Malvolio is a copper transporter in Drosophila melanogaster

The maintenance of copper (Cu) homeostasis at the cellular level is essential for all aerobic organisms and an excess of Cu is toxic as a result of the generation of reactive oxygen species(Linder and Hazegh-Azam,1996). Mechanisms for uptake, distribution, sequestration and efflux of Cu are well conserved among yeast, insect and mammalian systems(Camakaris et al., 1999). The Ctr family are the best characterised Cu uptake proteins, with Drosophila melanogaster having three, designated Ctr1A, B and C. In vivoand in vitro studies have demonstrated Ctr1A and Ctr1B are functional Cu transporters with different tissue and life stage expression profiles(Southon et al., 2004; Turski and Thiele, 2007; Zhou et al., 2003). Although human CTR1 (also known as SLC31A1 – HUGO) has traditionally been seen as the primary Cu uptake mechanism in gut epithelium, a recent study by Zimnicka et al. (Zimnicka et al., 2007)has found endogenous CTR1 to be localised to the basolateral surface of cultured intestinal and kidney cells, rather than the apical membrane,suggesting CTR1 may not be responsible for dietary Cu absorption. Divalent metal ion transporter 1 (DMT1; also known as SLC11A2 – HUGO, and Nramp2), which is capable of transporting Fe²⁺, Mn²⁺,Cu²⁺, Zn²⁺, Cd²⁺ and Pb²⁺(Gunshin et al., 1997) may also contribute to Cu uptake. DMT1 is expressed in multiple mammalian tissues,whereas its paralogue Nramp1 (SLC11A1) is expressed in macrophages(Gunshin et al., 1997; Vidal et al., 1995) as well as dendrites (Stober et al.,2007) and tertiary granules of neutrophils(Canonne-Hergaux and Gros,2002). The DMT1 family is evolutionarily conserved, with orthologues in several species including mouse, rat and yeast(Cellier et al., 1995). Although DMT1 is clearly a symporter that transports divalent cations across the plasma or endosomal membranes into the cytoplasm(Gunshin et al., 1997; Touret et al., 2003) the direction of Nramp1 transport is less clear, with some authors suggesting it is an antiporter, exporting divalent cations from the cytoplasm(Eichner Techau et al., 2007; Goswami et al., 2001). Eichner Techau et al. (Eichner Techau et al.,2007) propose that symporter activity was the ancestral function of this family and that following the gene duplication antiporter activity evolved for Nramp1.

Malvolio (Mvl) is the only member of the DMT1 family in the Drosophila genome and is orthologous to the human proteins NRAMP1(54.9% identity) and DMT1 (57.4%) (Cellier et al., 1995). Mvl was identified in a screen for mutants that affect taste behaviour (Rodrigues et al., 1995). A P-element induced partial loss-of-function mutant(Mvl^97f) causes flies to lose preference for sugar-supplemented medium. Rearing these flies on medium supplemented with Mn²⁺ or Fe²⁺ corrected this defect, whereas Ca²⁺, Mg²⁺ or Zn²⁺ did not, suggesting Mvl^97f flies may have a Mn²⁺ or Fe²⁺deficiency (Orgad et al.,1998). Recently, Folwell et al.(Folwell et al., 2006)examined Mvl localisation in Drosophila. Protein expression was detected in larval Malpighian tubules, testis and brain, in the amnioserosa of embryos and in the larval and adult alimentary canal, with intracellular localisation that was either punctate, cytoplasmic or at the plasma membrane. The authors concluded that Mvl must fulfil the roles of both Nramp1 and DMT1 given its distribution, however, metal homeostasis was not examined. Mvl, like DMT1, was recently shown to be a cation symporter(Eichner Techau et al., 2007),however it is still unclear if Mvl is a Cu transporter.

The model organism Drosophila melanogaster has recently been used to enhance our understanding of several aspects of Cu homeostasis including:Cu uptake by the Ctr family (Turski and Thiele, 2007; Zhou et al.,2003); Cu sequestration through the metallothionein (Mtn) system(Balamurugan et al., 2007; Egli et al., 2003; Selvaraj et al., 2005); and Cu efflux via ATP7 (Norgate et al.,2006). We now use this system, together with the Drosophila embryonic cell line S2, to demonstrate that Mvl is a functional Cu transporter, necessary for optimal Cu uptake and distribution.

Full-length Mvl cDNA lacking the C-terminal STOP codon (according to FlyBase annotation; The FlyBase Consortium, 2003; http://flybase.org/)was amplified from Drosophila S2 cell-derived cDNA using PCR. This was cloned into pAc3.1V5-HisA (Invitrogen, Mount Waverly, Vic, Australia)modified to contain an in-frame C-terminal FLAG epitope tag (MDYKDDDDKA) for expression in S2 cells and cloned into pUAST for expression in Drosophila using the UAS-Gal4 system. Pannier-Gal4/Sb, Ser (provided by Prof. E. Hafen, Institute of Molecular Systems Biology, University of Zurich, Zurich, Switzerland) were used to drive the expression of ATP7 in the pannier (pnr) domain of cuticular cells down the centre of the adult thorax and abdomen, as described previously(Norgate et al., 2006). To test for genetic interactions with ATP7, UAS-Mvl and Mvl^97f (Bloomington Stock # 5151, a viable P-element insertion mutation of Mvl) homozygous flies were crossed to pnr-Gal, UAS-ATP7/TM3,Sb flies and progeny examined. All flies were raised on standard Drosophila food medium containing approximately 5–10 μmol l^–1 Cu.

S2 cells were propagated in Drosophila serum free medium (SFM,Invitrogen) as previously reported(Southon et al., 2004). Cu was added to SFM as CuCl₂ at the concentration specified in the results. 10 μmol l^–1 Diamsar was added to SFM to deplete intracellular Cu as previously reported(Bellingham et al., 2004; Norgate et al., 2007). Mvl dsRNA was derived from Mvl cDNA bases 786–1251 and control dsRNA was derived from EYFP cDNA. dsRNA interference (dsRNAi) knockdown was conducted as previously reported (Southon et al., 2004; Worby et al.,2001). S2 cells maintaining stable overexpression of Mvl(55–60% of cells) were generated by co-transfecting pCoHygro with either pAcMvl or pAc empty vector control using Lipofectamine 2000 and cells were propagated in Schneider's Complete Medium (Invitrogen) with 10% foetal calf serum (Trace Scientific, Melbourne, Vic, Australia) supplemented with 300μg ml^–1 hygromycin B according to the manufacturer's instructions (Invitrogen). This medium was replaced with SFM for all experiments.

Total RNA for gene expression was extracted using the RNeasy kit (Qiagen,Doncaster, Victoria, Australia) and cDNA was transcribed using AMV reverse transcriptase (Promega, Annandale, NSW, Australia) as previously described(Southon et al., 2004). Primers for real-time PCR were designed using Primer3 software(Rozen and Skaletsky, 2000). Mvl forward and reverse primer sequences were: GCACCACCAGCAGATACTCA,and CCACAGCAAGACCCACAA, respectively. Actin 42A was used as a housekeeping gene and primer sequences have been described elsewhere(Southon et al., 2004). Real-time PCR was performed using the Rotor Gene 3000 (Corbett Research,Mortlake, NSW, Australia) using QuantiTect SYBR Green PCR Master Mix (Qiagen,Doncaster, Vic, Australia) and analysed as previously described(Southon et al., 2004).

Protein samples for western blotting were prepared from S2 cells with a lysis buffer containing: 1% Triton X-100, 20 mmol l^–1Tris–HCl (pH 7.5), 150 mmol l^–1 NaCl, 2 mmol l^–1 EDTA and 10% glycerol (Sigma, Castle Hill, NSW,Australia) with protease inhibitor cocktail added (Roche Diagnostics,Mannheim, Germany). Samples were lysed on ice for 2 h before clearing lysate by centrifugation. Protein (40 μg) was run on NuPAGE 4–12%Bis–Tris gels (Invitrogen), transferred to nitrocellulose membranes and probed with monoclonal mouse anti-FLAG antibody (1:2000, KM5–1C7 Walter and Eliza Hall Institute Biotechnology Centre, Melbourne, Vic, Australia) and polyclonal rabbit anti-mouse horseradish peroxidase coupled secondary antibody(1:7000, Dako, Glostrup, Denmark). Monoclonal mouse anti-α-tubulin was used as a loading control (Sigma).

Immunofluorescent detection of Mvl in S2 cells utilised a monoclonal mouse anti-FLAG antibody (1:200) and Alexa Fluor 488 anti-mouse secondary antibody(1:400, Invitrogen). DAPI (300 nmol l^–1; Invitrogen, Mount Waverly, Vic, Australia) was used to detect the nucleus. Images were recorded at 100× magnification using an Olympus FluoView 1000 confocal microscope with Olympus FluoView ver1.6a software (Olympus, Center Valley, PA, USA) and single sections are shown. An MtnA-EYFP reporter (provided by Prof. W. Schaffner, Institute of Molecular Biology, University of Zurich, Zurich,Switzerland) was used as a proxy measure of Cu distribution in dissected third instar larvae as previously described(Norgate et al., 2006; Selvaraj et al., 2005). Mvl^97f and w¹¹¹⁸ larval tissues were examined at 15× magnification (60× for higher magnification images) using the Olympus SZX12 dissecting microscope. Images were captured using the Olympus DP controller software using identical exposure settings(0.1 s exposure time). Twelve larvae (24 for basal food) were examined for each condition. Images of adult Drosophila were also recorded using this system. Representative images are shown.

⁶⁴Cu accumulation experiments were conducted as previously reported (Camakaris et al.,1995; Southon et al.,2004). Briefly cells were incubated with approximately 0.4 MBq ⁶⁴Cu (Australian Radioisotopes, Lucas Heights, NSW, Australia) and 2 μmol l^–1 non-radioactive Cu for 1 or 24 h. Cu accumulation was stopped by washing cells four times with 1 ml cold Hank's balanced salt solution containing 1 mmol l^–1 histidine(Sigma). Cells were then lysed in 100 μl of 0.1% SDS, containing 2 mmol l^–1 EDTA. Radioactivity was measured with a γ-counter(1282 CompuGamma, LKB Wallac, Turku, Finland) and Cu levels were standardised to total cellular protein, which was determined using Bio-Rad protein reagent according to the manufacturer's instructions (Bio-Rad, Gladesville, NSW,Australia). Cell viability following Cu chelation was determined by cell counting with a haemocytometer and Trypan Blue (Sigma) staining of dead cells.

Five replicates of 50 Drosophila where digested in 70%HNO₃ for 3 days at room temperature, with an additional 12 h at 60°C and then diluted to a final concentration of 10% HNO₃ with dH₂O. Metal levels were measured using a Vista-AX inductively coupled plasma atomic emission spectrometer (ICP-AES; Varian, Palo Alto, CA,USA). Five reading at 327.395 and 327.754 nmol l^–1 were averaged to give one value per biological replicate and expressed relative to the number of flies (ng/fly). A previous experiment showed no difference between normalisation of metal levels to either the number of flies or wet mass (data not shown).

Larval survival to adulthood was determined as reported previously(Norgate et al., 2007). Briefly, five replicates of 50 first instar larvae were scored for survival to adulthood on basal food (5–10 μmol l^–1 Cu) and 1 and 2 mmol l^–1 CuSO₄ (Sigma) as well as 100 μmol l^–1 bathocuproinedisulfonic acid (BCS; Sigma) to limit available Cu. Survival of Mvl^97f was compared to that of wild-type Armenia and w¹¹¹⁸ control Drosophila.

Statistical analysis was conducted using SPSS v11 (SPSS, Chicago, IL, USA). A one-sample Kolomogorov-Smirinov test was used to assess whether data was normally distributed. Statistical analyses are described in figure legends. P<0.05 was deemed statistically significant.

We utilised in vitro overexpression of Mvl as well as dsRNAi knockdown of endogenous Mvl in cultured S2 cells to demonstrate this gene encodes a functional Cu transporter. Mvlcontaining a C-terminal FLAG epitope tag (pAcMvl) was stably overexpressed in S2 cells. Western blotting using anti-FLAG detected a band of approximately 52 kDa corresponding to Mvl in lysate from pAcMvl cells(Fig. 1A), comparable to that seen with Mvl overexpression in Drosophila(Folwell et al., 2006). No significant signal was visible in lysate from cells expressing an empty vector control (pAc). Immunofluorescent confocal microscopy clearly demonstrated Mvl was localised to the plasma membrane of pAcMvl cells, with additional large punctate staining (Fig. 1B),whereas no significant staining was detected in pAc cells(Fig. 1C). Mvl localisation was not affected by either 24 h Cu limitation or 2.5 h exposure to 0.8 mmol l^–1 Cu (data not shown). The cellular localisation of Mvl is consistent with that seen in vivo with larval and adult Drosophila (Folwell et al.,2006) and is also similar to that of DMT1 in mammalian cells. DMT1 is localised to the apical membrane of mammalian intestinal and kidney cells where it is involved in Fe uptake and reabsorption, respectively(Canonne-Hergaux and Gros,2002; Knopfel et al.,2005), as well as in recycling endosomes to mediate Fe transport into the cytosol (Gruenheid et al.,1999; Lam-Yuk-Tseung et al.,2005). The punctate localisation of Mvl is likely to be a similar endosomal compartment.

We next measured Cu uptake using the radioisotope ⁶⁴Cu. S2 cells overexpressing Mvl or an empty vector control were exposed to 2μmol l^–1 Cu. Relative to control, overexpression of Mvl increased Cu levels by 9% and 35% after 1 h and 24 h exposure times, respectively (Fig. 2A). Conversely, dsRNAi knockdown of endogenous Mvl reduced Cu accumulation by 8% and 28% when cells were exposed to 2 μmol l^–1 Cu for 1 and 24 h, respectively(Fig. 2B). These results demonstrate Mvl is a functional Cu uptake protein in vitro, able to transport Cu across the plasma membrane under physiological conditions. To determine if Mvl-mediated Cu uptake is important for Cu homeostasis, we examined the viability of S2 cells following dsRNAi knockdown of Mvl. Cell viability was determined after 6 days following exposure to basal medium or Cu-limited medium using the Cu chelator Diamsar(Fig. 3). Viability under basal conditions was not affected by dsRNAi knockdown of Mvl. Control cells were sensitive to Cu limitation, with the number of viable cells reduced by 28% compared with cells maintained on basal medium. Cells were significantly more susceptible to Cu limitation following knockdown of Mvl, with viability reduced by 49%. Under Cu limiting conditions, relative to control cells, significantly fewer cells were viable when Mvl was knocked down. By contrast, dsRNAi knockdown of Mvl did not affect cell viability, relative to control, when cells were challenged with excess Cu for 48 h (data not shown).

The overexpression of Mvl in S2 cells significantly increased Cu levels, however, the effect was not as dramatic as that previously seen with Ctr1A and Ctr1B (Zhou et al., 2003), supporting the established view that the Cu-specific Ctr1 system is the primary Cu uptake mechanism. Nevertheless, impairment of Mvl function significantly decreases Cu uptake and reduces cell viability when Cu availability is low, whereas cell viability is unaffected when Cu levels are either adequate or elevated. This reduced Mvl-mediated Cu uptake suggests Ctr1A and Ctr1B are unable to elevate Cu uptake to compensate for reduction in Mvl activity. Similar results were seen when either Ctr1A or Ctr1B were suppressed in these cells(Southon et al., 2004)suggesting the Ctr1 and Mvl systems are independently regulated. Although speculative, the two systems may transport Cu to different cellular compartments. It is unclear if DMT1 can transport Cu to Atox1 as occurs with Ctr1 (Xiao and Wedd, 2002), or whether an alternative Cu chaperone is used for this non-Cu-specific transport system in mammals.

We sought to explore Mvl-mediated Cu transport in vivo utilising the homozygous viable, partial loss-of-function, Mvl^97fmutant flies (Rodrigues et al.,1995) as well as overexpression of Mvl with the UAS-Gal4 system. We examined viability, Cu uptake and distribution and markers of Cu-dependent enzyme activity.

Hypopigmentation seen with impairment of Ctr1B(Zhou et al., 2003) or with overexpression of ATP7 (Norgate et al., 2006) is indicative of a functional Cu deficiency and has previously been used to demonstrate in vivo Cu uptake and efflux,respectively, as tyrosinase is a Cu-dependent phenol oxidase involved in the production of biogenic amines needed for pigmentation(Wright, 1987). ATP7overexpression in the adult thorax also results in developmental disruption of the thorax, including absence of scutelar bristles and reduced thorax width(Norgate et al., 2006),presumably due to impaired activity of lysyl oxidase, the Cu-dependent enzyme involved the biogenesis of connective tissue matrices(Smith-Mungo and Kagan,1998).

We examined adult cuticle in homozygous Mvl^97f mutant flies under basal conditions, Cu-limited and Cu-excess conditions, but did not observe any significant change to pigmentation or cuticle morphology (data not shown). Similarly overexpression of Mvl in the pannierdomain under the control of the pnr-Gal4 driver did not have any observable effect (data not shown). These results demonstrate that impaired Mvl function does not cause a functional Cu deficiency and suggests other transporters, presumably Ctr1A or Ctr1B, are primarily responsible for Cu uptake into the developing epidermis.

To determine if any Mvl-mediated Cu transport occurs in the thorax, we next examined the effect of impaired Mvl function and Mvloverexpression when cuticle Cu levels are depleted by ATP7overexpression. Compared with w¹¹¹⁸ control flies(Fig. 4A), overexpression of ATP7 in the pannier domain(Fig. 4B) caused a severe Cu depletion, including hypopigmentation and reduced thoracic width (indicated by the distance between the two major posterior sensory bristles) as previously reported (Norgate et al.,2006). Overexpression of ATP7 in the pannierdomain in Mvl^97f/+ flies did not significantly affect pigmentation or thoracic development, relative to ATP7 overexpression in a wild-type background (Fig. 4C). By contrast, simultaneous overexpression of Mvl and ATP7 caused a more severe hypopigmentation phenotype and further reduction in thorax width, than ATP7 overexpression alone, indicative of a more severe functional Cu deficiency(Fig. 4D). These phenotypes were not affected by raising these flies on Cu-limited or Cu-excess food (data not shown).

The exacerbation of the functional Cu deficiency phenotype by co-overexpression of Mvl and ATP7 was unexpected, given that our previous studies found co-overexpression of Ctr1A and ATP7 almost completely rescued the hypopigmentation phenotype,despite no obvious pigmentation phenotype when Ctr1A was overexpressed alone (Norgate et al.,2006). These results are inconsistent with a simple model of Mvl acting at the plasma membrane to transport Cu into a tissue, and may be due to a more complex level of intracellular Cu regulation. In mammals, DMT1 not only functions at the plasma membrane but also transports Fe(Gruenheid et al., 1999; Lam-Yuk-Tseung et al., 2005)and Cd (Abouhamed et al., 2007)from an endosomal compartment into the cytosol. Therefore, it is reasonable to speculate that the punctate intracellular localisation of Mvl in S2 cells(Fig. 1) and Drosophila tissues (Folwell et al., 2006) represents an endosomal compartment and that Mvl can transport Cu from this organelle into the cytosol. It is possible that the overexpression of Mvl may worsen the hypopigmentation and cuticle phenotype through a redistribution of intracellular Cu, increasing transport of Cu out of an endosomal compartment and further restricting Cu delivery to Cu-dependent enzymes within the secretory pathway. The hypothesis that there are distinct pathways for Cu uptake and delivery to specific subcellular compartments and Cu-dependent proteins is supported by our in vitroresults, which show Ctr1A and Ctr1B are unable to maintain optimal Cu levels and prevent susceptibility to Cu limitation when Mvl is impaired(Fig. 3).

To assess the importance of Mvl to Cu homeostasis in live Drosophila we examined the viability of Mvl^97fflies in response to Cu limitation as well as Cu excess. First instar larvae were raised on basal food, on food containing the Cu chelator BCS, or on food with excess Cu, and survival to adulthood was determined for female(Fig. 5A) and male(Fig. 5B) flies separately. Compared to the control strains Armenia (wild-type) and w¹¹¹⁸,Mvl^97f males and females showed a strong sensitivity to excess Cu, with a significant reduction in viability at 1 mmol l^–1Cu and essentially no survival at 2 mmol l^–1 Cu. Under limiting Cu conditions Mvl^97f females again demonstrated reduced viability but, interestingly, Mvl^97f males were unaffected. Thus reduced Mvl function confers sensitivity to Cu excess and, in female flies, Cu limitation as well. It is unclear why male flies were unaffected by Cu limitation, but it is possible that Mvlexpression levels are gender specific. SEBIDA, the Drosophila sex bias database, cites seven different cDNA microarray studies that have compared male and female Drosophila melanogaster gene expression profiles and all have shown greater Mvl expression in male flies,with this difference statistically significant in four of the arrays(Gnad and Parsch, 2006). Given the Mvl^97f mutation is only a partial loss of function it is possible that male flies are able to produce sufficient Mvl protein for Cu uptake, whereas female flies are not.

The increased sensitivity of Mvl^97f flies to altered Cu conditions clearly demonstrates Mvl is an essential component of the Cu regulatory system, and is strikingly similar to that seen with Ctr1Bnull flies (Zhou et al.,2003). Absence of Ctr1B causes sensitivity to Cu limitation, suggestive of reduced transport from the midgut, as well as sensitivity to excess Cu due to reduced transport into a detoxification tissues such as the fat body or Malpighian tubules [analogous to the mammalian liver and kidneys, respectively (Cagan,2003; Sondergaard,1993)].

To explore why Mvl^97f flies are sensitive to Cu manipulation we used an MtnA-EYFP reporter as a marker for Cu distribution within third instar larvae. Excess Cu induces expression of the EYFP marker from the promoter of MtnA, and this technique has previously been used as a proxy measure of Cu distribution(Norgate et al., 2006; Selvaraj et al., 2005). We examined male and female larvae separately, but did not detect any differences between the sexes (data not shown). Mvl^97f/+/MtnA-EYFP larvae were dissected and compared to w¹¹¹⁸/MtnA-EYFP controls (Fig. 6). When larvae were raised on BCS to limit available Cu, EYFP could only be detected in the iron-cell region of the middle midgut and in the imaginal ring at the midgut–hindgut border of both control(Fig. 6A) and Mvl^97f/+ (Fig. 6B) larvae. The lack of EYFP expression in the remainder of the midgut and other tissues does not suggest that these tissues are not taking up Cu, but that all available Cu is being utilised. The EYFP detected in the iron-cell region and the imaginal ring suggests these were the only tissues where an excess of Cu was available to stimulate MtnA expression. Under basal conditions control larvae displayed significant EYFP in the proventriculus, gastric caecum, anterior midgut and Malpighian tubules(Fig. 6C,E). Importantly, Mvl^97f/+ larvae had dramatically reduced EYFP expression in the proventriculus and anterior midgut(Fig. 6D,F). We next exposed larvae to 1 mmol l^–1 Cu for 4 h prior to dissection to see if excess Cu altered this distribution. There was a dramatic increase in EYFP levels in all tissues including the posterior midgut, fatbody, hindgut and larval epidermis. However, there was no detectable difference between Mvl^97f/+ (Fig. 6H) and control (Fig. 6G) larvae. This saturation of EYFP in all tissues supports the hypothesis that Drosophila essentially take up all available Cu rather than risk a deficiency (Balamurugan et al., 2007).

We next raised Mvl^97f and w¹¹¹⁸control larvae under various Cu conditions and measured metal levels at pupation and adulthood by ICP-AES (Table 1). When raised on basal food, Fe levels were 40% lower in Mvl^97f pupae than controls. Cu levels were 7% lower,however, this was not statistically significant (P=0.09). Mn and Zn levels were unchanged. Raising larvae under Cu-limited conditions significantly reduced Cu levels in both control and Mvl^97fpupae. Cu levels were 10% lower in Mvl^97f pupae than control, however, again this was not statistically significant(P=0.10). Cu levels could not be assessed in Mvl^97f pupae following Cu exposure as viability was too severely reduced (Fig. 5). We therefore raised larvae under basal condition and exposed the adults to 6 mmol l^–1 excess Cu for 24 h. Cu levels were dramatically higher in Mvl^97f and w¹¹¹⁸ flies than that seen in pupae. Interestingly Mvl^97f flies also contained significantly more Mn and Zn than controls, and Fe levels were no longer lower in Mvl^97f, demonstrating a general elevation of metal levels. The elevated Cu seen in Mvl^97f flies exposed to short-term excess Cu is consistent with the hypothesis that reduced Cu transport into the fat body or Malpighian tubules leads to systemic accumulation and reduced viability. Unfortunately we could not detect any difference in MtnA-EYFP fluorescence between Mvl^97f/+ and control larvae at high Cu levels as the EYFP signal was saturated in all tissues because of a systemic upregulation of the Mtn Cu sequestration system.

The relatively small reduction in Cu levels seen in Mvl^97f pupae relative to controls, shows impaired Mvl-mediated Cu uptake does not result in a systemic Cu deficiency, presumably because of a compensatory Ctr1-mediated Cu uptake. The reduced MtnA-EYFP expression in the anterior midgut, a tissue known to express Mvl(Folwell et al., 2006), and the proventriculus of Mvl^97f larvae suggests that impaired Mvl-mediated Cu uptake in one or both of these regions is likely to be responsible for the reduced viability of female Drosophila under Cu limiting conditions. Ctr1A and Ctr1B are also expressed in the midgut (Chintapalli et al.,2007) and the loss of function of either of these Cu transporters confers sensitivity to Cu deficiency(Turski and Thiele, 2007; Zhou et al., 2003), suggesting these transporters are unable to fully compensate for the absence of another in this tissue.

These results demonstrate Malvolio is a physiologically important Cu transporter both in vitro and in vivo. The similarity between Mvl and Ctr1B mutant Drosophila with respect to their sensitivity to both Cu limitation and excess Cu, despite similar expression profiles, suggests each of these pathways are essential for optimal Cu uptake and distribution with only partial redundancy, and this is supported by in vitro experiments with dsRNAi knockdown in S2 cells. These results, together with the recent study redefining the role of human CTR1 (Zimnicka et al., 2007),suggest the role of DMT1 in mammalian dietary Cu uptake and distribution should be investigated further.

LIST OF ABBREVIATIONS

 
• BCS

  bathocuproinedisulfonic acid

 
• DMT1

  Divalent metal ion transporter 1

 
• dsRNAi

  dsRNA interference

 
• ICP-AES

  inductively coupled plasma atomic emission spectrophotometry

 
• Mtn

  metallothionein

 
• Mvl

  Malvolio

 
• SFM

  serum free medium

This work was supported by grants from the Australian Research Council and The Australian Institute of Nuclear Science and Engineering. The Olympus FluoView1000 confocal microscope was purchased with the assistance of the Rowden White Foundation. We thank Prof. Walter Schaffner for the gift of MtnA-EYFP Drosophila and Prof. Ernst Hafen for for the gift of Pnr-GAL4 Drosophila. We also thank Dr W. Scott Laidlaw for ICP-AES studies and Dr Ann Gaeth for assistance with confocal microscopy.