Chloroplast genes are expressed during intracellular symbiotic association of Vaucheria litorea plastids with the sea slug Elysia chlorotica.

The marine slug Elysia chlorotica (Gould) forms an intracellular symbiosis with photosynthetically active chloroplasts from the chromophytic alga Vaucheria litorea (C. Agardh). This symbiotic association was characterized over a period of 8 months during which E. chlorotica was deprived of V. litorea but provided with light and CO2. The fine structure of the symbiotic chloroplasts remained intact in E. chlorotica even after 8 months of starvation as revealed by electron microscopy. Southern blot analysis of total DNA from E. chlorotica indicated that algal genes, i.e., rbcL, rbcS, psaB, psbA, and 16S rRNA are present in the animal. These genes are typically localized to the plastid genome in higher plants and algae except rbcS, which is nuclear-encoded in higher plants and green (chlorophyll a/b) algae. Our analysis suggests, however, that similar to the few other chromophytes (chlorophyll a/c) examined, rbcS is chloroplast encoded in V. litorea. Levels of psbA transcripts remained constant in E. chlorotica starved for 2 and 3 months and then gradually declined over the next 5 months corresponding with senescence of the animal in culture and in nature. The RNA synthesis inhibitor 6-methylpurine reduced the accumulation of psbA transcripts confirming active transcription. In contrast to psbA, levels of 16S rRNA transcripts remained constant throughout the starvation period. The levels of the photosystem II proteins, D1 and CP43, were high at 2 and 4 months of starvation and remained constant at a lower steady-state level after 6 months. In contrast, D2 protein levels, although high at 2 and 4 months, were very low at all other periods of starvation. At 8 months, de novo synthesis of several thylakoid membrane-enriched proteins, including D1, still occurred. To our knowledge, these results represent the first molecular evidence for active transcription and translation of algal chloroplast genes in an animal host and are discussed in relation to the endosymbiotic theory of eukaryote origins.

Rad51 expression and localization in B cells carrying out class switch recombination.

Rad51 is a highly conserved eukaryotic homolog of the prokaryotic recombination protein RecA, which has been shown to function in both recombinational repair of DNA damage and meiotic recombination in yeast. In primary murine B cells cultured with lipopolysaccharide (LPS) to stimulate heavy chain class switch recombination, Rad51 protein levels are dramatically induced. Immunofluorescent microscopy shows that anti-Rad51 antibodies stain foci that are localized within the nuclei of switching B cells. Immunohistochemical analysis of splenic sections shows that clusters of cells that stain brightly with anti-Rad51 antibodies are evident within several days after primary immunization and that Rad51 staining in vivo is confined to B cells that are switching from expression of IgM to IgG antibodies. Following switch recombination, B cells populate splenic germinal centers, where somatic hypermutation and clonal proliferation occur. Germinal center B cells are not stained by anti-Rad51 antibodies. Rad51 expression is therefore not coincident with somatic hypermutation, nor does Rad51 expression correlate simply with cell proliferation. These data suggest that Rad51, or a highly related member of the conserved RecA family, may function in class switch recombination.

Molecular characterization of a FKBP-type immunophilin from higher plants.

Immunophilins are intracellular receptors for the immunosuppressants cyclosporin A, FK506, and rapamycin. In addition to their use in organ transplantation, these natural products have been used to investigate signaling pathways in yeast, plant, and mammalian cells. We have recently described the identification of an immunosuppressant-sensitive signaling pathway in and the purification of several immunophilins from Vicia faba plants. We now report the molecular characterization of a 15 kDa FK506- and rapamycin-binding protein from V. faba (VfFKBP15). The amino acid sequence deduced from the cDNA starts with a signal peptide of 22 hydrophobic amino acids. The core region of VfFKBP15 is most similar to yeast and mammalian FKBP13 localized in the endoplasmic reticulum (ER). In addition, VfFKBP15 has a carboxyl-terminal sequence that is ended with SSEL, a putative ER retention signal. These findings suggest that VfFKBP15 is a functional homolog of FKBP13 from other organisms. Interestingly, two distinct cDNAs corresponding to two isoforms of FKBP15 have been cloned from Arabidopsis and also identified from rice data base, suggesting that pFKBP15 (plant FKBP15) is encoded by a small gene family in plants. This adds to the diversity of plant FKBP members even with the same subcellular localization and is in contrast with the situation in mammalian and yeast systems in which only one FKBP13 gene has been found. Like the mammalian and yeast FKBP13, the recombinant VfFKBP15 protein has rotamase activity that is inhibited by both FK506 and rapamycin with a Ki value of 30 nM and 0.9 nM, respectively, illustrating that VfFKBP15 binds rapamycin in preference over FK506. The mRNA of VfFKBP15 is ubiquitously expressed in various plant tissues including leaves, stems, and roots, consistent with the ER localization of the protein. Levels of VfFKBP15 mRNA are elevated by heat shock, suggesting a possible role for this FKBP member under stress conditions.

Selection for genes encoding secreted proteins and receptors.

Extracellular proteins play an essential role in the formation, differentiation, and maintenance of multicellular organisms. Despite that, the systematic identification of genes encoding these proteins has not been possible. We describe here a highly efficient method to isolate genes encoding secreted and membrane-bound proteins by using a single-step selection in yeast. Application of this method, termed signal peptide selection, to various tissues yielded 559 clones that appear to encode known or novel extracellular proteins. These include members of the transforming growth factor and epidermal growth factor protein families, endocrine hormones, tyrosine kinase receptors, serine/threonine kinase receptors, seven transmembrane receptors, cell adhesion molecules, extracellular matrix proteins, plasma proteins, and ion channels. The eventual identification of most, or all, extracellular signaling molecules will advance our understanding of fundamental biological processes and our ability to intervene in disease states.

Expression of the CD36 homolog (FAT) in fibroblast cells: effects on fatty acid transport.

An adipocyte membrane glycoprotein, (FAT), homologous to human CD36, has been previously implicated in the binding/transport of long-chain fatty acids. It bound reactive derivatives of long-chain fatty acids and binding was specific and associated with significant inhibition of fatty acid uptake. Tissue distribution of the protein and regulation of its expression were also consistent with its postulated role. In this report, we have examined the effects of FAT expression on rates and properties of fatty acid uptake by Ob17PY fibroblasts lacking the protein. Three clones (P21, P22, and P25) were selected based on FAT mRNA and protein levels. Cell surface labeling could be demonstrated with the anti-CD36 antibody FITC-OKM5. In line with this, the major fraction of immunoreactive FAT was associated with the plasma membrane fraction. Assays of oleate and/or palmitate uptake demonstrated higher rates in the three FAT-expressing clones, compared to cells transfected with the empty vector. Clone P21, which had the highest protein levels on Western blots, exhibited the largest increase in transport rates. Fatty acid uptake in FAT-expressing P21 cells reflected two components, a phloretin-sensitive high-affinity saturable component with a Km of 0.004 microM and a basal phloretin-insensitive component that was a linear function of unbound fatty acid. P21 cells incorporated more exogenous fatty acid into phospholipids, indicating that binding of fatty acids was followed by their transfer into the cell and that both processes were increased by FAT expression. The data support the interpretation that FAT/CD36 functions as a high-affinity membrane receptor/transporter for long-chain fatty acids.

In vivo anergized CD4+ T cells express perturbed AP-1 and NF-kappa B transcription factors.

Anergy is a major mechanism to ensure antigen-specific tolerance in T lymphocytes in the adult. In vivo, anergy has mainly been studied at the cellular level. In this study, we used the T-cell-activating superantigen staphylococcal enterotoxin A (SEA) to investigate molecular mechanisms of T-lymphocyte anergy in vivo. Injection of SEA to adult mice activates CD4+ T cells expressing certain T-cell receptor (TCR) variable region beta-chain families and induces strong and rapid production of interleukin 2 (IL-2). In contrast, repeated injections of SEA cause CD4+ T-cell deletion and anergy in the remaining CD4+ T cells, characterized by reduced expression of IL-2 at mRNA and protein levels. We analyzed expression of AP-1, NF-kappa B, NF-AT, and octamer binding transcription factors, which are known to be involved in the regulation of IL-2 gene promoter activity. Large amounts of AP-1 and NF-kappa B and significant quantities of NF-AT were induced in SEA-activated CD4+ spleen T cells, whereas Oct-1 and Oct-2 DNA binding activity was similar in both resting and activated T cells. In contrast, anergic CD4+ T cells contained severely reduced levels of AP-1 and Fos/Jun-containing NF-AT complexes but expressed significant amounts of NF-kappa B and Oct binding proteins after SEA stimulation. Resolution of the NF-kappa B complex demonstrated predominant expression of p50-p65 heterodimers in activated CD4+ T cells, while anergic cells mainly expressed the transcriptionally inactive p50 homodimer. These alterations of transcription factors are likely to be responsible for repression of IL-2 in anergic T cells.

Redirecting the specificity of ubiquitination by modifying ubiquitin-conjugating enzymes.

Depletion of specific cellular proteins is a powerful tool in biological research and has many medical and agricultural benefits. In contrast to genetic methods currently available to attenuate protein levels, we describe an alternative approach that redirects the ubiquitin-dependent proteolytic pathway to facilitate specific proteolytic removal. Degradation via the ubiquitin pathway requires the prior attachment of multiple ubiquitins to the target protein. This attachment is accomplished, in part, by a family of enzymes designated E2s (or ubiquitin-conjugating enzymes), some of which use domains near their C termini for target recognition. Here, we demonstrate that E2 target recognition can be redefined by engineering E2s to contain appropriate protein-binding peptides fused to their C termini. In five dissimilar examples, chimeric E2s were created that recognized and ubiquitinated their respective binding partners with high specificity. We also show that ubiquitination of one protein targeted by this method led to its ATP-dependent degradation in vitro. Thus, by exploiting interacting domains derived from natural and synthetic ligands, it may be possible to design E2s capable of directing the selective removal of many intracellular proteins.

Implications physiopathologiques de la Nestine lors du remodelage pulmonaire et cardiaque à la suite de l’infarctus du myocarde, du diabète et de l’hypertension pulmonaire

Il est reconnu que la protéine filamenteuse intermédiaire Nestine est exprimée lors du processus de cicatrisation et du remodelage fibrotique. De plus, nous avons identifié l’expression de la Nestine au sein de deux populations distinctes qui sont directement impliquées dans les réponses de fibroses réparative et réactive. Ainsi, une population de cellules souches neurales progénitrices résidentes du coeur de rat adulte exprime la Nestine et a été identifiée à titre de substrat de l’angiogenèse et de la neurogenèse cardiaque. Également, la Nestine est exprimée par les myofibroblastes cicatriciels cardiaques et il a été établi que la protéine filamenteuse intermédiaire joue un rôle dans la prolifération de ces cellules. Ainsi, l’objectif général de cette thèse était de mieux comprendre les évènements cellulaires impliqués dans la réponse neurogénique des cellules souches neurales progénitrices résidentes cardiaques Nestine(+) (CSNPRCN(+)) lors de la fibrose réparative cardiaque et d’explorer si l’apparition de fibroblastes Nestine(+) est associée avec la réponse de fibrose réactive secondaire du remodelage pulmonaire. Une première publication nous a permis d’établir qu’il existe une régulation à la hausse de l’expression de la GAP43 (growth associated protein 43) et que cet événement transitoire précède l’acquisition d’un phénotype neuronal par les CSNPRCN(+) lors du processus de cicatrisation cardiaque chez le rat ayant subi un infarctus du myocarde. De plus, la surimposition de la condition diabétique de type 1, via l’injection unique de Streptozotocine chez le rat, abolit la réponse neurogénique des CSNPRCN(+), qui est normalement induite à la suite de l’ischémie cardiaque ou de l’administration de 6-hydroxydopamine. Le second article a démontré que le développement aigu de la fibrose pulmonaire secondaire de l’infarctus du myocarde chez le rat est associé avec une augmentation de l’expression protéique de la Nestine et de l’apparition de myofibroblastes pulmonaires Nestine(+). Également, le traitement de fibroblastes pulmonaires avec des facteurs de croissances peptidiques pro-fibrotiques a augmenté l’expression de la Nestine par ces cellules. Enfin, le développement initial de la condition diabétique de type 1 chez le rat est associé avec une absence de fibrose réactive pulmonaire et à une réduction significative des niveaux protéiques et d’ARN messager de la Nestine pulmonaire. Finalement, la troisième étude représentait quant à elle un prolongement de la deuxième étude et a alors examiné le remodelage pulmonaire chronique chez un modèle établi d’hypertension pulmonaire. Ainsi, les poumons de rats adultes mâles soumis à l’hypoxie hypobarique durant 3 semaines présentent un remodelage vasculaire, une fibrose réactive et une augmentation des niveaux d’ARN messager et de la protéine Nestine. De plus, nos résultats ont démontré que la Nestine, plutôt que l’alpha-actine du muscle lisse, est un marqueur plus approprié des diverses populations de fibroblastes pulmonaires activés. Également, nos données suggèrent que les fibroblastes pulmonaires activés proviendraient en partie de fibroblastes résidents, ainsi que des processus de transition épithélio-mésenchymateuse et de transition endothélio-mésenchymateuse. Collectivement, ces études ont démontré que des populations distinctes de cellules Nestine(+) jouent un rôle majeur dans la fibrose réparative cardiaque et la fibrose réactive pulmonaire.

Étude du polymorphisme rs3846662 de l’HMGCR dans le cerveau et le système périphérique

Dans cette thèse, l’impact du polymorphisme rs3846662 sur l’épissage alternatif de la 3-hydroxy-3-méthylglutaryl coenzyme A réductase (HMGCR) a été investigué in vivo, chez des patients atteints d’hypercholestérolémie familiale (HF) ou de maladie d’Alzheimer (MA). Le premier manuscrit adresse la problématique de la normalisation de la quantification relative des ARNm par PCR quantitative. Les découvertes présentées dans ce manuscrit nous ont permis de déterminer avec un haut niveau de confiance les gènes de référence à utiliser pour la quantification relative des niveaux d’ARNm de l’HMGCR dans des échantillons de sang (troisième manuscrit) et de tissus cérébraux post-mortem (quatrième manuscrit). Dans le deuxième manuscrit, nous démontrons grâce à l’emploi de trois cohortes de patients distinctes, soit la population canadienne française du Québec et les deux populations nord américaines « Alzheimer’s Disease Cooperative Study (ADCS) » et « Alzheimer’s Disease Neuroimaging Initiative (ADNI) », que le génotype AA au locus rs3846662 confère à ces porteurs une protection considérable contre la MA. Les femmes porteuses de ce génotype voient leur risque de MA diminuer de près de 50% et l’âge d’apparition de leurs premiers symptômes retarder de 3.6 ans. Les porteurs de l’allèle à risque APOE4 voient pour leur part leurs niveaux de plaques séniles et dégénérescences neurofibrillaires diminuer significativement en présence du génotype AA. Enfin, les individus atteints de déficit cognitif léger et porteurs à la fois de l’allèle APOE4 et du génotype protecteur AA voient leur risque de convertir vers la MA chuter de 76 à 27%. Dans le troisième manuscrit, nous constatons que les individus atteints d’HF et porteurs du génotype AA ont, contrairement au modèle établi chez les gens normaux, des niveaux plus élevés de cholestérol total et de LDL-C avant traitement comparativement aux porteurs de l’allèle G. Le fait que cette association n’est observée que chez les non porteurs de l’APOE4 et que les femmes porteuses du génotype AA présentent à la fois une augmentation des niveaux d’ARNm totaux et une résistance aux traitements par statines, nous indique que ce génotype influencerait non seulement l’épissage alternatif, mais également la transcription de l’HMGCR. Comme une revue exhaustive de la littérature ne révèle aucune étude abondant dans ce sens, nos résultats suggèrent l’existence de joueurs encore inconnus qui viennent influencer la relation entre le génotype AA, l’épissage alternatif et les niveaux d’ARNm de l’HMGCR. Dans le quatrième manuscrit, l’absence d’associations entre le génotype AA et les niveaux d’ARNm Δ13 ou de protéines HMGCR nous suggère fortement que ce polymorphisme est non fonctionnel dans le SNC affecté par la MA. Une étude approfondie de la littérature nous a permis d’étayer cette hypothèse puisque les niveaux de HNRNPA1, la ribonucléoprotéine influencée par l’allèle au locus rs3846662, sont considérablement réduits dans la MA et le vieillissement. Il est donc proposé que les effets protecteurs contre la MA associés au génotype AA soient le résultat d’une action indirecte sur le processus physiopathologique.

Impaired renal endothelial nitric oxide synthase and reticulocyte production as modulators of hypertension induced by recombinant human erythropoietin in the rat

INTRODUCTION AND AIMS: Hypertension is a common side effect of recombinant human erythropoietin (rHuEPO) therapy; however, the exact pathways remain to be elucidated. The discovery of non-hematopoietic actions of rHuEPO increased the number of patients that could putatively benefit from this therapy; however, to achieve those effects higher doses are usually needed, which increase the risk and incidence of adverse events. Our aim was to study the effect of a broad range of rHuEPO doses on hematological and biochemical parameters, blood pressure and renal function and damage in the rat, focusing on endothelial nitric oxide synthase (eNOS) and hypoxia-inducible factors (HIFs). METHODS: Male Wistar rats were divided in 5 groups receiving different doses of rHuEPO (100, 200, 400 and 600 IU/kg body weight (BW)/week) and saline solution (control), during 3 weeks. Blood and 24h urine were collected to perform hematological and biochemical analysis. Blood pressure (BP) was measured by the tail-cuff method. The kidney tissue was collected to mRNA and protein expression assays and to characterize renal lesions. RESULTS: A dose-dependent increase in red blood cells count, hematocrit and hemoglobin levels was found with rHuEPO therapy, in rHuEPO200, rHuEPO400 and rHuEPO600 groups. Increased reticulocyte count was found in the rHuEPO400 and rHuEPO600 groups. BP raised in all groups receiving rHuEPO. The rHuEPO200 and rHuEPO600 groups presented increased kidney protein levels of HIF2α and a reduction in kidney protein levels of eNOS, along with the highest grade of vascular and tubular renal lesions. CONCLUSIONS: Our study showed that rHuEPO-induced hypertension might involve indirect (hematological) and direct (renal) effects which varies according to the dose used. Thus, rHuEPO therapy should be performed rationally and under adequate surveillance, as hypertension develops even with lower doses. Especial caution with higher doses should be taken, as rHuEPO-induced hypertension leads to early renal damage without alterations in traditional markers of renal function, thus masking the serious adverse effects and risks.

Acute cadmium chloride administration induces hepatic and renal cytochrome P450 2A5 in the mouse

Modulation of the cytochrome P450 (CYP) monooxygenase system by cadmium was investigated in male, adult DBA/2J mice treated with a single dose (16 Amol/kg body weight, i.p.) of cadmium chloride (CdCl2) at various time points. The total CYP content of kidney microsomes started to decrease 4 hours earlier than in the liver (P < 0.05), with maximal decreases at 24 hours of 56% and 85% in the liver and kidney, respectively. In contrast, both hepatic and renal coumarin 7-hydroxylase (COH) activity (indicative of CYP2A5 activity) relative to total CYP content started to progressively increase at 8 hours, with renal activity 61 times higher than the hepatic activity. Maximum increases were observed, 15-fold in the liver and 64-fold in the kidney after 24 hours. Liver and kidney CYP2A5 mRNA levels increased maximally 12 and 4 hours after treatment, respectively and decreased to almost half 6 hours later. In contrast, kidney and liver CYP2A5 protein levels increased maximally at 18 and 24 hours. This study demonstrates that hepatic and renal CYP2A5 is upregulated by cadmium with a faster response in the kidney than in the liver. This observation is concordant with the fact that kidney is the target organ for cadmium toxicity. The observed increase in the mRNA but not in protein levels after maximal induction suggests involvement of post-transcriptional mechanisms in the regulation of CYP2A5 expression by cadmium.

Short-term peaks in glucose promote renal fibrogenesis independently of total glucose exposure

Postprandial hyperglycemia is implicated as a risk factor predisposing to vascular complications. This study was designed to assess recurrent short-term increases in glucose on markers of renal fibrogenesis. Human renal cortical fibroblasts were exposed to fluctuating short-term (2 h) increases to 15 mM D-glucose, three times a day over 72 h, on a background of 5 mM D-glucose. To determine whether observed changes were due to fluctuating osmolality, identical experiments were undertaken with cells exposed to L-glucose. Parallel experiments were performed in cells exposed to 5 mM D-glucose and constant exposure to either 15 or 7.5 mM D-glucose. Fluctuating D-glucose increased extracellular matrix, as measured by proline incorporation ( P < 0.05), collagen IV ( P < 0.005), and fibronectin production ( P < 0.001), in association with increased tissue inhibitor of matrix metalloproteinase (MMP) ( P < 0.05). Sustained exposure to 15 mM D-glucose increased fibronectin ( P < 0.001), in association with increased MMP-2 ( P = 0.01) and MMP-9 activity ( P < 0.05), suggestive of a protective effect on collagen matrix accumulation. Transforming growth factor-beta(1) (TGF-beta(1)) mRNA was increased after short-term (90 min) exposure to 15 mM glucose (P < 0.05) and after 24-h exposure to 7.5 mM ? ( P < 0.05). Normalization of TGF-beta(1) secretion occurred within 48 h of constant exposure to an elevated glucose. Fluctuating L-glucose also induced TGF-beta(1) mRNA and a profibrotic profile, however, to a lesser extent than observed with exposure to fluctuating D-glucose. The results suggest that exposure to fluctuating glucose concentrations increases renal interstitial fibrosis compared with stable elevations in D-glucose. The effects are, in part, due to the inherent osmotic changes.

RNA interference against human papillomavirus oncogenes in cervical cancer cells results in increased sensitivity to cisplatin

Targeted inhibition of oncogenes in tumor cells is a rational approach toward the development of cancer therapies based on RNA interference (RNAi). Tumors caused by human papillomavirus (HPV) infection are an ideal model system for RNAi-based cancer therapies because the oncogenes that cause cervical cancer, E6 and E7, are expressed only in cancerous cells. We investigated whether targeting HPV E6 and E7 oncogenes yields cancer cells more sensitive to chemotherapy by cisplatin, the chemotherapeutic agent currently used for the treatment of advanced cervical cancer. We have designed siRNAs directed against the HPV E6 oncogene that simultaneously targets both E6 and E7, which results in an 80% reduction in E7 protein and reactivation of the p53 pathway. The loss of E6 and E7 resulted in a reduction in cellular viability concurrent with the induction of cellular senescence. Interference was specific in that no effect on HPV-negative cells was observed. We demonstrate that RNAi against E6 and E7 oncogenes enhances the chemotherapeutic effect of cisplatin in HeLa cells. The IC50 for HeLa cells treated with cisplatin was 9.4 mu M, but after the addition of a lentivirus-delivered shRNA against E6, the IC50 was reduced almost 4-fold to 2.4 mu M. We also observed a decrease in E7 expression with a concurrent increase in p53 protein levels upon cotreatment with shRNA and cisplatin over that seen with individual treatment alone. Our results provide strong evidence that loss of E6 and E7 results in increased sensitivity to cisplatin, probably because of increased p53 levels.

Roles of heterogeneous nuclear ribonucleoproteins A and B in cell proliferation

Overexpression of heterogeneous nuclear ribonucleoproteins (hnRNPs) A2 and B1 has been observed in a variety of tumour types, however, it is unknown whether this dysregulation is a consequence of, or a driving force for, unregulated cell proliferation. We have shown that the levels of hnRNPs A1, A2 and B1, but not A3, are modulated during the cell cycle of Colo16 squamous carcinoma cells and HaCaT immortalized keratinocytes, suggesting that A1, A2 and B1 are needed at particular cell cycle stages. However, the levels of hnRNP A1, A2 and B1 mRNAs were constant, indicating that regulation of protein levels was controlled at the level of translation. RNAi suppression of hnRNP At or A3 alone did not affect the proliferation of Colo16 cells but the proliferation rate was significantly reduced when both were suppressed simultaneously, or when either was suppressed together with hnRNP A2. Reducing hnRNP A2 expression in Colo16 and HaCaT cells by RNAi led to a non-apoptotic-related decrease in cell proliferation, reinforcing the view that this protein is required for cell proliferation. Suppression of hnRNP A2 in Colo16 cells was associated with increased p21 levels but p53 levels remained unchanged. In addition, expression of BRCA1 was downregulated, at both mRNA and protein levels. The observed effects of hnRNP A2 and its isoforms on cell proliferation and their correlation with BRCA1 and p21 expression suggest that these hnRNP proteins play a role in cell proliferation.

Decreased hephaestin activity in the intestine of copper-deficient mice causes systemic iron deficiency

Copper and iron metabolism intersect in mammals. Copper deficiency simultaneously leads to decreased iron levels in some tissues and iron deficiency anemia, whereas it results in iron overload in other tissues such as the intestine and liver. The copper requirement of the multicopper ferroxidases hephaestin and ceruloplasmin likely explains this link between copper and iron homeostasis in mammals. We investigated the effect of in vivo and in vitro copper deficiency on hephaestin (Heph) expression and activity. C57BL/6J mice were separated into 2 groups on the day of parturition. One group was fed a copper-deficient diet and another was fed a control diet for 6 wk. Copper-deficient mice had significantly lower hephaestin and ceruloplasmin (~50% of controls) ferroxidase activity. Liver hepcidin expression was significantly downregulated by copper deficiency (~60% of controls), and enterocyte mRNA and protein levels of ferroportin1 were increased to 2.5 and 10 times, respectively, relative to controls, by copper deficiency, indicating a systemic iron deficiency in the copper-deficient mice. Interestingly, hephaestin protein levels were significantly decreased to ~40% of control, suggesting that decreased enterocyte copper content leads to decreased hephaestin synthesis and/or stability. We also examined the effect of copper deficiency on hephaestin in vitro in the HT29 cell line and found dramatically decreased hephaestin synthesis and activity. Both in vivo and in vitro studies indicate that copper is required for the proper processing and/or stability of hephaestin.