Influence of HFE variants and cellular iron on monocyte chemoattractant protein-1
Polymorphisms in the MHC class 1-like gene known as HFE have been proposed as genetic modifiers of neurodegenerative diseases that include neuroinflammation as part of the disease process. Variants of HFE are relatively common in the general population and are most commonly associated with iron overload, but can promote subclinical cellular iron loading even in the absence of clinically identified disease. The effects of the variants as well as the resulting cellular iron dyshomeostasis potentially impact a number of disease-associated pathways. We tested the hypothesis that the two most common HFE variants, H63D and C282Y, would affect cellular secretion of cytokines and trophic factors.
We screened a panel of cytokines and trophic factors using a multiplexed immunoassay in human neuroblastoma SH-SY5Y cells expressing different variants of HFE. The influence of cellular iron secretion on the potent chemokine monocyte chemoattractant protein-1 (MCP-1) was assessed using ferric ammonium citrate and the iron chelator, desferroxamine. Additionally, an antioxidant, Trolox, and an anti-inflammatory, minocycline, were tested for their effects on MCP-1 secretion in the presence of HFE variants.
Expression of the HFE variants altered the labile iron pool in SH-SY5Y cells. Of the panel of cytokines and trophic factors analyzed, only the release of MCP-1 was affected by the HFE variants. We further examined the relationship between iron and MCP-1 and found MCP-1 secretion tightly associated with intracellular iron status. A potential direct effect of HFE is considered because, despite having similar levels of intracellular iron, the association between HFE genotype and MCP-1 expression was different for the H63D and C282Y HFE variants. Moreover, HFE genotype was a factor in the effect of minocycline, a multifaceted antibiotic used in treating a number of neurologic conditions associated with inflammation, on MCP-1 secretion.
Our results demonstrate that HFE polymorphisms influence the synthesis and release of MCP-1. The mechanism of action involves cellular iron status but it appears there could be additional influences such as ER stress. Finally, these data demonstrate a pharmacogenetic effect of HFE polymorphisms on the ability of minocycline to inhibit MCP-1 secretion.
KeywordsAmyotrophic Lateral Sclerosis Minocycline Ferric Ammonium Citrate Labile Iron Pool Cellular Iron Status
amyotrophic lateral sclerosis
analysis of variance
beta 2 microglobulin
central nervous system
Dulbecco's modified Eagles Medium
enzyme-linked immunosorbent assay
ferric ammonium citrate
hypoxia inducible factor
heat shock protein
monocyte chemoattractant protein-1
major histocompatibility complex
nuclear factor-kappa B
quantitative reverse-transcription polymerase chain reaction
standard error of the mean
Iron accumulation in various brain regions is associated with the pathogenesis of several neurodegenerative diseases. The relationship between loss of brain iron homeostasis and the diseases appears to be generation of oxidative stress and exacerbation of inflammatory responses. The degree to which loss of iron homeostasis contributes to cell dysfunction and loss is not clear. Cellular iron homeostasis involves the complex interaction of numerous proteins, and genetic variants in genes encoding these proteins may influence this process and hence pathogenesis and disease progression. Polymorphisms in the major histocompatibility complex class 1-like gene known as HFE have been proposed as genetic modifiers of amyotrophic lateral sclerosis (ALS) , Alzheimer's disease (AD) , Parkinson's disease (PD) , and stroke . The HFE protein is primarily thought to regulate transferrin-mediated cellular iron intake through its interaction with the transferrin receptor , but a recent study from our laboratory suggests a potentially broader role in cell function . The H63D HFE variant appears deficient in reducing the affinity of the transferrin receptor for transferrin, while the C282Y HFE variant aggregates in the endoplasmic reticulum [5, 7]. These HFE variants are thought to underlie the iron overload disorder known as hemochromatosis, but appear to have low penetration rates . Nonetheless, there is little disagreement that HFE variants impact iron status at the cellular level independent of clinical disease [5, 6, 7, 9].
Cellular iron dysregulation has the potential to indirectly impact numerous disease-associated pathways. Our study is focused on the link between cellular iron regulation and secretion of soluble mediators of inflammatory reactions [10, 11]. Iron dysregulation in neuronal cells induces toxicity , potentially resulting in recruitment of inflammatory cells . One factor proposed to contribute to the relentless progression associated with many neurodegenerative diseases is the recruitment and activation of microglia in the regions of neuronal death . As mediators of innate immunity in the central nervous system, microglia respond to a variety of cytokines and chemokines and in turn secrete a variety of inflammatory mediators and trophic factors which likely contribute to neurodegenerative disease pathogenesis.
By influencing cellular iron homeostasis, we hypothesize that HFE variants may differentially impact secretion of cytokines and trophic factors, which may explain the association of these alleles with various diseases. In the current study we determine the influence of HFE variants and iron regulation on cellular secretion of factors relevant to neurodegenerative disease and explore mechanisms underlying these relationships.
SH-SY5Y human neuroblastoma cell lines and human astrocytoma cell lines U251 and U138 MG, were obtained from American Type Culture Collection (Manassas, VA). Mouse microglial BV-2 cells were a kind gift of Dr. Steven W. Levison (University of Medicine & Dentistry of New Jersey). Cell culture reagents including DMEM/F12, DMEM, penicillin/streptomycin, penicillin/streptomycin/glutamine and Geneticin, as well as calcein-AM were purchased from Invitrogen (Carlsbad, CA). Fetal bovine serum was purchased from Gemini Bio-Products (West Sacramento, CA). MTS reagent was purchased from Promega (Chatsworth, CA). PCR primers for MCP-1 were purchased from SuperArray (Frederick, MD). The DNeasy Tissue kit and RNeasy Mini kit were purchased from Qiagen (Valencia, CA). The DC protein assay and the Bio-Plex Human 27-plex panel of cytokines and growth factors were obtained from Bio-Rad (Hercules, CA). Anti-Hsp70 and anti-Hsp90 antibodies were purchased from Stressgen (Ann Arbor, MI). All other reagents were purchased from Sigma-Aldrich (St. Louis, MO).
Astrocytoma cell lines were maintained in Dulbecco's Modified Eagle's Medium supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin. BV-2 cells were maintained in DMEM supplemented with 10% heat-inactivated FBS and 1% penicillin/streptomycin. Cells were generally used after seven to ten (maximum) passages.
Genomic DNA was extracted from astrocytoma cell lines using the DNeasy Tissue kit according to the manufacturer's instructions. The PCR-amplified DNA samples were digested by restriction enzymes to determine the HFE genotype . DNA (50 ng) was amplified by PCR using the primers forward: 5'-ACA TGG TTA AGG CCT GTT GC-3', and reverse: 5' CTT GCT GTG GTT GTG ATT TTC C-3' for detection of the H63D allele, and primers forward: 5' CAA GTG CCT CCT TTG GTG AAG GTG ACA CAT-3', and reverse: 5' CTC AGG CAC TCC TCT CAA CC-3' for the C282Y allele. The PCR reaction was performed in a total volume of 50 μL containing 100 pmol of each primer in a PTC-200 Peltier Thermal Cycler (MJ Research, Woburn, MA). The PCR reaction was initiated at 95°C for 15 min, followed by 40 cycles of denaturation at 95°C for 45 s, annealing at 58°C for 45 s, and extension at 72°C for 1 min 30 s. A 294 bp product was obtained for the H63D polymorphism. Following digestion with MboI, the 63D allele gave restriction fragments of 237 and 57 bp, while the H63 allele resulted in three fragments (138, 99 and 57 bp). For the C282Y polymorphism, the primers resulted in a 343 bp product. Following digestion with RsaI, the 282Y allele resulted in products of 203, 111 and 29 bp, whereas the C282 allele yielded restriction fragments of 203 and 140 bp. Polymorphisms were detected by restriction fragment length analysis in 5% TBE polyacrylamide gel.
Cell viability assays
Differentiated SH-SY5Y cell lines were grown in 96-well plates until 70–80% confluent. Cells were treated with various concentrations of desferroxamine (DFO), ferric ammonium citrate (FAC), Trolox and minocycline. Cell viability was assessed with the colorimetric agent MTS and compared to untreated cells.
SH-SY5Y cells were grown as separate cultures until approximately 80% confluent. Cells were lysed with a RIPA buffer supplemented with 1% Triton X-100 and protease inhibitor cocktail. Forty micrograms total protein was separated by electrophoresis in a 4–20% Criterion polyacrylamide gel. Protein was then transferred to a nitrocellulose membrane and blocked for 1 hr at room temperature in TBS-T with 5% nonfat milk. Membranes were then probed with anti-FLAG (1:2000), anti-Hsp70 (1:1000) and anti-Hsp90 (1:1000) primary antibodies in TBS-T with 5% nonfat milk overnight at 4°C. HRP-conjugated secondary antibodies (1:5000) were then added in 5% nonfat milk for 1 hr at room temperature. Signals were obtained by chemiluminescence and visualized by CCD camera. Each blot shows results obtained from three independent cell cultures.
Intracellular labile iron
Intracellular labile iron was determined in the SH-SY5Y cells using the cell-permeable dye calcein-AM. Cells were grown in 24-well plates to 80% confluency. Calcein-AM was then added in medium at a final concentration of 0.25 μM and cells were incubated at 37°C for 30 minutes. Cells were then washed three times in ice-cold Hanks' balanced salt solution and lysed in 200 μL RIPA buffer plus 1% Triton X-100. The fluorescence of the lysate was measured and normalized to the protein content of the lysate, determined by BioRad DC protein assay, after subtracting a blank consisting of cells incubated with medium lacking the dye.
Multiplex cytokine bead assay
Cytokines and trophic factors assayed with bead-based immunoassay (Bio-Rad; Hercules, CA)
fibroblast growth factor (FGF)
granulocyte colony stimulating factor (G-CSF)
monocyte chemoattractant protein-1 (MCP-1)
granulocyte-macrophage colony stimulating factor (GM-CSF)
macrophage inflammatory protein-1α (MIP-1α)
IFN-γ induced protein-10 (IP-10)
platelet derived growth factor (PDGF) bb
regulated on activation normal T-cell expressed and presumably secreted (RANTES)
IL-1 receptor antagonist (IL-1ra)
tumor necrosis factor-alpha (TNF-α)
vascular endothelial growth factor (VEGF)
Measurement of MCP-1 release
Cells were plated in 24-well plates and grown to ~70% confluency. The medium was replaced and cells were incubated for 48 hrs, after which time medium was collected, cells were washed two times in Hanks' balanced salt solution, and cells were lysed in RIPA buffer plus 1% Triton X-100. Protein content of the cell lysate was determined by DC protein assay. MCP-1 concentrations in the conditioned medium were determined using commercially available anti-human (SH-SY5Y or astrocytoma cells) or anti-mouse (BV-2 cells) MCP-1 ELISA (GE Healthcare, Piscataway, NJ), and reported as MCP-1 normalized to the protein content of the cell lysate. The impact of iron loading and iron deprivation on MCP-1 release was determined in each cell line. Each cell line was treated with varying concentrations of ferric ammonium citrate (FAC) or desferroxamine (DFO) at levels below toxic concentrations, determined by MTS cytotoxicity assay. Additionally, MCP-1 release was determined in transfected SH-SY5Y cells treated with 50 or 100 μM Trolox or 25 μM minocycline, all concentrations determined to be below toxic levels. For FAC, DFO, Trolox, or minocycline treatment, agents were added to cell culture medium when medium was changed after cells reached ~70% confluency, resulting in treatment for 48 hours.
Quantitative reverse-transcription polymerase chain reaction (qRT-PCR)
We performed qRT-PCR for MCP-1 and GAPDH mRNA expression in SH-SY5Y cells using PCR primers commercially available from SuperArray. Total RNA was extracted from five independent cell cultures using the RNeasy Mini kit (Qiagen) and 5 μg of total RNA was used in a 20 μL cDNA reaction mixture using the ReactionReady First Strand cDNA synthesis kit (SuperArray). PCR was performed on cDNA using the RT2 Real-Time SYBR Green assay (SuperArray) with an ABI 7300 (Applied Biosystems, Foster City, CA) real-time PCR system. The reaction was initiated at 95°C for 10 min, followed by 40 cycles of denaturation at 95°C for 15 s, and annealing and extension at 60°C for 1 min. Expression levels of MCP-1 were normalized to expression of GAPDH mRNA.
Nuclear fractions were isolated from four independent cultures of each stably-transfected SH-SY5Y cell line using a nuclear extraction kit (Millipore, Billerica, MA). Nuclear extracts were analyzed for NF-κB (p50) content by commercially-available colorimetric assay (Cayman Chemical, Ann Arbor, MI), and results are expressed as optical density per milligram of nuclear protein.
Data are shown as mean ± SEM., with each n representing an independent cell culture. The data were analyzed by one-way ANOVA with Tukey's post-hoc analysis or Student's t test, as appropriate. Differences among means were considered statistically significant when the p-value was < 0.05.
Stably-transfected SH-SY5Y human neuroblastoma cells expressing different variants of HFE were generated as previously reported . Prior to use, cells were differentiated with 10 μM all-trans retinoic acid to induce a neuronal phenotype. Expression of FLAG-tagged HFE was confirmed in the cells, as shown in Figure 1A. Relative iron levels were determined in cells using the iron sensitive dye calcein-AM, whose fluorescence is quenched by binding iron. The cells transfected with vector alone do not contain HFE and have iron levels similar to the cells carrying the H63D alleles. Intracellular iron is decreased by the presence of the Wt HFE, consistent with the known functions of the HFE protein, compared to the vector alone group as well as the cells carrying the HFE allelic variants. To determine the effects of HFE polymorphisms, the comparisons are limited to the cells expressing HFE and thus do not include the vector alone transfected cells. The intracellular level of labile iron was affected by the presence of HFE in these cells in the pattern of C282Y (p < 0.001 vs. Wt) > H63D (p < 0.05 vs. Wt) > Wt (Figure 1B). These data demonstrate that the cells in our study had differences in the labile iron pool associated with the different polymorphisms and are generally consistent with our previous report .
NF-κB is known to have a prominent role in the transcriptional regulation of many inflammatory mediators including MCP-1 . Analysis of the nuclear content of NF-κB in the SH-SY5Y cell lines, however, revealed no differences associated with any of the HFE variants (Figure 3B).
The analysis of the panel of cytokines and trophic factors revealed that MCP1 is the most sensitive to cellular iron status and HFE expression. MCP-1 is known to recruit and enhance the activation of cells of the monocyte lineage . Higher levels of MCP-1 have been found in CSF and plasma in association with diseases characterized by excessive inflammation including ALS and AD [23, 24] and in both of these diseases multiple reports exist that HFE gene variants, particularly the H63D polymorphism, are risk factors for the disease [1, 2, 25]. Additionally, previous studies from our group and others [23, 26] have determined a positive correlation between age and expression of MCP-1, which may play a role in the higher incidence of neurodegenerative diseases at advanced ages. Furthermore, Lawless et al.  demonstrated that individuals with hemochromatosis who were homozygous for the C282Y allele had lower median plasma levels of MCP-1. Those hemochromatosis patients with at least one H63D allele, however, had higher median plasma levels of MCP-1 compared to Wt/Wt control subjects. These findings are consistent with the data using neuroblastoma cells reported herein.
In this study we have utilized a stably-transfected human neuroblastoma cell culture model to elucidate the consequences of HFE gene variants on MCP-1 expression. There are two key observations in this study. First is that H63D-expressing neuroblastoma cells have more iron than Wt HFE cells and secrete more MCP1. Secondly, the relationship between MCP1 secretion and iron is altered in the presence of the C282Y HFE variant via a mechanism which appears to be post-transcriptional or post-translational. The decreased expression of MCP-1 in the C282Y HFE cells may be related to the increased expression of Hsp70 in these cells. Our results on MCP-1 secretion in the C282Y cells are not consistent with a report using transiently-transfected HEK-293 cells . Our data suggest that iron could be a driving factor behind the secretion of MCP-1 and the transient transfection of cells already expressing Wt HFE may not compare well to our model. The expression of both Wt and C282Y forms of HFE in a cell model may have a confounding effect on MCP-1 secretion. It is also possible that the effect on MCP-1 secretion, as shown with the non-transfected astrocytoma cells in our study, is cell specific.
To determine the effect of cellular iron status on MCP-1 secretion, the cells were treated with either an iron chelator or iron was added to the media. In both the neuroblastoma cells and astrocytoma cells, the Wt and H63D cells and the non-HFE expressing vector control neuroblastoma cells, chelating the iron with DFO resulted in a decrease in MCP-1 expression. Iron chelation failed to decrease MCP-1 expression in the C282Y HFE cells. The addition of iron increased MCP-1 expression only in Wt neuroblastoma cells and both genotypes in the astrocytoma cells. Thus in general, it appears that the expression of MCP-1 can be manipulated by altering iron status. There is, however, an interaction of genotype and cell type that influences the MCP-1 expression independent of iron status. For example, the Wt neuroblastoma cells have less iron and less MCP-1 expression than the H63D-expressing cells. However, the reverse occurs in the astrocytoma cell lines; the Wt cells have more iron in the labile iron pool and more MCP-1 expression than the cells carrying the H63D allele. The astrocytoma cells are heterozygotic for the H63D variant and thus the baseline differences in iron status from Wt/Wt and the effect on MCP-1 secretion may reflect a heterozygotic effect. The astrocytoma cells also differ in genetic background. In the context of our study, regardless of cell type and HFE genotype, MCP-1 secretion was increased by increasing cellular iron and decreased by iron chelation, with the exception of cells expressing C282Y HFE. Thus it appears that C282Y HFE has an independent effect on secretion of MCP-1 that negates the effects of iron.
These data suggest that there is an additional function of the C282Y HFE mutant protein that is impacting MCP-1 secretion. Effects of the C282Y HFE variant contributing to the iron overload condition, hereditary hemochromatosis, have been suggested to result from induction of the ER overload and unfolded protein responses . Subsequent effects of these latter responses include activation of NF-κB  and potential activation of apoptosis , which have been associated with overexpression of C282Y HFE in one cell model . However, we found no difference in nuclear localization of NF-κB in our cell lines and have suggested a post-transcriptional or post-translational effect on MCP-1 that may be directly influenced by the mutant HFE protein.
The relationship between iron status, HFE status, and MCP-1 secretion prompted analysis of the known regulatory mechanisms for MCP-1 in these cells. MCP-1 is thought to be regulated mainly at the transcriptional level, responding to a variety of transcription factors. In neuroblastoma cell lines, the presence of the H63D and C282Y forms of HFE were associated with higher levels of MCP-1 mRNA than in the wildtype cells. The levels of MCP-1 mRNA were similar to that seen in the vector only cells. A large number of studies have explored the upregulation of MCP-1 by NF-kB, an oxidative stress-sensitive transcription factor which might explain the relationship between increased iron and MCP-1. However, nuclear localization of NF-κB did not vary by HFE genotypes, and thus iron status, in our model, consistent with a previous report that iron status did not affect baseline nuclear NF-κB in microglial cells .
A likely mechanism through which cellular iron status affects cell stress and MCP-1 regulation is through the formation of reactive oxygen species. In the present study the antioxidant Trolox had no significant effect on MCP-1 secretion in any of the SH-SY5Y cell lines. It is possible that the length of treatment (72 hours) was not sufficient to impact MCP-1 synthesis or release; however, longer duration of treatment showed evidence of cellular toxicity (data not shown).
An additional pharmaceutical intervention we examined was minocycline. Minocycline has been explored for use in treating a number of neuroinflammatory conditions including ALS, AD, PD, multiple sclerosis, and ischemic stroke. This compound has iron chelation , antioxidant , anti-inflammatory , caspase-inhibitory , and mitochondria-protective properties . In the present study, effects of treating cells with minocycline varied with HFE genotype, significantly decreasing MCP-1 release in the two SH-SY5Y cell lines with the highest MCP-1 secretion.
The multifaceted nature of minocycline causes uncertainty of the specific contribution of each mechanism of action. Minocycline, like other tetracyclines, has iron chelation properties . The ability to reduce intracellular iron levels may explain the benefit of minocyline in reducing MCP-1 secretion in the relatively iron-laden H63D HFE cells. The lack of effect seen in Wt HFE cells, which responded to iron chelation with DFO, may result from other effects of minocycline, or simply reflect the relative differences in chelator strengths between DFO and minocycline. An alternative explanation for the differential effects of minocycline involves its actions on mitochondria. Minocycline has been shown to have mitochondrial protective properties, including reduction of mitochondrial swelling and permeability transition pore opening . Mitochondrial dysfunction in the presence of HFE gene variants has been suggested for both the C282Y allelic variant  and the H63D HFE variant . Additionally, we have determined that the H63D HFE variant is associated with elevated cytosolic calcium levels (manuscript under review), that may result from altered mitochondrial function. Mitochondrial dysfunction and elevated calcium levels are mechanisms proposed to induce transcription of MCP-1; however, the association of mitochondrial function with MCP-1 regulation is mostly regulated through NF-κB . Thus, in contrast to its other actions, the data herein suggest that the effects of minocycline on MCP-1 secretion involve iron chelation. Although the exact mechanism of minocycline is not identified in these cells, these data strongly suggest that HFE genotype should be considered when assessing treatment efficacy of minocycline, and perhaps other pharmacological agents. The lack of benefit of minocycline in the treatment of humans with ALS, for example , despite its benefit in the mouse model , may be due to patient heterogeneity for HFE alleles. Additional studies of minocycline in patients with ALS may be warranted if they include stratification by HFE genotype because approximately 30% of ALS patients carry at least one H63D or C282Y HFE allele [1, 25]. Given the broad effects of HFE on regulation of MCP-1, such stratification may be justified in other clinical trials as well.
Supported in part by the Paul and Harriett Campbell Fund for ALS Research; the Zimmerman Family Love Fund; the ALS Association Greater Philadelphia Chapter; and The Judith & Jean Pape Adams Charitable Foundation.
- 5.Feder JN, Penny DM, Irrinki A, Lee VK, Lebron JA, Watson N, Tsuchihashi Z, Sigal E, Bjorkman PJ, Schatzman RC: The hemochromatosis gene product complexes with the transferrin receptor and lowers its affinity for ligand binding. Proc Natl Acad Sci USA. 1998, 95: 1472-1477. 10.1073/pnas.95.4.1472.PubMedCentralCrossRefPubMedGoogle Scholar
- 7.Waheed A, Parkkila S, Zhou XY, Tomatsu S, Tsuchihashi Z, Feder JN, Schatzman RC, Britton RS, Bacon BR, Sly WS: Hereditary hemochromatosis: effects of C282Y and H63D mutations on association with beta2-microglobulin, intracellular processing, and cell surface expression of the HFE protein in COS-7 cells. Proc Natl Acad Sci USA. 1997, 94: 12384-12389. 10.1073/pnas.94.23.12384.PubMedCentralCrossRefPubMedGoogle Scholar
- 20.Yrjanheikki J, Tikka T, Keinanen R, Goldsteins G, Chan PH, Koistinaho J: A tetracycline derivative, minocycline, reduces inflammation and protects against focal cerebral ischemia with a wide therapeutic window. Proc Natl Acad Sci USA. 1999, 96: 13496-13500. 10.1073/pnas.96.23.13496.PubMedCentralCrossRefPubMedGoogle Scholar
- 25.Sutedja NA, Sinke RJ, Van Vught PW, Linden Van der MW, Wokke JH, Van Duijn CM, Njajou OT, Schouw Van der YT, Veldink JH, Berg Van den LH: The association between H63D mutations in HFE and amyotrophic lateral sclerosis in a Dutch population. Arch Neurol. 2007, 64: 63-67. 10.1001/archneur.64.1.63.CrossRefPubMedGoogle Scholar
- 32.Fernandez-Gomez FJ, Galindo MF, Gomez-Lazaro M, Gonzalez-Garcia C, Cena V, Aguirre N, Jordan J: Involvement of mitochondrial potential and calcium buffering capacity in minocycline cytoprotective actions. Neuroscience. 2005, 133: 959-967. 10.1016/j.neuroscience.2005.03.019.CrossRefPubMedGoogle Scholar
- 33.Kim JM, Kim JS, Lee JY, Kim YJ, Youn HJ, Kim IY, Chee YJ, Oh YK, Kim N, Jung HC, Song IS: Vacuolating cytotoxin in Helicobacter pylori water-soluble proteins upregulates chemokine expression in human eosinophils via Ca2+ influx, mitochondrial reactive oxygen intermediates, and NF-kappaB activation. Infect Immun. 2007, 75: 3373-3381. 10.1128/IAI.01940-06.PubMedCentralCrossRefPubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.