Role of Hypolipidemic Drug Clofibrate in Altering Iron Regulatory Proteins IRP1 and IRP2 Activities and Hepatic Iron Metabolism in Rats Fed a Low-Iron Diet

(1)

Role of Hypolipidemic Drug Clofibrate in Altering Iron Regulatory

Proteins IRP1 and IRP2 Activities and Hepatic Iron Metabolism

in Rats Fed a Low-Iron Diet

Hui-Ling Huang and Ning-Sing Shaw

Department of Agricultural Chemistry, National Taiwan University, Taipei, Taiwan

Received September 10, 2001; accepted January 23, 2002

Role of Hypolipidemic Drug Clofibrate in Altering Iron Regu-latory Proteins IRP1 and IRP2 Activities and Hepatic Iron Me-tabolism in Rats Fed a Low-Iron Diet. Huang, H.-L., and Shaw, N.-S. (2002). Toxicol. Appl. Pharmacol. 180, 118 –128.

In addition to reducing the expression of transferrin and ceru-loplasmin genes, hypolipidemic peroxisome proliferators may al-ter iron homeostasis in the liver. Therefore, this study investigates the effects of clofibrate on proteins related to liver iron metabolism in a rat model using a 2ⴛ 2 experimental design: two dose levels of clofibrate in diet (0 and 0.5%) and two dietary iron levels (35 ppm as normal level and 15 ppm as low-iron diet). Twenty-four Wistar rats were assigned to the four diets and fed for 6 weeks. Subsequent measurements of iron parameters in the blood and the liver indicated that, in addition to mild anemia and the reduction in serum iron and total iron-binding capacity, clofibrate treatment altered IRP1 and IRP2 activities differentially and increased mitochondrial aconitase both at activity and protein levels. At both normal and low-iron intakes, clofibrate caused a 50% reduc-tion in serum iron and TIBC with a corresponding reducreduc-tion in transferrin mRNA. The RNA-binding activities of IRP1 were selectively activated by clofibrate treatment even though liver iron concentration was not depleted. The RNA-binding activity of IRP2 was selectively activated by the low iron intake and corre-lated with an increase of transferrin receptor mRNA, while clofibrate treatment offset the effects of the low iron intake. © 2002 Elsevier Science (USA)

Key Words: clofibrate; iron regulatory proteins; mitochondrial

aconitase; transferrin receptor; rats.

Fibrates such as clofibrate, fenofibrate, and gemfibrozil are

an established class of drugs for treating hypertriglyceridemia

and combined hyperlipidemia. Fibrates are peroxisome

prolif-erators (PPs), which cause hepatomegaly, proliferation of

per-oxisomes, and induction of many enzymes involved in

␤-oxi-dation and

␻-oxidation of fatty acids in rodents (Schoonjans et

al., 1996). Although incapable of inducing peroxisome

prolif-eration in humans, fibrates alter the expression of enzymes

regulating serum cholesterol and lipid homeostasis (Auwerx,

1992; Schoonjans et al., 1996).

PPs also caused transcriptional repression in the liver. Genes

down-regulated by PPs include those related to lipid

metabo-lism, such as apoE, apoAI, apo AIV, and apo CIII (Schoonjans

et al., 1996), and those not obviously involved in lipid

ho-meostasis, such as transferrin (Tf) (Hertz et al., 1996),

cerulo-plasmin (Cp), and P-type ATPase 7B (ATP7B) (Eagon et al.,

1999). Transferrin is highly expressed in the adult mammalian

liver and is secreted by hepatocytes into the serum, where it

functions as an iron-transport protein (Crichton and

Charlo-teaux-Wauters, 1987). A lack of transferrin causes iron

accu-mulation in the liver (Hayashi et al., 1993). Meanwhile,

ceru-loplasmin is a copper-containing serum protein exhibiting

ferroxidase activity, which accelerates the incorporation of iron

into apotransferrin (Harris et al., 1999). Aceruloplasminemic

subjects (Harris, 1998) and Cp-null mice (Harris et al., 1999)

have marked iron accumulation in hepatocytes and

reticuloen-dothelial cells.

At the cellular level, iron storage and uptake are

coordi-nately regulated posttranscriptionally by interactions between

cytoplasmic factors, iron regulatory proteins 1 and 2 (IRP1 and

IRP2), and the cis-regulating iron-responsive element (IRE)

located in the 3

⬘-untranslated region (UTR) of transferrin

re-ceptor (TfR) mRNA as well as in the 5

⬘-UTRs of mRNAs for

ferritin, mitochondrial aconitase (m-ACO), and

erythroid-spe-cific 5-aminolevulinic acid synthase (Hentz and Ku¨hn, 1996).

When cellular iron becomes limiting, IRP1 is recruited into the

high-affinity binding state, and IRP2 is synthesized in active

form. The binding of IRPs to the IRE in the 5

⬘-UTR represses

the translation of ferritin and m-ACO, while an association of

IRPs with IREs in the 3

⬘-UTR stabilizes the TfR mRNA

transcript. On the other hand, the expansion of the labile iron

pool causes IRP1 inactivation and IRP2 degradation to result in

an efficient translation of ferritin and m-ACO mRNA and rapid

degradation of TfR mRNA.

Powanda et al. (1978) were probably the first to describe the

effects of clofibrate on iron and copper metabolism; serum Tf

concentration and Cp ferroxidase activity were both reduced

after 1 week of treatment, while the concentration of hepatic

iron increased, but no copper accumulation occurred.

Mean-while, Hertz et al. (1996) demonstrated that other fibrate

com-118

(2)

pounds caused a 50% reduction of serum iron and total

iron-binding capacity (TIBC), corresponding to a reduced Tf

mRNA level in rat as well as in a hepatoma cell line. PPs also

caused a reduction of liver Cp and ATP7B mRNA and serum

Cp activity (Anderson et al., 1999; Eagon et al., 1999). In

contrast to the outcome of 1 week of treatment (Powanda et al.,

1978), clofibrate caused a 3.3-fold increase of liver copper, but

did not affect liver iron after a 60-day exposure (Eagon et al.,

1999).

Adverse outcomes associated with fibrate treatment include

hepatotoxicity and anemia. Abnormal liver function tests

includ-ing elevation of serum glutamate oxaloacetate transaminase

(GOT) and glutamate pyruvate transaminase (GPT) have occurred

during fibrate administration (Sirtori et al., 1992). Increases to

more than three times the upper limit of normal values have been

recorded in controlled multiple-dose trials of fenofibrate

(Schre-fer, 2001). As for anemia, mild to moderate reduction of

hemo-globin and hematocrit has been documented. In The Coronary

Drug Project (The Coronary Drug Project Research Group, 1975),

the clofibrate treatment group recorded a significant lowering of

hematocrit (Hct) values (from 46 to 44.8%) compared to the

placebo group (from 46.2 to 45.9%) and an elevated prevalence of

anemia indicated by abnormal hematocrit values (Hct

⬍ 38%, 7.9

vs 4.1%; Hct

⬍ 36%, 3.4 vs 1.7%) after 5 years of follow-up.

Therefore, we hypothesized that the onset of PP-induced anemia

may depend on dietary iron level and that PP may alter hepatic

iron homeostasis. Consequently, we investigated the effect of

clofibrate on iron status, hepatic IRP activities, and the expression

of hepatic ferritin, TfR, and m-ACO in rats fed a semisynthetic

low-iron diet.

MATERIALS AND METHODS

Chemicals. The following chemicals were used: Clofibrate (ethyl 2-(4-chlorophenoxy)-2-methylpropionate) (Fluka Chemie, Germany), Serum iron/ UIBC kit and Serum Copper kit (RANDOX, Antrim, UK), horseradish per-oxidase-conjugated goat anti-rabbit IgG (Vectastain, Vector Laboratories), PVDF transfer membrane (PolyScreen, NEN Life Science, Boston, MA), ␣-[32

P]CTP (3000 Ci/mmol or 111 TBq/mmol, NEN–DuPont, Boston, MA), positively charged nylon membrane for hybridization, GeneScreen Plus (NEN Life Science), X-OMAT AR film and Bio-Max film (Kodak), reagents for in vitro transcription, DNA labeling kit, CTP4A Western blotting kit, ECL Western blotting detection system, and Micro Spin G-50 column (Amersham Pharmarcia Biotech Inc.). Other chemicals were obtained from Sigma Chem-ical Co. (St. Louis, MO) and Merck (Taiwan).

Animals and treatment. Twenty-four weanling, male Wistar rats (Labo-ratory Animal Center, College of Medicine, National Taiwan University), weighing 148⫾ 10 g, were housed individually in stainless-steel cages with wire-mesh floors in a temperature-, humidity-, and light-controlled room. Animal care and handling conformed to the NSC Guide for the Care and Use of Laboratory Animals (National Science Council, 1993). The basal diet was AIN-76 formula (AIN, 1977) modified to contain 15% soybean oil (Table 1). Rats (six each group) were assigned to the four diets with treatment of clofibrate (0 and 0.5%) and dietary iron levels (35 and 15 ppm) in 2⫻ 2 experimental design. The 35 ppm Fe is the normal dietary level for rats, while the 15 ppm Fe is a suboptimal level that will deplete liver Fe store and may cause mild anemia (Chen et al., 1997). Food and deionized water were freely available. After 42 days of dietary treatment, rats were killed by carbon dioxide

asphyxiation, and blood and liver were collected for analysis. Serum was separated by low-speed centrifugation and was frozen at⫺70°C until analyzed. Blood analyses. Hemoglobin concentration was determined by the cyan-methemoglobin method using Drabkin’s solution (Oser, 1965). Serum iron and TIBC were determined using the serum iron/UIBC kit. Serum copper was determined using the Serum Copper kit. Serum ceruloplasmin was assessed as ferroxidase activity by measuring the oxidation of o-diansidine dihydrochlo-ride, and one unit of activity was defined as oxidation of 1␮mol substrate per minute (Schosinsky et al., 1974).

Liver Fe and Cu. Fractions of liver were digested with nitric acid and hydrogen peroxide in a microwave digestion oven (Milestone, Microwave Laboratory System, Italy). Iron and copper were measured by atomic absorp-tion spectrometry (Hitachi 180-30). The steady concentraabsorp-tion of desferriox-amine (DFO)-chelatable iron was measured by electron paramagnetic reso-nance spectroscopy (EPR) as described by Cairo et al. (1995) and Yegorov et al. (1993). Briefly, 0.05 ml of 10 mM desferrioxamine mesylate was added to 0.5 ml of liver homogenate (50% w/v in 20 mM Tris–HCI and 135 mM KCl, pH 7.4), incubated at room temperature for 20 min, and then placed in a quartz tube, frozen in liquid nitrogen, and stored at ⫺70°C before analysis. EPR spectra were recorded on a Brucker 200 spectrometer (NTHU Instrument Center, National Tsing Hua University, Hsin-chu, Taiwan) at liquid nitrogen temperature under the following conditions: Klystron frequency, 9.12 GHz; power, 20 mW; modulation amplitude, 2.0 mT. The intensity of the EPR signal at g⫽ 4.3 was measured to estimate DFO-chelatable Fe concentration. A calibration plot was obtained by adding to a series of control homogenate incremental volumes of a freshly prepared FeSO4solution of predetermined concentration, and the homogenate was treated as described above.

Preparation of cellular fractions. Liver cytosol, mitochondria, and mi-crosome were obtained by differential centrifugation modified from Chen et al. (1997). Liver homogenate (25% w/v) was prepared by homogenizing a fraction of liver in ice-cold HDSC buffer (50 mM Hepes, pH 7.4, 1 mM dithiothreitol, 0.25 M sucrose, 2 mM trisodium citrate, 0.5 mg/L leupeptin and 0.2 mM

TABLE 1

Compositions of Semisynthetic Normal and Low-Iron Dietsa Containing Clofibrate Ingredients of diets Composition (%) No clofibrate 0.5% Clofibrate 35 ppm Fe 15 ppm Fe 35 ppm Fe 15 ppm Fe Casein 20 20 20 20 Corn starch 55 55 55 55 Soybean oil 15 15 15 15 Vitamin mixture AIN-76a 1 1 1 1 Mineral mixturea (iron omitted) 3.5 3.5 3.5 3.5 Cellulose 5 5 5 5 Choline 0.2 0.2 0.2 0.2 Methionine 0.3 0.3 0.3 0.3 Clofibrateb — — 0.5 0.5 Total amount 100 100 100 100 Iron contentc (ppm) 35 15 35 15 a

Based on AIN-76 formula (AIN, 1977), in which the 35 ppm Fe is normal dietary level for rats, while the 15 ppm Fe is a suboptimal level that will deplete liver Fe store and may cause mild anemia.

b

Ethyl 2-(4-chlorophenoxy)-2-methylpropionate (Fluka Chemie).

c

(3)

phenylmethylsulfonyl fluoride) using a Potter–Elvehiem homogenizer. The mitochondria pellet was obtained from the postnuclear supernatant by centrif-ugation at 15,000g for 30 min, washed three times to reduce cytosolic con-tamination, resuspended in HDSC, and stored at⫺70°C. The postmitochon-drial supernatant was further centrifuged at 105,000g for 1 h (Ultra-centrifuge, 50.4 Ti Rotor, Beckman). The supernatant was designated as the cytosol fraction and the pellet was designated as the microsomes. The microsome pellet was suspended in 0.05 M phosphate buffer containing 1 mM EDTA at pH 7.7, stored at⫺70°C, and used for CYP4A analysis.

Electrophoretic mobility shift assay of IRPs activity. IRPs–IRE binding activity was determined by gel retardation analysis using a32_P-labeled IRE-containing RNA probe of the first 73 nucleotides from ratL-ferritin 5⬘-UTR in the absence or presence of 2-mercaptoethanol (2-ME) (Chen et al., 1997; Schalinske and Eisenstein, 1996). RatL-ferritin cDNA (kindly provided by Dr. R. S. Eisenstein, University of Wisconsin, Madison, WI) was linearized by SmaI digestion and used as template.32_{P-labeled IRE probe was produced by} in vitro transcription using␣-[32_{P]CTP, Riboprobe Germini II Core System} (Promega) and T7 RNA polymerase (Boeringer Mannheim, Germany) and then purified by a 10% acrylamide/8 M urea gel. The specific radioactivity of the probe was estimated to be about 4700 – 4900 cpm/fmol.

For the binding assay, 7.5␮g of cytosol protein (measured with the Lowry method; Lowry et al., 1951) was incubated with a saturating level of 32 P-labeled IRE probe (final concentration at 1 nmol/L) in the presence of reaction buffer (10 mM Hepes, pH 7.6, 3 mM MgCl2, 40 mM KCl, 1 mM DTT, and 5% glycerol) in a final volume of 20␮l for 10 min at room temperature and then 2␮l heparin (5 mg/mL) was added and incubated for another 5 min (Barton et al., 1990). When measuring total IRP1 in the cytosol, 1.5␮g of cytosol protein was used in the binding reaction and incubation was carried out in the presence of 2-ME.

The IRP-bound probe was separated from the free probe on a 4% nonde-naturing polyacrylamide gel (60:1 acrylamide/bisacrylamide) in 0.5⫻ Tris borate–EDTA buffer and electrophoresed (Hofer SE-600, with a cooling system) at 300 V for 1.5 h at 5°C. The gel was then dried on filter paper (Gel drier Model 283, Bio-Rad), and an autoradiogram on X-OMAT AR film was prepared. The bands corresponding to IRE-bound IRP1 and IRP2 were cut out and counted in a liquid scintillation counter (Beckman, LS 5000 CE).

Western blot analysis of cytochrome P450 4A, ferritin, and mitochondrial aconitase. For the analysis of CYP 4A, the microsomal fraction containing 5 ␮g of protein was resolved on a 12% SDS–polyacrylamide gel, transferred to PVDF membranes, detected by a commercial CYP4A Western blotting kit, and visualized with an ECL Western blotting detection system. For the analysis of ferritin, liver cytosol was heated at 75°C for 10 min and then was centrifuged at 10,000g for 20 min, and protein in the supernatant was resolved on a 15% SDS–polyacrylamide gel, transferred to PVDF membranes, detected with rabbit anti-rat ferritin polyclonal antibodies (Shaw and Chen, 1993) and horseradish peroxidase-conjugated goat anti-rabbit IgG, and the immunoreac-tive, 19 –21-kDa protein was visualized with an enchanced chemiluminescence of ECL detection regents (Amersham Pharmarcia Biotech Inc.). For the anal-ysis of m-ACO, the mitochondrial fraction containing 10␮g of protein was

resolved on a 10% SDS–polyacrylamide gel, transferred to PVDF membranes, detected with rabbit anti-human m-ACO polyclonal antibody (kindly provided by Dr. H.-H. Zhuang, Memorial University of Chang Gung) and alkaline phosphatase-conjugated goat anti-rabbit IgG (Tropix), and visualized with an enhanced chemiluminescence substrate CDP-star (Tropix). The results of the Western blotting analyses were recorded on Bio-Max films and quantified by densitometry (Microcomputer image device-M4 V3.0).

RNA preparation and Northern blot analysis of transferrin and trans-ferrin receptor. Total RNA was extracted from liver using TRIZOL reagent (GIBCO-BRL, Grand Island, NY). All the cDNAs used as probes were obtained by cloning RT–PCR products of total RNA prepared from the liver of control rats, and their identities were confirmed by sequencing. The synthetic oligonucleotides used to amplify respective cDNA probe are listed in Table 2. The cDNA probes was prepared using a DNA labeling kit and purified by a Micro Spin G-50 column. Ten and 40␮g of total liver RNA was used for Tf and TfR analysis, respectively. RNA was separated on a formaldehyde– agarose gel (1%) and transferred to a nylon membrane. Following UV/cross-linking (Hoefer), membrane was hybridized to32_{P-labeled probes (⬎5 ⫻ 10}8 cpm/␮g DNA) using ␤-actin as an internal standard. All the results were recorded on X-OMAT AR films and quantified by densitometry.

Aconitase enzyme assay. Activity of m-ACO was determined by the coupled reaction of aconitase (EC 4.2.1.3) and isocitrate dehydrogenase (EC 1.1.142) (Rose and O’Connell, 1967). One unit of enzyme activity was defined as the reduction of 1␮mol NADP⫹per minute.

Statistical analysis. The main effects of dietary iron level and clofibrate were evaluated by two-way ANOVA. The differences among the four groups were analyzed by Duncan’s Multiple Range test. The statistical significance level was set at p⬍ 0.05. All statistics were carried out on the SAS System (Version 6.12, SAS Institute, Cary, NC).

RESULTS

Growth Repression, Hepatomegaly, and Liver Enzymes

Clofibrate treatment reduced body weight and weight gain in

the normal iron group by 21 and 30%, respectively, and in the

low-iron group by 14 and 21%, respectively (Table 3). Growth

suppression by clofibrate treatment was greater at normal iron

intakes than at low iron intakes. Since food intakes were

reduced by less than 10%, the suppression occurred mainly due

to reduced feed efficiency, indicating inefficient energy

utili-zation.

Clofibrate-treated rats displayed marked hepatomegaly

(Ta-ble 4). Liver weights increased 10 and 30% and relative liver

weights increased 40 and 50% at normal and low iron intakes,

respectively. The hepatomegaly in terms of final liver weight

TABLE 2

Oligonucleotides for Reverse Transcriptase–PCR Cloning

Protein

Sequence of 5⬘ sense (1) and 5⬘ antisense (2) strands

Probe size

(bp) Sequence no.

GenBank

access no. Reference

Transferrin (1) 5⬘ TggCT CaggA ACACT TTggc 658 893–1550 D38380 Hoshino et al. (1996) (2) 5⬘ ggAgC ACAgC CTTgA CTgAA

Transferrin receptor (1) 5⬘ ggTAA ACTgg TCCAT gcTAA T 451 275–725 M58040 Roberts and Griswold (1990) (2) 5⬘ CATTG TTCAC AgTgA gcTTC

␤-Actin (1) 5⬘ gTggg CCgCT CTAgg CACCA 540 103–642 55574 Nudel et al. (1983)

(4)

was comparable between the two clofibrate-treated groups,

regardless of iron levels. While maintained within normal

ranges, serum GOT and GPT were significantly elevated by

clofibrate treatment, and this phenomenon was aggravated by

low iron intake. As expected, clofibrate treatment induced

significant expression of CYP 4A in microsomes (Fig. 1),

which was not compromised by the low iron intake.

Indices of Iron Status

Both clofibrate treatment and low iron intake caused

reduc-tion in hemoglobin concentrareduc-tion (significant main effects at

p

⬍ 0.05, Table 5). Low iron intake or clofibrate alone

reduced hemoglobin by 8%, while the decrease in hemoglobin

in the clofibrate groups was not significantly influenced by

dietary iron levels. In contrast, clofibrate, but not iron intake,

significantly affected serum iron and TIBC, both of which were

reduced by 50% regardless of dietary iron levels. Meanwhile,

transferrin saturation remained normal for all four groups and

was not influenced by clofibrate or iron intakes (Table 5).

Hepatic ferritin (Fig. 2) and iron (Table 6) levels changed with

dietary iron intakes and were significantly lower at low iron

intakes. Clofibrate treatment caused total liver iron to increase by

25 and 86% in the normal and low-iron groups, respectively

(Table 6). Meanwhile, clofibrate caused an 114 and 35% increase

of DFO-chelatable iron at normal and low iron intakes,

respec-tively; while low iron intake caused a 54 and 27% decrease with

and without clofibrate treatment, respectively. At normal iron

intake, the clofibrate-treated rats exhibited the highest

concentra-tion of DFO-chelatable iron, which was 2.2 times that of the

untreated counterpart. Furthermore, the clofibrate treatment

caused a twofold increase in m-ACO activity at both normal and

low iron intakes (Table 6). The expression of m-ACO protein

increased with clofibrate treatment and was significantly higher at

low iron intakes (Fig. 3).

Transferrin and Transferrin Receptor

Clofibrate treatment significantly reduced hepatic Tf mRNA

level by approximately 50% at both normal and low iron

TABLE 3

Growth and Feed Efficiency of Rats as Affected by Low Iron Intake and Clofibrate Treatment

Treatment and Fe groupa

No clofibrateb _{0.5% Clofibrate}b

p value of two-way ANOVA

35 ppm Fe 15 ppm Fe 35 ppm Fe 15 ppm Fe Fe Clofibrate

Final body weight (g/rat) 417⫾ 411 ₃₇₈_{⫾ 24}2 ₃₃₀_{⫾ 15}2 ₃₂₅_{⫾ 22}2 _0.43 _0.0001

Body weight gain (g/day/rat) 6.7⫾ 11 _5.8_{⫾ 0.7}2 _4.7_{⫾ 0.3}3 _4.5_{⫾ 0.5}3 _0.043 _0.0001 Food intake (g/rat/day) 19.1⫾ 2.31 _17.3_{⫾ 1.3}1,2 _17.4_{⫾ 1.4}2 _16.1_{⫾ 1.5}2 _0.26 _0.047 Feed efficiencyc,d _0.35_{⫾ 0.01}1 _0.33_{⫾ 0.02}1 _0.27_{⫾ 0.01}2 _0.27_{⫾ 0.01}2 _0.17 _0.0001

a_{The 35 ppm Fe is normal dietary level for rats, while the 15 ppm Fe is a suboptimal level that will deplete liver Fe store and may cause mild anemia.} b_{Each value represents group mean}_{⫾ SD of six rats. Different superscript numbers in each row indicate significant difference by Duncan’s Multiple Range}

test at p⬍ 0.05.

c_{Feed efficiency}_{⫽ body weight gain (g/day/rat) ⫼ feed intake (g/day/rat).} d_{Log-transformed data were analyzed.}

TABLE 4

Liver Weights and Liver Function Indicators in Rats as Affected by Low Iron Intake and Clofibrate Treatment

Treatment and Fe groupa No clofibrateb _{0.5% Clofibrate}b p value of two-way ANOVA 35 ppm Fe 15 ppm Fe 35 ppm Fe 15 ppm Fe Fe Clofibrate Liver weight (g)c _23.8_{⫾ 2.4}1 _20.4_{⫾ 1.4}2 _26.2_{⫾ 0.9}3 _26.9_{⫾ 1.0}3 _0.06 _0.0001 Relative liver weight

(% body weight) 5.7⫾ 0.21 _5.4_{⫾ 0.2}1 _7.9_{⫾ 0.3}2 _8.3_{⫾ 0.5}2 _0.24 _0.0001

GOTd_(U/L) ₅₈_{⫾ 8}1 ₆₄_{⫾ 5}1,2 ₆₁_{⫾ 6}1,2 ₇₃_{⫾ 6}2 _0.04 _0.0238

GPTd_(U/L) ₂₅_{⫾ 4}1 ₂₇_{⫾ 5}1 ₂₈_{⫾ 4}1 ₃₇_{⫾ 9}2 _0.04 _0.0065

a_{The 35 ppm Fe is normal dietary level for rats, while the 15 ppm Fe is a suboptimal level that will deplete liver Fe store and may cause mild anemia.} b_{Each value represents sample mean}_{⫾ SD of six rats. Different superscript numbers in each row indicate significant difference by Duncan’s Multiple Range}

test at p⬍ 0.05.

c_{Iron and clofibrate interaction was statistically significant by two-way ANOVA at p}_{⬍ 0.05.} d_{Log-transformed data were analyzed.}

(5)

intakes (Fig. 4A). Both dietary iron intake and clofibrate

sig-nificantly affected hepatic TfR mRNA level; which was

re-duced by clofibrate, while increased by low iron intake.

With-out clofibrate, low iron intake caused TfR mRNA to increase

by twofold, but this effect disappeared with clofibrate

treat-ment (Fig. 4B).

Hepatic IRP1 and IRP2 Activities

The main effects of iron intake or clofibrate treatment on

IRP1 activity were not significant (Fig. 5A). Nevertheless, the

clofibrate-treated group had significantly higher IRP1 activity

than the untreated counterpart at normal iron intake (Fig. 5B).

Without clofibrate treatment, IRP2 activity was significantly

enhanced by low iron intake, but clofibrate treatment offset this

effect (Fig. 5C).

Serum and Hepatic Copper

Clofibrate caused serum copper to decrease by 30 and 18%

at normal and low iron intakes, respectively. Serum

ceruloplas-min ferroxidase activity decreased with serum copper levels

(Table 7). In contrast, clofibrate caused hepatic copper to

increase by more than twofold (Table 7). The low iron intake

did not affect serum ceruloplasmin and copper levels, but it

increased hepatic copper accumulation by approximately 69%

with or without clofibrate treatment (Table 7).

DISCUSSION

Previous studies (Anderson et al., 1999; Eagon et al., 1999;

Hertz et al., 1996; Powanda et al., 1978) have demonstrated

FIG. 1. A representative gel electrophoresis and immunoblotting of mi-crosomal CYP4A in the rat liver as affected by a low-iron diet and clofibrate treatment. The 35 ppm Fe is normal dietary level for rats, while the 15 ppm Fe is a suboptimal level which will deplete liver Fe store and may cause mild anemia. CYP4A abundance was determined by Western blot analysis using 5 ␮ g microsomal protein. Each lane represents one rat and results of six rats were quantified by densitometry and statistically analyzed with two-way ANOVA (significant level set at p⬍ 0.05).

TABLE 5

Hematological Indices of Iron Status of Rats as Affected by Low Iron Intake and Clofibrate Treatment

No clofibrateb _{0.5% Clofibrate}b _{p value of two-way ANOVA}

35 ppm Fe 15 ppm Fe 35 ppm Fe 15 ppm Fe Fe Clofibrate

Hemoglobin (mmol/L) 2.4⫾ 0.11 _2.2_{⫾ 0.1}2 _2.2_{⫾ 0.1}2 _2.1_{⫾ 0.1}2 _0.0295 _0.0062

Serum Fe (␮ mol/L) 42⫾ 31 ₄₂_{⫾ 6}1 ₂₁_{⫾ 4}2 ₂₂_{⫾ 2}2 _0.76 _0.0001

TIBC (␮ mol/L) 98⫾ 71 ₁₀₃_{⫾ 6}1 ₅₂_{⫾ 8}2 ₅₅_{⫾ 8}2 _0.60 _0.0001

Transferrin saturation (%) 43⫾ 4 41⫾ 5 40⫾ 9 42⫾ 9 0.95 0.79

a_{The 35 ppm Fe is normal dietary level for rats, while the 15 ppm Fe is a suboptimal level that will deplete liver Fe store and may cause mild anemia.} b_{Each value is sample mean}_{⫾ SD of six rats. Different superscript numbers in each row indicate significant difference by Duncan’s Multiple Range test at}

p⬍ 0.05.

FIG. 2. Ferritin in the rat liver as affected by a low-iron diet and clofibrate treatment. The 35 ppm Fe is normal dietary level for rats, while the 15 ppm Fe is a suboptimal level which will deplete liver Fe store and may cause mild anemia. Ferritin was determined by Western blot analysis using 70 ␮ g cytosolic protein (A) and results were quantified by densitometry (B). In A, lane f represents rat ferritin purchased from Sigma Co. In B, the values are mean⫾ SD of six rats and bars not sharing the same superscript letters are significantly different from one another by Duncan’s Multiple Range test ( p⬍ 0.05).

(6)

that PPs suppress the expression of transferrin, ceruloplasmin,

and ATP7B and reduce the concentrations of serum iron and

copper. Complementing previous results, this study reveals that

clofibrate alters internal iron exchange and liver iron

homeosta-sis. The physiological implication of the adaptation can be

partially explained.

Clofibrate and Internal Iron Exchange

Since plasma Tf level is dominated by liver Tf expression,

the reduction in TIBC owing to clofibrate treatment can be

fully accounted for by the reduction in Tf mRNA. Recycling of

scavenged hemoglobin iron to erythropoietic marrow via

plasma transferrin represents the single largest iron flux in the

body (Fairbanks, 1994). Although liver iron was not depleted,

and the labile iron pool represented by DFO-chelatable Fe was

expanded, clofibrate treatment resulted in anemia herein,

indi-cating that the 50% reduction of serum Tf actually limits iron

efflux from the storage sites as well as the availability of iron

for erythropoiesis. The blood outcome observed in rats is

applicable to PP-treated patients, who frequently displayed a 2

to 4% reduction in hemoglobin concentrations (Schrefer,

2001). Ceruloplasmin is also essential for hepatic iron efflux

(Harris et al., 1999); however, Cp at 70% of the normal levels

was adequate for iron mobilization in the clofibrate-treated rats

(Cohen et al., 1983; Roeser et al., 1970). Therefore, the

reduc-tion in transferrin is fully responsible for the clofibrate-induced

anemia.

Two questions relevant to PP-induced anemia are raised:

whether this mild anemia worsens at suboptimal iron intake and

whether iron supplement alleviates it. Herein, the result of a

two-way ANOVA (p

⫽ 0.0295 for Fe) indicates a significant

main effect of iron intake on hemoglobin concentration (Table 5).

Nevertheless, the hemoglobin level of the clofibrate-treated

groups was not reduced by low iron intake, most likely due to

up-regulation of the number of transferrin receptors and increased

iron uptake hence in the iron-depleted erythroblasts, thus partially

compensating for the reduced iron supply from the liver.

How-ever, the fraction of the recycling hemoglobin iron is reduced with

decreased TIBC and will not match the rate of hemoglobin

deg-radation, thus providing an explanation for the elevated anemia

rate in the clinical follow-up (The Coronary Drug Project, 1975).

TABLE 6

Hepatic Iron Content and Mitochondrial Aconitase Activity of Rats as Affected by Low Iron Intake and Clofibrate Treatment

No clofibrateb _{0.5% Clofibrate}b p value of two-way ANOVA 35 ppm Fe 15 ppm Fe 35 ppm Fe 15 ppm Fe Fe Clofibrate Hepatic Fe (␮mol/g liver) 0.84⫾ 0.191 _0.36_{⫾ 0.11}2 _0.94_{⫾ 0.20}1 _0.47_{⫾ 0.11}2 _0.0001 _0.068

(␮mol/ whole liver) 20⫾ 42 ₇_{⫾ 2}3 ₂₅_{⫾ 5}1 ₁₃_{⫾ 3}4 _0.0001 _0.0019

DFO-chelatable Fec_{(nmol/g liver)} _11.3_{⫾ 3.8}2 _8.3_{⫾ 2.1}2 _24.2_{⫾ 8.5}1 _11.2_{⫾ 2.7}2 _0.0008 _0.0008 Mitochondrial aconitase activityd

(mU/mg mitochondrial protein) 19.1⫾ 1.62 _17.0_{⫾ 1.5}2 _41.8_{⫾ 4.8}1 _42.1_{⫾ 3.8}1 _0.45 _0.0001

p⬍ 0.05.

c_{Iron and clofibrate interaction was statistically significant by two-way ANOVA at p}_{⬍ 0.05.}

d_{One enzyme unit}_{⫽ 1}_{␮mol NADP}⫹_{consumed per min in a coupling reaction of aconitase and isocitrate dehydrogenase.}

FIG. 3. Mitochondrial aconitase (m-ACO) protein level in the rat liver as affected by a low-iron diet and clofibrate treatment. The 35 ppm Fe is normal dietary level for rats, while the 15 ppm Fe is a suboptimal level which will deplete liver Fe store and may cause mild anemia. Abundance was determined by Western blot analysis using 10␮ g mitochondrial protein (A) and results were quantified by densitometry (B). In A, each lane represents one rat. In B, the values are mean⫾ SD of six rats and bars not sharing the same superscript letters are significantly different from one another by Duncan’s Multiple Range test ( p⬍ 0.05).

(7)

Iron supplement does not eliminate the limit of the reduced TIBC,

therefore, it does not correct PP-induced anemia.

Clofibrate and Liver Iron Accumulation

Previous studies are inconsistent regarding PP-induced iron

accumulation in the liver; short-term drug exposure increased

liver iron content (Powanda et al., 1978), while prolonged

exposure did not (Eagon et al., 1999). The 42-day exposure

herein actually caused iron loading, as indicated by the

per-centage increase in total iron (25– 86%) exceeding that of the

proportion of hepatomegaly (10 –30%). On the other hand,

tumors induced by chronic PP administration are strongly

FIG. 4. Abundance of (A) transferrin, Tf, and (B) transferrin receptor, TfR mRNAs in the rat liver as affected by a low-iron diet and clofibrate treatment. The 35 ppm Fe is normal dietary level for rats, while the 15 ppm Fe is a suboptimal level which will deplete liver Fe store and may cause mild anemia. Abundance was determined by Northern blot analysis (total RNA 10␮ g for Tf and 40 ␮ g for TfR) and results were quantified by densitometry normalized to ␤-actin mRNA. The values are mean⫾ SD of six rats and bars in each panel not sharing the same superscript letters are significantly different from one another by Duncan’s Multiple Range test ( p⬍ 0.05).

FIG. 5. Cytosolic IRP1 and IRP2 activities in the rat liver as affected by a low-iron diet and clofibrate treatment. IRPs activities were determined by EMSA and bound IRP1 and IRP2 as well as free probe are indicated (A). The 35 ppm Fe is normal dietary level for rats, while the 15 ppm Fe is a sub-optimal level which will deplete liver Fe store and may cause mild anemia. Activities were quantified by scintillation counting (B and C). The values are mean⫾ SD of six rats and bars not sharing the same superscript letters are significantly different from one another by Duncan’s Multiple Range test ( p⬍ 0.05).

(8)

resistant to iron accumulation even given very high levels of

dietary iron (Rao et al., 1986). Taking together, a

time-depen-dent effect of PP on hepatic iron homeostasis is indicated in

that liver iron increases transiently in the initial stage of

PP-induced liver hyperplasia preceding the development of

hepato-cellular carcinomas (HCC). The excess iron may work for

functional and synthetic purposes, because numerous

PP-in-duced enzymes (cytochrome P450s, peroxisomal oxidases, and

mitochondrial enzymes) are Fe-containing proteins. Clofibrate

treatment did not reduce liver ferritin levels, suggesting that

intracellular iron is conserved by the reduction of

transferrin-mediated iron efflux. Therefore, the altering of iron

homeosta-sis by PPs is a physiological adaptation to conserve iron in liver

cells for metabolic functions.

Differential Regulation of IRPs

As an essential nutrient, cellular iron is closely regulated

through the action of TfR and ferritin, the proteins for iron

uptake and storage, respectively. Ferritin and TfR are regulated

posttranscriptionally by the IRP/IRE system (Cairo and

Pi-etrangelo, 2000). It is well established that RNA binding

ac-tivities of both IRP1 and IRP2 in the liver are influenced by

dietary iron intakes (Chen et al., 1997, 1998) and generally

change with similar kinetics when activated by iron deficiency

or deactivated by iron excess (Guo et al., 1994; Henderson and

Ku¨hn, 1995). Herein, we demonstrated that IRP1 and IRP2

reacted differentially, with IRP2 activity increasing, while

IRP1 activity remained unchanged, in response to the low-iron

diet (15 ppm Fe), which provided 42% of the recommended

dietary iron level for rats and caused mild iron-deficiency

symptoms as indicated by an 8% reduction in hemoglobin

concentration and a 14% reduction in body weight gain. This

result supports the observation by Chen et al. (1997) that IRP2

is more sensitive than IRP1 to in vivo changes in iron level.

Additionally, we provide the first in vivo observation that the

change in TfR mRNA was connected with the change in IRP2

activity, both increasing at the low iron intake and being

simultaneously suppressed by clofibrate treatment. A similar

correlation has been documented in NO-treated RAW 264.7

cells previously (Kim and Ponka, 1999). Such a connection

does not exist between IRP1 activity and TfR mRNA, since

TfR mRNA did not change as IRP1 activity increased in the

clofibrate-treated, normal iron group herein. Thus,

corroborat-ing the previous findcorroborat-ings that IRP2 is more effective than IRP1

in modulating ferritin synthesis in vivo (Chen et al., 1997), the

present study proves a similar IRP2 effect on the stability of

TfR mRNA.

Herein, clofibrate treatment selectively activated IRP1

de-spite the expanded liver iron and DFO-chelatable Fe

concen-trations. This activation cannot be attributed to cellular iron

depletion and is best explained by some H

2

O

2

-triggered

path-ways. Recent investigations have proven that H

2

O

2

treatment

causes IRP1, but not IRP2, to bind IREs rapidly in cell models

(Pantopoulos and Hentz, 1995; Pantopoulos et al., 1996) and in

perfused liver (Mueller et al., 2001). Furthermore, liver IRP1 is

activated by H

2

O

2

leaking from the mitochondrial respiratory

chain (Pantopoulos et al., 1997). In PP-treated livers, H

2

O

2

increases because of elevated activity of peroxisomal fatty

acyl-CoA oxidase and microsomal cytochrome P450 enyzems,

while the activity of H

2

O

2

-degrading enzymes, peroxisomal

catalase and cytosolic glutathione peroxidase, is relatively

in-sufficient (Yeldandi et al., 2000). The leakage of H

2

O

2

from

peroxisomes and undegraded H

2

O

2

in cytosol may serve as a

signal to activate IRP1. Additionally, the H

2

O

2

activation of

IRP1 was followed by the suppression of ferritin synthesis and

an increase in TfR numbers (Pantopoulos and Hentze, 1995;

Caltagirone et al., 2001), implying an expansion of the

intra-cellular free-iron pool, which is now demonstrated in the

clofibrate-treated, normal iron group herein, indicating that

TABLE 7

Serum and Hepatic Copper Indices of Rats as Affected by Low Iron Intake and Clofibrate Treatment

No clofibrateb _{0.5% Clofibrate}b p value of two-way ANOVA 35 ppm Fe 15 ppm Fe 35 ppm Fe 15 ppm Fe Fe Clofibrate Serum Cu (␮ mol/L)c ₁₇_{⫾ 1}1 ₁₅_{⫾ 1}2 ₁₂_{⫾ 1}3 ₁₂_{⫾ 1}3 _0.11 _0.0001 Ceruloplasmin (U/L)d ₁₂₁_{⫾ 16}1 ₁₁₁_{⫾ 16}1 ₉₁_{⫾ 5}2 ₈₅_{⫾ 8}2 _0.12 _0.0001

Hepatic Cu (nmol/g liver)e ₅₂_{⫾ 15}2 ₈₈_{⫾ 36}2 ₁₂₅_{⫾ 88}1,2 ₂₁₀_{⫾ 139}1 _0.0255 _0.0039 Cu:Fe molar ratioe _0.06_{⫾ 0.02}2 _0.27_{⫾ 0.15}1 _0.12_{⫾ 0.08}2 _0.51_{⫾ 0.41}1 _0.0001 _0.0489

p⬍ 0.05.

c_{Iron and clofibrate interaction was statistically significant by two-way ANOVA at p}_{⬍ 0.05.}

d_{One unit}_{⫽ 1}_{␮mol o-diansidine dihydrochloride oxidized per minute by the ferroxidase reaction of ceruloplasmin.} e_{Data were log transformed for statistical analysis.}

(9)

peroxisome-derived oxidative stress can activate IRP1 via

H

2

O

2

production and alter iron homeostasis.

Metabolic Adaptations of m-ACO

The mRNA of m-ACO, the TCA enzyme converting citrate

to isocitrate, contains an IRE in the 5

⬘-UTR and is subjected to

translational regulation by IRP/IRE binding (Schalinske et al.,

1998). Dietary iron deficiency reduces both the level and the

activity of m-ACO protein in rat liver in a time- and

dose-dependent manner (Chen et al., 1997, 1998). However, neither

the activity nor the protein level of m-ACO decreased when

IRP2 was activated by the low-iron diet herein, which is the

first in vivo demonstration of differential regulation between

m-ACO and TfR mRNA by IRP2, probably due to differing

affinities between IRP2 and IRE isoforms from different

mRNAs (Theil and Eisenstein, 2000). It is noted that the two

clofibrate-treated groups displayed similar m-ACO activities,

while differing in protein levels. An explanation for this

dis-crepancy is that the assembly of holo-m-ACO was limited at

low iron intake with an excess of nonfunctional apoprotein in

the cell.

Regardless of iron status, clofibrate treatment caused

m-ACO activity to increase by 133%, while clofibrate also

enhanced the activity of three rate-limiting enzymes in the

TCA cycle, citrate synthase, nicotinamilinked isocitrate

de-hydrogenase, and

␣-ketoglutarate dehydrogenase, by 24, 54,

and 153%, respectively (Prager et al., 1993; Schon et al.,

1994). In the clofibrate-treated liver, the production of

acetyl-CoA is elevated and leads to increased citrate formation via

citrate synthase (Ball et al., 1979). Accumulating citrate acts as

an allosteric activator of acetyl-CoA carboxylase and favors

fatty acid synthesis. The enhanced activities of m-ACO and

rate-limiting enzymes makes it possible to accommodate

higher acetyl-CoA influx and citrate flux, owing to PP-induced

␤-oxidation of fatty acids.

Clinical Implication of Liver Copper Accumulation

Clofibrate-induced copper accumulation in the liver was not

significant during a 5-day exposure to PP (Powanda et al.,

1978), while it increased markedly with prolonged exposure

(Eagon et al., 1999). The copper loading was specific as

verified by the twofold increase in Cu to Fe mole ratio of the

treated groups (Table 7). It is well established that excess

copper acts as a prooxidant to promote the formation of free

radicals and the harmful oxidation of lipids and proteins. In the

hereditary Wilson’s disease, excess copper causes

hepatocel-lular necrosis of the liver, as well as portal and periportal

inflammation, and ultimately fibrosis (Pfeil and Lynn, 1999).

Since elevated serum GOT and GPT have been observed

occasionally in patients taking PP drugs (Sirtori et al., 1992),

FIG. 6. Iron homeostasis in the liver affected by a low-iron diet and clofibrate treatment. When iron intake is low, IRP2 is activated by the depleted free-iron pool in the cell, which will increase the stability of transferrin receptor (TfR) mRNA and the number of transferrin receptors. When clofibrate is administered, iron efflux from the liver decreases due to suppressed expression of transferrin (Tf) gene and decreased iron transport in the serum. Clofibrate elevates peroxisomal oxidase, microsomal CYP and mitochondrial aconitase (m-ACO) activities, which may increase intracellular H2O2level. IRP1 activated by H2O2 suppresses ferritin synthesis, thus expanding the intracellular free iron pool, which allows more iron to be available for the biosynthesis of iron-containing enzymes in response to clofibrate treatment on the one hand, and inactivates IRP2 to reduce TfR mRNA and receptor numbers on the other hand. Thus, a new steady state of iron homeostasis will be established.

(10)

copper-induced oxidative stress may be a major contributor to

PP-induced liver toxicity. Few clinical trials have monitored

copper status, mainly because of a lack of feasible blood

markers. Since Cp and liver copper change inversely during PP

treatment, serum Cp may be used to check the risk of hepatic

copper loading.

In conclusion, the present study demonstrates that changes

in hepatic iron homeostasis by clofibrate involve a differential

regulation of IRP1 and IRP2 with changes in m-ACO both at

protein and enzyme activity levels, and a model is proposed to

depict these effects (Fig. 6). Clofibrate suppresses Tf

expres-sion and reduces hepatic iron efflux to maintain an expanded

free-iron pool in the liver cell via differential regulation of

IRP1 and IRP2 activities, in order to conserve iron for the

synthesis of a variety of Fe-containing enzymes, including

m-ACO, which are up-regulated by hypolipidemic

peroxiso-mal proliferators for metabolic functions. We also observe that

TIBC, Cp, and liver copper levels rapidly returned to normal

while clofibrate was discontinued (data not shown); periodic

discontinuation of fribrate drugs needs to be considered in

order to alleviate PP-induced anemia and liver copper toxicity.

ACKNOWLEDGMENTS

We are grateful to Dr. R. S. Eisenstein from the Department of Nutritional Sciences, University of Wisconsin, for the gift of ratL-ferritin cDNA, and to

Dr. H.-H. Zhuang from Memorial University of Chang Gung for the gift of rabbit anti-human mitochondrial aconitase antibody. This project is supported by Grant NSC89-2320-B002-165 from the National Science Council in Taiwan.

REFERENCES

American Institute of Nutrition (AIN) (1977). Report of American Institute of Nutrition Ad hoc committee on standards for nutritional studies. J. Nutr. 107, 1340 –1344.

Anderson, S. P., Cattley, R. C., and Corton, J. C. (1999). Hepatic expression of acute-phase protein genes during carcinogenesis induced by peroxisome proliferators. Mol. Carcinogen. 26, 226 –238.

Auwerx, J. (1992). Regulation of gene expression by fatty acids and fibric acid derivatives: An integrative role for PPARs. Horm. Res. 38, 269 –277. Ball, M. R., Gumaa, K. A., and MeLean, P. (1979). Effect of clofibrate on the

CoA thioester profile in rat liver. Biochem. Biophys. Res. Commun. 87, 489 – 496.

Barton, H. A, Eisenstein, R. S., Bomford, A. B., and Munro, H. N. (1990). Determinants of the interaction of the iron-responsive element-binding protein and its binding site in ratL-ferritin mRNA. J. Biol. Chem. 265, 7000 –7008.

Cairo, G., and Pietrangelo, A. (2000). Iron regulatory proteins in pathobiology. Biochem. J. 352, 241–250.

Cairo, G., Tacchini, L., Pogliaghi, G., Anzon, E., Tomasi, A., and Bernelli-Zazzera, A. (1995). Induction of ferritin synthesis by oxidation stress, transcriptional and post-transcriptional regulation by expension of the “free-iron pool.” J. Biol. Chem. 270, 700 –703.

Caltagirone, A., Weiss, G. and Pantopoulos K. (2001). Modulation of cellular iron metabolism by hydrogen peroxide. J. Biol. Chem. 276, 19738 –19745. Chen, O. S., Blemings, K. P., Schalinske, K. L., and Eisenstein, R. S. (1998). Dietary iron intake rapidly influences iron regulatory proteins, ferritin sub-units and mitochondrial aconitase in rat liver. J. Nutr. 128, 525–535.

Chen, O. S., Schalinske, K. L., and Eisenstein, R. S. (1997). Dietary iron intake modulates the activity of iron regulatory proteins and the abun-dance of ferritin and mitochondrial aconitase in rat liver. J. Nutr. 127, 238 –248.

Cohen, N. L., Keen, C. L., Lonnerdal, B., and Hurley, L. S. (1983). The effect of copper chelating drugs on liver iron mobilization in the adult rat. Bio-chem. Biophys. Res. Commun. 113, 127–134.

Crichton, R. R., and Charloteaux-Wauters, M. (1987). Iron transport and storage. Eur. J. Biochem. 164, 485–506.

Eagon, P. K., Teepe, A. G., Elm, M. S., Tadic, S. D., Epley, M. J., Beiler, B. E., Shinozuka, H., and Rao, K. N. (1999). Hepatic hyperplasia and cancer in rats: Alterations in copper metabolism. Carcinogenesis 20, 1091–1096 Fairbanks, V. F. (1994). Iron in medicine and nutrition. In Modern Nutrition in

Health and Disease, 8th ed. (M. E. Shils, J. A. Olson, and M. Shike, Eds.), pp. 185–213. Lea & Febiger, Malvern, PA.

Guo, B., Yu, Y., and Leibold, E. A. (1994). Iron regulates cytoplasmic levels of a noveliron-responsive element-binding protein without aconitase activ-ity. J. Biol. Chem. 269, 24252–24260.

Harris, Z. L., Durley, A. P., Man, K. T., and Gitlin, J. D. (1999). Targeted gene disruption reveals and essential role for ceruloplasmin in cellular iron efflux. Proc. Natl. Acad. Sci. USA 96, 10812–10817.

Harris, Z. L., Klomp, L. W. J., and Gitlin, J. D. (1998). Aceruloplasminemia: An inherited neurodegenerative disease with impairment of iron homeosta-sis. Am. J. Clin. Nutr. 67, (Suppl.), 972S–977S

Hayashi, A., Wada, Y., Suzuki, T., and Shimizu, A. (1993). Studies on familial hypotransferrinemia: Unique clinical course and molecular pathology. Am. J. Hum. Genet. 53, 201–213.

Henderson, B. R., and Ku¨hn, L. (1995). Differential modulation of the RNA-binding proteins IRP-1 and IRP-2 in response to iron. J. Biol. Chem. 270, 20509 –20515.

Hentz, M. W., and Ku¨hn, L. C. (1996). Molecular control of vertebrate iron metabolism: mRNA-based regulatory circuits operated by iron, nitric oxide, and oxidative stress. Proc. Natl. Acad. Sci. USA 93, 8175– 8182. Hertz, R., Seckbach, M., Zakin, M. M., and Bar-Tana, J. (1996).

Transcrip-tional suppression of the transferrin gene by hypolipidemic peroxisome proliferator. J. Biol. Chem. 271, 218 –224.

Hoshino, A., Hisayasu, S., and Shimada, T. (1996). Complete sequence anal-ysis of rat transferrin and expression of transferrin but not lactoferrin in the digestive glands. Comp. Biochem. Physiol. 113B, 491– 497.

Kim, S., and Ponka, P. (1999). Control of transferring receptor expression via nitric oxide-mediated modulation of iron-regulatory protein 2. J. Biol. Chem. 274, 33035–33042.

Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. T. (1951). Protein measurement with Folin phenol reagent. J. Biol. Chem. 193, 265– 275.

Mueller, S., Pantopoulos, K., Hubner C. A., Stremmel W., and Hentz, W. M. (2001). IRP1 activation by extracellular oxidative stress in the perfused rat liver. J. Biol. Chem. 276, 23192–23196.

National Science Council (1993). NSC Guide for the Care and Use of Labo-ratory Animals. National Science Council, Taipei, Taiwan.

Nudel, U., Zakut, R., Shani, M., Neuman, S., Levy, Z., and Yaffe, D. (1983). The nucleotide sequence of the rat cytoplasmic beta-actin gene. Nucleic Acids Res. 11, 1759 –1771.

Oser, B. L. (1965). Hawk’s Physiological Chemistry, 14th ed., McGraw–Hill, New York.

Pantopoulos, K., and Hentze, M. W. (1995). Rapid responses to oxidative stress mediated by iron regulatory protein. EMBO J. 14, 2917–2924. Pantopoulos, K., Mueller, S., Atzberger, A., Ansorge, W., Stremmel, W., and

(11)

protein-1 (IRP-1) by extra- and intracellular oxidative stress. J. Biol. Chem. 272, 9802–9808.

Pantopoulos, K., Weiss, G., and Hentze, M. W. (1996). Nitric oxide and oxidative stress (H2O2) control mammalian iron metabolism by different pathways. Mol. Cell. Biol. 16, 3781–3788.

Pfeil, S. A., and Lynn, D. J. (1999). Wilson’s disease, copper unfettered. J. Clin. Gastroenterol. 29, 22–31.

Powanda, M. C., Blackburn, B. S., Bostian, K. A., Fowler, J. P., Hauer E. C., and Pekarek, R. S. (1978). Clofibrate-induced alterations in zinc, iron and copper metabolism. Biochem. Pharmacol. 27, 125–127.

Prager, C., Schon, H. J., Nikfardjam, M., Schmid, D., Untersalmberger, M., Kremser, K., and Kramar, R. (1993). Clofibrate elevates enzyme activities of the tricarboxylic acid cycle in rat liver. J. Lipid Res. 34, 359 –364.

Rao, M. S., Subbarao, V., and Reddy, J. K. (1986). Increased resistance of peroxisome proliferator-induced hepatic lesions to iron overload in rats. Cancer Lett. 32, 33–39.

Roberts, K. P., and Griswold, M. D. (1990). Characterization of rat transferrin receptor cDNA: The regulation of transferrin receptor mRNA in testes and in Sertoli cells in culture. Mol. Endocrinol. 4, 531–542.

Roeser, H. P., Lee, G. R., Nacht, S., and Cartwright, G. E. (1970). The role of ceruloplasmin in iron metabolism. J. Clin. Invest. 49, 2408 –2417. Rose, I. A., and O’Connell, E. L. (1967). Mechanism of aconitase action. I.

The hydrogen transfer reaction. J. Biol. Chem. 242, 1870 –1879. Schalinske, K. L., Chen, O. S., and Eisenstein, R. S. (1998). Iron differentially

stimulates translation of mitochondrial aconitase and ferritin mRNAs in mammalian cells. J. Biol. Chem. 273, 3740 –3746.

Schalinske, K. L., and Eisenstein, R. S. (1996). Phosphorylation and activation

of both iron regulatory proteins 1 and 2 in HL-60 cells. J. Biol. Chem. 271, 7168 –7176.

Schon, H. J., Grgurin, M., Klune, G., Prager, C., Marz, R., Legenstein, E, Bock, P., and Kramar, R. (1994). Effects of hypolipidaemics cetaben and clofibrate on mitochondrial and peroxisomal enzymes of rat liver. J. Pharm. Pharmacol. 46, 144 –147.

Schoonjans, K., Stael, B., and Auwerx, J. (1996). Role of the peroxisome proliferator-activated receptor (PPAR) in mediating the effects of fibrates and fatty acids on gene expression. J. Lipid Res. 37, 907–925.

Schosinky, K. H., Lehmann, H. P., and Beeler, M. F. (1974). Measurement of ceruloplasmin from its oxidase activity in serum by use of o-dianisdine dihydrochloride. Clin. Chem. 20, 1556 –1563.

Schrefer, J. (2001). Mosby’s GenRx, 11th ed., Mosby, St. Louis.

Shaw, N. S., and Chen, S. H. (1993). Effect of zinc deficiency on hepatic iron and ferritin in rats. J. Chin. Agric. Chem. Soc. 31, 633– 641.

Sirtori, C. R., Calabresi, L., Werba, J. P., and Franceschini, G. (1992). Tolerability of fibric acids: Comparative data and biochemical bases. Phar-macol. Res. 26, 243–260.

The Coronary Drug Project Research Group (1975). Clofibrate and niacin in coronary heart disease. J. Am. Med. Assoc. 231, 360 –381.

Theil, E. C., and Eisenstein, R. S. (2000). Combinatorial mRNA regulation: Iron regulatory proteins and iso-iron-responsive elements (Iso-IREs). J. Biol. Chem. 275, 40659 – 40662.

Yegorov D. Y., Kozlov, A. V., Azizova, O. A., and Vladimirov, Y. A. (1993). Simultaneous determination of Fe(III) and Fe(II) in water solutions and tissue homogenates using desferal and 1,10-phenanthroline. Free Radical Biol. Med. 15, 565–574.

Yeldandi, A. V., Rao, M. S., and Reddy, J. K. (2000). Hydrogen peroxide generation in peroxisome proliferator-induced oncogenesis. Mutat. Res. 448, 159 –177.