ATP7A and ATB7B are copper ATPases of the P-type 1B family. Mutations in the genes that encode these copper transporters are responsible for Menkes disease (MD) and Wilson disease (WD), respectively. MD is a severe X-linked copper deficiency disorder which typically leads to death before the age of 3 years. The copper deficiency is caused by a defective ATP7A-mediated transport of copper from the intestine into the portal circulation. As a consequence, the clinical features of MD are a malfunction of a large number of copper-requiring enzymes. WD is an autosomal copper toxicity disorder. Typically, copper accumulates in the liver as a result of a disrupted ATP7B-driven copper export from the hepatocytes. Both transporters are responsible for copper export, but in different cell types. The ATP7A is expressed in almost all cell types and tissues, but its expression levels differ among the cells and tissues and are age-dependent (Lenartowicz et al., 2010). In contrast, ATP7B is predominantly expressed in the adult liver Lenartowicz et al., 2015).

Fibroblasts from MD patients accumulate large amounts of copper as a result of impaired copper extrusion in contrast to fibroblasts from WD patients, in which no accumulation of copper is observed (own unpublished results). Both ATP7A and ATP7B are expressed in fibroblasts, but the expression level of ATP7A is approximately 10 times higher than that of ATP7B (own unpublished results).

Female carriers are mosaics of wild-type and mutant cells due to random X-inactivation, and are rarely affected. A likely explanation is that the majority (more than 90%) of the females have a skewed inactivation of the X-chromosome bearing the ATP7A mutation and therefore, predominantly express the normal allele (Møller et al., 2012). However, from a large number of fibroblast cultures we were able to identify cultures from three MD carriers, including one manifesting carrier with MD, with an even or almost even X-inactivation pattern (Møller et al., 2012).

Not much is known about the possible transcriptional regulation of ATP7A and ATP7B. It seems that regulation and copper homeostasis occur predominantly in post-translational mechanisms (van den Berghe and Klomp, 2010). This study was designed to test the transcriptional effect on ATP7A and ATP7B in fibroblast cultures after short (3 days) or longer (10 days) term treatment with iron and copper and chelators thereof, and insulin. Short-term treatment might exhibit a rapid effect on transcriptional levels, whereas long-time treatment may also involve cellular selection. We treated fibroblasts obtained from controls with both intact ATP7A and intact ATP7B, fibroblasts from MD with only intact ATP7B, and fibroblasts from WD with only intact ATP7A.

Furthermore, we looked at clonal growth selection as an effect of functional ATP7A gene expression, and the cellular concentration of copper. Mosaic cultures obtained from female MD carriers with an even X-inactivation pattern, were treated with copper or the copper chelator (Bathocuproine disulphonate, BCS).

Both P-type ATPases (ATP7A and ATP7B), are involved in copper efflux and intracellular sequestration. The overall topologies and structures of ATP7A and ATP7B are similar and they are approximately 60% homologous. They have eight transmembrane domains with six copper-binding motifs in the cytoplasmic N-terminal region. The N-terminal metal-binding domains interact with ATOX1 and receive copper in the form of Cu⁺. The transmembrane hydrophobic domains (TMDs) anchor the protein in the membrane and form a channel for the ATP-dependent transport of copper ions across the plasma or intracellular membranes. Cellular ATP7A and ATP7B play a dual role depending on the copper status. Under normal physiological conditions they are TGN-resident proteins and mediate copper loading of copper-dependent enzymes in the secretory pathway. When the intracellular copper concentration is elevated, ATP7A and ATP7B – after their re-localization/re-distribution to the plasma membrane or to vesicles – are essential for copper efflux or secretion. ATP7A and ATP7B contain similar catalytic domains and functional motifs that govern their intracellular cycling between the TGN and the plasma membrane (La Fontaine and Mercer, 2007; Lutsenko et al., 2007; Gupta and Lutsenko, 2009; Barry et al., 2010).

However, whereas both ATP7A and ATP7B preserve cellular copper homeostasis by expelling copper ions from the cell’s interior to prevent toxic accumulation, there are some differences. When the cells are exposed to a rise in copper concentrations, ATP7A is translocated to the basolateral membrane in polarized cells, whereas ATP7B is transported to the apical membrane (Mercer et al., 2003; La Fontaine and Mercer, 2007; Gupta and Lutsenko, 2009; Veldhuis et al., 2009; Barry et al., 2010). Furthermore, whereas ATP7A in the secretory pathway, delivers copper to lysyl oxidase, tyrosinase, dopamine-β-hydroxylase, and peptidylglycine-α-amidating monooxygenase among others (Petris et al., 2000; Hardman et al., 2007a; Lutsenko et al., 2007), the ATP7B protein participates in copper-loading of apo-ceruloplasmin (Barry et al., 2010).

We found both similarities and differences in the transcriptional regulation of ATP7A and ATP7B. A summary is shown in Table 2. We found that treatment with insulin led to a transient increase in ATP7A transcript, but had no effect on ATP7B transcript in control fibroblasts. This observation is in agreement with previous observations by Hardman et al. (2007b) who reported that treating Jeg-3 cells with insulin for 72 h (3 days) leads to an increase in the ATP7A transcript but has no effect on the ATP7B transcript. Insulin (and ostrogen) also changed the localization of ATP7A in the Jeg-3 cells; ATP7A was more diffusely distributed in the cytoplasm with increased accumulation at the basolateral membrane relative to the untreated cells (Hardman et al., 2007a). Our results indicate that the effect of transcriptional regulation of the ATP7A gene by insulin is not restricted to the placental cells and is closely related to the control of copper transport from mother to fetus, but it is also observed in other cell types.

Iron deficiency led to an increased expression of both ATP7A and ATP7B, but copper-deficiency only led to a decreased expression of ATP7A. The addition of copper led to an increased expression of ATP7A, but a decreased expression of ATP7B. The changes in transcripts (increase or decrease) were less than twofold. We have previously shown that the relative transcriptional level of ATP7A in control fibroblasts vary by almost a factor 2 (Paulsen et al., 2006) making it unlikely that the observed transcriptional regulation has any physiological significance.

Interactions between iron and copper metabolism were documented over a century ago, but the homeostatic effects at the molecular level and the involvement of ATP7A and ATP7B in this regulation have not been fully elucidated. Interaction between iron and copper is essential for metal homeostasis because, e.g., copper-dependent ferroxidases like ceruloplasmin, hephaestin, and zyclopen facilitate dietary iron absorption by oxidizing ferrous Fe⁺⁺ to ferric Fe⁺⁺⁺ iron (Vulpe et al., 1999; Collins et al., 2010; Gambling et al., 2011; Vashchenko and MacGillivray, 2013). Any impairment to the incorporation process of copper results in the secretion of an apo-ceruloplasmin that lacks any ferroxidase activity, and is rapidly degraded in the plasma (Hellman and Gitlin, 2002; Kono, 2013). Atp7b -/- mice, which constitute an animal model of WD, display disturbances in iron metabolism related to low serum ceruloplasmin oxidase activity (Merle et al., 2010). Furthermore, a complex interplay between these two metals seems to involve multiple copper-dependent regulatory mechanisms that act in concert to control the expression of ferroportin, the only identified ferrous iron transporter that exports iron from cells to plasma (Xie and Collins, 2011; Starzyński et al., 2013).

An increased expression of ATP7A has previously been identified in intestinal epithelial cells of iron-deficient rats and confirmed by in vivo and in vitro analyses (Collins et al., 2005, 2009; Collins, 2006). During iron deficiency many tissues become hypoxic. Cellular responses to hypoxia are mainly regulated by the activation of transcription factors called hypoxia-inducible factors (HIFs) (Hashimoto and Shibasaki, 2015). Iron deficiency in the intestinal epithelial cells leads to the upregulation of the ATP7A gene and iron-homeostasis related genes: the cell surface ferrireductase gene Dtcyb, the iron ion transporter gene DMT1 and the basolateral iron exporter ferroportin gene (FPN) (Xie and Collins, 2011, Xie and Collins, 2013). Results of previous studies indicate that HIF2α binds to the promotor region of the ATP7A gene, the analysis of which revealed that this region contains three phylogenetically conserved, putative hypoxia-response elements (HRE; 5′-NCGTGN-3′) (Xie and Collins, 2011, 2013).

Fibroblasts incubated for 3–10 days and mosaic cultures incubated for 6–8 weeks revealed that a large amount of functional ATP7A is advantagous in high concentrations of copper, whereas a low amount/zero of functional ATP7A is advantageous in low concentrations of copper. Both after 3 and 10 days of treatment, the effect on the transcription level could be visualized, whereas in the mosaic cell culture the effect was seen on clonal selection. The lack of decreased expression of ATP7A in WD fibroblasts after BCS treatment might be a result of the absence of functional ATP7B protein in these cells, thereby reducing the demand for a downregulation of ATP7A. Furthermore, in both types of experiments we found that cultures grown in basic, control medium also led to an increased expression of ATP7A and increased the growth of cells with the wild-type ATP7A. An increase in the ATP7B transcript level after long-term culturing was only observed for the WD fibroblasts. This might indicate that even under normal conditions, an increased amount of ATP7A is an advantage, whereas an increased amount of ATP7B is an advantage only if the level of functional ATP7B is lower than normal, e.g., as a result of mutations.

A large fraction of the cells died in a mosaic culture with cells from a female carrier grown in medium supplemented with 100–400 μM CuCl₂ or 10–50 μM BCS, respectively, for either 6 or 8 weeks. This suggests that growth selection is a likely explanation for the observed skewed X-inactivation pattern in these cells. Our results indicate that the two different cell types have complementary growth characteristics. In the presence of a low concentration of copper, the selection went against cells with the active normal X-chromosome, whereas in the presence of a high concentration of copper, the selection went against cells with the active mutant X-chromosome.

Amplification of the ATP7A gene has previously been observed in selected CHO cells after stepwise increments of copper in the growth medium (Camakaris et al., 1995). The copper concentration was increased from 142 μM to 205 μM over 27 days, and the cells were cultured for an additional 99 days at the latter concentration. Both mechanisms: ATP7A gene amplification, and growth advantage of the cells that express the normal ATP7A gene show that the ATP7A gene is essential for the cells under conditions with high concentrations of copper. In contrast, the growth advantage of the cells that express the mutant ATP7A gene indicates that the absence of a functional ATP7A gene is more advantageous if no copper is present. To our knowledge, a positive selection of cells without a functional ATP7A gene an effect of the lack of copper has not been reported previously.

Affected carriers of MD are rare. In a total of 517 families we know of 9 confirmed affected females. However, we are aware of the fact that there are mildly affected females who have not been characterized as manifesting carriers. We know of several girls with pili torti, hypopigmentation of the skin or other mild symptoms of MD. The total number of females with symptoms is unknown. It is possible that copper treatment during pregnancy decreases the brain damage in heterozygous female fetuses both indirectly by clonal cell selection of the developing cells that express the normal ATP7A gene, and directly by the transport of copper to copper-requiring enzymes. Thus the serum copper level in pregnant heterozygous females should remain within the upper level of the reference interval in order to diminish the symptoms of MD and, in particular, brain damage in both hemi- and heterozygous fetuses. Copper supplements during pregnancy should be considered.

The X-inactivation pattern in a MD mice model should be determined. It would reveal the effect of prenatal copper treatment on the X-inactivation pattern in brain tissue in heterozygous female mouse fetuses.

In X-linked recessive diseases, females are rarely affected, probably due to skewed inactivation of the X-chromosome harboring the mutant allele. Identification of small molecules leading to negative selection against cells deficient in the disease-related gene might be relevant for other X-linked diseases where the phenotype in females could be hampered due to selection of cells with the active normal X-chromosome.

Ethics approvals were obtained from the Danish Bioethics Committees for the Capital Region (KF 01-169/03 and H-1-2011-FSP-31). Written informed consent was obtained from the WD patient after information was given about the project “investigation of copper and iron metabolism”. MD fibroblasts were collected for diagnostic purpose, and since it was not possible to get any contact to the patients or the patients guardians, dispensation for obtaining informed consent was given from the Danish Bioethics Committee for the Capital Region.

ML, TJ, and LM, designed the experiments. ML and LM performed the experiments. ML and LM drafted the article MO made Figure 2. TM and TJ revised the manuscript critically for its significant intellectual content, and the experiment design.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.