ATOX1: A novel copper-responsive transcription factor in mammals?

The International Journal of Biochemistry & Cell Biology 41 (2009) 1233–1236

Contents lists available at ScienceDirect

The International Journal of Biochemistry & Cell Biology journal homepage: www.elsevier.com/locate/biocel

Molecules in focus

ATOX1: A novel copper-responsive transcription factor in mammals? Patricia A.J. Muller, Leo W.J. Klomp ∗ Laboratory for Metabolic and Endocrine Diseases, UMC Utrecht, Lundlaan 6, 3584 EA Utrecht, The Netherlands

a r t i c l e

i n f o

Article history: Received 15 July 2008 Received in revised form 31 July 2008 Accepted 1 August 2008 Available online 7 August 2008 Keywords: Copper ATOX1 Transcription CCND1 CABE

a b s t r a c t As the trace element copper is essential, but extremely toxic in high concentrations, intracellular copper concentrations are tightly controlled. Once in the cell, copper is distributed by metallochaperones, including the small cytoplasmic protein ATOX1. ATOX1 plays an important role in the transfer of copper to the copper export P-type ATPases ATP7A and ATP7B to facilitate copper excretion. Recently, a novel function for Atox1 as a transcription factor (TF) regulating Ccnd1 was proposed. Crystal structures of ATOX1 reveal copper-dependent homodimerization of ATOX1. As many TFs regulate gene expression as a dimer and bind to DNA repeats, we investigated the promotor region of CCND1 and detected a direct repeat sequence in the Atox1 binding site (tentatively referred to as CABE, copper-responsive Atox1 binding element). We therefore propose copper-dependent homodimerization to be an essential step in the regulation of ATOX1-dependent transcription. © 2008 Elsevier Ltd. All rights reserved.

1. Introduction Copper is an essential trace element that is required in various biological processes. At the same time, excess copper concentrations are extremely toxic. A balance between copper necessity and toxicity is therefore crucial for survival. Disturbances in this balance are the primary manifestations of the genetic disorders Menkes disease and Wilson disease, respectively. To maintain a balance, copper import and export are tightly controlled by several proteins. In cells, copper import is mediated by CTR1 (copper transporter 1) and CTR2 (copper transporter 2) (van den Berghe et al., 2007; Zhou and Gitschier, 1997). To ensure that the total pool of free copper is less than 1 ion per cell, copper-binding metallochaperones specifically deliver copper to cellular compartments and copper-dependent enzymes. This function requires the transient copper-dependent interaction between metallochaperone and target enzyme. Among these chaperones is ATOX1 (anti-oxidant 1, previously HAH1) (Klomp et al., 1997). ATOX1 was originally characterized as the human homologue of Atx1p that prevents oxidative injury in Sod1 yeast (superoxide dismutase 1) (Lin and Culotta, 1995). ATOX1 interacts with ATP7A (Menkes protein) and ATP7B (Wilson protein) in a copper-dependent manner. This interaction facilitates the transfer of copper from ATOX1 to ATP7A or ATP7B, thus permitting copper transport into the secretory pathway and copper export from the cell. A recent study, however, reveals a novel

∗ Corresponding author. Tel.: +31 88 7555318; fax: +31 88 7554295. E-mail address: [email protected] (L.W.J. Klomp). 1357-2725/$ – see front matter © 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.biocel.2008.08.001

role for ATOX1 as a copper-dependent transcription factor (TF) (Itoh et al., 2008). In the present work, we will discuss the implications of this finding and provide additional in silico evidence for ATOX1 to function as a copper-dependent transcription regulator. 2. Structure and expression 2.1. The structure of the ATOX1 gene and the ATOX1 protein The ATOX1 gene is located on chromosome 5 and contains 4 exons and 3 introns. ATOX1 encodes for a ubiquitously expressed protein of 68 amino acids and contains a MBS (metal-binding site) and a potential NLS (nuclear localization signal) (Itoh et al., 2008; Klomp et al., 1997) (Fig. 1A). The MBS, MXCXXC, is highly conserved in metallochaperones and also occurs six times in the N-terminus of ATP7A and ATP7B. Biochemical and spectroscopic data have revealed binding of Cu(I) to the MBSs of ATOX1, ATP7A and ATP7B (Anastassopoulou et al., 2004). A conserved KKTGK motif in the C-terminus of ATOX1 was originally identified as important in the anti-oxidant function of yeast Atx1, but recent data suggest that this domain might also function as a NLS involved in recruiting Atox1 to the nucleus (Hamza et al., 2001; Itoh et al., 2008). 2.2. Dimerization of ATOX1 Several studies revealed an interaction between ATOX1 and ATP7A or ATP7B. Incubation with a copper chelator strongly diminished these interactions (van Dongen et al., 2004; Wernimont et al., 2000). In addition, ATOX1 in which the Cys residues of the

1234

P.A.J. Muller, L.W.J. Klomp / The International Journal of Biochemistry & Cell Biology 41 (2009) 1233–1236

Fig. 1. Structure of ATOX1 and structure of the CCND1 promotor region. (A) The ATOX1 protein is 68 amino acids in size and contains two characterized domains; A MBS (metal-binding site) and a NLS (nuclear localization signal). N, N-terminus; C, C-terminus. (B) Only in the presence of copper ATOX-1 forms dimers in which the cys residues of MBSs interact with Cu(I). In the dimer two Cys residues of each monomer (ATOX1 or ATP7A/B) form a three-coordinate Cu(I) complex with a loose interaction with the fourth Cys residue. (C) ATOX1 interacts with the CCND1 promotor region, which is underlined. Part of this interaction region belongs to a conserved repeat sequence that is indicated by arrows. (D) Conservation of the promotor region of CCND1 in several species. The indicated conserved region was used to screen for the presence of these repeat sequences in the promotor regions of other genes (Table 1).

MBS were mutated (mutant ATOX1) was unable to interact with ATP7B (Strausak et al., 2003), indicating that the MBSs coordinate these interactions. Yeast atx1 exhibited impaired transfer of copper into the secretory pathway. This phenotype could be complemented by overexpression of ATOX1, but not mutant ATOX1, illustrating the importance and conservation of this interaction in copper metabolism (Klomp et al., 1997). X-ray crystallography and biochemical analyses also revealed copper-dependent ATOX1 homodimerization (Wernimont et al., 2000). In these homodimers, two Cys residues of each monomer formed a three-coordinate Cu(I) complex with a loose interaction with the fourth Cys residue (Fig. 1B). Remarkably, glutathione bound to ATOX1 and facilitated the formation of ATOX1-Cu(I) homodimers (Ralle et al., 2003). These studies propose that ATOX1 homodimers could function as copper storage molecules, but the precise function and nature of these homodimers remains elusive (Fig. 2).

development (Hamza et al., 2001). A lack of Atox1 in cells resulted in accumulation of copper and a failure of cells to excrete copper. 3.2. ATOX1 as a copper-dependent TF Until recently, only one gene was characterized to encode a copper-responsive TF in the genome of mammals, MTF1 (metal

3. Biological function 3.1. ATOX1 as a copper chaperone protein Disruption of Atox1 in mice resulted in severe phenotypic alterations, including a failure to thrive, growth retardation, congenital eye defects, hypopigmentation and seizures (Hamza et al., 2001). In part, these characteristics resembled the phenotypes of mice that experience copper-deprivation (Pena et al., 1999; Prohaska, 1983). The observed impaired copper transport from the mother to the embryo in Atox−/− mice was consistent with a copper-depletion phenotype, indicative for an important role of Atox1 in embryonic

Fig. 2. Cellular model of ATOX1 function. ATOX1 can function as a copper chaperone that delivers copper to the export pumps ATP7A or ATP7B. Alternatively, ATOX1 can function as a copper-dependent transcription factor. We here propose a model in which copper-dependent homodimerization of ATOX1 drives the expression of CCND1 and possibly other genes. ATOX1 might shuttle between the nucleus and might form homodimers to activate transcription. Alternatively, copperhomodimers of ATOX1 might be formed in the cytosol, possibly mediated by other molecules including GSH (glutathione) and may subsequently be recruited to the nucleus to drive gene expression.

P.A.J. Muller, L.W.J. Klomp / The International Journal of Biochemistry & Cell Biology 41 (2009) 1233–1236

transcription factor 1). MTF1 regulates the metal-dependent expression of metallothionein genes (Heuchel et al., 1994). During the last two decades several attempts, including whole-genome cDNA microarrays, have failed to identify other mammalian copperdependent TFs (Huster et al., 2007; Muller et al., 2007). The general idea was therefore that mammalian copper homeostasis was primarily regulated post-transcriptionally. Recent exciting data by Itoh et al, however, revealed that Atox1 represents a second mammalian copper-dependent TF that drives the expression of Ccnd1 (cyclin D1) (Itoh et al., 2008). Copper incubation normally increased proliferation of MEFs (mouse embryonic fibroblasts), but in Atox1-deficient MEFs (Atox1−/− cells), copper-dependent and copper-independent proliferation was decreased. These effects were attributed to an Atox1-dependent increase in the mRNA and protein expression of Ccnd1, an important regulator of transition of G1 to S-phase. Using state-of-the-art ChIP assays, GST-pulldown experiments and electromobility shift assays, Itoh et al established a direct copper-dependent interaction of Atox1 to the promotor region of Ccnd1 and characterized a GAAAGA region approximately 500 bp from the transcription start site of Ccnd1 as the Atox1-interaction region. In addition, copper incubation resulted in nuclear accumulation of Atox1 for which both the MBS and the NLS of Atox1 were important (Itoh et al., 2008). These data thus suggest that Atox1 functions as a novel copper-dependent mammalian TF regulating cell proliferation. This conclusion is further supported by the remarkable resemblance between the phenotypes of Ccnd1−/− and Atox1−/− mice: both mouse models display severe growth retardation, neurological problems and congenital eye defects (Fantl et al., 1995; Sicinski et al., 1995).

1235

3.3. ATOX-homodimers as copper-dependent TFs Many TFs bind to DNA repeats as dimers to regulate transcription of target genes. Precedence for copper-responsive transcription regulators that bind to DNA repeats exists in bacteria. In E. coli, the DNA interaction sites that are bound by the copper-responsive TFs CusR, CueR and CpxR contain inverted or direct repeat sequences. Of these proteins, CueR forms copper-dependent homodimers and belongs to the family of MerR TFs that are ligand-dependent activators of transcription. To assess whether Atox1 could function as a copper-responsive homodimeric TF, we examined the region of the Ccnd1 promotor to which Atox1 binds in detail (Itoh et al., 2008). A perfect direct repeat sequence (TGAAAnTGAAA) was detected and tentatively named CABE (copper-responsive Atox1 binding element) (Fig. 1C). The importance of the repeat was confirmed by conservation in several mammals at 400–1000 bp upstream of the transcription start site of Ccnd1 (Fig. 1D). ATOX1 subunits interact with each other in a head-to-head fashion, suggestive for an interaction of the dimers with inverted repeats on the DNA. However, precendence of other ligand-dependent TFs that interact with direct DNA repeats in a head-to-head fashion exists for various nuclear hormone receptor dimers, including HNF4␣, PPAR-␣ and LXR/RXR, illustrative for the possibility of an ATOX1-dimer to bind to a direct repeat. 3.4. Could ATOX1 regulate other genes? The observation that Atox1 drives the expression of Ccnd1 opens the possibility that other genes might also be regulated by Atox1. We therefore performed a whole-genome search for puta-

Table 1 Genes that contain a perfect potential ATOX1 DNA interaction site in the promotor region Name

ID

Chrom

Gene description

Position

Sequence

CCND1 LOC442229 PIK3CB RAGE LOC731788 H2AFY MAP2K1IP1 SAP30BP PPA2 NARF SPATA18 STON1 LPHN3 TAS2R50 LOC653801 VSX1 TMEM45A RPS6KA5 CRYBA4 PNPT1 PI4K2B PHLPPL C3orf27 C12orf60 UFM1 FLJ38482 NDUFB3 ETAA16 FNDC6 VRK1 NAT5 TRA

595 442229 5291 5891 731788 9555 8649 29115 27068 26502 132671 11037 23284 259296 653801 30813 55076 9252 1413 87178 55300 23035 23434 144608 51569 201931 4709 54465 152028 7443 51126 6955

11 6 3 14 21 5 4 17 4 17 4 2 4 12 3 20 3 14 22 2 4 16 3 12 13 4 2 2 3 14 20 14

Cyclin D1 Similar to mitochondrial carrier triple repeat 1 Phosphatidylinositol 3-kinase catalytic, beta polypeptide MAPK/MAK/MRK overlapping kinase Similar to neurofibromin H2A histone family, member Y isoform 2 Mitogen-actvated protein kinase kinase 1 interacting protein Transcription regulator protein Inorganic pyrophosphatase 2 isoform 1 precursor Nuclear prelamin A recognition factor isoform A Spermatogenesis associated 18 homolog Stoned B/TFIIA-alpha/beta-like factor Latrophilin 3 precursor Taste receptor, type 2, member 50 Hypothetical protein Visual system homeobox 1 protein isoform a Transmembrane protien 46 Ribosomal protein S6 kinase, 90 kDa, polypeptide 5 isoform a Crystallin, beta A4 Polyribonucleotide nucleotidyltrasferase 1 Phosphatidylinositol 4-kinase type-II beta PH domain and leucin-rich repeat protein phosphatase like Putative GR6 protein LOC144608 Ubiquitin-fold modifier 1 Hypothetical protein loc201931 NADH dehydrogenase (ubiquinone) 1 beta subcomplex, 3, 12 kDa ETAA16 protein Fibronectin type III domain containing 6 Vaccinia related kinase 1 N-Acetyltransferase 5 isoform a

495 15 38 228 275 494 585 592 660 1021 1188 1202 1310 1508 1548 1696 1936 2055 2070 2130 2174 2197 2374 2394 2439 2523 2685 2715 2749 2976 2985 2991

TTTACACGTGTAATGAAAATGAAAGAAGA ATGAAAATGAAAGA AATGAAAATGAAA TTAACGAAAATGAAAGAA AATGAAAATGAAAGA TGAAAATGAAAGA TTAATGAAAATGAAA AATGAAAATGAAA ATGAAAATGAAAG AATGAAAATGAAA TAATGAAAATGAAAG AATGAAAATGAAA AATGAAAATGAAA AATGAAAATGAAA TGTTAATGAAAATGAAA AATGAAAATGAAA AATGAAAATGAAAAAAGA AATGAAAATGAAA ATGAAAATGAAAG TGAAAATGAAAGA AATGAAAATGAAA AATGAAAATGAAA ATGAAAATGAAAGA AATGAAAATGAAA TGAAATGAAAGAA TAATGAAAATGAAA AATGAAAATGAAA AATGAAAATGAAA ATGAAAATGAAAG TGAAAATGAAAGA AATGAAAATGAAAGA AATGAAAATGAAAGAAGA

Using the conserved region indicated in Fig. 1D, we screened the whole genome for the repeat sequence, including the flanking region, in promotor regions. Only genes that had a perfect repeat sequence within 3000 bp from the transcription start site are depicted in this table. The name of the gene, the Entrez identifier (ID), the chromosome the gene is located on (chrom), a gene description, the position from the transcription start site and the sequence are shown.

1236

P.A.J. Muller, L.W.J. Klomp / The International Journal of Biochemistry & Cell Biology 41 (2009) 1233–1236

tive CABEs within 3000 bp from transcription start sites (Fig. 1D). From this analysis, we retrieved 32 genes that all contained a perfect repeat sequence in the promotor region and could thus potentially be regulated by ATOX1 (Table 1). Remarkably, two of these genes, VSX1 (visual system homeobox 1) and CRYBA4 (crystalline beta A4) are associated with abnormalities in eye development and visual perception. As congenital eye defects are one of the phenotypic characteristics of the Atox1−/− mouse (Hamza et al., 2001), an ATOX1-dependent regulation of these genes might possibly underlie these defects. Notably, congenital eye defect have not been reported in other copper-deprived mice (Pena et al., 1999; Prohaska, 1983), suggesting that Atox1 could drive the expression of these genes under limited copper concentrations. As these genes might be target genes of ATOX1 after increased copper exposure, we screened the data of several cDNA microarray studies performed in various tissues after copper exposure for increased expression of these 32 genes (Huster et al., 2007; Muller et al., 2007). Increased expression of Ccnd1 in copper-overloaded rat livers was reported by Kita et al. (2004). We, however, did not detect any of these 32 genes among the differentially expressed genes after copper exposure in these microarray experiments. This could indicate that the regulation of copper-mediated gene expression via ATOX1 is variable in different cellular systems, organs or developmental stages. Monitoring of the gene expression of these 32 genes in ATOX1-depleted cells compared to wildtype cells after copper exposure could validate whether these 32 genes are copperdependent ATOX1 target genes. 3.5. TF versus copper chaperone From these data the question arises how ATOX1 can exert a dual function as a transcription regulator and as a copper chaperone. Here, we propose two hypothetical models. Whereas ATP7A and ATP7B are Golgi-resident membrane proteins that are not localized to the nuclear membrane, ATOX1 is small enough to diffuse freely across the nuclear membrane. At high nuclear copper concentrations, ATOX1 might then form copper-dependent TF homodimers without any competition of the MBSs of these P-type ATPases and drive transcription. Alternatively, ATOX1 homodimerization within the cytosol might prevent interaction with ATP7A/B and simultaneously trigger ATOX1 to translocate to the nucleus to initiate transcription. Such a model potentially suggests the presence of a copper-dependent affinity threshold of ATOX1 hetero- or homodimerization. Consistent with this idea, kinetic and dynamic analyses of the interaction between MBSs of two ATOX1 molecules or the ATOX1 MBS with one of the six MBSs of ATP7A and ATP7B reveal distinct binding affinities among these domains, dependent on the available copper concentrations (Strausak et al., 2003; van Dongen et al., 2004). In addition, ATOX1-dimerization after increased GSH concentrations could be indicative for a role of additional molecules or mechanisms in regulating ATOX1 homo- or heterodimerization (Ralle et al., 2003). To functionally discriminate between the two models, the localization and occurrence of copper-dependent ATOX1 homodimers could be investigated by fluorescence resonance energy transfer (FRET). The feasibility of using a FRET-based sensor for copper-dependent ATOX1 homodimerization has previously been established in fluorescence spectrophotometry experiments (van Dongen et al., 2006). Co-expression of differently labeled ATOX1 could reveal the localization of the homodimers in living cells. NLS mutants of differently labeled ATOX1 could further demonstrate the absence or presence of cytosolic dimers.

4. Possible medical and biological implications ATOX1 is important in regulating cell proliferation (Itoh et al., 2008). In Wilson disease, copper overload in the liver can eventually result in HCC (hepatocellular carcinoma). Furthermore, copper overload is observed in various tumors, and cancer patients often benefit from copper chelation therapy. A characterization of the precise mechanisms of ATOX1-mediated transcriptional regulation of the cell cycle would therefore be potentially beneficial in designing intervention strategies in cancers and Wilson disease. Acknowledgements We would like to thank Dr. E. Kalkhoven for helpful comments. This work is funded by the WKZ Fund (40-00812-98-03106). References Anastassopoulou I, Banci L, Bertini I, Cantini F, Katsari E, Rosato A. Solution structure of the apo and copper(I)-loaded human metallochaperone HAH1. Biochemistry 2004;43:13046–53. Fantl V, Stamp G, Andrews A, Rosewell I, Dickson C. Mice lacking cyclin D1 are small and show defects in eye and mammary gland development. Genes Dev 1995;9:2364–72. Hamza I, Faisst A, Prohaska J, Chen J, Gruss P, Gitlin JD. The metallochaperone Atox1 plays a critical role in perinatal copper homeostasis. Proc Natl Acad Sci USA 2001;98:6848–52. Heuchel R, Radtke F, Georgiev O, Stark G, Aguet M, Schaffner W. The transcription factor MTF-1 is essential for basal and heavy metal-induced metallothionein gene expression. EMBO J 1994;13:2870–5. Huster D, Purnat TD, Burkhead JL, Ralle M, Fiehn O, Stuckert F, et al. High copper selectively alters lipid metabolism and cell cycle machinery in the mouse model of Wilson disease. J Biol Chem 2007;282:8343–55. Itoh S, Kim HW, Nakagawa O, Ozumi K, Lessner SM, Aoki H, et al. Novel role of antioxidant-1 (Atox1) as a copper-dependent transcription factor involved in cell proliferation. J Biol Chem 2008;283:9157–67. Kita Y, Masaki T, Funakoshi F, Yoshida S, Tanaka M, Kurokohchi K, et al. Expression of G1 phase-related cell cycle molecules in naturally developing hepatocellular carcinoma of Long-Evans Cinnamon rats. Int J Oncol 2004;24: 1205–11. Klomp LW, Lin SJ, Yuan DS, Klausner RD, Culotta VC, Gitlin JD. Identification and functional expression of HAH1, a novel human gene involved in copper homeostasis. J Biol Chem 1997;272:9221–6. Lin SJ, Culotta VC. The ATX1 gene of Saccharomyces cerevisiae encodes a small metal homeostasis factor that protects cells against reactive oxygen toxicity. Proc Natl Acad Sci USA 1995;92:3784–8. Muller P, van Bakel H, van de Sluis B, Holstege F, Wijmenga C, Klomp LW. Gene expression profiling of liver cells after copper overload in vivo and in vitro reveals new copper-regulated genes. J Biol Inorg Chem 2007;12:495–507. Pena MM, Lee J, Thiele DJ. A delicate balance: homeostatic control of copper uptake and distribution. J Nutr 1999;129:1251–60. Prohaska JR. Changes in tissue growth, concentrations of copper, iron, cytochrome oxidase and superoxide dismutase subsequent to dietary or genetic copper deficiency in mice. J Nutr 1983;113:2048–58. Ralle M, Lutsenko S, Blackburn NJ. X-ray absorption spectroscopy of the copper chaperone HAH1 reveals a linear two-coordinate Cu(I) center capable of adduct formation with exogenous thiols and phosphines. J Biol Chem 2003;278:23163–70. Sicinski P, Donaher JL, Parker SB, Li T, Fazeli A, Gardner H, et al. Cyclin D1 provides a link between development and oncogenesis in the retina and breast. Cell 1995;82:621–30. Strausak D, Howie MK, Firth SD, Schlicksupp A, Pipkorn R, Multhaup G, et al. Kinetic analysis of the interaction of the copper chaperone Atox1 with the metal binding sites of the Menkes protein. J Biol Chem 2003;278:20821–7. van den Berghe PV, Folmer DE, Malingre HE, van Beurden E, Klomp AE, van de Sluis B, et al. Human copper transporter 2 is localized in late endosomes and lysosomes and facilitates cellular copper uptake. Biochem J 2007;407:49–59. van Dongen EM, Dekkers LM, Spijker K, Meijer EW, Klomp LW, Merkx M. Ratiometric fluorescent sensor proteins with subnanomolar affinity for Zn(II) based on copper chaperone domains. J Am Chem Soc 2006;128:10754–62. van Dongen EM, Klomp LW, Merkx M. Copper-dependent protein-protein interactions studied by yeast two-hybrid analysis. Biochem Biophys Res Commun 2004;323:789–95. Wernimont AK, Huffman DL, Lamb AL, O’Halloran TV, Rosenzweig AC. Structural basis for copper transfer by the metallochaperone for the Menkes/Wilson disease proteins. Nat Struct Biol 2000;7:766–71. Zhou B, Gitschier J. hCTR1: a human gene for copper uptake identified by complementation in yeast. Proc Natl Acad Sci USA 1997;94:7481–6.