Evidence that translation reinitiation leads to a partially functional Menkes protein containing two copper-binding sites

Abstract

Menkes disease (MD) is an X-linked recessive disorder of copper metabolism. It is caused by mutations in the ATP7A gene encoding a copper-translocating P-type ATPase, which contains six N-terminal copper-binding sites (CBS1-CBS6). Most patients die in early childhood. We investigated the functional effect of a large frameshift deletion in ATP7A (including exons 3 and 4) identified in a patient with MD with unexpectedly mild symptoms and long survival. The mutated transcript, ATP7A(Delta ex3+ex4), contains a premature termination codon after 46 codons. Although such transcripts are generally degraded by nonsense-mediated mRNA decay (NMD), it was established by real-time PCR quantification that the ATP7A(Delta ex3+ex4) transcript was protected from degradation. A combination of in vitro translation, recombinant expression, and immunocytochemical analysis provided evidence that the ATP7A(Delta ex3+ex4) transcript was protected from degradation because of reinitiation of protein translation. Our findings suggest that reinitiation takes place at two downstream internal codons. The putative N-terminally truncated proteins contain only CBS5 and CBS6. Cellular localization and copper-dependent trafficking of the major part of endogenous and recombinant ATP7A(Delta ex3+ex4) proteins were similar to the wild-type ATP7A protein. Furthermore, the ATP7A(Delta ex3+ex4) cDNA was able to rescue a yeast strain lacking the homologous gene, CCC2. In summary, we propose that reinitiation of the NMD-resistant ATP7A(Delta ex3+ex4) transcript leads to the synthesis of N-terminally truncated and at-least-partially functional Menkes proteins missing CBS1-CBS4. This finding--that a mutation that would have been assumed to be null is not--highlights the need to examine the biochemical phenotype of patients to deduce the efficacy of copper therapy.

Figure 1.

Characterization of patient ATP7A mRNA. PCR amplification of exons 1–6 was performed on cDNA obtained from RNA isolated from fibroblasts from patient A.B. (P) and from an unaffected control person (C). The PCR products were separated on a 1% agarose gel and were visualized by ethidium bromide staining. M, Sizes (in bp) of the marker DNA fragments.

Figure 2.

Expected cDNA sequence of ATP7A^Δex3+ex4. Lowercase letters indicate part of the 5′ UTR, and uppercase letters indicate exons 1, 2, and 5. The predicted normal amino acid sequence (one letter) is indicated below the DNA sequence. Italic letters above the DNA sequence indicate predicted amino acids as a consequence of frameshifting. The vertical lines indicate the exon 1/2 and exon 3/5 junctions. The normal translation-initiation ATG codon in exon 2 is underlined and highlighted. Alternative ATG codons in exon 2 are underlined, and the ATG461 and ATG475 codons in exon 5 are underlined and indicated by number. The position of the T7-containing forward primer T7c.-134-(-111) is indicated. The numbering is in accordance with the normal ATG1 start codon in exon 2.

Figure 3.

A coupled in vitro transcription/translation of the ATP7A^Δex3+ex4 transcript indicates reinitiation of translation at the two internal codons AUG461 and AUG475. On A.B. (P) and control (C) cDNA, different primer pairs were used to PCR-amplify sequences encoding N-terminal fragments. The T7-containing PCR products were transcribed and translated in vitro, in the presence of ³⁵S methionine, and were analyzed by SDS-PAGE. The used primer pairs were T7-M1/cATP7A642 (lanes 2 and 3), T7-M461/cATP7A642 (lanes 4 and 5), T7-kozak-M461/cATP7A642 (lanes 6 and 7), and T7-kozak-M475/cATP7A642 (lanes 8 and 9). Rainbow ¹⁴C-methylated proteins (Amersham) were used as molecular-weight markers (M).

Figure 4.

No ATP7A^Δex3+ex4 protein could be identified by western blotting analysis in cell extract from A.B. fibroblasts. Proteins in total cell lysates from fibroblasts were separated by SDS-PAGE, followed by immunoblotting with antibody to the ATP7A protein. The results of lysates obtained from positive control cells (C), negative control cells (c.121-?-8333+?del) (N), and A.B. cells (P) are shown. Kaleidoscope Prestained standard was used as a molecular marker (Biorad).

Figure 5.

The endogenously and recombinantly synthesized ATP7AΔ^ex3+ex4 protein shows a copper-dependent trafficking. Fibroblasts and transiently transfected CHO cells were cultured in the presence of CuCl₂ or BCS, as indicated, to test the effect of the deletion of exons 3 and 4 on the copper-dependent trafficking. Fibroblasts from A.B. (P), positive control fibroblasts from unaffected persons (C), and negative control fibroblasts from a patient with MD with deletion of exons 3–23 (c.121-?-8333+?del) (N) were analyzed. The analyzed CHO cells were transfected with pCEP4ATP7A^Δex3+ex4 and with different constructs optimized to express the truncated ATP7A proteins initiated at p.M461 (pCEP4ATP7A^Δ1−460 and pCEP4ATP7A^{Δ1−460,M475L}) and at p.M475 (pCEP4ATP7A^Δ1−474). CHO cells transfected with pCEP4ATP7A^wt and empty vector (EV) pCEP4 were used as positive and negative controls, respectively. Fixed cells were subjected to double-IF staining for the ATP7A protein (green) and the TGN marker GS28 (red). The combined images are shown (yellow). Images were viewed using a 100× oil objective.

Figure 6.

Quantification of the effect of CBS1–CBS4 on cellular localization by IF microscopy. A, Percentage of positive cells with ATP7A signal concentrated in TGN, compared with positive cells with more diffuse staining. The cells were treated with CuCl₂ or BCS, as indicated. The total number of positive cells is defined as 100%. A.B. fibroblasts (P), positive control fibroblasts (C), and negative control fibroblasts (N), in addition to CHO cells transfected with pCEP4ATP7A^wt, pCEP4ATP7A^Δex3+ex4, pCEP4ATP7A^Δ1−460, pCEP4ATP7A^Δ1−474, pCEP4ATP7A^{Δ1−460,M475L}, and EV pCEP4 were investigated. More than 100 cells of each type were analyzed. B, Percentages of cells with a positive ATP7A signal, a less-positive signal, and no signal were determined. A.B. fibroblasts (P) and positive control fibroblasts (C) treated with BCS were investigated. More than 100 cells of each type were analyzed.

Figure 7.

Schematic diagram of the ATP7A constructs. Exons 1–6 of ATP7A are depicted as separate small boxes. Exons 7–23 are depicted as a big black box. The predicted N-terminal amino acids are indicated. All the constructs encode an inframe C-terminal Myc tag.

Figure 8.

Western blotting shows that a protein of ∼120 kDa is synthesized from pCEP4ATP7A^Δex3+ex4, which is consistent with reinitation at ATG461 and ATG475. Cell lysates from CHO cells transiently transfected with pCEP4ATP7A^Δex3+ex4, pCEP4ATP7A^Δ1−460, pCEP4ATP7A^Δ1−474, and pCEP4ATP7A^{Δ1−460,M475L} were subjected to SDS-PAGE and immunoblotting, with the use of antibody to the Myc epitope tag. CHO cells transfected with pCEP4ATP7A^wt and EV pCEP4 were used as positive and negative controls, respectively. Aliquots of 3–15 μl total cell lysate (∼6–30 μg total protein) were loaded to obtain approximately equal signal intensity in all lanes.

Figure 9.

N-terminally truncated Menkes proteins complement a ccc2 yeast strain by restoring high-affinity iron uptake. The Δccc2 strain 8 was transformed with EV (pRS316GPD), pRS316GPDATP7A^wt, pRS316GPDATP7A^Δex3+ex4, pRS316GPDATP7A^Δ1−460, pRS316GPDATP7A^Δ1−474, and pRS316GPDATP7A^{Δ1−460,M475L}. The wild-type strain 7 was transformed with EV. The transformed colonies were cultured as described in the “Material and Methods” section. At optical density OD₆₀₀=0.1, an aliquot of 5 μl was streaked on plates with iron-limited medium (A and D). Rescue of the Δccc2 yeast can also be accomplished by addition of either a high concentration of copper (B and E) or iron (C and F) to the iron-limited medium.^, Plates were photographed after 4 d at 30°C or 37°C, as indicated. Schematic illustration of the plates is shown.

Figure 10.

Complementation efficiency of N-terminal–truncated Menkes proteins synthesized in Δccc2 yeast. The growth in iron-limited medium of wild-type strain 7 harboring pRS316GPD (EV) (n=5) and of Δccc2 strain 8 harboring pRS316GPDATP7A^wt (n=4), pRS316GPDATP7A^Δex3+ex4 (n=3), pRS316GPDATP7A^Δ1−460 (n=3), pRS316GPDATP7A^Δ1−474 (n=3), and pRS316GPDATP7A^{Δ1−460,M475L} (n=3) was determined by OD₆₀₀ measurements. OD₆₀₀ (log scale) was plotted in comparison with time, and, in accordance with the equation describing exponential growth (Y=a^*b^x), the doubling time was calculated as log2/logb. SDs are shown.