![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Department of Nutritional Sciences, University of Wisconsin-Madison, Madison, WI 53706
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGMENTS
LITERATURE CITED
ABSTRACT 
Iron regulatory protein 1 (IRP1) and IRP2 are cytoplasmic RNA binding proteins that coordinate cellular iron homeostasis in mammals. We investigated the effect of dietary iron intake on rat liver IRP activity in relation to the abundance of two targets of IRP action, ferritin and mitochondrial aconitase (m-aconitase). Rats were fed diets containing 2, 11, 20, 37 (control), 72 or 107 mg iron/kg diet for 3 wk. RNA binding activity of IRP1 and IRP2 was enhanced one- to twofold in rats fed 11 or 2 mg iron/kg diet compared with control rats. IRP RNA binding activity was inversely correlated to blood hemoglobin levels (r = 
0.787; P < 0.0001). Compared with control rats, liver ferritin levels were depressed in rats fed 20 mg iron/kg diet and were undetectable in rats ingesting diets with 11 or 2 mg iron/kg diet. Ferritin concentrations were biphasically related to IRP RNA binding activity with the regulation of IRP occurring before the onset of ferritin accumulation. Iron deficiency caused up to a 50% decline in m-aconitase abundance. IRP RNA binding activity and m-aconitase abundance were inversely correlated (r = 0.751; P < 0.0001). Our results indicate that (1) liver IRP activity is responsive to a range of dietary iron levels, (2) there appears to be a differential effect of IRPs on ferritin and m-aconitase abundance, and (3) activation of IRPs may contribute to the alterations in energy metabolism in iron deficiency through an impairment of m-aconitase synthesis.
INTRODUCTION
Iron is an essential nutrient but potentially toxic element. The nutritional requirement for iron is well documented through its function as an essential component of proteins that participate in oxygen delivery, electron transport, DNA synthesis as well as a variety of other biochemical reactions required for cell viability (Dallman 1986
, Eisenstein et al. 1997, Hentze and Kühn 1996, Klausner et al. 1993). However, iron can also be detrimental to cell function because of its potential to induce formation of damaging reactive oxygen intermediates (Aisen et al. 1990). Cellular iron homeostasis is maintained through the use of specific proteins to promote the beneficial uses of iron while simultaneously minimizing the inappropriate actions of the mineral. Modulation of the synthesis rates of proteins involved in the cellular uptake, utilization and storage of iron is a central mechanism responsible for the maintenance of cellular iron homeostasis.
Two of the proteins that perform a key role in controlling the availability of iron within the cell are transferrin receptor (TfR)5 and ferritin (Eisenstein et al. 1997, Klausner et al. 1993, Hentze and Kühn 1996). TfR is present on the cell surface where it binds the iron transport protein diferric transferrin (Fe2-Tf), and the TfR-(Fe2-Tf) complex is internalized by receptor-mediated endocytosis. Iron is released from Tf into the cytoplasm where it can be used for synthesis of various essential iron-containing proteins or deposited for storage in the multisubunit macromolecule ferritin. Apo-Tf complexed with TfR is recycled to the cell surface where apo-Tf is released into the circulation to acquire more iron. Cellular uptake and storage of iron are controlled through changes in the synthesis rate and therefore abundance of TfR and ferritin. The synthesis rates of TfR and the H- and L-subunits of ferritin are coordinately but divergently regulated by cellular iron status.
Iron modulates TfR and ferritin synthesis posttranscriptionally (Eisenstein et al. 1997, Klausner et al. 1993, Hentze and Kühn 1996). The mRNAs encoding H- and L-ferritin and TfR contain similar stem-loop structures, called iron responsive elements (IREs), in their 5
and 3 untranslated regions (UTR), respectively. Iron modulates the translation of ferritin mRNA and stability of TfR mRNA due to the regulated interaction of IREs with iron regulatory proteins (IRPs). In response to low levels of intracellular iron, IRPs bind to the single IRE in the 5UTR of ferritin mRNA, thereby blocking translation of the message which ultimately leads to a reduction in cellular iron storage capacity. Concomitantly, IRPs bind to multiple IREs in the 3UTR of TfR mRNA, enhancing stability of the mRNA which leads to increased production of the iron uptake protein. Conversely, when iron levels are high, IRPs dissociate from ferritin and TfR mRNAs, resulting in enhanced translation of ferritin mRNAs and decreased stability and abundance of the mRNA encoding TfR.
Two IRPs are known to exist, IRP1 and IRP2, which bind IREs with similar affinity but differ in the mechanism by which iron regulates their function. IRP1 was found to be similar in amino acid sequence to mitochondrial aconitase (m-aconitase), a 4Fe-4S iron-sulfur protein, which is a component of the tricarboxylic acid cycle (Hentze and Kühn 1996, Klausner et al. 1993). It is now apparent that IRP1 is a bifunctional protein with activities as a sequence-specific RNA binding protein and as the cytoplasmic isoform of aconitase, c-aconitase (Kennedy et al. 1992). Iron down-regulates the RNA binding function by inducing assembly of a 4Fe-4S cluster in IRP1, converting it to c-aconitase. Thus, iron can modulate the RNA binding activity of IRP1 by influencing the amount of the protein that is in the RNA binding or c-aconitase form without changing the total amount of the binding protein present in the cell. IRP2 is similar in amino acid sequence to IRP1 with the notable addition of a 73 amino acid insertion in IRP2 (Guo et al. 1995, Iwai et al. 1995). This 73 amino acid sequence is a key determinant of the iron-dependent degradation of IRP2 (Guo et al. 1995, Iwai et al. 1995). Thus, in contrast to IRP1, cellular iron status modulates the steady-state level of IRP2. Taken together, it is apparent that IREs and IRPs are components of a regulatory and sensory network that performs a central role in the maintenance of cellular iron homeostasis.
Other cellular mRNAs also contain an IRE in their 5UTR. In mammalian systems, two other mRNAs, erythroid 5-aminolevulinate synthase (eALAS) mRNA and m-aconitase mRNA have an IRE in their 5UTR (referenced in Hentze and Kühn 1996). In the case of eALAS, the rate-limiting enzyme in heme formation, it appears that IRPs modulate translation of its mRNA. The presence of a putative IRE in the 5UTR of m-aconitase indicates a potential role for IRPs in modulating energy metabolism. When linked to a heterologous mRNA, the m-aconitase IRE can bind IRP1, and together the RNA/protein complex represses translation of the chimeric mRNA in an in vitro protein synthesis system (Gray et al. 1996). IREs in m-aconitase and eALAS differ from the ferritin IRE with respect to the size of a bulged nucleotide region on the 5 side of the upper stem region, and it has been suggested that this may be a basis for functional differences among various IREs (Bettany et al. 1992, Theil 1994). It has also been suggested that the greater affinity of IRPs for ferritin IREs, compared with IREs in eALAS or m-aconitase, might lead to differential regulation of the synthesis of these proteins in response to changes in intracellular iron levels (Bettany et al. 1992, Cox et al. 1991, Theil 1994).
Our current understanding of the role of IRPs in mediating iron metabolism is based predominantly on cell culture studies. Although much has been learned about IRP function using cell culture approaches, very few studies have examined the regulation of IRPs in vivo, and these have been limited to evaluations of the impact of severe dietary iron deficiency or the acute effects of large loads of iron administered by injection on IRP function (Rabie et al. 1995, Ward et al. 1994, Yu et al. 1992). No studies have been conducted to examine the extent to which graded physiological changes in iron intake affect IRP RNA binding activity and the manner in which such potential changes in IRP activity relate to manifestations of IRP function such as ferritin subunit abundance. Furthermore, because liver ferritin is the main depot for storage of body iron and levels of liver ferritin are influenced by dietary iron intake, it is of interest to examine the relationship between liver ferritin, IRP activity and dietary iron intake. Finally, given the apparent difference in affinity of IRPs for IREs in a number of mRNAs, it is of interest to determine if differences exist in the effect of dietary iron intake on the various targets of IRP action under physiological conditions. The extent to which m-aconitase protein abundance is affected by iron status has not been investigated.
We have examined in rats the effects of varying dietary iron intake on hepatic IRP1 and IRP2 RNA binding activity as well as on the abundance of two targets of IRP action, ferritin subunits and m-aconitase. By varying dietary iron from levels well below the dietary requirement to levels that exceed it, we were able to examine the relationship between large variations in ferritin subunit levels and changes in IRP function. Our results demonstrate that as dietary iron intake increases from very low levels to levels approaching the dietary requirement for iron, large decreases in IRE binding activity occur before there is a significant induction of ferritin subunit abundance. Similar to the ferritin subunits, m-aconitase protein abundance was decreased significantly in iron deficiency. Ferritin subunit abundance was much more strongly repressed in iron deficiency than was m-aconitase abundance, supporting the concept that greater binding affinity of IRPs to ferritin IREs promotes more effective repression of the synthesis of ferritin. Furthermore, by demonstrating a significant relationship between liver IRP RNA binding activity and m-aconitase protein abundance, we provide support for the hypothesis that IRPs may also influence cellular energy production.
MATERIALS AND METHODS
Table 1.
Composition of iron-deficient diet1
80°C. For measurement of RNA binding activity in RBC, the cell pellet was lysed in 10 volumes of the buffer containing 20 mmol/L Hepes pH 7.4, 10 mmol/L sodium pyrophosphate, 50 mmol/L sodium fluoride, 50 mmol/L 
-glycerophosphate, 5 mmol/L EDTA, 1 mmol/L sodium orthovanadate, 2 mmol/L benzamidine, 100 mg/L of leupeptin and pepstatin, 250 mg/L soybean trypsin inhibitor, 0.2 mmol/L phenylmethylsulfonyl fluoride (PMSF), 25 mg/L of p-nitroguanidinobenzoate, and 0.5% Nonidet P-40, centrifuged at 16,000 × g for 10 min; the resulting supernatant was divided into equal portions and stored at 80°C.
Subcellular fractionation of liver mitochondria, cytosol and microsomes.
Fresh liver was minced with scissors and homogenized in 3 volumes of HDGC (20 mmol/L Hepes pH 7.5, 1 mmol/L DTT, 10 % glycerol and 2 mmol/L trisodium citrate, 0.5 mg/L leupeptin, 0.7 mg/L pepstatin and 0.2 mmol/L PMSF). Citrate was included because it stabilizes the iron-sulfur cluster of aconitases (Kennedy et al. 1992). Liver mitochondria, cytosol and microsomes were obtained by differential centrifugation. The homogenate was centrifuged at 600 × g for 10 min to remove nuclei, and mitochondria were obtained from the postnuclear supernatant by centrifugation at 10,000 × g for 10 min. This crude mitochondrial pellet was washed with HDGC three times to remove any cytosolic contamination. Cytosolic contamination was determined using the marker enzyme, lactate dehydrogenase [LDH (EC 1.1.1.27)] (Schmidt 1974). About 0.3 % of LDH activity measured in the initial homogenate was found in the isolated mitochondria. Recovery of mitochondria was determined by glutamate dehydrogenase [GDH (EC 1.4.1.3)] enzyme activity (Bergmeyer and Bernt 1974) and ranged from 10 to 35%. Mitochondria were resuspended in HDGC and stored at 80°C. Under these conditions, aconitase activity was stable for at least 1 mo (results not shown). Mitochondria were lysed by the addition of Triton X-100 to a final concentration of 1% (v/v) prior to the measurement of aconitase activity and abundance. Liver cytosol was obtained from the postmitochondrial supernatant by ultracentrifugation at 100,000 × g for 60 min. The purity and recovery of the cytosol were determined using GDH and LDH, respectively. Mitochondrial contamination ranged from 0.5 to 2% and the recovery of cytosol ranged from 65 to 73%. Microsomes free of cytosol were obtained by resuspending and washing the microsomal pellet three times with HDGC buffer. LDH activity was undetectable in the washed microsomes. The washed microsomes were resuspended in an equal volume of HDGC containing 1.0 mol/L potassium chloride to dissociate IRPs from endogenous mRNA/IRP complex associated with polyribosomes (Seiser et al. 1995). The released IRPs were separated from microsomes by centrifugation at 100,000 × g for 60 min and the supernatant was dialyzed against 500 volumes of HDGC buffer. Protein concentration was determined by the Bradford assay using bovine serum albumin as a standard (Pierce, Rockford, IL).
Enzyme assays.
GDH (Bergmeyer and Bernt 1974) and LDH (Schmidt 1974) enzyme activity was assayed with standard methods. Aconitase activity was determined by coupled reaction of aconitase (EC 4.2.1.3) and isocitrate dehydrogenase (EC 1.1.1.42) (Rose and O'Connell 1967).
RNA binding activity assay.
RNA binding activity was performed by gel shift analysis using a [32P]labeled RNA of the first 73 nucleotides of the rat L-ferritin 5UTR (Schalinske and Eisenstein 1996). Briefly, cell extracts to be analyzed for IRE binding activity were incubated with saturating levels of [32P]labeled RNA (1 nmol/L) in the presence of a final concentration of 5% glycerol, 1 mmol/L magnesium acetate, 20 mmol/L of Hepes pH 7.6, 75 mmol/L potassium chloride and 20 mg/L of nuclease free bovine serum albumin (Sigma, St. Louis, MO) in a final volume of 30 µL. Binding reactions were started by addition of RNA and were performed at room temperature for 10 min. Then 3 µL of heparin (5 g/L in water) was added and the sample incubated for another 5 min after which 25 µL of the sample was loaded onto a native 4% polyacrylamide (60:1 acrylamide/bisacrylamide) gel in 0.5X Tris borate-EDTA buffer (Barton et al. 1990). The samples were electrophoresed at 300 V for 2 h. Gels were kept cooled to between 4 and 10°C to minimize dissociation of the RNA/protein complex. The optimal amount of protein used for the gel shift assay was determined specifically for each subcellular fraction and this amounted to 20 µg of RBC lysate protein, 10 µg liver cytosolic protein and 5 µg protein of microsomal supernatant. The amount of IRP present in active and inactive (i.e., c-aconitase) forms was assessed using 2-mercaptoethanol (2-ME) (referenced in Klausner 1993). Twenty micrograms protein of RBC lysates, 0.5 µg protein of cytosol and 5 µg protein of salt-washed microsomal supernatant were used for assay of 2-ME-induced RNA binding activity. RNA binding activity was quantified by liquid scintillation counting (Schalinske and Eisenstein 1996). The coefficient of variation for RNA binding determinations for a given sample was <10%. Additional details of the gel shift assay have been described previously (Barton et al. 1990).
Gel electrophoresis and Western blot analysis.
Tissue ferritin and m-aconitase levels were determined by SDS-PAGE followed by Western blot analysis using rabbit polyclonal antibodies raised against rat liver ferritin and bovine heart mitochondrial aconitase, respectively. The rabbit anti-bovine m-aconitase antibodies cross-reacted with rat liver m-aconitase (see Results). For separation of ferritin H and L subunits, the discontinuous Tricine-SDS PAGE system was used (Schägger and von Jagow 1987), whereas the Tris-SDS PAGE method (referenced in Schalinske and Eisenstein 1996) was used for m-aconitase. Ferritin was analyzed using 50 µg liver homogenate protein, and 10 µg of liver mitochondrial protein was used for determination of m-aconitase levels. After proteins were electrophoretically separated in 10% polyacrylamide gels, they were electrophoretically transferred to nitrocellulose membranes. Membranes were incubated with polyclonal antibodies (1/1000 dilution) in BLOTTO-Tween [5% (wt/v) nonfat dry milk, 10 mmol/L sodium phosphate (pH 7.5), 0.9% NaCl, 0.5% Tween-20] followed by incubation with goat anti-rabbit IgG conjugated with alkaline phosphatase (1/1000 dilution in BLOTTO-Tween) (Southern Biotechnology, Bir- mingham, AL). Using the NBT/BCIP substrate kit (Pierce), the identity of ferritin subunits and m-aconitase was confirmed by their comigration with the purified rat liver ferritin and bovine heart m-aconitase, respectively (results not shown). Purified bovine heart m-aconitase was a kind gift of Claire Kennedy (Medical College of Wisconsin) and Helmut Beinert (University of Wisconsin-Madison). Immunoblots were quantified by computerized densitometry (PDI, Huntington Station, NY) to determine the relative abundance of ferritin and m-aconitase across dietary treatment groups. Relative amounts of ferritin were expressed as a percentage of the control group. For m-aconitase, the absolute amount was determined using a standard curve of purified m-aconitase that ranged from 2 to 20 ng protein (r = 0.98).
Hepatic ferritin iron analysis.
Hepatic ferritin iron was determined by a two-step analysis: extraction of tissue ferritin as described by Gabrio et al. (1953) followed by acid hydrolysis of nonheme iron as described by Weinfield (1964). Fresh liver homogenate, obtained as described above, was immediately diluted with 2.5 volumes HDGC buffer, and extraction of tissue ferritin was performed at 4°C for 1 h. Following centrifugation at 1500 × g for 15 min, the pellets were further extracted two times and the pooled supernatants were designated liver ferritin extract. No ferritin was detectable by immunoblotting of proteins in the final pellet, and the amount of ferritin in the combined supernatants did not differ from the amount present in the original homogenate (results not shown). The liver ferritin extract was heated at 75°C for 15 min with frequent rotation as described (Weinfield 1964), centrifuged at 1500 × g for 15 min and the resulting supernatants were stored at 20°C. The liver ferritin extract prepared was subjected to acid hydrolysis with 2.8 mol/L HCl at 90°C for 1 h after which precipitated proteins were removed by centrifugation at 1500 × g for 15 min. One milliliter of the resulting supernatant was incubated with 20 µL of 2% hydroquinone and 20 µL of 1% o-phenanthroline (both from Fluka Chemical, Ronkonkoma, NY) and the optical density at 505 nm was then determined. A standard curve was generated based on the absorbance of a standard solution of ferrous sulfate at pH 3.0. Like the liver ferritin extract, the standard iron solution was carried through the acid hydrolysis procedure as well. All glassware for the ferritin iron assay was acid washed and all chemicals and reagents were ultrapure.
Statistical analysis.
Mean values across all dietary treatment groups were subjected to a one-way ANOVA (P < 0.05) (GraphPad InStat Software) and compared using the least significant difference procedure (Snedecor and Cochran 1980).
RESULTS
|
Table 2. Weight gain, blood hemoglobin and hepatic ferritin iron of rats as a function of dietary iron intake1 |
, Klausner et al. 1993, Schalinske and Eisenstein 1996). The forms of IRPs that are considered to be active with respect to regulation of the utilization of IRE-containing mRNAs exhibit high affinity for IREs (KD 
50-200 pmol/L). The high affinity pool of IRPs can be further divided between a free pool, not bound to RNA, and a bound pool which represents IRPs bound to ferritin, transferrin receptor, m-aconitase and perhaps other mRNAs (Barton et al. 1990). The forms of IRPs that are considered to be functionally inactive with respect to IRE binding include the c-aconitase form of IRP1 and an oxidized form of IRP2 (Hentze and Kühn 1996, Klausner et al. 1993, Schalinske and Eisenstein 1996). We determined the effect of dietary iron intake on the active and inactive pools of IRP1 and IRP2 in liver.
DISCUSSION 
Fig. 1.
Effect of dietary iron intake on RNA binding activity of iron regulatory proteins (IRPs) in rat liver cytosol. A representative autoradiogram of spontaneous RNA binding activity of cytosolic IRP1 and IRP2 is shown in panel A. Spontaneous [no 2-mercaptoethanol (2-ME) added] RNA binding activity (panel B) of IRP1, IRP2 and total RNA binding activity as well as 2-ME-induced RNA binding activity (panel C) were determined in liver cytosol from rats fed diets containing 2, 11, 20, 37, 72 or 107 mg iron/kg diet for 3 wk. Rats pair-fed the control diet at the level of food intake exhibited by groups fed 2 and 107 mg Fe/kg diet for 3 wk (pf 2 and pf 107, respectively) are also shown. RNA binding activity was determined by gel shift analysis as described in Materials and Methods. Quantitative data are means ± SEM (n = 6). Vertical bars for a given measurement that have a different letter are significantly different (P < 0.05).
[View Larger Version of this Image (33K GIF file)]
, Klausner et al. 1993, Schalinske and Eisenstein 1996). We examined the effect of iron intake on 2-mercaptoethanol (2-ME) inducible IRE binding activity to determine if the abundance of the inactive forms of IRPs responded to dietary iron. In rat liver cytosol, we have found that 5% 2-ME gives the maximal induction of IRP1 RNA binding activity and that, in this tissue, IRP2 is largely unaffected by 2-ME (results not shown). Thus, because IRP2 is not affected by 2-ME, this assay provides a measure of the amount of IRP1 in the low affinity RNA binding form (c-aconitase). In cytosolic extracts from livers of control rats, IRE binding activity in the presence of 2-ME (5.1 ± 0.23 pmol/mg protein) was observed to be elevated by about 30-fold compared with the binding measured in the absence of 2-ME (Fig. 1, compare panels B and C). The amount of 2-ME inducible IRE binding activity for the rats ingesting the diets containing 2 or 11 mg iron/kg diet was significantly greater than that of controls (Fig. 1, panel C). In rats fed the diet containing 2 mg iron/kg diet, RNA binding in the presence of 2-ME was 22% greater than the level of 2-ME inducible RNA binding activity present in control rats. Thus, iron deficiency increased the abundance of IRP1 protein in liver. Relative to the level of binding activity in the control groups, the absolute increase in 2-ME inducible RNA binding activity in rats fed the diet containing 2 mg iron/kg diet was 1.17 pmol/mg protein. In comparison, the spontaneous (measured without 2-ME) RNA binding activity in rats fed the diet containing 2 mg iron/kg diet was 0.26 pmol/mg protein greater than the level of spontaneous IRE binding activity measured in rats fed the control diet.
). Microsomal IRE binding activity in rats ingesting the control diet was 0.159 ± 0.011 pmol/mg protein. IRE binding activity in the microsomal fraction increased progressively as the iron concentration of the diet decreased. In rats ingesting the diet containing 2 mg iron/kg diet microsomal IRE RNA binding activity was 111% greater than the activity found in the fraction from control rats and was 265% greater than the activity observed in rats pair-fed the control diet (pf 2 group) (Fig. 2). IRE binding activity in the microsomal fraction was also greater in rats consuming 11 and 20 mg iron/kg diet compared with the control group, although the value for the 20 mg iron/kg diet group was not significantly different from the control group (Fig. 2). Addition of 2-ME to the microsomal fraction increased RNA binding activity by 5-10%, indicating that the IRPs present in this fraction were largely in a spontaneously high affinity state for RNA binding (results not shown). Because there appeared to be little if any IRE binding activity in an inactive state in the microsomal fraction, our results suggest that the amount of IRP bound to mRNAs such as ferritin, TfR and m-aconitase increases significantly in iron deficiency.
Fig. 2.
Effect of dietary iron intake on RNA binding activity of microsome-associated iron regulatory proteins (IRPs) in rat liver. RNA binding activity of microsome-associated IRPs was determined in liver from rats fed various levels of dietary iron as well as pair-fed controls as described in the legend of Figure 1. Quantitative data are means ± SEM (n = 6). Vertical bars that have a different superscript letter are significantly different (P < 0.05).
[View Larger Version of this Image (18K GIF file)]
). The level of total IRE binding activity in RBC, which was between 0.050 and 0.055 pmol/mg protein (Fig. 3), was much lower than in liver cytosol. Spontaneous IRP activity in RBC extracts was not affected by dietary iron intake (Fig. 3). In contrast, the level of 2-ME inducible IRE binding activity was significantly greater in the 2 and 11 mg iron/kg diet groups compared with the control group (Fig. 3).
Fig. 3.
Effect of dietary iron intake on spontaneous and 2-mercaptoethanol (2-ME)-induced RNA binding activity of iron regulatory proteins (IRPs) in RBC of rats. Spontaneous and 2-ME-induced RNA binding activity of IRPs was determined in RBC from rats fed various levels of dietary iron as well as pair-fed controls as described in the legend of Figure 1. The RNA binding activity was determined by gel shift analysis as described in Materials and Methods. Values illustrated by vertical bars are means ± SEM (n = 6). Values that do not share a superscript letter are significantly different (P < 0.05).
[View Larger Version of this Image (25K GIF file)]
UTR and hence is a putative target for the translational inhibitory action of IRP1 and/or IRP2. The effect of iron intake on the cellular aconitases was examined for two reasons. First, by determining the effect of iron intake on the activity of c-aconitase, we could obtain a direct measure of the abundance of the inactive RNA binding form of IRP1. Second, it would be possible to determine whether the changes in m-aconitase activity were consistent with its abundance being regulated by IRPs.
Fig. 4.
Effect of dietary iron intake on cytosolic and mitochondrial aconitase activity in rat liver. Aconitase activity was determined in liver cytosol (panel A) and isolated mitochondria (panel B) from rats fed various levels of dietary iron as well as pair-fed controls as described in Figure 1. Aconitase activity was determined by the coupled reaction of aconitase and isocitrate dehydrogenase as described (Rose and O'Connell 1967). One unit of enzyme activity is defined as 1 µmol of NADP+ reduced per minute. Data are means ± SEM (n = 6). Values that have a different superscript letter are significantly different (P < 0.05).
[View Larger Version of this Image (22K GIF file)]
, Hentze and Kühn 1996, Theil 1994), we wished to evaluate if there was a differential effect of iron intake on m-aconitase protein in relation to changes in ferritin subunit abundance.
).
Fig. 5.
Effect of dietary iron intake on relative abundance of m-aconitase or ferritin in rat liver. Immunoblotting of m-aconitase (panel A, lanes a through f) and ferritin (panel B, lanes g through l) was performed in isolated liver mitochondria and crude liver homogenate, respectively, obtained from rats fed various levels of dietary iron as well as pair-fed controls as described in Figure 1. Gel electrophoresis and immunblotting of m-aconitase and ferritin were performed as described in Materials and Methods. Identification of m-aconitase and H- and L-subunits of ferritin (denoted by the arrows) was based on their cross-reactivity with their specific antibody and by their comigration with purified bovine heart m-aconitase and rat liver ferritin (results not shown). Relative abundance of m-aconitase (panel C) or ferritin (the sum of the H and L-subunits) (panel D) across dietary groups was determined by densitometry and is expressed as percentage of the amount of m-aconitase or ferritin present in the extract from rats fed the control diet. Data are means ± SEM (n = 6). Values that have a different superscript letter are significantly different (P < 0.05). In panel D, ferritin protein was undetectable in rats fed 2 and 11 mg iron/kg diet and thus the two groups were not included in the statistical analysis.
[View Larger Version of this Image (33K GIF file)]
0.787, P < 0.0001) between liver cytosolic RNA binding activity and blood Hb levels (Fig. 6, panel A). The relationship between liver RNA binding activity and liver ferritin iron was biphasic (Fig. 6, panel B). The same relationship was observed between ferritin subunit levels and cytosolic RNA binding activity (results not shown). Most of the regulation of liver RNA binding activity occurred over a range of iron intake in which there was little change in ferritin iron or ferritin subunit levels. Ferritin iron levels did not increase significantly until RNA binding activity had reached its minimum level. In contrast, there was a strong inverse linear relationship (r = 0.751, P < 0.001) between liver RNA binding activity and m-aconitase protein levels (Fig. 6, panel C).
Fig. 6.
Relationship between cytosolic RNA binding activity in liver and blood hemoglobin concentration, liver ferritin iron concentration and m-aconitase abundance obtained from rats fed various levels of dietary iron as well as pair-fed controls as described in Figure 1. Individual dietary treatment groups of rats are denoted by the symbols indicated in panel A. The correlation coefficient (r) for panels A and C was 0.787 and 0.751, respectively. All correlations were significant (P <0.0001).
[View Larger Version of this Image (20K GIF file)]
). The evidence of a weaker interaction of IRPs with the IRE in eALAS (Cox et al. 1991) and m-aconitase (Zheng et al. 1992) mRNAs compared with their interaction with the ferritin IRE supports the application of this principle to proteins encoded by IRE-containing mRNAs. A similar hypothesis has been proposed for the regulation of ferritin synthesis and heme formation in erythroid cells (Cox et al. 1991).
, Zheng et al. 1992) mRNAs are a major determinant of the differential response of these proteins to changes in dietary iron intake. Thus, our results provide direct support for a hierarchical action of IRPs on different IRE-containing mRNAs.
).
). Somewhat surprisingly, we found that iron deficiency in vivo was associated with a slight increase in c-aconitase activity (Fig. 4, panel A). This was supported by results of the analysis of 2-ME inducible IRE RNA binding activity (Fig. 1, panel C). Similarly, Rabie et al. (1995) found that iron deficiency failed to depress c-aconitase in rat small intestine. Our observations that iron deficiency failed to decrease the levels of c-aconitase raises the issue of whether the increase in RNA binding activity of IRP1 observed under these conditions arose due to disassembly of the Fe-S cluster in pre-existing c-aconitase or if newly synthesized IRP1 was the basis for the increase in RNA binding activity in iron deficiency. The fact that the amount of 2-ME inducible RNA binding activity significantly increased in iron deficiency indicates that there is an increase in the amount of IRP1 in the cell. Perhaps more of the newly synthesized binding protein is diverted to the RNA binding form in iron deficiency. However, it is apparent from our results that only 1-3% of IRP1 in rat liver is present in the RNA binding form (Fig. 1, panels B and C). Thus, small changes in the abundance of the c-aconitase form of IRP1 could result in large changes in the abundance of the RNA binding form.
, Beutler 1959, Cartier et al. 1986). What role might m-aconitase perform in such situations? Are there additional examples of iron-dependent changes in m-aconitase activity and does this has an impact on cellular metabolism? Iron deficiency is associated with a decrease in pyruvate oxidation by hepatocyte mitochondria (Klempa et al. 1989), and there are increased lactate concentrations in blood and muscle (Finch et al. 1979, McLane et al. 1981). The decrease in pyruvate oxidation by hepatocyte mitochondria occurred in the presence of no change in citrate synthase activity measured in isolated mitochondria, indicating that metabolic steps including and/or beyond aconitase are impaired in iron deficiency (Klempa et al. 1989). To our knowledge, no studies have simultaneously examined the effect of iron intake, or other factors, on both the cytoplasmic and mitochondrial aconitases, but some studies have measured total cellular aconitase activity without specifically examining one of the two aconitase isoforms. There is a decline in total aconitase activity in kidney and in m-aconitase activity in muscle in iron deficiency (Beutler 1959, Ohira et al. 1987). In skeletal muscle, there is actually an increase in the activity of many of the tricarboxylic acid (TCA) cycle enzymes other than m-aconitase and the other iron-containing enzyme in the TCA cycle, succinate dehydrogenase, both of which decline in activity (Ohira et al. 1987). The results of Ohira et al. (1987) support the notion that m-aconitase activity may limit oxidation of acetyl CoA in iron deficiency. Although m-aconitase is not normally considered a rate-limiting component of the TCA cycle, small decreases in aconitase activity, in the presence of no change in other TCA cycle enzymes, have been associated with significant decreases in acetate oxidation by cells (Janero and Hreniuk 1996). Thus, it seems reasonable to suggest that iron-dependent reductions in m-aconitase abundance contribute to a reduction in TCA cycle capacity.
, Sherman 1978) which may occur due to an impaired ability to oxidize acetyl-CoA. Taken together, it seems reasonable to hypothesize that a reduction in m-aconitase activity induced by IRPs represents an adaptive response such that when oxidation of mitochondrial reducing equivalents is impaired, then acetyl CoA can be diverted into fatty acid synthesis by reducing mitochondrial metabolism of citrate. A second potential reason for the decline in m-aconitase abundance in iron deficiency may relate to the overall functional impairment of mitochondria in iron deficiency and the effects this may have on free radical generation (Cartier et al. 1986). Mitochondria from liver or muscle of iron-deficient rats exhibit a greater degree of uncoupling than do mitochondria from normal rat muscle (Maguire et al. 1982, Masini et al. 1994, Willis and Dallman 1989). Because mitochondria are a major cellular source of oxygen radicals under normal circumstances, increased uncoupling may contribute to a greater likelihood of incomplete reduction of oxygen, resulting in increased cellular levels of free radicals. By reducing m-aconitase abundance, IRPs may be acting in part to diminish the supply of reducing equivalents to the electron transport chain in order to lower production of oxygen radicals. It is of interest that IRP1 responds to oxidative stress, and it has been hypothesized that this IRP may act as one level of the cellular antioxidant defense systems (referenced in Hentze and Kühn 1996).
-untranslated region of mammalian mitochondrial aconitase. J. Biol. Chem. 271: 24226-24230.
FOOTNOTES
Manuscript received 23 July 1996. Initial reviews completed 10 September 1996. Revision accepted 9 October 1996.
ACKNOWLEDGMENTS
We thank Dan Steffen for excellent technical assistance and Ken Blemings for helpful comments.
LITERATURE CITED
This article has been cited by other articles:
|
|
 |
|
  M. A. Khan, W. E. Walden, D. J. Goss, and E. C. Theil Direct Fe2+ Sensing by Iron-responsive Messenger RNA{middle dot}Repressor Complexes Weakens Binding J. Biol. Chem., October 30, 2009; 284(44): 30122 - 30128. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. M. Deck, A. Vasanthakumar, S. A. Anderson, J. B. Goforth, M. C. Kennedy, W. E. Antholine, and R. S. Eisenstein Evidence That Phosphorylation of Iron Regulatory Protein 1 at Serine 138 Destabilizes the [4Fe-4S] Cluster in Cytosolic Aconitase by Enhancing 4Fe-3Fe Cycling J. Biol. Chem., May 8, 2009; 284(19): 12701 - 12709. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. B. Zumbrennen, M. L. Wallander, S. J. Romney, and E. A. Leibold Cysteine Oxidation Regulates the RNA-Binding Activity of Iron Regulatory Protein 2 Mol. Cell. Biol., April 15, 2009; 29(8): 2219 - 2229. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. J. Niles, M. S. Clegg, L. A. Hanna, S. S. Chou, T. Y. Momma, H. Hong, and C. L. Keen Zinc Deficiency-induced Iron Accumulation, a Consequence of Alterations in Iron Regulatory Protein-binding Activity, Iron Transporters, and Iron Storage Proteins J. Biol. Chem., February 22, 2008; 283(8): 5168 - 5177. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. E. Schroeder, M. B. Reddy, and K. L. Schalinske Retinoic Acid Modulates Hepatic Iron Homeostasis in Rats by Attenuating the RNA-Binding Activity of Iron Regulatory Proteins J. Nutr., December 1, 2007; 137(12): 2686 - 2690. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A.-S. Zhang, S. A. Anderson, K. R. Meyers, C. Hernandez, R. S. Eisenstein, and C. A. Enns Evidence That Inhibition of Hemojuvelin Shedding in Response to Iron Is Mediated through Neogenin J. Biol. Chem., April 27, 2007; 282(17): 12547 - 12556. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. E. Walden, A. I. Selezneva, J. Dupuy, A. Volbeda, J. C. Fontecilla-Camps, E. C. Theil, and K. Volz Structure of Dual Function Iron Regulatory Protein 1 Complexed with Ferritin IRE-RNA Science, December 22, 2006; 314(5807): 1903 - 1908. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Pondarre, B. B. Antiochos, D. R. Campagna, S. L. Clarke, E. L. Greer, K. M. Deck, A. McDonald, A.-P. Han, A. Medlock, J. L. Kutok, et al. The mitochondrial ATP-binding cassette transporter Abcb7 is essential in mice and participates in cytosolic iron-sulfur cluster biogenesis Hum. Mol. Genet., March 15, 2006; 15(6): 953 - 964. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Rincker, S. L. Clarke, R. S. Eisenstein, J. E. Link, and G. M. Hill Effects of iron supplementation on binding activity of iron regulatory proteins and the subsequent effect on growth performance and indices of hematological and mineral status of young pigs J Anim Sci, September 1, 2005; 83(9): 2137 - 2145. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-F. Liew and N.-S. Shaw Mitochondrial Cysteine Desulfurase Iron-Sulfur Cluster S and Aconitase Are Post-transcriptionally Regulated by Dietary Iron in Skeletal Muscle of Rats J. Nutr., September 1, 2005; 135(9): 2151 - 2158. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Robb and M. Wessling-Resnick Regulation of transferrin receptor 2 protein levels by transferrin Blood, December 15, 2004; 104(13): 4294 - 4299. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. S. Eisenstein and K. L. Ross Novel Roles for Iron Regulatory Proteins in the Adaptive Response to Iron Deficiency J. Nutr., May 1, 2003; 133(5): 1510S - 1516. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. C. Theil Ferritin: At the Crossroads of Iron and Oxygen Metabolism J. Nutr., May 1, 2003; 133(5): 1549S - 1553. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Erlitzki, J. C. Long, and E. C. Theil Multiple, Conserved Iron-responsive Elements in the 3'-Untranslated Region of Transferrin Receptor mRNA Enhance Binding of Iron Regulatory Protein 2 J. Biol. Chem., November 1, 2002; 277(45): 42579 - 42587. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. S. Chen, S. Hemenway, and J. Kaplan Inhibition of Fe-S cluster biosynthesis decreases mitochondrial iron export: Evidence that Yfh1p affects Fe-S cluster synthesis PNAS, September 17, 2002; 99(19): 12321 - 12326. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. L. Ross and R. S. Eisenstein Iron Deficiency Decreases Mitochondrial Aconitase Abundance and Citrate Concentration without Affecting Tricarboxylic Acid Cycle Capacity in Rat Liver J. Nutr., April 1, 2002; 132(4): 643 - 651. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. B. Walter, M. D. Knutson, A. Paler-Martinez, S. Lee, Y. Xu, F. E. Viteri, and B. N. Ames Iron deficiency and iron excess damage mitochondria and mitochondrial DNA in rats PNAS, February 19, 2002; 99(4): 2264 - 2269. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. J. Pinero, N. Li, J. Hu, J. L. Beard, and J. R. Connor The Intracellular Location of Iron Regulatory Proteins Is Altered as a Function of Iron Status in Cell Cultures and Rat Brain J. Nutr., November 1, 2001; 131(11): 2831 - 2836. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. K. Nanda and J. L. Leibowitz Mitochondrial Aconitase Binds to the 3' Untranslated Region of the Mouse Hepatitis Virus Genome J. Virol., April 1, 2001; 75(7): 3352 - 3362. [Abstract] [Full Text]
|
|
|
|
|
|
C. L. Kwik-Uribe, D. Gietzen, J. B. German, M. S. Golub, and C. L. Keen Chronic Marginal Iron Intakes during Early Development in Mice Result in Persistent Changes in Dopamine Metabolism and Myelin Composition J. Nutr., November 1, 2000; 130(11): 2821 - 2830. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. S. Eisenstein and K. P. Blemings Iron Regulatory Proteins, Iron Responsive Elements and Iron Homeostasis J. Nutr., December 1, 1998; 128(12): 2295 - 2298. [Abstract] [Full Text]
|
|
|
|
|
|
Y. Ke, J. Wu, E. A. Leibold, W. E. Walden, and E. C. Theil Loops and Bulge/Loops in Iron-responsive Element Isoforms Influence Iron Regulatory Protein Binding. FINE-TUNING OF mRNA REGULATION? J. Biol. Chem., September 11, 1998; 273(37): 23637 - 23640. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. S. Chen, K. P. Blemings, K. L. Schalinske, and R. S. Eisenstein Dietary Iron Intake Rapidly Influences Iron Regulatory Proteins, Ferritin Subunits and Mitochondrial Aconitase in Rat Liver J. Nutr., March 1, 1998; 128(3): 525 - 535. [Abstract] [Full Text]
|
|
|
|
|
|
K. L. Schalinske, O. S. Chen, and R. S. Eisenstein Iron Differentially Stimulates Translation of Mitochondrial Aconitase and Ferritin mRNAs in Mammalian Cells. IMPLICATIONS FOR IRON REGULATORY PROTEINS AS REGULATORS OF MITOCHONDRIAL CITRATE UTILIZATION J. Biol. Chem., February 6, 1998; 273(6): 3740 - 3746. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Ke and E. C. Theil An mRNA Loop/Bulge in the Ferritin Iron-responsive Element Forms in Vivo and Was Detected by Radical Probing with Cu-1,10-phenantholine and Iron Regulatory Protein Footprinting J. Biol. Chem., January 18, 2002; 277(4): 2373 - 2376. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Caltagirone, G. Weiss, and K. Pantopoulos Modulation of Cellular Iron Metabolism by Hydrogen Peroxide. EFFECTS OF H2O2 ON THE EXPRESSION AND FUNCTION OF IRON-RESPONSIVE ELEMENT-CONTAINING mRNAs IN B6 FIBROBLASTS J. Biol. Chem., June 8, 2001; 276(23): 19738 - 19745. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Narahari, R. Ma, M. Wang, and W. E. Walden The Aconitase Function of Iron Regulatory Protein 1. GENETIC STUDIES IN YEAST IMPLICATE ITS ROLE IN IRON-MEDIATED REDOX REGULATION J. Biol. Chem., May 19, 2000; 275(21): 16227 - 16234. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. C. Theil and R. S. Eisenstein Combinatorial mRNA Regulation: Iron Regulatory Proteins and Iso-iron-responsive Elements (Iso-IREs) J. Biol. Chem., December 22, 2000; 275(52): 40659 - 40662. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
