 ATP7A missense mutations on mRNA splicing, mRNA accumulation, protein accumulation, Cu-ATPase activity, cellular localization and copper-dependent trafficking. The observed effects of the 36 ATP7A missense mutations are summarized in Figs 2 and 11.

Figure 11
figure11

Distribution and effect of ATP7A disease-causing missense mutations. The mutations, presented in this study, are plotted as spheres in a previously established homology model of ATP7A7, based on the crystal structure of the homologous protein LpCopA from Legionella pneumophila 3. The various domains of ATP7A are colored as in Fig. 1. However, metal-binding domains 1 to 4 (H1–H4) are not depicted in the figure, as their structural localization is unknown, and regions with major insertions relative to LpCopA are shown in black. The approximate position of the (TGN- or plasma-) membrane is shown in wheat. The color of spheres represent the mutational effect on ATP7A localization: Copper dependent trafficking of the protein (blue); retention in post-Golgi compartments (grey); retention in the trans-Golgi network (orange); not certain (green); no detectable protein (purple). Underlined missense mutations have been reported to confer a non-classical phenotype. A1325V leads to classical MD in some patients (see Fig. 2).

For eight of the 36 predicted amino acid substitutions - E628VC, K633RC, K802NC, R844HC, G876RC, K1037NC, G1369RC and A1373PC - no ATP7A protein was detected, and RT-PCR analyses revealed that all eight variants, except R844HC, affected splicing of the ATP7A transcript. The reason for the decreased amount of R844HC transcript is puzzling. It is possible that this mutation enhances mRNA degradation by affecting the mRNA structure. The mal-splicing of the transcripts encoding E628VC or K633RC leads to the skipping of exon 8, while the mal-splicing of the transcripts that encodes G876RC leads to the skipping of exon 12. These events result in a frameshift and a premature termination codon (PTC) after 34, or seven aberrant amino acids, respectively. The reduced amount of transcript observed in these cells may therefore be explained by the activation of nonsense-mediated decay (NMD)38. NMD is a translation-coupled quality control system that recognizes and degrades mRNA containing a PTC. The NMD complex is anchored to the exon junction complex (EJC), which is removed from the mRNA by the elongating ribosomes.

Skipping of exon 21 in G1369RC and A1373PC also leads to an out-of-frame transcript, but the PTC is located 25 nt upstream from the last exon, exon 23. The normal amount of transcript observed for these two mutants therefore concur with the notion that only PTC’s located more than 50–55 nt upstream from the last exon mediate NMD38. NMD is most likely not activated for K802NC and K1037FC as the skipping of exons 10 and 15, respectively, preserves the reading frame (Table 1).

As substitutions E628VC, K633RC, K802NC, R844HC, K1037NC, G1369RC and A1373PC were able to complement the ccc2Δ yeast strain, the absence or insufficient amount of correctly spliced full-length transcript is therefore the most likely cause of the classical MD phenotype observed in these 7 patients. Of the eight mutants, only G876RC was unable to complement ccc2Δ, indicating that this substitution also affects the catalytic activity of ATP7A if some residual protein is properly produced; G876 is located in the TGE-loop of the A-domain which is responsible for dephosphorylation39, 40.

The two mutations predicting the substitution G860VC and G1255RC, lead to a decreased accumulation of ATP7A protein in fibroblasts, despite the presence of normal amounts of correctly spliced mRNA, thus pointing to compromised protein folding. Indeed, the more bulky valine and arginine side chains may prevent formation of the central β-sheet of the A-domain, and decrease flexibility in a turn of the P-domain adjacent to the phosphorylation site, respectively. The reduced amount of accumulated G860VC could be rescued at 30 °C and by inhibiting proteasomal degradation, indicating that the reduced amount of G860VC seems to be controlled through proteasomal degradation. This was not the case for G1255RC, as neither reduced temperature nor inhibition of proteasomal degradation rescued the reduced accumulation of this variant.

The intracellular localization of ATP7A was affected in the majority of the investigated fibroblasts. Copper-dependent trafficking was only obtained for Q924RO and S761PA. Q924 is adjacent to the amino-terminus of TMA where the MBDs are likely to interact through charge-complementation. The glutamine-to-arginine substitution may thus compromise the interaction of the MBD-complex with the catalytic core of the protein. S761 is structurally located near D782, which corresponds to E189 that is implicated in copper release in L. pneumophila CopA5. Considering that the two mutations result in a non-classical phenotype and OHS, respectively, and that they display some complementation, these protein variants are quite likely to be partly functional in the patients.

21 of the mutants (S653YA, G727RC, G728DA, C1000RA, G1005RC, A1007VA, D1044GC, G1255RC, K1282EA, G1300EC, D1301GC, G1302VC, G1302EC, N1304KC, D1305AC, D1305GC, A1308DC, G1315RC, A1362VA, M1393TC, and S1397FC) prevented copper-induced trafficking from the TGN to cytoplasmic vesicles or the plasma membrane, whereas only four of the mutants (G666RA, G876EA, G1015DA, A1325VA/O/C), led to permanent post-TGN, cytoplasmic localization.

The D1044E substitution has previously been shown to inhibit trafficking from TGN12, 13 while a mutated 875TGE motif (replaced by AAA) leads to permanent localization of the ATP7A protein to the plasma membrane, regardless of the copper concentration13. These observations together with the fact that D1044 is essential for phosphorylation and 875TGE is required for de-phosphorylation, led to the hypothesis that copper-regulated trafficking of ATP7A from TGN to the cell membrane is associated with the formation of the phosphorylated catalytic intermediate, whereas retro-trafficking from the cell membrane to TGN is associated with dephosphorylation13. Furthermore it has been shown that mutation of the conserved aspartic acid residue in ATP7B blocks redistribution from TGN, whereas mutation of the conserved TGE phosphatase domain traps ATP7B in cytosolic vesicular compartments. Nevertheless, this putative coupling between phosphorylation/dephosphorylation and cellular localization is still being debated41, 42.

We observed that mutants with retained TGN localization, which are typically linked to the classical MD phenotype, and with no or poor ccc2Δ yeast complementation (note that among these, mutants with some capacity to complement represent exceptions with less severe phenotypes), generally cluster within domains that are crucial for phosphorylation (Figs 2 and 11). One group may impair phosphorylation indirectly as the delivery of copper to the essential TM copper-binding site(s) is required for the completion of phosphorylation, whereas other residues compromise phosphorylation more directly. Specifically, substitutions S653YA, G727RC, G728DA, C1000RA, G1005RC, A1007VA, A1362VA, M1393TC and S1397FC target the TM-domain and may affect copper-uptake at the GG- platform (S653YA, G727RC and G728DA, located in TMA and TMB) or copper-transport directly by affecting ion-binding to the intramembranous site(s) (C1000RA, M1393TC, S1397FC and A1362VA)5. The observed cellular location of A1362VA concurs with a previous publication27. The mutants D1044GC, G1255RC, G1300EC, D1301GC, G1302VC, G1302EC, N1304KC, D1305AC, D1305GC, A1308DC and G1315RC all target the P-domain and may thus prevent formation of the acyl-phosphate intermediate with D1044.

In contrast, none of the four substitutions G666RA, G876EA, G1015DA, A1325VA/O/C were detected in TGN under any conditions, although WB confirmed the presence of normal levels of ATP7A protein in the cells. These mutants have less severe effects on patients and are able to complement in ccc2Δ yeast (except in the case of A1325VA/O/C). This may indicate hyper- phosphorylation of the proteins with permanent localization to the cytoplasmic vesicles or plasma membrane. Indeed, the structural localization of these residues supports this notion to some degree. G666 is located in TMA adjacent to the copper exit site formed partly by D782, which corresponds to E189 in L. pneumophila CopA5, and possibly hampers copper-release and de-phosphorylation (Fig. 12a). G876 and A1325 are located near the phosphorylation site (Figs 11 and 12b) and may influence de-phosphorylation negatively; in fact, G876 is, as described above, the second amino acid of the conserved TGE-motif in the A-domain, responsible for de-phosphorylation. Finally, G1015 is located in the cytoplasmic linker between M4 and the P-domain and might abolish de-phosphorylation by preventing the coupling of the TM- and P-domains which is essential for the catalytic cycle. However, we cannot exclude the possibility that G1015DA and A1325A/O/C might obstruct phosphorylation as they are positioned in or close to the P-domain.

Figure 12
figure12

The post-Golgi arrested mutations G666RA and G876EA. The domains of ATP7A are colored as in Fig. 1, and functionally relevant residues shown as sticks using a previously established homology model of ATP7A (6) based on the crystal structure of the homologous protein LpCopA from Legionella pneumophila (2). (A) close-view of G666R at the possible release pathway from the CPC-motif (C1000 and C1002) and via D782 to the non-cytosolic side. The mutation may directly prevent copper passage through sterical hindrance or influence of the local environment at the membrane interface. (b) Close-view of G876E at the catalytic Aspartate (D1044). G876 is located in the TGE-loop of the A-domain which is responsible for dephosphorylation (AlF4 is a phosphate mimic used for structure determination of LpCopA). The mutation is likely to prevent dephosphorylation and turn-over.

It is probably not the phosphorylation status per se that determines the localization of ATP7A, but rather the concomitant conformation (according to the above mentioned E1-E1P-E2P-E2 reaction) with considerably altered three-dimensional structures throughout the reaction cycle (re-visit Fig. 1). The reason why E2 states (associated with de-phosphorylation) may be dominant when ATP7A is in TGN, and E1 conformations (associated with phosphorylation) appear linked to post-TGN localization, could be that distinct conformations have different affinities to trafficking-assisting molecules; this would be highly interesting considering that the catalytic turnover is relatively fast41, whereas trafficking takes much longer, but is in agreement with the novel concept of resting functional states of wild-type P-type ATPase transporters43.

It is noteworthy that all mutations leading to copper dependent trafficking or plasma membrane localization and/or trapping of the protein in cytoplasmic vesicles have been identified in patients with an atypical phenotype or OHS (6/6), indicating residual protein activity. The presence of residual activity of these mutants was supported by complementation of the ccc2Δ yeast strain, which showed that five out of six were able to complement the ccc2Δ yeast strain. In contrast, most of the mutations leading to permanent TGN localization of the protein conferred classical MD (71%; 15/21). Furthermore, only seven out of 21 (33%) TGN-trapped variants were able to complement the ccc2Δ yeast strain, and four of these conferred a mild phenotype. All seven had reduced activity; growth was only visible after many hours of incubation. Altogether, the ccc2Δ yeast complementation results support the notion that development of a mild phenotype is associated with a partially active protein.

In MD, copper accumulates in intestinal cells leading to copper deficiency in the body and subsequently to reduced cuproenzyme activity. Some cuproenzymes are located in the mitochondria (cytocrome c oxidase), others in the cytosol (tyrosinase and Cu/Zn superoxide dismutase) and some are secretory enzymes loaded with copper in TGN (lysyl oxidase, ceruloplasmin, dopamine ß-hydroxylase)44. One explanation for the milder phenotype of patients with an ATP7A variant located at the plasma-membrane or trapped in cytoplasmic vesicles, as opposed to a TGN-located variant, could be that the plasma membrane/vesicles located variants are able to transport some copper from the intestine to the blood stream, thereby enabling some copper-loading of cuproenzymes located in the mitochondria or the cytosol. The reason why e.g. OHS patients with small amounts of normal protein that stem from splice site mutations, have pronounced connective tissue defects, due to the absence of active lysyl oxidase, is puzzling. As proposed by Mercer45, it could also be explained by the ability of the small amount of ATP7A to export copper to the blood stream, whereas transport of copper to TGN might be hampered. MD fibroblasts accumulate copper, and the small amount of wild-type protein might be occupied at the plasma membrane to excrete copper, due to high cellular copper.

Whereas the cellular localization might predict the severity of the disease, it is unlikely that it correlates with the effect of copper treatment. Early treatment rather than the exact character of the mutation seems to be the most important factor for the outcome. A significant effect of early copper treatment of patients even with predicted severe mutations has been demonstrated in several publications46,47,48. The treatment of three patients with the frequent G727R variant led to a near normal outcome in two of the three patients who were treated within 25 days after birth, in contrast to the poor outcome in the third patient who began treatment at 7.6 months48. Similarly, 35 MD patients with different mutations, who were treated before the age of 1 month led to a significant improvement and a higher survival rate in 25 of the 35 patients (71.2%) after the age of 3 years. Poorer improvement and a lower survival rate of only 50% after the age of 3 years48 was observed in a group of 22 patients whose treatment was initiated later than 1 month after birth. The 35 patients who were treated early had a mixture of some expected mild and some severe mutations (e.g Q724H, del Ex1, del Ex2–14, del Ex20-23, G666R, G728D, 2233delT, Q724X and A629P). Similar cases of mixed mutations were observed in the 22 patients who were treated later (e.g Q724H, del Ex1, del Ex13-14, K1034N and Q1383X)48.

In summary, most substitutions that lead to TGN-trapping have no or low catalytic activity, and predispose to a severe classical MD phenotype, whereas most variants that lead to copper dependent trafficking or to cytoplasmic localization, have residual catalytic activity and render a milder phenotype. The mechanism behind the trafficking is very complex. We found that several ATP7A substitutions assumed to affect the catalytic phosphorylation and located in two different domains, led to permanent TGN localization, whereas substitutions that may influence dephosphorylation impair TGN-retrieval. These results support previous studies indicating that phosphorylation is crucial for the exit of ATP7A from TGN, while dephosphorylation is crucial for recycling from the cytoplasmic compartments back to TGN13. In contrast to the previously published paper that studied cells transfected with plasmids encoding ATP7A variants13 we used a setup based on a large number of endogenously expressed ATP7A variants in fibroblasts from MD patients. The present results verify and add to previous suggestions.

Furthermore, in contrast to other studies, we also investigated the effect of missense mutations on the ATP7A transcript and demonstrate that a substantial number of predicted missense mutations found in MD patients prohibit correct exon-intron splicing, while the amino acid substitution per se does not abolish Cu-ATPase activity. These are new observations which contribute significantly to our understanding of MD.

Methods

Patients

The 36 patients were referred to the Kennedy Center for diagnostic confirmation of Menkes disease. We divided the patients into three sub-classes based on phenotype and progression: Classical Menkes disease (C) for patients with severe symptoms and death before 5 years of age; Atypical MD for patients with milder symptoms and longer survival (more than 5 years). The Occipital Horn Syndrome (OHS) for patients with mainly connective tissue manifestations. We have previously29 described the patient with the K1283E mutation as classic. However, we now know that the patient survived for at least 15 years and have therefore re-assigned this case to Atypical MD.

Fibroblast culturing

Fibroblast cell cultures obtained from patient-skin biopsies (and one control) were cultured as described previously49, 50.

Proteasomal inhibition

Fibroblast cell cultures were grown in the presence of 20 µM Bortezombib (added to the normal medium) for 20 hours before harvesting.

Copper-uptake (64Cu) and –retention

The copper-uptake (64Cu) and retention profile of the fibroblast cultures were performed as described previously50.

Immunofluorescence

Immunofluorescence (IF) was performed as described previously29. In brief, fibroblasts grown on glass coverslips were incubated in pure F-10 with 50 µM Bathocuproine disulfonic acid (BCS; Sigma; www.sigmaaldrich.com) for 2 hours. The cells were subsequently fixed in freshly prepared 4% paraformaldehyde in PBS for 20 min at 4 °C, washed and permeabilized with 0.2% Triton-X100 in PBS. Nonspecific signals were blocked with 3% BSA in PBS with 0.2% Triton and subsequently incubated with chicken anti-ATP7A antibodies against amino acids 1407–1500 (ab 13995, Abcam, www.abcam.com). In parallel, the cells were incubated with mouse antibodies against the 28-kDa SNARE protein (GS28) located in the Golgi compartment (BD transduction laboratory, www.bdbiosciences.com). Alexa 488 conjugated goat anti-chicken secondary and tertiary Alexa 488 rabbit anti-goat antibodies were used to detect the ATP7A protein, whereas Alexa 546 conjugated donkey anti-mouse secondary antibodies were used to detect GS28. The localization of ATP7A was analyzed with confocal microscopy (Olympus 1000 Fluoview) and the degree of co-localization was analyzed by Coloc 2 (Fiji/http://imagej.net/ImageJ).

Western blotting

Cultured patient fibroblasts were harvested by trypsination and lysed in lysis-buffer (50 mM Hepes pH 7.6; 250 mM NaCl; 0.1% NP40; 5 mM EDTA) for 30 minutes on ice. The protein concentration in the lysates was determined using the BCA protein assay kit (Pierce, Rockford, IL, USA). A sample corresponding to 25 µg protein was mixed with loading-buffer and DTT, and loaded on 12-well RunBlue SDS gels (Expedeon, www.expedeon.com/). SDS-PAGE was carried out at 130 V and the Western blot was performed according to the manufacturer's protocol (Expedeon, www.expedeon.com/). After blotting, the membranes were blocked for one hour in blocking-buffer (ATP7A: 25 mM Tris-buffer (pH 7.4); 5% skimmed milk; 0.5% Tween20. GAPDH: 25 mM Tris-buffer (pH 7.4); 5% skimmed milk; 0.1% Tween20; 0.25 M NaCl). The membranes were cut in two so that the upper part contained the ATP7A protein (168 KD) and the lower part contained the GAPDH protein (38 KD). Thereafter, the membranes were incubated with primary antibody diluted in incubation buffer (25 mM Tris-buffer (pH 7.4); 5% skimmed milk; 0.1% Tween20) overnight at 4 °C. The primary antibodies were chicken polyclonal anti-ATP7A (ab13995, abcam; 1:1,000) and rabbit polyclonal anti-GAPDH (NB300-327, Novus biological; 1:5000), respectively. The next day the membranes were washed for 8 × 5 minutes in washing buffer (10 mM Tris-buffered saline (50 mM Tris, 150 mM NaCl, pH 7.6), 0.1% Tween20). Secondary antibody diluted 1:5,000 in washing buffer was then added to the membranes which were left to incubate for one hour at room temperature on a shaking stand. The secondary antibodies used were donkey anti-chicken (HRP) (DAKO, www.dako.com/dk/) and swine anti-rabbit (HRP) (DAKO, www.dako.com/dk/) to target the primary antibodies against ATP7A and GADPH, respectively. The membranes were washed again for 8 × 5 minutes in washing buffer. The HRP signal was detected using the Super Signal West Dura kit according to the manufacturer’s protocol (Pierce; Rockford, IL, USA) and visualized by autoradiography using Amersham Hyperfilm ECL (Amersham, www.gelifesciences.com).

For Western blotting of yeast-produced ATP7A, cells were inoculated at 30 °C to OD450 = 0.05 in minimal medium with galactose as the sole carbon source. Cells were harvested at OD450 = 1.0 and crude membranes were prepared as described previously51. SDS-PAGE and Western blotting using the same polyclonal anti-ATP7A antibody and HRP conjugated donkey anti-chicken antibody used above were performed as described52 and visualized using the chemiluminescent Immobilon kit from Millipore and the LAS4500 imager from GE Healthcare.

Quantitative Real-time PCR

Real-time PCR was performed on cDNA. A Taq-Man 6-carboxy-fluorescein (FAM) labeled probe and primer pairs against the boundary between exon 1 and exon 2 (part number Hs00921963_m1) or against the boundary between exon 22 and exon 23 (part number Hs00921966_m1) in ATP7A cDNA were used to detect the total amount of ATP7A transcript. A FAM labeled probe and primers for the human GAPDH transcript (part number 4352934E) were used as an endogenous control. Relative quantification of GAPDH transcript was carried out on parallel samples. The cDNA samples obtained from the RNA preparations were assayed in triplicate, in 100 ng/sample, in a total volume of 25 μl. All probes were purchased from Applied Biosystems. PCR amplification and detection were performed with an ABI7500 (Applied Biosystems, www.invitrogen.com) in accordance with the manufacturer’s instructions. The threshold cycle (CT) is defined as the fractional cycle number at which the fluorescence passes a fixed threshold. Standard curves for CT values compared with log cDNA concentration were prepared by assaying five-fold serial dilutions of control cDNA: from 100 ng/sample to 0.16 ng/sample, with the GAPDH and ATP7A probes, respectively. In all experiments, the amount of ATP7A and GAPDH mRNA were calculated by linear regression of the lines generated by the standard curves, log cDNA concentration against CT. The amount was measured as the relative amount compared to a control sample. The normalized ATP7A N value is calculated by dividing the ATP7A mRNA value with the GAPDH value (SDs are shown).

Investigation of ATP7A transcript by RT-PCR

Total RNA from approximately 5 × 106 cultured skin fibroblasts was isolated with the RNAeasy kit (Qiagen, www.qiagen.com), and single-stranded cDNA was synthesized with the High-Capacity cDNA Archive Kit in accordance with the manufacturer's instructions (Applied Biosystems, www.invitrogen.com). PCR amplification of fragments spanning the cDNA fragments of interest - including the affected and flanking exons - was performed with sequence-specific primers as noted in the legends to Fig. 5. PCR products were subsequently separated and visualized upon agarose-gel electrophoresis in the presence of Ethidium Bromide, excised from the gel and purified with the QIAquick gel extraction kit (Qiagen, www.qiagen.com) and sequenced using the PCR amplification primers. Sequencing was performed using an ABI3100 sequencer (Applied Biosystems, www.invitrogen.com). The sequence of used primers is shown in Supplementary Table 2.

Yeast expression plasmids and complementation test

Site-directed mutagenesis of human Menkes cDNA was performed according to ref. 53. Menkes expression plasmids were constructed by homologous recombination by transformation of HindIII, SalI, BamHI digested pEMBLyex454 and wild-type or mutated human Menkes cDNA into strain PAP6064 (mat a his3 Δ1::UPR-lacZ HIS3 leu20 met15 Δ0 ura3v Δ0 ccc2::kanMx4) as described previously55. The nucleotide sequences of all constructs were confirmed by DNA sequencing. Complementation tests were performed by spotting 5 µl of exponentially growing cells at OD450 = 0.5 onto SG (minimal medium with galactose) plates with 1 mM Ferrozine, 1 µM CuCl2 and 135 µM Fe(NH4)2SO4. To confirm viability, 5 µl of each yeast culture was also spotted on non-selective medium (SD minimal medium). The plates were incubated at 30 °C and inspected daily. For determining complementation in liquid culture, cells were grown on micro plates in media containing 1 mM Ferrozine, 1 µM CuCl2 and Fe2+ concentrations ranging from 0–105 µM (in the form of Fe(NH4)2SO4) as indicated in Fig. 9.

Study approval

All experiments were carried out in accordance with relevant guidelines and regulations. The cell lines and patient information were previously published5, 29.

References

  1. 1.

    Tumer, Z. & Moller, L. B. Menkes disease. Eur. J. Hum. Genet. 18, 511–518 (2010).

  2. 2.

    Menkes, J. H., Alter, M. & Steigleder, G. K. et al. A sex-linked recessive disorder with retardation of growth, peculiar hair, and focal cerebral and cerebellar degeneration. Pediatrics 29, 764–779 (1962).

  3. 3.

    Lenartowicz, M., Krzeptowski, W., Koteja, P., Chrzascik, K. & Møller, L. B. Prenatal Treatment of Mosaic Mice (Atp7a mo-ms) Mouse Model for Menkes Disease, with Copper Combined by Dimethyldithiocarbamate (DMDTC). PLoS One. 7, e40400 (2012).

  4. 4.

    Kennerson, M. L. et al. Missense mutations in the copper transporter gene ATP7A cause X-linked distal hereditary motor neuropathy. Am J Hum Genet 86, 343–52, doi:10.1016/j.ajhg.2010.01.027 (2010).

  5. 5.

    Gourdon, P., Liu, X. Y., Skjorringe, T., Morth, J. P., Moller, L. B., Pedersen, B. P. & Nissen, P. Crystal structure of a copper-transporting PIB-type ATPase. Nature 475, 59–64 (2011).

  6. 6.

    Mattle, D., Sitsel, O., Autzen, H. E., Meloni, G., Gourdon, P. & Nissen, P. On allosteric modulation of P-type Cu(+)-ATPases. J. Mol. Biol. 425, 2299–2308 (2013).

  7. 7.

    Gonzalez-Guerrero, M., Eren, E., Rawat, S., Stemmler, T. L. & Arguello, J. M. Structure of the two transmembrane Cu+ transport sites of the Cu+-ATPases. J. Biol. Chem. 283, 29753–29759 (2008).

  8. 8.

    Gourdon, P., Sitsel, O., Lykkegaard, K. J., Birk, M. L. & Nissen, P. Structural models of the human copper P-type ATPases ATP7A and ATP7B. Biol. Chem 393, 205–216 (2012).

  9. 9.

    Larin, D., Mekios, C., Das, K., Ross, B., Yang, A. S. & Gilliam, T. C. Characterization of the interaction between the Wilson and Menkes disease proteins and the cytoplasmic copper chaperone, HAH1p. J. Biol. Chem. 274, 28497–28504 (1999).

  10. 10.

    Arguello, J. M. Identification of ion-selectivity determinants in heavy-metal transport P1B-type ATPases. J. Membr. Biol. 195, 93–108 (2003).

  11. 11.

    Mandal, A. K., Yang, Y., Kertesz, T. M. & Arguello, J. M. Identification of the transmembrane metal binding site in Cu+-transporting PIB-type ATPases. J. Biol. Chem. 279, 54802–54807 (2004).

  12. 12.

    Huster, D. et al. Diverse functional properties of Wilson disease ATP7B variants. Gastroenterology 142, 947–956 (2012).

  13. 13.

    Petris, M. J. et al. Copper-regulated trafficking of the Menkes disease copper ATPase is associated with formation of a phosphorylated catalytic intermediate. J. Biol. Chem 277, 46736–46742 (2002).

  14. 14.

    Voskoboinik, I., Mar, J., Strausak, D. & Camakaris, J. The regulation of catalytic activity of the menkes copper-translocating P-type ATPase. Role of high affinity copper-binding sites. J. Biol. Chem 276, 28620–28627 (2001).

  15. 15.

    Arguello, J. M., Eren, E. & Gonzalez-Guerrero, M. The structure and function of heavy metal transport P1B-ATPases. Biometals 20, 233–248 (2007).

  16. 16.

    Lutsenko, S., Tsivkovskii, R. & Walker, J. M. Functional properties of the human copper-transporting ATPase ATP7B (the Wilson's disease protein) and regulation by metallochaperone Atox1. Ann. N. Y. Acad. Sci 986, 204–211 (2003).

  17. 17.

    Moller, J. V., Juul, B. & Le, M. M. Structural organization, ion transport, and energy transduction of P-type ATPases. Biochim. Biophys. Acta 1286, 1–51 (1996).

  18. 18.

    Paulsen, M., Lund, C., Akram, Z., Winther, J. R., Horn, N. & Moller, L. B. Evidence that translation reinitiation leads to a partially functional Menkes protein containing two copper-binding sites. Am. J. Hum. Genet 79, 214–229 (2006).

  19. 19.

    Petris, M. J., Mercer, J. F., Culvenor, J. G., Lockhart, P., Gleeson, P. A. & Camakaris, J. Ligand-regulated transport of the Menkes copper P-type ATPase efflux pump from the Golgi apparatus to the plasma membrane: a novel mechanism of regulated trafficking. EMBO J. 15, 6084–6095 (1996).

  20. 20.

    Camakaris, J. et al. Gene amplification of the Menkes (MNK; ATP7A) P-type ATPase gene of CHO cells is associated with copper resistance and enhanced copper efflux. Hum. Mol. Genet. 4, 2117–2123 (1995).

  21. 21.

    Francis, M. J., Jones, E. E., Levy, E. R., Ponnambalam, S., Chelly, J. & Monaco, A. P. A Golgi localization signal identified in the Menkes recombinant protein. Hum Mol Genet 7, 1245–52 (1998).

  22. 22.

    Strausak, D., La Fontaine, S., Hill, J., Firth, S. D., Lockhart, P. J. & Mercer, J. F. The role of GMXCXXC metal binding sites in the copper-induced redistribution of the Menkes protein. J Biol Chem 274, 11170–11177 (1999).

  23. 23.

    Zhu, S., Shanbhag, V., Hodgkinson, V. L. & Petris, M. J. Multiple di-leucines in the ATP7A copper transporter are required for retrograde trafficking to the trans-Golgi network. Metallomics 8, 993–1001, doi:10.1039/c6mt00093b (2016).

  24. 24.

    Yi, L. & Kaler, S. ATP7A trafficking and mechanisms underlying the distal motor neuropathy induced by mutations in ATP7A. Ann N Y Acad Sci 1314, 49–54, doi:10.1111/nyas.12427 (2014).

  25. 25.

    Veldhuis, N. A. et al. Phosphorylation regulates copper-responsive trafficking of the Menkes copper transporting P-type ATPase. Int J Biochem Cell Biol 41, 2403–12, doi:10.1016/j.biocel.2009.06.008 (2009).

  26. 26.

    La Fontaine, S., Firth, S. D., Lockhart, P. J., Brooks, H., Camakaris, J. & Mercer, J. F. Intracellular localization and loss of copper responsiveness of Mnk, the murine homologue of the Menkes protein, in cells from blotchy (Mo blo) and brindled (Mo br) mouse mutants. Hum Mol Genet 8, 1069–75 (1999).

  27. 27.

    Ambrosini, L. & Mercer, J. F. Defective copper-induced trafficking and localization of the Menkes protein in patients with mild and copper-treated classical Menkes disease. Hum Mol Genet 8, 1547–55 (1999).

  28. 28.

    Lenartowicz, M., Grzmil, P., Shoukier, M., Starzyński, R., Marciniak, M. & Lipiński, P. Mutation in the CPC motif-containing 6th transmembrane domain affects intracellular localization, trafficking and copper transport efficiency of ATP7A protein in mosaic mutant mice–an animal model of Menkes disease. Metallomics 4, 197–204, doi:10.1039/c1mt00134e (2012).

  29. 29.

    Moller, L. B. et al. Identification and analysis of 21 novel disease-causing amino acid substitutions in the conserved part of ATP7A. Hum. Mutat. 26, 84–93 (2005).

  30. 30.

    Qi, M. & Byers, P. H. Constitutive skipping of alternatively spliced exon 10 in the ATP7A gene abolishes Golgi localization of the menkes protein and produces the occipital horn syndrome. Hum. Mol. Genet. 7, 465–469 (1998).

  31. 31.

    Desmet, F. O., Hamroun, D., Lalande, M., Collod-Beroud, G., Claustres, M. & Beroud, C. Human Splicing Finder: an online bioinformatics tool to predict splicing signals. Nucleic Acids Res 37, e67 (2009).

  32. 32.

    Mount, S. M. A catalogue of splice junction sequences. Nucleic Acids Res 10, 459–472 (1982). Raponi, M. et al. Prediction of single-nucleotide substitutions that result in exon skipping: identification of a splicing silencer in BRCA1 exon 6. Hum. Mutat. 32 436–444 (2011).

  33. 33.

    Møller, L. B. et al. A silent nucleotide substututionin the ATP7A gene in a child with Menkes disease. Mol. Genet. and Metab 110, 490–492 (2013).

  34. 34.

    Vonk, W. I. et al. The copper-transporting capacity of ATP7A mutants associated with Menkes disease is ameliorated by COMMD1 as a result of improved protein expression. Cell Mol. Life Sci. 69, 149–163 (2012).

  35. 35.

    van den Berghe, P. V. et al. Reduced expression of ATP7B affected by Wilson disease-causing mutations is rescued by pharmacological folding chaperones 4-phenylbutyrate and curcumin. Hepatology 50, 1783–1795 (2009).

  36. 36.

    van der Velden, L. M. et al. Folding defects in P-type ATP 8B1 associated with hereditary cholestasis are ameliorated by 4-phenylbutyrate. Hepatology 51, 286–296 (2010).

  37. 37.

    Tang, J., Donsante, A., Desai, V., Patronas, N. & Kaler, S. G. Clinical outcomes in Menkes disease patients with a copper-responsive ATP7A mutation, G727R. Mol Genet Metab 95, 174–81, doi:10.1016/j.ymgme.2008.06.015 (2008).

  38. 38.

    Kurosaki, T. & Maquat, L. E. Nonsense-mediated mRNA decay in humans at a glance. J Cell Sci 129, 461–467 (2016).

  39. 39.

    Lewis, D., Pilankatta, R., Inesi, G., Bartolommei, G., Moncelli, M. R. & Tadini-Buoninsegni, F. Distinctive features of catalytic and transport mechanisms in mammalian sarco-endoplasmic reticulum Ca2+ ATPase (SERCA) and Cu+(ATP7A/B) ATPases. J. Biol. Chem. 287, 32717–32727 (2012).

  40. 40.

    Clausen, J. D., Vilsen, B., McIntosh, D. B., Einholm, A. P. & Andersen, J. P. Glutamate-183 in the conserved TGES motif of domain A of sarcoplasmic reticulum Ca2+-ATPase assists in catalysis of E2/E2P partial reactions. Proc. Natl. Acad. Sci. USA 101, 2776–2781 (2004).

  41. 41.

    Usta, J., Barakeh, H., Mahfouz, H. & Cortas, N. Copper-stimulated adenosine triphosphatase from rat liver. Isolation and kinetic characterization. Ann. N. Y. Acad. Sci 834, 475–478 (1997).

  42. 42.

    Pase, L., Voskoboinik, I., Greenough, M. & Camakaris, J. Copper stimulates trafficking of a distinct pool of the Menkes copper ATPase (ATP7A) to the plasma membrane and diverts it into a rapid recycling pool. Biochem. J. 378, 1031–1037 (2004).

  43. 43.

    Veshaguri, S. et al. Direct observation of proton pumping by a eukaryotic P-type ATPase. Science 35, 1469–73 (2016).

  44. 44.

    Kodama, H., Fujisawa, C. & Bhadhprasit, W. Inherited copper transport disorders:Biochemical mechanisms, diagnosis, and treatment. Current Drug Metabolism 13, 237–250 (2012).

  45. 45.

    Mercer, J. F. B. The molecular basis of copper-transport diseases. Trends in Molecular Medicine 7, 64–69 (2001).

  46. 46.

    Christodoulou, J. et al. Early treatment of Menkes disease with parenteral copper-histidine: long-term follow-up of four treated patients. Am J Med Genet 76, 154–64 (1998).

  47. 47.

    Tümer, Z., Birk Møller, L. & Horn, N. Screening of 383 unrelated patients affected with Menkes disease and finding of 57 gross deletions in ATP7A. Hum Mutat. 22, 457–64 (2003).

  48. 48.

    Kaler, S. G. Neurodevelopment and brain growth in classic Menkes disease is influenced by age and symptomatology at initiation of copper treatment. J Trace Elem Med Biol 28, 427–30, doi:10.1016/j.jtemb.2014.08.008 (2014).

  49. 49.

    Skjorringe, T., Tumer, Z. & Moller, L. B. Splice site mutations in the ATP7A gene. PLoS. One. 6, e18599 (2011).

  50. 50.

    Moller, L. B. et al. Diagnosis of copper transport disorders. Curr. Protoc. Hum. Genet. Chapter 17, Unit17 (2011).

  51. 51.

    Jørgensen, J. R. & Pedersen, P. A. Role of phylogenetically conserved amino acids in folding of Na,K-ATPase. Biochemistry 40, 7301–8 (2001).

  52. 52.

    Pedersen, P. A., Rasmussen, J. H. & Jørgensen, P. L. Expression in high yield of pig alpha 1 beta 1 Na,K-ATPase and inactive mutants D369N and D807N in Saccharomyces cerevisiae. J Biol Chem 271, 2514–22 (1996).

  53. 53.

    Ho, S. N., Hunt, H. D., Horton, R. M., Pullen, J. K. & Pease, L. R. Site-directed mutagenesis by overlap extension using the polymerase chain reaction. Gene 77, 51–59 (1989).

  54. 54.

    Cesareni, G. & Murray, J. Plasmid vectors carrying the replication origin of filamentous single-stranded phages. In Setlow J. K., Hollaender, A. editors Genetic engineering, Principles and methods. New York: Plenum Press. Pp 135–149 (1987).

  55. 55.

    Bomholt, J., Helix-Nielsen, C., Scharff-Poulsen, P. & Pedersen, P. A. Recombinant production of human Aquaporin-1 to an exceptional high membrane density in Saccharomyces cerevisiae. PLoS. One. 8, e56431 (2013).

Download references

Acknowledgements

We thank Jette Rune Rasmussen for assistance with generating the figures, Susan Peters for proof-reading the manuscript and David Sorensen for excellent technical assistance. The research was funded by the Lundbeck Foundation.

Author information

Affiliations

Authors

Contributions

T.S., S.S.T., L.B.M. and P.A.P. performed the Western Blots. L.B.M. performed the Real-time P.C.R. and the biochemical diagnosis. T.S. and L.B.M. performed the investigation of transcript by RT-PCR and sequencing. T.S. and L.B.M. performed the IF. P.A.P. conducted the yeast expression plasmids and complementation tests. P.G. created figures 1, 11 and 12. T.S., L.B.M., P.A.P., P.G. and P.N. wrote the manuscript. L.B.M. initiated the project with the help of all the above mentioned.

Corresponding author

Correspondence to Lisbeth Birk Møller.

Ethics declarations

Competing Interests

The authors declare that they have no competing interests.

Additional information

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Skjørringe, T., Amstrup Pedersen, P., Salling Thorborg, S. et al. Characterization of ATP7A missense mutants suggests a correlation between intracellular trafficking and severity of Menkes disease. Sci Rep 7, 757 (2017). https://doi.org/10.1038/s41598-017-00618-6

Download citation

Further reading

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.