| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
* Department of Neurosurgery, G.M. Leader Family Laboratory for Alzheimer’s Disease Research,
Department of Pharmacology, The Pennsylvania State University College of Medicine, Hershey, Pennsylvania, USA
2Correspondence: Department of Neurosurgery, H110, Pennsylvania State University College of Medicine, 500 University Dr. (H110), Hershey, PA 17033-0850, USA. E-mail: jconnor{at}psu.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: gene expression • iron-responsive elements • iron regulatory proteins • quantitative real-time polymerase chain reaction • oxidative stress
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, and it is the most common Mendelian inherited disorder in northern Europeans, with a prevalence of 1:200 to 1:500 (2
, 3)
. An even higher prevalence of 1:100 is likely in the Irish population (4
5
6
7)
.
The HFE gene encodes for a novel 343 amino acid major histocompatibility complex class 1 molecule. Initially, two missense mutations were identified in the HFE gene, which is located on the short arm of chromosome 6 (8)
. The most common mutations are known as C282Y and H63D mutations (9
, 10)
. More than 80% of hemochromatosis patients are homozygous for the C282Y mutation (8
, 11
, 12)
. Although the C282Y mutation occurs less frequently in the general population than the H63D mutation (1.9% vs. 8.9%), the C282Y mutation is more frequently associated with clinical HH (1
, 13)
. The frequency of the H63D mutation is reportedly increased in patients with Alzheimer’s disease (AD) and amyotrophic lateral sclerosis (ALS) (14
15
16
17
18
19
20
21)
, though there are also studies that did not find an increased frequency (22
23
24
25)
. This discrepancy may reflect a gene/environment interaction. Therefore, we developed stable cell lines carrying the C282Y or H63D mutation as a possible model in which to test hypotheses about how the different mutations affect the intracellular environment and the effect of epigenetic influences on the cells with mutations.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cell culture and cloning
Human neuroblastoma SH-SY5Y cells were maintained in DMEM/F12 media supplemented with 10% FBS, 1% antibiotics, 1x nonessential amino acid, and 1.8 g/L sodium bicarbonate. Human astrocytoma U87-MG and embryonic kidney HEK293 cell lines were ordered from American Type Culture Collection (ATCC; Rockville, MD, USA) and maintained in DMEM with 4 mM L-glutamine, 100 U/ml penicillin, 100 µg/ml streptomycin, and 10% FBS. Human epithelial HeLa cells were obtained from Dr. Shao-Cong Sun (Pennsylvania State University, Hershey, PA, USA) and cultured in DMEM as described above. All experiments were performed at 37°C in 5% CO2 atmosphere culture conditions.
Transfection
The HFE cDNA was fused to the FLAG octapeptide sequence to more easily detect the HFE protein. H63D and C282Y mutant cDNA were generated from WT HFE cDNA using a site-directed mutagenesis kit (Stratagene, San Diego, CA, USA) (20)
. The description of the transfection protocol for the SH-SY5Y cell line has been reported previously (20)
. Stable cell lines were obtained following transfection with WT, H63D or C282Y forms of HFE using the lipid transfection reagent, Lipofectamine (Invitrogen). The clones were confirmed for expression of HFE gene by PCR and FLAG/HFE antibody by immunoblotting. Western blot analysis on whole cell homogenates was used to identify differences in protein expression.
Assay of labile iron pool
The status of the labile iron pool was monitored with the iron-sensitive dye calcein (26
27
28)
. After culture of HFE-transfected cells in 96-well plates for 2 days, cells were incubated with 0.25 µM calcein acetoxymethyl ester dye (Molecular Probes, Inc., Carlsbad, CA, USA) at 37°C for 20 min, then rinsed twice with Hanks’ buffer to remove extra calcein. The fluorescence intensity was measured using a fluorescent plate reader and compared with samples without calcein treatment. The cells were subsequently harvested and analyzed for total protein concentration.
IRP activity
One of the most sensitive indicators of cellular iron status is iron regulatory protein (IRP) activity. IRPs bind to an iron response element (IRE) on select mRNAs (29)
. To assay this activity, we used two methods. First, we generated an IRE-driven fluorescent probe using red fluorescent protein (RFP) based on an IRE-cyan fluorescent protein construct reported previously (30)
. This method allows for detection of IRP activity in living cells. To use the fluorescent probe, 2 µg of pDsRed2-N1 vector or IRE-RFP construct was transfected into the different stable cell lines (2x106 cells) using the nucleofection system (Amaxa BioSystems, Gaithersburg, MD, USA), then cultured in 96-well plates. Fluorescence was determined during a 4 day period.
The second method for determining IRP activity was a RNA band shift assay. A synthetic RNA transcript containing the IRE was generated from the oligonucleotide template 5'-ttatgctgagtgatatccctctcctaggacgaagttgtcacgaacctgcctagg-3' with T7 polymerase in the presence of [32P]cytidine triphosphate. Binding reactions were carried out as described previously (31)
. Briefly, 5 µg of protein was incubated with the synthetic radiolabeled IRE probe. To measure total IRP binding activity, samples were treated with 2% ß-mercaptoethanol. The RNA-protein complexes were separated on a 4% native polyacrylamide gel and visualized by autoradiography. The autoradiograms were digitized using a Bio-Rad G5800 Calibrated Densitometer. The densitometric image obtained was then analyzed using Quantity One software. The results of the RNA band shift assays were expressed as active IRP binding from each treatment normalized to the IRP binding obtained in SW1088 cells that were used as a control.
Gene array experiment
We used the GEArray Focused DNA Microarray system (SuperArray Co., Frederick, MD, USA) to observe altered gene expression in HFE stably transfected cells. The microarrays were performed according to the manufacturer’s instructions. Briefly, RNA was extracted from cultured cell lines using an RNA purification kit (Qiagen, Chatsworth, CA, USA), and 5 µg of RNA was used in a 20 µl cDNA reaction mixture. Intensity of gene expression was quantified by web-based GEAnalysis Suite software (SuperArray). Differences in expression were based on normalization to 18S rRNA. Each experiment was performed at least twice.
Quantitative real-time PCR
For quantitative real-time PCR, SYBR Green PCR Master Mix (SuperArray) and ROX reference dye (Invitrogen) were used. For signal detection, the ABI Prism 7700 Sequence Detector System was programmed with an initial sterilization step of two minutes at 50°C, followed by 15 min of denaturation at 95°C and 40 cycles for 30 s at 95°C, 30 s at 55°C and 30 s at 72°C. Each reaction sample was run in triplicate, and reaction efficiency estimates were derived from standard curves that were generated using serial dilutions of the corresponding plasmid for each quantitative real-time PCR set. A gene expression normalization factor was calculated for each sample based on the geometric mean of user-defined housekeeping genes.
PCR and RT-PCR
To confirm alterations in gene expression, some genes were selected for RT-PCR. For these studies, total RNA was extracted using a RNA purification kit (Qiagen) and 5 µg of total RNA was used in a 20 µl cDNA reaction mixture. PCR was performed on cDNA and the sham control reactions. The PCR reaction was initially denatured at 94°C for 3 min, then subjected to 30 or 40 cycles of denaturation at 94°C for 45 s, annealing at 55°C for 45 s, and extension at 72°C for 90 s before a final extension at 72°C for 10 min. The PCR product was analyzed in 2% agarose gel electrophoresis and NIH image software.
To determine the HFE mRNA expression level in HFE stably transfected SH-SY5Y cell lines, the plasmid DNA was purified from HFE stably transfected SH-SY5Y cell lines by modified QIAprep procedure (Qiagen). The PCR was performed using either cDNA or plamid DNA to make an 1101 bp full-length HFE PCR product (20)
, then the PCR products were analyzed in 1% agarose gel.
Western blot
The expression of HFE, transferrin receptor, and ferritin at the protein level in the human cell lines was performed on gradient gels (4–20%) following standard procedures as described earlier (28)
. The antibodies used in this study were FLAG (1:2000) for detection of HFE, Tf receptor (1:1000), or H-ferritin (HS-59, 1:500) and ß-actin (1:5000). Differences in protein expression were determined using chemiluminescent reagents (KPL).
Cell viability assay
Cells were plated at a density of 2 x 104 cells/well in 96-well flat-bottomed microtiter plates, then cultured for 2 days. The cells were exposed to DFO (1 or 10 µM) and/or hydrogen peroxide for 24 h. All studies were performed at least in triplicate. The concentration of hydrorgen peroxide for the cell viability assays was determined empirically. Cell viability was assessed with the colorimetric MTS assay (Promega, Chatsworth, CA, USA).
Detection of cell stress
The degree of lipid peroxidation in the various cell lines was determined by measuring the concentrations of 8-isoprostane (EIA Kit, Cayman Labs, Ann Arbor, MI, USA). The concentration of oxidatively modified proteins was detected with an OxyBlot kit (Chemicon, El Segundo, CA, USA). The loss of mitochondrial membrane potential was measured by the JC-1 assay kit (BioCarta, San Diego, CA, USA).
Statistical analysis
The data were analyzed by one-way ANOVA and the Student’s t test. Differences among means were considered statistically significant when the P value was < 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. These results were confirmed and extended to the mRNA level in this study (Fig. 1
A, B). In addition, U87-MG, HeLa, and HEK293 cells were included as positive controls. HeLa cells and the human astrocytoma U87-MG cells expressed similar levels of HFE mRNA (Fig. 1A
). The level of mRNA for HFE in the HEK293 was barely detectable. HFE cDNA and protein expression can be demonstrated following stable transfection of the SH-SY5Y cells with WT HFE or the H63D or C282Y HFE mutants (Fig. 1C, D
). The HFE gene was not detected in vector and mock control transfected cells with PCR. At the protein level, expression of the C282Y mutant form of HFE was routinely greater per milligram of total cell protein than the H63D mutant or the WT HFE. The latter two proteins were expressed in similar amounts. HFE protein was not detected in the vector control (Fig. 1D
).
|
Effect of HFE expression on cellular iron status
Compared with cells carrying the WT HFE, H63D mutant forms of HFE have 30% higher levels of Tf receptor, whereas cells carrying the C282Y mutation have 96% less expression of Tf receptor (Fig. 1D, E
). Ferritin expression in cells transfected with WT HFE or H63D are similar (85.33±2.09 vs., 83.03±1.83); however, ferritin is 90% less in cells transfected with C282Y (Fig. 1D, F
).
Because of the differences in baseline expression of HFE protein among the different cell lines, we normalized the expression of Tf receptor and ferritin to HFE expression to ensure the effects were not linked to the level of HFE expression (Fig. 1G, H
). Cells carrying the H63D mutant form of HFE have a 40% increase in Tf receptor expression when normalized to HFE, whereas cell lines carrying the C282Y mutation have a 150% decrease (Fig. 1G
). Ferritin expression in cells transfected with WT HFE or H63D is similar (1.19±0.03 vs. 1.04±0.02); however, it is decreased by 95% in cells transfected with C282Y (Fig. 1H
). These results suggest that the differences in expression of these proteins are not based on the amount of HFE expression.
To directly examine the effect of HFE expression on the cellular iron status, we determined the relative amounts of the labile iron pool using the iron-sensitive dye calcein. Calcein fluorescence is quenched by iron binding in the assay. Therefore, the increase in calcein fluorescence indicates a decrease in the labile iron pool. As shown in Fig. 2
, both cell lines carrying the mutant HFE gene have increased labile iron pools compared with the WT cell line (45.7±6.2 vs. 7.7±1.2 for H63D mutant cell line and 45.7±6.2 vs. 0.2±0.1 for C282Y mutant cell line). Cells transfected with WT HFE have less iron in the labile iron pool than in vector-transfected cells (45.7±6.2 vs. 27.5±3.2). The amount of iron in the labile iron pool of vector-transfected cells and nontransfected cells is not different (27.5±3.2 vs. 30.1±3.3).
|
To determine whether the differences in the labile iron pool are sufficient to alter the regulatory system, we determined IRP activity. To directly assess IRP activity, cell homogenates were probed with 32P-IRE. The IRP/IRE binding activity was similar for the cell lines with WT and H63D HFE as well as the vector control and the nontransfected control cells. Those cells carrying the C282Y HFE mutant gene, however, had significantly increased activity (Fig. 3
A). The total amount of IRP was similar in all cell lines (Fig. 3B
). The IRP activity was also evaluated using an IRE-driven RFP to reveal the difference in IRP activity in living cells. The expression of the RFP vector control and the IRE-RFP vector was similar over the 4 days examined in all cell lines except those carrying the C282Y mutation (Fig. 3C-G
). The presence of the C282Y mutation was associated with less RFP than the RFP vector control (Fig. 3G
) which is consistent with higher IRP binding activity. It was noted that the baseline expression of RFP was 5-fold higher in cells carrying the C282Y mutation than in the other transfected cells, suggesting that transcription rates for the C282Y cells are relatively higher.
|
Gene expression profiling of the HFE-transfected SH-SY5Y cells
There were a number of differences in gene expression patterns between the cell lines transfected with WT, H63D, or C282Y when compared with vector-transfected cells (Table 1
). We set a minimum limit of > 2-fold change for the expression of a gene to be considered significantly different. When the gene arrays were compared for the mutant-carrying cell lines against the cell lines transfected with WT HFE, there were fewer differences. Only one gene, GM-CSF (colony-stimulating factor 2), was changed between H63D cell lines and WT. On the other hand, the expression of nine genes was different between the C282Y and WT cell lines. Of these nine, six were in the signal transduction pathway array. Four of the genes in this pathway were up-regulated including Cox-2, whose change in expression was greater than for any other gene, followed by PKCß (protein kinase C, beta 1). The two down-regulated genes in this pathway were CRBP1 (retinol binding protein 1, cellular) and ß-casein. The second greatest difference in expression between C282Y and WT was found in the lipoxygenase (LOX) gene (lysyl oxidase, 11.2-fold increase). This was the only gene that was altered in the metal transport and homeostasis array when the mutant containing cells were compared with WT. HFE mRNA expression was not altered when the mutant cell lines were compared with WT.
|
In comparing the HFE-transfected cell lines to the vector transfection a broader change in expression profiling was noted. Nine genes in the stress and toxicity pathway were increased in the WT cell line compared with vector; none of the genes in this pathway were changed when H63D was compared with vector, and only three were altered in the C282Y cell line compared with vector. In contrast, the expression of six genes in the metal transport and homeostasis pathway was altered by the presence of the C282Y mutation vs. vector, but only one gene (HFE) was changed in this pathway in the WT and two genes in the H63D.
The most consistent differences in gene expression patterns were seen when the HFE-transfected cell lines were compared with vector in the stress response to cellular damage array. The three genes in this pathway were down-regulated by the presence of HFE (WT or mutant).
Genes in the signal transduction pathway were responsive to the presence of HFE and the response was clearly dependent on the type of HFE. The C282Y cell line was associated with more changes in genes in the signal transduction pathway (6/10) than the other two cell lines. Only one gene, SWIP-1 was decreased, and Cox-2 mRNA was elevated in the C282Y cell line 10X.
PCR confirmation of gene arrays
The gene array results were validated using quantitative real-time PCR and RT-PCR. The oligonucleotide sequences used for PCR are summarized in Table 2
. The most consistently expressed gene among the cell lines was 18S rRNA (data not shown), so this gene was chosen for normalization of expression of the other genes. Using this quantitative approach we observed a significant increase in lysyl oxidase gene (LOX, 38.08±6.25) from the metal transport and homeostasis array and Cox-2 gene (57.44±16.25) from the signal transduction array. We also confirmed that retinol binding protein 1 gene (CRBP1, 15.14±1.82) was significantly decreased (Fig. 4
A). With RT-PCR, RNA expression of lysyl oxidase gene was increased 16.73 ± 3.81-fold, and Cox-2 mRNA increased 68.46 ± 23.25-fold; retinol binding protein 1 (CRBP1) gene decreased 13.30 ± 3.10-fold (Fig. 4B
).
|
|
Effect of HFE mutation on the oxidative stress
The altered gene profiles in the HFE-transfected cells suggested that the presence of HFE could result in different responses to cell stress. The C282Y-transfected cells were more resistant to hydrogen peroxide exposure than the other cell lines (Fig. 5
A).
|
The basal rates of oxidative stress were different for the mutant-carrying cell lines compared with control. The amount of lipid peroxidation (Fig. 5B
) and oxidatively modified proteins (Fig. 5C
) was decreased in the C282Y-carrying cells compared with the other cell lines. H63D cell lines have higher levels of oxidatively modified proteins compared with WT HFE cells (81.0±3.1 vs. 74.9±3.9, Fig. 5C
). The H63D cell line also has a lower resting level of mitochondrial membrane potential than the C282Y cell line or any of the controls (Fig. 5D-F
).
Effect of Iron chelation on cellular toxicity and iron status
Because iron chelation is capable of protecting cells from oxidative stress (32
, 33)
and HFE mutations altered intracellular iron status, we exposed different cells lines to an iron chelator, Desferal (DFO), to test the hypothesis that the HFE mutant cells were more resistant to protection by iron chelation than the WT HFE cells. First, we exposed the cell lines to high or low concentrations of DFO to determine the effect of iron chelation on cell viability. The low concentration of DFO 1 µM had no effect on any of the cells. The higher concentration (10 µM) had no effect on the WT HFE cells (Fig. 6
A), provoked a modest but significant increase in the number of H63D cells (Fig. 6B
), and was toxic to the C282Y cells (Fig. 6C
). DFO protected the WT HFE (Fig. 6A
) and H63D (Fig. 6B
) -carrying cell lines from exposure to 100 µM hydrogen peroxide but not the C282Y cell line (Fig. 6C
). Concentrations of DFO did not change the labile iron pool in the WT cells but decreased the labile iron pool in the H63D cell line (Fig. 6E
). An increase in the labile iron pool was observed when the C282Y cells were exposed to 10 µM DFO. Treatment with 10 µM DFO decreased the labile iron pool in the presence of hydrogen peroxide in each cell line (Figs. 6D-F
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, the response to cytokines (33)
, and vulnerability to oxidative stress (35
36
37)
. Consistent with this latter idea, the present study shows that the presence of the HFE mutation alters the basic levels of oxidative stress, mitochondrial membrane potential, and response to iron chelation.
We selected the human neuroblastoma cell line because HFE expression was not detected in the untransfected cells at the RNA or protein level and because of our interest in the role of HFE in neurodegenerative disorders (21
, 38)
. Most studies examining the functional significance of HFE in cell models have used the human epithelial HeLa and embryonic kidney HEK293 cell lines. HEK293 cells reported do not express detectable levels of HFE protein (39)
, but we could detect HFE mRNA expression with RT-PCR as shown in Fig. 1A
. The effect of WT HFE on Tf receptor and ferritin in the neuroblastoma cells are consistent with those reported for HeLa but not consistent with the changes reported for IRP activity (40
41
42)
. These data suggest signaling from the labile iron pool, or the utilization of iron in the labile iron pool could be different between the two cell types. The decrease in ferritin in the neuroblastoma cells is consistent with that found in HEK293 cells (39
, 43)
but the Tf receptor expression was unchanged in the HEK293 cells, indicating further cell specific differences.
The effect of expressing the mutant forms of HFE on intracellular iron status and the iron management proteins has been determined in HeLa, HEK293, and human lung cancer H1299 cells (39
, 43
, 44)
. The impact of the mutations on iron parameters in these cells and our neuroblastoma cells appears to be both mutation and cell specific. For example, in HEK293 cells transfected with the HFE mutants, the Tf receptor was not changed in association with either mutation (43)
. In H1299 cells, the expression of H63D or WT HFE was associated with an increase in Tf receptors, but the C282Y mutation did not impact the Tf receptor expression (44)
. In the neuroblastoma cells, Tf receptor expression was increased in the presence of the H63D mutation but decreased in the cell lines carrying the C282Y allele.
Similar mutation and cell specific findings have been reported for ferritin and IRP activity. In HEK 293 cells neither mutation in HFE is associated with a change in ferritin expression. In H1299 cells, ferritin synthesis and concentration were not affected by the C282Y mutation but are decreased when the H63D mutation is present (44)
. In the neuroblastoma cell lines, the expression of ferritin was not different in the cells carrying the H63D mutation or WT HFE, but ferritin was dramatically decreased in cells carrying the C282Y mutation. WT and H63D HFE cells had increased IRP activity in H1299, but the C282Y mutation had no effect (44)
. In the neuroblastoma cell lines, the opposite finding occurred: the H63D mutation had no effect but the C282Y mutation was associated with a 100% increase in binding activity.
The explanation for cell-specific effects of the mutation is hampered by the lack of complete understanding of the function of the HFE protein. The most parsimonious explanation is that the different cell types have differences in iron metabolism that direct intracellular iron utilization in different ways. The iron storage capacity and relative concentration of ferritin in the different cell lines could also play a role in the outcome of these studies. For example, in vivo studies have shown that neurons express relatively low levels of ferritin and predominantly H-ferritin (45)
. It is also possible that some of the differences between our results and that for the non-neuronal cells were influenced by the endogenous expression of WT HFE in the other cell types.
To explore the impact of the HFE mutation on basic cell functions, we began by examining the iron regulatory system in the cells. Expression of the Tf receptor and ferritin are regulated by IRPs that are responsive to the labile iron pool. To monitor IRP activity, we used an IRE-driven reporter system and a gel shift assay. There were no differences in IRP activity between vector control cells and the WT cells or H63D cell lines with either measure. IRP activity was higher in the C282Y cell lines and the expression of the IRE-driven reporter was lower, which is consistent with more IRP activity and a low labile iron pool (46)
. The C282Y-carrying cells, however, had a much higher level of the labile iron pool than the other cell lines, and the baseline expression of the IRE-driven probe was also higher. These observations coupled with the decreased expression of ferritin in the C282Y cell lines suggest that the C282Y cells may be more transcriptionally and metabolically active than the WT or H63D cell lines. The decrease in the Tf receptor is consistent with the higher labile iron pool, but not with the increased IRP activity. The apparent disconnection between the labile iron pool and Tf receptor expression suggests that the labile iron pool is chronically elevated (perhaps by decreasing ferritin) as a result of the C282Y mutation and that the expression of the Tf receptor may have adjusted accordingly. Any changes that occurred in the level of expression in Tf receptor mRNA were below the 2-fold cutoff we had established as necessary to be considered significant. Therefore, changes in expression at the protein level may reflect changes at the post-transcriptional level and IRP activity. At this time, however, the data indicate that the C282Y mutation can disrupt the relationship between the Tf receptor mRNA and Tf receptor protein. This observation may be fundamentally important to understanding the consequence of the C282Y mutation on iron overload disorders and cellular compartmentalization of iron.
The differences in the expression levels of HFE in our stable cell lines are noteworthy. Similar levels of expression between WT and H63D and the relative overexpression of C282Y mutant HFE suggest a feedback system that has been fundamentally altered in the presence of the C282Y mutation. Alternatively, the elevated expression of HFE in the C282Y cells could be associated with a generalized increase in transcription in the C282Y cells. This idea is supported by the relative increase in baseline expression of the IRE-RFP probe in the C282Y cells. The notion of increased transcription in the C282Y mutant-carrying cells is under investigation, and if proven to be true, could be relevant to the increased frequency of the C282Y mutation reported in a number of cancers (47)
.
To more broadly assess the effect of the HFE protein or its mutant forms on a cell, we used targeted gene arrays. We previously demonstrated that changing the cellular iron status can have a significant impact on gene expression patterns in cells (48)
. A targeted gene array approach was taken because of the relation of HFE to metal homeostasis, the hypothesis that HFE mutations may effect cell stress response, and the possibility that signal transduction could be altered because HFE protein is thought to be a membrane protein. In the metal transport and homeostasis array, a number of genes were up-regulated when HFE-containing cells were compared with vector controls; most of these were associated with the C282Y mutation. Only the LOX gene (LOX, lysyl oxidase) was increased when C282Y cells were compared with WT or H63D cell lines. Lysyl oxidase, a copper-responsive gene product, is an extracellular enzyme that catalyzes the synthesis of lysine and hydroxylysine. Lysyl oxidase is localized within the cytoplasm of neurons (49)
. Elevated expression of lysyl oxidase is found in ALS patients and in the transgenic G93A SOD1 (mSOD1) mouse (49
, 50)
. The elevated expression of LOX in the HFE mutant cell line may be relevant to the studies of HFE mutations in ALS.
One gene in this pathway, ß-2M (ß-2-microglobulin), was increased in both the H63D and C282Y cell lines, but not in the cell line expressing WT HFE. ß-2M forms a complex with the HFE protein and the Tf receptor on the cell membrane (51)
, and the increase in expression of this gene in the presence of the mutant forms of HFE suggests a possible feedback relationship. Reportedly, the C282Y mutant form of HFE fails to migrate to the cell surface and complex with ß-2M and the Tf receptor whereas the H63D mutant does form the complex (40
, 52)
. Perhaps the elevated expression of ß-2M suggests there is a feedback system between HFE and ß-2M that fails in the presence of a mutation in the HFE gene.
Other genes in the metal transport and homeostasis pathway that were increased include an ABCB7 (ATP binding cassette subfamily B member 7) and thioredoxine peroxidase (peroxiredoxin 4) genes. The ABCB7 gene product is expressed in the mitochondria and may play a role in heme transport along with the regulation of iron-sulfur cluster containing proteins. The other increased gene is thioredoxine peroxidase (Peroxiredoxin 4). Peroxiredoxin 4 is involved in the NF-
B signaling pathway through phosphorylation of IB
, and also plays a role in protecting protein free thiol groups against oxidative damage and thioredoxin-dependent peroxidase activity. That increased ABCB7 and thioredoxine peroxidase gene expression occurred only in the C282Y cell lines is further evidence that the presence of the HFE mutation induces specific changes in cell energy production, which is perhaps why the C282Y cell line is relatively intolerant to iron chelation. The gene product of thioredoxine peroxidase is also involved in drug resistance (53)
suggesting that the relative increased expression of these genes could affect pharmacological-based treatment strategies.
Within the stress and toxicity pathway, almost all of the genes responding occurred in the cell line expressing WT HFE, and all were elevated in comparison to the vector control. This was an unexpected response in the context of iron cell biology because the data indicate (Fig. 2)
there is a decrease in iron in the labile iron pool as the result of expressing WT HFE. In general, a decrease in the labile iron pool is associated with a reduction in cell stress (32)
. Based on the data, it appears that expression of HFE is associated with an increase in cell stress. However, the FLICE gene (caspase 8) is not up-regulated in the WT HFE cell line, which indicates that although cell stress may be elevated, the caspase system was not activated. Caspase 8 mRNA expression was one of three genes whose expression was elevated in the C282Y cell line compared with the vector control. The difference in the mRNAs expressed in the C282Y cell line and WT HFE cell line clearly suggest a difference in baseline stress and possibly the response to stress. The elevation of these genes may also suggest that the cell lines carrying the WT HFE may have heightened ability to handle stress and thus may be less vulnerable than the H63D cell line in which no change in gene expression was noted compared with vector control. Possibly consistent with this line of reasoning is the decrease in genes related to response to cell damage in all of the HFE-carrying cell lines compared with vector control.
The functional cellular analyses support the novel idea presented in this study and in the gene array data that the presence of the HFE mutation fundamentally alters the stress levels in the cells. Furthermore, the data indicate that the different HFE mutations have different effects. The C282Y-carrying cells have less lipid peroxidation and oxidatively modified proteins than the other cell lines and show greater resistance to the presence of a stress factor. In contrast, the other HFE mutant cell lines carrying H63D have more oxidatively modified proteins and a relative decrease in mitochondrial membrane potential. These data may be relevant to the observations that H63D mutations have an increased frequency in neurodegenerative diseases such as Alzheimer’s and amyotrophic lateral sclerosis (17
, 20
, 54)
.
The highest number of genes whose expression was changed was in the signal transduction pathway. Of particular note is the robust and specific increase in the mRNA for Cox-2 (PTGS2) seen in the C282Y cell lines compared with vector control or the WT HFE cell line. COX-2 is the inducible isoform of cyclooxygenase and catalyzes the rate-limiting step in prostaglandin synthesis. COX-2 contains a heme center and therefore the changes in gene expression could be directly related to the cellular iron changes associated with the HFE mutations. If this latter idea is correct, however, it would be expected that a similar change in expression would have been seen in the H63D cell line for the Cox-2 gene because the labile iron pool was also elevated relative to the WT cell line. Therefore, the elevation of Cox-2 gene is not simply a reflection of the iron in the labile iron pool. COX-2 promotes cell growth, inhibits apoptosis, enhances cell motility and adhesion, and is a key step in carcinogenesis. The increased expression of Cox-2 in C282Y-carrying cells could have a key role in the increased risk of cancer in people carrying this mutation (55
56
57)
. Another gene in the signal transduction pathway that was relatively robust in its response in the C282Y cell lines and is also related to cancer was the retinal binding protein gene (58)
.
HFE reportedly has two functions. One is binding to the Tf receptor in competition with transferrin and the other is inhibition of iron release (59)
. However, the data presented here strongly suggest that HFE protein has a potentially much greater effect on cell function, changing energy levels, susceptibility to stress, and possibly transcription activation. Our cell culture model system will serve as a valuable tool in identifying the cellular functions that are altered as a result of carrying these common mutations, and therefore may serve as a first line of study to determine the role of HFE mutations in neurodegenerative diseases and cancer.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication May 12, 2006. Accepted for publication September 22, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. M. Mitchell, W. M. Freeman, W. T. Randazzo, H. E. Stephens, J. L. Beard, Z. Simmons, and J. R. Connor A CSF biomarker panel for identification of patients with amyotrophic lateral sclerosis Neurology, January 6, 2009; 72(1): 14 - 19. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) or the RFP without an IRE (
). The fluorescent RFP protein expression was determined for each group over 4 days in normal culture media and the amount of fluorescence was compared. An asterisk (*) indicates a significant difference compared with mock (RFP vector)-transfected cells (P<0.05).
