The core control system of intracellular iron homeostasis: a mathematical model.

Iron is a metal essential for cellular metabolism. However, excess iron available for reactions contributes to the formation of dangerous reactive oxygen species, such as the hydroxyl radical, via the Fenton reaction. Therefore, intracellular iron levels are tightly constrained by a control system of proteins. This paper contains a mathematical model, in the form of a system of five ordinary differential equations, of the core of this control system, including the labile iron pool as well as proteins that regulate uptake, storage, and export and are connected through negative feedback loops. The model is validated using data from an overexpression experiment with cultured human breast epithelial cells. The parameters in the mathematical model are not known for this particular cell culture system, so the analysis of the model was done for a generic choice of parameters. Through a mixture of analytical arguments and extensive simulations it is shown that for any choice of parameters the model reaches a unique stable steady state, thereby ruling out oscillatory behavior. It is shown furthermore that the model parameters are identifiable through suitable experiments.

Modulation of adipocyte differentiation by tumor necrosis factor and transforming growth factor beta.

Cultured TA1 adipocytes treated with tumor necrosis factor alpha (TNF) lose intracytoplasmic lipid and, over a period of days, come to resemble their predifferentiated progenitors (preadipocytes). To examine the extent to which this phenotypic reversion represents a return to a less differentiated cell, we examined three major characteristics that distinguish preadipocytes from adipocytes: (a) pattern of gene expression; (b) hormonal requirement for accelerated adipogenesis; and (c) pattern of protein synthesis. We found that within hours of TNF addition to adipocytes, mRNAs for genes whose expression is augmented during adipogenesis decreased to predifferentiated levels; in addition, like preadipocytes, TNF-treated adipocytes required exposure to hormones to accelerate adipogenesis. Further, the pattern of protein synthesis seen on polyacrylamide gels reverted to that seen before differentiation. Transforming growth factor-beta (TGF-beta) also caused a rapid decrease in expression of adipose genes when added to fully differentiated cells, an effect that was achieved by treatment with either TGF-beta 1 or TGF-beta 2. These effects were seen in the absence of a demonstrable proliferative response to either TNF or TGF-beta. Thus characteristics that define the "terminally" differentiated state in adipocytes are subject to modulation by environmental influences.

Coordinate transcriptional and translational regulation of ferritin in response to oxidative stress.

The global increase in transcription of cytoprotective genes induced in response to oxidative challenge has been termed the antioxidant response. Ferritin serves as the major iron-binding protein in nonhematopoietic tissues, limiting the catalytic availability of iron for participation in oxygen radical generation. Here we demonstrate that ferritin is a participant in the antioxidant response through a genetically defined electrophile response element (EpRE). The EpRE of ferritin H identified in this report exhibits sequence similarity to EpRE motifs found in antioxidant response genes such as those encoding NAD(P)H:quinone reductase, glutathione S-transferase, and heme oxygenase. However, the EpRE of ferritin H is unusual in structure, comprising two bidirectional motifs arranged in opposing directions on complementary DNA strands. In addition to EpRE-mediated transcriptional activation, we demonstrate that ferritin is subject to time-dependent translational control through regulation of iron-regulatory proteins (IRP). Although IRP-1 is initially activated to its RNA binding (ferritin-repressing) state by oxidants, it rapidly returns to its basal state. This permits the translation of newly synthesized ferritin transcripts and ultimately leads to increased levels of ferritin protein synthesis following oxidant exposure. Taken together, these results clarify the complex transcriptional and translational regulatory mechanisms that contribute to ferritin regulation in response to prooxidant stress and establish a role for ferritin in the antioxidant response.

FER-1, an enhancer of the ferritin H gene and a target of E1A-mediated transcriptional repression.

Ferritin, the major intracellular iron storage protein of eucaryotic cells, is regulated during inflammation and malignancy. We previously reported that transcription of the H subunit of ferritin (ferritin H) is negatively regulated by the adenovirus E1A oncogene in mouse NIH 3T3 fibroblasts (Y. Tsuji, E. Kwak, T. Saika, S. V. Torti, and F. M. Torti, J. Biol. Chem. 268:7270-7275, 1993). To elucidate the mechanism of transcriptional repression of the ferritin H gene by E1A, a series of deletions in the 5' flanking region of the mouse ferritin H gene were constructed, fused to the chloramphenicol acetyltransferase (CAT) gene, and transiently cotransfected into NIH 3T3 cells with an E1A expression plasmid. The results indicate that the E1A-responsive region is located approximately 4.1 kb 5' to the transcription initiation site of the ferritin H gene. Further analyses revealed that a 37-bp region, termed FER-1, is the target of E1A-mediated repression. This region also serves as an enhancer, augmenting ferritin H transcription independently of position and orientation. FER-1 was dissected into two component elements, i.e., a 22-bp dyad symmetry element and a 7-bp AP1-like sequence. Insertion of these DNA sequences into a ferritin H-CAT chimeric gene lacking an E1A-responsive region indicated that (i) the 22-bp dyad symmetry sequence by itself has no enhancer activity, (ii) the AP1-like sequence has moderate enhancer activity which is repressed by E1A, and (iii) the combination of the dyad symmetry element and the AP1-like sequence is required for maximal enhancer activity and repression by E1A. Gel retardation assays and cotransfection experiments with c-fos and c-jun expression vectors suggested that members of the Fos and Jun families bind to the AP1-like element of FER-1 and contribute to its regulation. In addition, gel retardation assays showed that E1A reduces the ability of nuclear proteins to bind to the AP1-like sequence without affecting the levels of nuclear factors that recognize the 22-bp dyad symmetry element. Taken together, these results demonstrate that FER-1 serves as both an enhancer of ferritin H transcription and a target for E1A-mediated repression.

Iron addiction: a novel therapeutic target in ovarian cancer.

Ovarian cancer is a lethal malignancy that has not seen a major therapeutic advance in over 30 years. We demonstrate that ovarian cancer exhibits a targetable alteration in iron metabolism. Ferroportin (FPN), the iron efflux pump, is decreased, and transferrin receptor (TFR1), the iron importer, is increased in tumor tissue from patients with high grade but not low grade serous ovarian cancer. A similar profile of decreased FPN and increased TFR1 is observed in a genetic model of ovarian cancer tumor-initiating cells (TICs). The net result of these changes is an accumulation of excess intracellular iron and an augmented dependence on iron for proliferation. A forced reduction in intracellular iron reduces the proliferation of ovarian cancer TICs in vitro, and inhibits both tumor growth and intraperitoneal dissemination of tumor cells in vivo. Mechanistic studies demonstrate that iron increases metastatic spread by facilitating invasion through expression of matrix metalloproteases and synthesis of interleukin 6 (IL-6). We show that the iron dependence of ovarian cancer TICs renders them exquisitely sensitive in vivo to agents that induce iron-dependent cell death (ferroptosis) as well as iron chelators, and thus creates a metabolic vulnerability that can be exploited therapeutically.

Effect of verapamil on doxorubicin cardiotoxicity: altered muscle gene expression in cultured neonatal rat cardiomyocytes.

Verapamil reverses multidrug resistance acquired by cancer cells during treatment with chemotherapeutic agents such as doxorubicin by inhibiting the function of P-glycoprotein. Verapamil has also been suggested to potentiate the cardiotoxicity of doxorubicin. We have recently demonstrated that selective inhibition of cardiac muscle gene expression is among the earliest events in doxorubicin cardiotoxicity. To explore the influence of verapamil on doxorubicin cardiotoxicity, we evaluated [14C]-doxorubicin accumulation, cardiac muscle gene expression by Northern blot analysis, and ultrastructural changes in cultured cardiomyocytes in the presence and absence of verapamil. Treatment with a combination of doxorubicin and verapamil for 24 h did not augment doxorubicin accumulation in cardiomyocytes, although substantial augmentation of doxorubicin accumulation by verapamil in cardiac fibroblasts was observed. Further, treatment with verapamil for 24 h did not augment the decrease in expression of muscle genes induced by doxorubicin (myosin light chain 2 slow, troponin I, M isoform creatine kinase). However, we found that verapamil reduced alpha-actin gene expression in a direct, doxorubicin-independent manner. Furthermore, the effect of doxorubicin plus verapamil on alpha-actin gene expression was additive over a wide range of doxorubicin and verapamil concentrations, resulting in a selective augmentation of doxorubicin-induced inhibition of gene expression for this single muscle protein gene. This was reflected in a substantial increase in cardiac myocyte damage when treatment with verapamil and doxorubicin was compared to treatment with doxorubicin alone by thin section electron microscopy. This suggests a possible mechanism by which verapamil may potentiate doxorubicin cardiotoxicity.

Effects of Anti-repulsive Guidance Molecule C (RGMc/Hemojuvelin) Antibody on Hepcidin and Iron in Mouse Liver and Tumor Xenografts.

OBJECTIVE: Hepcidin is a peptide hormone produced by the liver that regulates systemic iron homeostasis. Hepcidin is also synthesized by tumors, where it contributes to tumor growth by increasing the tumoral retention of iron. Targeted reduction of hepcidin may therefore be useful in reducing tumor growth. H5F9-AM8 is an antibody in preclinical development for the anemia of chronic disease that reduces hepcidin synthesis by binding to RGMc, a co-receptor involved in the transcriptional induction of hepcidin by BMP6. We explored the ability of H5F9-AM8 to act as an anti-tumor agent. METHODS: Effects of anti-hemojuvelin antibody on hepcidin synthesis were assessed by qRTPCR in tissue culture and in tumor xenografts and livers of mice treated with H5F9-AM8 or saline. Tumor growth was assessed using caliper measurements. Serum iron was measured colorimetrically and tissue iron was measured using western blotting and inductively coupled mass spectrometry. RESULTS: In tissue culture, the anti-hemojuvelin antibody H5F9-AM8 significantly reduced BMP6-stimulated hepcidin synthesis in HepG2 and other cancer cells. In mice, H5F9-AM8 reduced hepcidin in the liver and increased serum iron, total liver iron, and liver ferritin. Although hepcidin in tumors was also significantly decreased, H5F9-AM8 did not reduce tumor iron content, ferritin, or tumor growth. CONCLUSION: Anti-hemojuvelin antibody successfully reduces hepcidin in both tumors and livers but has different effects in these target organs: it reduces iron content and ferritin in the liver, but does not reduce iron content or ferritin in tumors, and does not inhibit tumor growth. These results suggest that despite their ability to induce hepcidin in tumors, the anti-tumor efficacy of systemic, non-targeted hepcidin antagonists may be limited by their ability to simultaneously elevate plasma iron. Tumor-specific hepcidin inhibitors may be required to overcome the limitations of drugs that target the synthesis of both systemic and tumor hepcidin.

Murine ferritin heavy chain: isolation and characterization of a functional gene.

A mouse liver genomic library screened with a full-length cDNA encoding murine ferritin heavy chain (mFHC) [Torti et al., J. Biol. Chem. 263 (1988) 12638-12644] yielded a functional genomic clone mFHC. The genomic clone isolated included a region of approximately 3 kb containing four exons and three introns. Sequence comparisons of the mouse genomic clone with other genomic clones from rat, human and chicken showed a high degree of similarity among species in the coding regions. Introns and flanking sequences were less conserved. However, comparison of mFHC promoter elements with FHC genes from other species revealed common elements. Analysis of the genomic structure of FHC suggested the presence of pseudogenes. S1 nuclease analysis, however, confirmed that this mouse clone, when transfected into human MRC-5 fibroblasts, was transcribed, indicating that this clone contains an FHC functional gene.

Adenovirus E1A blocks oxidant-dependent ferritin induction and sensitizes cells to pro-oxidant cytotoxicity.

Ferritin is a protein that oxidizes and sequesters intracellular iron in a mineral core. We have reported that the E1A oncogene selectively represses ferritin H transcription, resulting in reduced levels of the ferritin H protein. Here we demonstrate that cells respond to pro-oxidant challenge by inducing ferritin mRNA and protein, and that this response is completely blocked by E1A. Concordantly, E1A sensitized cells to the cytotoxic effects of oxidative stress and enhanced the accumulation of reactive oxygen species in response to pro-oxidant challenge. These results demonstrate that expression of E1A impedes the cellular response to oxidative stress, including the induction of ferritin.

N-acetyl-L-cysteine protects endothelial cells but not L929 tumor cells from tumor necrosis factor-alpha-mediated cytotoxicity.

The effect of N-acetyl-L-cysteine on the cytotoxicity of tumor necrosis factor-alpha was investigated in cultured bovine pulmonary artery endothelial cells and L929 mouse tumor cells. In endothelial cells, a 72-h incubation with tumor necrosis factor-alpha (100 ng/ml) reduced the number of viable cells to 27% of control. Simultaneous incubation with N-acetyl-L-cysteine (0.5-5 mmol/l) protected endothelial cells from tumor necrosis factor-alpha-mediated cytotoxicity and increased viability in a concentration-dependent fashion to 69% of control. Under the same conditions, a 72-h incubation with tumor necrosis factor-alpha (100 ng/ml) reduced the number of viable L929 tumor cells to 31% of control. However, this cytotoxic response remained unaltered in the presence of N-acetyl-L-cysteine (0.5-5 mmol/l). Similar results were obtained when using a lower concentration of tumor necrosis factor-alpha (50 ng/ml). These findings demonstrate protection from tumor necrosis factor-alpha-mediated toxicity by N-acetyl-L-cysteine in endothelial cells but not in a tumor cell line. It is concluded that N-acetyl-L-cysteine might serve as a therapeutic agent to limit the vascular toxicity of tumor necrosis factor-alpha without affecting its antineoplastic activity.

Curcumin: from ancient medicine to current clinical trials.

Curcumin is the active ingredient in the traditional herbal remedy and dietary spice turmeric (Curcuma longa). Curcumin has a surprisingly wide range of beneficial properties, including anti-inflammatory, antioxidant, chemopreventive and chemotherapeutic activity. The pleiotropic activities of curcumin derive from its complex chemistry as well as its ability to influence multiple signaling pathways, including survival pathways such as those regulated by NF-kappaB, Akt, and growth factors; cytoprotective pathways dependent on Nrf2; and metastatic and angiogenic pathways. Curcumin is a free radical scavenger and hydrogen donor, and exhibits both pro- and antioxidant activity. It also binds metals, particularly iron and copper, and can function as an iron chelator. Curcumin is remarkably non-toxic and exhibits limited bioavailability. Curcumin exhibits great promise as a therapeutic agent, and is currently in human clinical trials for a variety of conditions, including multiple myeloma, pancreatic cancer, myelodysplastic syndromes, colon cancer, psoriasis and Alzheimer's disease.

Human H-kininogen is a ferritin-binding protein.

H-kininogen is a multifunctional protein: it inhibits cysteine proteases, plays a role in contact activation of the coagulation cascade, and is the precursor of the potent proinflammatory peptide bradykinin. In the experiments described here, we identify H-kininogen as a ferritin-binding protein. Ferritin is a cellular and serum protein that is elevated in acute and chronic inflammation and many cancers. Despite numerous reports of ferritin-binding protein(s) in human serum, the nature and function of these proteins remain unclear. As a first step in characterizing the interaction between ferritin and its binding protein(s), we devised a ligand blot assay and used it to guide purification of a ferritin-binding protein from human serum. Edman degradation of the purified protein determined the sequence HNLGHGHK(H)ERDQGHG, a sequence with identity to residues 421-436 of human H-kininogen. These results were confirmed by demonstrating that commercially purified H-kininogen possessed ferritin binding activity and that ferritin binding could not be detected in plasma from kininogen-deficient individuals. Ligand blot assays mapped the ferritin binding domain to the light chain of H-kininogen chain, and revealed that both H and L recombinant ferritins possess H-kininogen binding activity. The unexpected identification of H-kininogen as a ferritin-binding protein may link ferritin in the complex chain of interactions by which H-kininogen mediates its multiple effects in contact activation and inflammation.

The role of iron in the cytotoxicity of tumor necrosis factor.

The effect of iron on the cytotoxicity of tumour necrosis factor-alpha (TNF, cachectin) was examined in L929 cells, which are killed by TNF at low concentrations. In L929 cells, the addition of iron (FeNTA) either prior to or concurrent with the addition of TNF markedly augmented the cytotoxicity of TNF over a wide range of TNF concentrations. Iron alone was not cytotoxic to the cells and did not inhibit protein synthesis. The iron chelator deferoxamine was able to protect against TNF cytotoxicity in L929 cells. Iron chelators also protected L929 cells from the cytotoxicity of TNF plus cycloheximide, suggesting that iron plays a role in this mode of TNF killing as well. TNF did not induce the synthesis of ferritin heavy chain in L929 cells. These experiments demonstrate that cellular iron status affects the ability of TNF to kill sensitive target cells.

Role of H and L subunits in mouse ferritin.

Ferritin is an iron-binding protein composed of two subunits, H and L. Twenty-four of these subunits assemble to form apoferritins whose subunit composition varies in a characteristic way in different tissues. Using recombinant proteins, we have assessed the role of H and L subunits in mouse ferritin function and compared these to human ferritin subunits. We report that mouse ferritin subunits exhibit considerable functional similarity to their human counterparts, including a prominent role of the H subunit in the facilitation of rapid iron uptake, and a key role of amino acid residues Glu-62 and His-65 in this process. In addition, amino acid residues important to assembly of the protein are conserved between mouse and human, permitting the formation of fully functional hybrid proteins containing both mouse and human subunits. However, murine and human ferritin H subunits also evidenced substantial functional differences; murine ferritin H showed a consistent reduction in iron uptake activity relative to human ferritin H. Creation of chimeric human/mouse ferritin H subunits by "helix swapping" mapped the domain of the protein critical to this activity difference to the DE helix. These findings suggest a novel functional role for carboxyl-terminal domains of the ferritin H subunit.

Recombinant ferritin: modulation of subunit stoichiometry in bacterial expression systems.

We describe a strategy for the creation of recombinant ferritin heteropolymers which mimic the natural heterogeneity of this protein. This method entailed the co-expression of cDNA for both ferritin H and ferritin L subunits in a single bacterium using either a bicistronic vector, in which both cDNAs were expressed from the vector, or a dual vector expression strategy, in which each subunit was expressed from a separate compatible plasmid in a single bacterial host. Electron microscopy and sucrose density gradient centrifugation demonstrated that ferritin assembled spontaneously in such bacteria to form catalytically active proteins of the expected size and shape. Isoelectric focusing revealed that protein isolated from any of these bacteria exhibited a restricted heterogeneity in subunit composition. Such multi-subunit recombinant ferritins spontaneously assembled in bacteria may be useful in further studies of ferritin assembly and function. Our results further suggest that varying expression levels is a simple way to alter levels of individual components within a multi-subunit recombinant protein, and that this approach may be of general utility in assessing the contribution of individual components to the function of multi-subunit proteins or protein complexes.

Activation of the ferritin H enhancer, FER-1, by the cooperative action of members of the AP1 and Sp1 transcription factor families.

We have previously reported that the adenovirus E1A oncogene represses the transcription of the H subunit of the mouse ferritin gene. Subsequent analyses defined FER-1, a 37-nucleotide sequence located 4.1 kilobases proximal to the start site of transcription, as the target of E1A-mediated transcriptional repression and as an enhancer of the ferritin H gene. FER-1 is composed of an AP1-like sequence followed by an element with dyad symmetry. To achieve maximal enhancer activity and transcriptional repression by E1A, both elements were essential. Using gel retardation assays, we now demonstrate that the binding complex for the AP1-like sequence of FER-1 contains JunD, FosB, and ATF1. Furthermore, JunD and FosB were able to activate FER-1 enhancer activity by transient cotransfection with ferritin H-chloramphenicol acetyltransferase reporter constructs. This augmented enhancer activity was inhibited by E1A. In addition, we have defined the minimal sequence in the dyad element of FER-1 required for protein interaction. This was determined to be a C-rich sequence to which Sp1 and Sp3 bind. Experiments with recombinant proteins indicate that members of both transcription factor families simultaneously bind FER-1. Taken together, these results elucidate molecular mechanisms involved in the transcriptional regulation of a pivotal gene in iron metabolism and provide insights into the contribution of the Sp1 family to the activation of AP1-dependent enhancers.

Iron-independent induction of ferritin H chain by tumor necrosis factor.

Iron increases the synthesis of the iron-storage protein, ferritin, largely by promoting translation of preexisting mRNAs for both the H and L ferritin isoforms (H, heavy, heart, acidic; L, light, liver, basic). We have recently cloned and sequenced a full-length cDNA to murine ferritin H and identified ferritin H as a gene induced by tumor necrosis factor alpha (TNF-alpha, cachectin). Using primary human myoblasts, we have now examined the relationship between TNF-alpha and iron in regulating ferritin. Four lines of evidence suggest that TNF-alpha regulates ferritin independently of iron. First, evaluation of mRNA showed that TNF-alpha increased ferritin H chain specifically, provoking no change in steady-state levels of ferritin L mRNA; iron, in contrast, increased the mRNA of both isoforms. Second, the increase in ferritin H protein synthesis observed during TNF-alpha treatment was dependent on an increase in ferritin H mRNA: actinomycin D blocked the TNF-alpha-induced changes in ferritin H but did not inhibit the translational induction of ferritin seen with iron treatment. Third, equal ferritin mRNA induction was observed in iron-loaded cells and in cells depleted of iron by a permeant chelator, 2,2'-dipyridyl. Fourth, ferritin H induction by TNF-alpha and iron was additive over the entire range of iron concentrations, even at TNF-alpha doses known to maximally stimulate ferritin H mRNA levels. Nonetheless, the role of iron in translational regulation of ferritin was retained in TNF-alpha-treated cells; effective biosynthesis of TNF-alpha-induced, H-subunit-predominant ferritin protein required iron and could be enhanced by treatment of the cells with additional iron or blocked by 2,2'-dipyridyl. Finally, we observed that the TNF-alpha-mediated increase in ferritin synthesis peaked at 8 hr and was followed by a decrease in both H and L isoferritin synthesis; the addition of iron, however, reversed the late-occurring depression in ferritin synthesis. This suggests that TNF-alpha-induced synthesis of H-rich ferritin may reduce the regulatory pool of intracellular iron, secondarily inhibiting iron-mediated translation of ferritin mRNA. We conclude that TNF-alpha acts independently of iron in its induction of ferritin H mRNA but requires the presence of iron for this effect to be fully expressed at the protein level.

Selective loss of CDC2 and CDK2 induction by tumor necrosis factor-alpha in senescent human diploid fibroblasts.

Normal human diploid fibroblasts undergo a finite number of doublings in culture. This process of senescence is accompanied by a loss in the ability to respond to proliferative stimuli and is therefore distinct from the quiescent state induced by nutrient deprivation. We have studied changes in gene expression induced in these cells following exposure to the cytokine, tumor necrosis factor-alpha (TNF). We observed that TNF induced CDC2 and CDK2 expression in early-passage quiescent WI-38 fibroblasts. However, as cells approached senescence, their ability to induce CDC2 and CDK2, as well as stimulate DNA synthesis in response to TNF, progressively declined, with minimal to absent induction in senescent cells. This occurred despite the TNF-dependent induction of such proliferation-independent genes as manganese superoxide dismutase and interleukin-6 in senescent and quiescent cells. Serum was similarly unable to induce CDC2 or CDK2 expression in senescent cells. These results demonstrate that senescent cells are selectively deficient in TNF-mediated induction of CDC2 and CDK2, genes crucial to DNA synthesis and mitosis.

Interleukin 1 induces ferritin heavy chain in human muscle cells.

Interleukin 1 alpha (IL-1) and tumor necrosis factor alpha (TNF) are two monokines which play a prominent role in the response to inflammation and injury. We recently observed that TNF leads to an increase in the synthesis of the heavy chain of ferritin, suggesting that TNF may be involved in iron homeostasis (Torti et al. (1988) J. Biol. Chem. 263, 12638-12644). The experiments reported here demonstrate that in cultured human muscle cells, IL-1 induces ferritin H mRNA and protein as effectively as TNF. TNF and IL-1 were additive in their effects on ferritin H expression, and IL-1 induction of ferritin H was not blocked by anti-TNF antibodies. Ferritin H induction was a specific response not observed with beta or gamma interferon, nor with transforming growth factor beta. Both differentiated myotubes as well as myoblasts responded to IL-1 with the induction of ferritin H. These results suggest that monokine-mediated alterations in the subunit composition of the ferritin molecule may be of biological relevance in the response to inflammation and injury.