Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Gene-gene interactions among coding genes of iron-homeostasis proteins and APOE-alleles in cognitive impairment diseases

  • Veronica Tisato ,

    Contributed equally to this work with: Veronica Tisato, Giovanni Zuliani

    Roles Conceptualization, Data curation, Formal analysis, Supervision, Writing – original draft, Writing – review & editing

    Affiliation Department of Morphology, Surgery and Experimental Medicine and LTTA Centre, University of Ferrara, Ferrara, Italy

  • Giovanni Zuliani ,

    Contributed equally to this work with: Veronica Tisato, Giovanni Zuliani

    Roles Formal analysis, Methodology, Supervision, Validation

    Affiliation Department of Morphology, Surgery and Experimental Medicine, University of Ferrara and Section of Internal and Cardiorespiratory Medicine, Azienda Ospedaliero-Universitaria of Ferrara, Ferrara, Italy

  • Marco Vigliano,

    Roles Conceptualization, Data curation, Methodology, Software

    Affiliation Department of Biomedical and Specialty Surgical Sciences, and Section of Medical Biochemistry, Molecular Biology & Genetics and Ctr. Hemostasis & Thrombosis, University of Ferrara, Ferrara, Italy

  • Giovanna Longo,

    Roles Data curation, Formal analysis, Methodology

    Affiliation Department of Biomedical and Specialty Surgical Sciences, and Section of Medical Biochemistry, Molecular Biology & Genetics and Ctr. Hemostasis & Thrombosis, University of Ferrara, Ferrara, Italy

  • Eugenia Franchini,

    Roles Methodology

    Affiliation Department of Biomedical and Specialty Surgical Sciences, and Section of Medical Biochemistry, Molecular Biology & Genetics and Ctr. Hemostasis & Thrombosis, University of Ferrara, Ferrara, Italy

  • Paola Secchiero,

    Roles Conceptualization

    Affiliation Department of Morphology, Surgery and Experimental Medicine and LTTA Centre, University of Ferrara, Ferrara, Italy

  • Giorgio Zauli,

    Roles Conceptualization

    Affiliation Department of Morphology, Surgery and Experimental Medicine and LTTA Centre, University of Ferrara, Ferrara, Italy

  • Elvezia Maria Paraboschi,

    Roles Data curation, Methodology

    Affiliation Department of Biomedical Sciences, Humanitas University, Rozzano (Milan) and Humanitas Clinical and Research Center, Rozzano, Milan, Italy

  • Ajay Vikram Singh,

    Roles Data curation, Methodology

    Affiliation Department of Physical Intelligence Max Planck Institute, Stuttgart, Germany

  • Maria Luisa Serino,

    Roles Investigation

    Affiliation Department of Medical Sciences and Ctr. Hemostasis & Thrombosis, University of Ferrara, Ferrara, Italy

  • Beatrice Ortolani,

    Roles Formal analysis

    Affiliation Department of Morphology, Surgery and Experimental Medicine, University of Ferrara and Section of Internal and Cardiorespiratory Medicine, Azienda Ospedaliero-Universitaria of Ferrara, Ferrara, Italy

  • Amedeo Zurlo,

    Roles Data curation

    Affiliation Operative Unit of Geriatrics, Azienda Ospedaliero-Universitaria S. Anna, Ferrara, Italy

  • Cristina Bosi,

    Roles Methodology

    Affiliation Department of Morphology, Surgery and Experimental Medicine, University of Ferrara and Section of Internal and Cardiorespiratory Medicine, Azienda Ospedaliero-Universitaria of Ferrara, Ferrara, Italy

  • Antonio Greco,

    Roles Data curation, Investigation

    Affiliation Complex Structure of Geriatrics, Research Laboratory, Department of Medical Sciences, I.R.C.C.S. Casa Sollievo della Sofferenza, San Giovanni Rotondo, Foggia, Italy

  • Davide Seripa,

    Roles Data curation, Writing – review & editing

    Affiliation Complex Structure of Geriatrics, Research Laboratory, Department of Medical Sciences, I.R.C.C.S. Casa Sollievo della Sofferenza, San Giovanni Rotondo, Foggia, Italy

  • Rosanna Asselta,

    Roles Data curation, Investigation, Methodology, Writing – original draft, Writing – review & editing

    Affiliation Department of Biomedical Sciences, Humanitas University, Rozzano (Milan) and Humanitas Clinical and Research Center, Rozzano, Milan, Italy

  •  [ ... ],
  • Donato Gemmati

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Supervision, Validation, Writing – original draft, Writing – review & editing

    Affiliation Department of Biomedical and Specialty Surgical Sciences, and Section of Medical Biochemistry, Molecular Biology & Genetics and Ctr. Hemostasis & Thrombosis, University of Ferrara, Ferrara, Italy

  • [ view all ]
  • [ view less ]

Gene-gene interactions among coding genes of iron-homeostasis proteins and APOE-alleles in cognitive impairment diseases

  • Veronica Tisato, 
  • Giovanni Zuliani, 
  • Marco Vigliano, 
  • Giovanna Longo, 
  • Eugenia Franchini, 
  • Paola Secchiero, 
  • Giorgio Zauli, 
  • Elvezia Maria Paraboschi, 
  • Ajay Vikram Singh, 
  • Maria Luisa Serino


Cognitive impairments of different aetiology share alterations in iron and lipid homeostasis with mutual relationships. Since iron and cholesterol accumulation impact on neurodegenerative disease, the associated gene variants are appealing candidate targets for risk and disease progression assessment. In this light, we explored the role of common single nucleotide polymorphisms (SNPs) in the main iron homeostasis genes and in the main lipoprotein transporter gene (APOE) in a cohort of 765 patients with dementia of different origin: Alzheimer’s disease (AD) n = 276; vascular dementia (VaD), n = 255; mild cognitive impairment (MCI), n = 234; and in normal controls (n = 1086). In details, four genes of iron homeostasis (Hemochromatosis (HFE: C282Y, H63D), Ferroportin (FPN1: -8CG), Hepcidin (HAMP: -582AG), Transferrin (TF: P570S)), and the three major alleles of APOE (APOE2, APOE3, APOE4) were analyzed to explore causative interactions and synergies. In single analysis, HFE 282Y allele yielded a 3-fold risk reduction in the whole cohort of patients (P<0.0001), confirmed in AD and VaD, reaching a 5-fold risk reduction in MCI (P = 0.0019). The other iron SNPs slightly associated with risk reduction whereas APOE4 allele resulted in increased risk, reaching more than 7-fold increased risk in AD homozygotes (P = 0.001), confirmed to a lower extent in VaD and MCI (P = 0.038 and P = 0.013 respectively) as well as in the whole group (P<0.0001). Comparisons of Mini Mental State Examination (MMSE) among AD showed appreciable lowering in APOE4 carriers (P = 0.038), confirmed in the whole cohort of patients (P = 0.018). In interaction analysis, the HFE 282Y allele completely extinguished the APOE4 allele associated risk. Conversely, the coexistence in patients of a substantial number of iron SNPs accrued the APOE4 detrimental effect on MMSE. Overall, the analysis highlighted how a specific iron-allele burden, defined as different combinations of iron gene variants, might have different effects on cognitive impairment and might modulate the effects of established genetic risk factors such as APOE4. Our results suggest that established genetic risk factors might be affected by specific genetic backgrounds, making patients differently suited to manage iron accumulation adding new genetic insights in neurodegeneration. The recently recognized interconnections between iron and lipids, suggest that these pathways might share more than expected. We therefore extended to additional iron gene variants the newly proposed influencing mechanisms that HFE gene has on cholesterol metabolism. Our results have a strong translational potential promoting new pharmacogenetics studies on therapeutic target identification aimed at optimally tuning brain iron levels.


Alzheimer’s disease (AD), vascular dementia (VaD) and mild cognitive impairment (MCI) represent the most common forms of cognitive impairment affecting elderly people [13]. The comprehension of the mechanisms leading to neurodegeneration and cognitive impairment is a challenging task and inflammatory mechanisms [46] and iron-driven oxidative stress are key recognised physiopathological processes. Iron is essential for neuronal life but when exceeding it might result deleterious for brain cell functions [710] producing reactive oxygen species (ROS) and pro-inflammatory proteins [11, 12] not optimally handled in AD and MCI patients [1315].

Increased iron levels have been reported in several neurodegenerative diseases when compared to brain of elderly healthy individuals [3, 9] and iron manipulation has been proposed as therapeutic approach for neuro-protection/restoration [16, 17]. In vitro and in vivo studies showed that one or more genetic variants or low key molecules levels might cause cell death and iron-driven oxidative stress, affecting local iron balance and healing [1821] in different pathological settings [22, 23] rising the interest on specific single nucleotide polymorphisms (SNPs) in iron management for (neuro)degenerative diseases [2428]. Among the genes and SNPs of iron homeostasis, only few have been widely investigated.

The High FErrum (HFE; HGNC:4886) hemochromatosis gene codes for a MHC class I-type protein involved in iron absorption via regulation of the interaction transferrin-transferrin receptor [29]. The most prevalent disease-associated gene variants in the HFE gene are the missense mutations C282Y and H63D, both involved in oxidative-stress and iron-dependent inflammation [23, 30]. In neurodegenerative diseases, HFE gene variants have been differentially linked to multiple sclerosis (MS) pathophysiology with controversial data [3133], and we recently reported on HFE variants as markers of both poorer MS prognosis and higher MS aggressiveness [28]. The presence of HFE variants has also been associated with AD and other neurodegenerative conditions with the identification of novel connections and synergisms with the lipoprotein metabolism [23, 27, 3436]. Though controversial results have been published so far, a recent metanalysis on AD reported no association for C282Y and protection for H63D [37].

The FerroPortiN1 (FPN1, also known as SLC40A1; HGNC:10909) gene codes for a multiple transmembrane domain protein the main cellular iron exporter in mammals [38]. FPN1 can be regulated at many different levels including (post)-transcriptionally regulation through mRNA stability by the iron responsive element (IRE) in the 5’untranslated region (5’UTR) of mRNA [39]. FPN1 alterations in level/function may lead to either iron deficiency or iron overload and its overexpression has been demonstrated in AD brains [40]. A recent role of FPN1 and Amyloid Precursor Protein (APP) in relation to iron/cupper homeostasis has been reported in AD supporting their combined investigation [41]. Several polymorphisms have been identified in FPN1 gene and studied in relation to HFE [42] including -8CG and -98GC SNPs, close to IRE element and in complete linkage disequilibrium. FPN1 gene variants have been investigated in pathological conditions characterized by local iron-overload [10, 27] and in different neurodegenerative diseases [28, 43].

The Hepcidin (HAMP, also known as Anti-Microbial Peptide; HGNC:15598) gene codes for a 25 amino acid peptide hormone that inhibits iron entry into the blood by inhibiting FPN1 protein at the cell membrane level [44]. Hepcidin is a key contributor to iron overload in cerebral ischemia [45], and an interesting hepcidin-ferroportin cooperation on iron balance has been recently demonstrated in a human in vitro neuronal model via enhanced APP-translation [46]. However, data about the impact of HAMP gene variants in neurodegenerative disease and cognitive impairments are still lacking. We recently demonstrated that a polymorphism in the HAMP promoter region (−582AG) plays a significant role in MS progression [28], suggesting that it might be also involved in other neurodegenerative conditions. This could be of particularly importance considering FPN1 internalization and lysosomal break down due to HAMP action with striking effects on APP-mRNA translation activity [40, 47].

The TransFerrin (TF; HGNC:11740) gene, codes for an iron transporter in circulating blood. The iron cellular uptake requires the binding of TF to its receptor (TFR1) at cell membrane level followed by endocytosis. While normal HFE is able to combine with TFR1 tuning its affinity for TF, HFE variants do not correctly bind the TFR, thus limiting TF-mediated iron uptake leading to both uncontrolled TF activity and increased iron uptake. Similarly, TF may result overloaded by iron influx/efflux changes due to anomalous interactions between APP and FPN1 [41, 48]. The P570S (TF, C1C2) genetic variant has been assessed in the context of iron balancing [49, 50] and in several neurodegenerative diseases [28, 5153], but a synergistic effect of the HFE and TF genes also suggests possible gene-gene and gene-environment interactions [54].

Considering connections among iron metabolism and lipoprotein metabolism, we decided to investigate potential interactions between iron homeostasis SNPs and Apo-lipoproteinE (APOE; HGNC:613), the major recognised gene associated with AD. In particular, we included in our analysis the three most common isoforms of APOE: APOE2 (C112/C158); APOE3 (C112/R158); APOE4 (R112/R158) by allele variant discrimination. APOE is a multifunctional protein with key roles in lipid homeostasis and neurobiology [55] and it has been demonstrated that APOE2, APOE3, and APOE4 differentially stimulate APP transcription and amyloid-beta (Aβ) secretion [56]. In particular, whilst E4 is considered the most important genetic risk factor for AD, E2 is protective and E3 can be considered neutral. The pathobiological phenotype of APOE4 is characterized by age-dependent cerebral Aβ-peptide deposition, with E4 isoform promoting cerebral Aβ accumulation and the E2 involved on its clearance [57, 58]. It is of note that deregulated iron homeostasis and amyloidogenesis have been hypothesised to be synergistically connected during neurodegeneration in AD [59]. The iron dysregulation hypothesis and the amyloid cascade hypothesis coined in the context of AD, might share more than expected also in other neurodegenerative diseases and new insights from the genetic side should be taken into account [5962].

In this work, in a cohort of elderly individuals with cognitive impairments of different origin, we evaluated the impact and role of the above described gene polymorphisms belonging to the iron and lipid homeostasis pathways, having established and mutually dependent roles in cognitive impairment pathogenesis.


Patient population

For patient recruitment, the study was approved by the Local Ethic Committee of “Casa Sollievo della Sofferenza”, San Giovanni Rotondo, Italy (Protocol number #3877/DS). Control samples were obtained from established anonymous collection. The Local Ethics Committee of the University-Hospital of Ferrara, Italy reviewed the use of this collection for a previous study stating that there was no need of approval since only anonymous blood samples were collected and it was impossible to identify subjects involved in the study as the link between samples and subject data was irreversibly erased (Protocol number #160519). This study has been conducted according to the Declaration of Helsinki, the Guidelines for Good Clinical Practice and the guidelines for Strengthening the Reporting of Observational Studies in Epidemiology, and it was approved by the local ethics committee for human experimentation. All patients and/or their relatives or a legal guardian in case of critically disabled demented patients were informed in detail about the research project and research protocol. Written informed consent for research was obtained from each patient, or from relatives or a legal guardian in case of critically disabled demented patients under the approval of the local ethics committee. Iron SNPs frequencies show different geographical distributions, such as in the case of the most investigated HFE C282Y variant that has also strong north-south and east-west gradients within the different European countries [63]. For this reason, we included only individuals with Italian ancestry that was verified/confirmed by asking the surname of parents and grandmothers. Patients consecutively attending the Complex Structure of Geriatrics, Ospedale “Casa Sollievo della Sofferenza”, San Giovanni Rotondo, Foggia were enrolled and divided in three different groups:

  • AD patients (n = 276; mean age 78.46±5.7 years, males: 30.5%) as defined by the NINCDS-ADRDA criteria [64], the Global Deterioration Scale ranged from stage 3 to stage 6 and only patients with “probable” AD were included.
  • VaD patients (n = 255; mean age 78.69±6.8 years; males: 35.3%) as defined by the NINDS–AIREN criteria [65], the Global Deterioration Scale ranged from stage 4 to stage 6 and only patients with “probable” VaD were included.
  • MCI patients (n = 234; mean age 78.44±5.7 years, males: 43.5%) defined by the presence of short/long-term memory impairment with/without impairment in other single or multiple cognitive domains in individuals who did not meet the standardized criteria for dementia [66] and still independent in instrumental activities of daily living. Most of patients showed amnestic multi-domain MCI.

For patient recruitment, the study was approved by the Local Ethic Committee of “Casa Sollievo della Sofferenza”, San Giovanni Rotondo, Italy (protocol number # 3877/DS). As controls, we selected a total of 1086 anonymous samples age and gender matched with cases group (mean age 78.27±6.5 years, males: 36.8%) belonging to an already established anonymous collection of 3500 blood banked samples. The procedure followed the local Institutional Review Board guidelines for investigation on human biological samples. Subjects signed informed consent stating that the biological samples were collected and banked anonymously for research purposes. The Local Ethics Committee of the University-Hospital of Ferrara, Italy reviewed the use of this collection for a previous study [67] stating that there was no need of approval since only anonymous blood samples were collected and it was impossible to identify subjects involved in the study as the link between samples and subject data was irreversibly erased. The original cohort of samples did not include those with a diagnosis of established cognitive impairment, presence of severe congestive heart failure, bleeding tendency, severe liver or kidney disease, severe chronic obstructive pulmonary disease and cancer. There were no evidences of acute illnesses at the time of blood sampling.

Clinical evaluations

At the time of enrollment personal data and clinical history were collected by trained geriatricians by a structured interview of patients and caregivers. General and neurological examinations were routinely performed. The neuropsychological assessment was performed through a battery of tests as previously described [6]. The Mini Mental State Examination (MMSE; normal score >27) was used to estimate the severity and progression of cognitive impairment and expressed as median and interquartile range [68]. All subjects underwent brain CT using a 64 volumetric scanner (GE LightSpeed VCT). The CT scan information was used to support the clinical diagnosis and to evaluate the presence of brain pathologies associated with cognitive impairment (e.g. cerebrovascular disease and normal pressure hydrocephalus). When necessary, subjects underwent brain MRI (Philips Achieva 1.5T). To exclude other causes of cognitive impairment patients were routinely monitored and clinical-chemistry analyses including serum B-12 vitamin, serum folate, liver function tests including ammonia, kidney function tests, thyroid function tests, complete blood cell count and arterial oxygen saturation were performed. Exclusion criteria: severe congestive heart failure (New York Heart Association class III-IV), severe liver or kidney disease, severe chronic obstructive pulmonary disease and cancer. There were no evidences of acute illnesses at the time of clinical observation and blood sampling.

DNA extraction, PCR conditions, and sequencing

Venous whole blood was collected from subjects upon an overnight fast and immediately frozen at -80°C. DNA was isolated from whole blood by the automated DNA extraction and purification robot (BioRobot EZ1 system QIAGEN; Hilden, Germany), which performs purification of nucleic acids using a magnetic bead technology. HFE, FPN1, HAMP, TF and APOE SNPs were genotyped in the entire case-control cohorts by PCR, amplifying the relevant genomic region using specific couple of primers [28, 69]. The primers sequences together with the PCR conditions are shown in S1 Table, PCR were performed in a PTC-200 thermal cycler (M. J. Research, Inc., Watertown, MA, USA). SNPs detection was performed by pyrosequencing using the Pyromark ID System (Biotage AB Uppsala, Sweden) according to the standard procedures for amplicon denaturation, purification, and sequencing. The list of primers used to amplify/sequence the target genes is shown in S1 Table. According to our internal quality control procedures, we confirmed haplotypes by re-genotyping about 20% of randomly selected samples among each different genotype group for each specific polymorphism by means of enzymatic restriction of PCR amplicons. The restriction enzymes utilized (New England Biolabs Inc., Hitchin, UK) are listed in S1 Table together with the DNA fragments obtained after enzymatic digestion and the specific temperature for each different restricted amplicon. All digestion reactions were carried out according to the Supplier’s instructions. There were no discrepancies between genotypes determined in duplicate and/or by different methods. Already known genotypes were used as internal control references.

Statistical analysis

Statistical differences among groups were assessed by Student’s t-test and Chi-squared test for mean values and genotype distribution comparisons respectively. Yates' correction or Fisher's exact test were applied when appropriate. Odds Ratio (OR) and 95% confidence interval (95% CI) were used to estimate the risk associated to different SNPs and combinations. ORs and P-values were calculated comparing cases versus controls by means of the dominant model [--/(+-/++)] and the allelic distribution [-/+]. All the remaining genetic model comparisons were also computed, though they did not show any significant differences (data not shown). All the models were considered to have a potential role on OR association and the main comparisons were further stratified by presence/absence of APOE4 allele in the whole group and in the different groups. Finally, the genotype distribution (--/+-/++) between cases and controls were additionally compared and P-value calculated. For APOE, presence/absence of the E4-risk-allele was the main considered discriminant in MMSE comparisons whilst all the possible genotypes were utilized for crude ORs computations. A multivariate logistic regression model was applied to calculate adjusted ORs for SNPs by means of dominant model. P-values ≤ 0.05 were considered significant. Deviation from Hardy-Weinberg equilibrium was calculated for each polymorphism in both cases and controls. All analyses were performed by using SPSS Statistical Package (Version 22; SPSS Inc., Chicago, IL, USA) and MedCalc statistical software (Version 17.9.7).


Clinical and demographic characteristics of patients and controls

Table 1 shows the main features of patients, considered as whole group or as different disease-specific subgroups, and controls. There were no differences between the whole group of patients and controls in terms of gender distribution and mean age with the exception of a very slight under-representation of males within the AD groups (P = 0.049). Although at different extent, the APOE4 allele was significantly over-represented in all case subgroups compared to controls, ascribing to AD the highest rate (P = 0.00001). Our cohort of patients did not show systemic iron overload or deficiency as whole group and in the different disease subgroups as well as within genders. Finally, as expected, elderly had often lower mean iron circulating levels and the MCI subgroup showed the highest median MMSE score among patients.

Table 1. Main clinical and demographic characteristics of patients and controls.

SNPs genotype distribution and crude ORs evaluation

Table 2 shows the allele and genotype distributions of the five iron-related SNPs in the whole cohort of patients, in the three groups with different types of dementia as well in controls. Each group of cases and subgroups were statistically compared with controls for all the SNPs frequencies analyzed and P-values are reported. Finally, ORs were computed comparing cases versus controls by dominant [--/(+-/++)] and allele [-/+] distribution models.

Table 2. SNPs genotype distribution in patients and controls and crude ORs.

As for HFE C282Y, the 282Y-allele yielded a 3-fold risk reduction in the whole cohort of patients (OR = 0.34; 0.19–0.59; P<0.0001). The protective effect was confirmed at the same extent in the AD and VaD subgroups and almost reaching a 5-fold risk reduction among the MCI subgroup (OR = 0.2; 0.06–0.66; P = 0.0019). Among the other iron SNPs investigated, a significant risk reduction was also observed in the VaD subgroup either for FPN1 -8G-carriers (OR = 0.63; 0.46–0.85; P = 0.0033) and HAMP -582G-carriers (OR = 0.74; 0.56–0.98; P = 0.04) computing together homozygous and heterozygous cases. Both associations showed a similar strength in the whole cohort of patients (OR = 0.82; 0.67–0.99; P = 0.047 and OR = 0.8; 0.67–0.97; P = 0.025 respectively).

The presence of the APOE4 allele gave significant ORs in all the subgroups and in the whole cohort when performing an indiscriminate comparison of any APOE4 allele condition versus the remaining genotypes. However, in consideration of the recognised protective effect of APOE2 against AD and other neurodegenerative diseases, we calculated in detail APOE-allele specific ORs considering the E3/E3 as reference class (S2 Table). As expected, those AD patients homozygous for the APOE4 allele showed the highest risk association (OR = 7.56; 2.27–24.15; P = 0.001). Although at different extent, appreciable increasing risks were also observed in the VaD and MCI subgroups and in the whole cohort of cases.

Since ORs did not change considering or not the E2/E4 genotype from the global E4(+) computation (OR: 2.79; 2.18–3.57 versus 2.58; 2.02–3.28, respectively), we left the E2/E4 genotype in all subsequent analyses.

All the genotypes of the polymorphisms investigated were distributed according to the Hardy-Weinberg equilibrium.

Multivariate logistic regression analyses

To support the above described results, and to exactly weight the role of each iron variable considered, we performed a multivariate logistic regression model taking into account the variables associated to dementia. As shown in Table 3, HFE C282Y gave the strongest association in each subgroup considered confirming the crude ORs showed in Table 2, though AD yielded a borderline value. FPN1 -8CG confirmed its highest association found in the VaD subgroup as well as HAMP -582AG did in VaD and in the whole cohort. The remaining SNPs did not reach significant associations.

MMSE score in patients stratified by type of dementia and SNPs genotype

MMSE comparisons, stratified by the different iron SNPs genotypes, did not reveal any significant result neither within the different subgroups of patients or within the whole group of cases. Slight differences that deserve to be further investigated are those observed among AD and VaD subgroups for FPN1 -8CG (P = 0.062 and P = 0.08 respectively), for HFE C282Y and H63D (P = 0.074 and P = 0.099 respectively). Similar MMSE analysis, stratified by presence/absence of APOE4-allele, yielded different MMSE score just in the whole group and in AD subgroup. Accordingly, AD patients carrying the APOE4-allele showed significant lower MMSE median and interquartile range values compared with those carrying the remaining APOE-alleles (18.7, 15.8–21.3 versus 20.4, 16.5–23.4; P = 0.038). Even though at a lower extent, a similar trend was observed in the whole group of patients (P = 0.018) (S3 Table).

Effect on ORs evaluation of iron gene SNPs-APOE4 allele (coupled analysis)

The observation that HFE C282Y SNP and APOE4-allele were the only ones strongly associated in all the subgroups of patients respectively with risk reduction and risk increasing, prompted us to perform a risk assessment by coupling each polymorphic allele in the iron genes with the APOE4 allele to disclose potential mutual influences. As shown in Table 4, no additive or multiplicative effects were observed; on the contrary the coexistence of the APOE4 allele with any other among the iron SNPs led to a decreased risk with respect to the APOE4 risk obtained in single analysis (S2 Table). The effect was maximum among those co-carrying the HFE 282Y allele, which completely extinguished the APOE4 associated risk, while a partial but still appreciable effect was observed with the remaining iron SNPs according to their strength showed in the single analysis (Table 2).

Table 4. ORs evaluation in patients carrying both APOE4 allele and any other polymorphic allele among the iron SNPs.

Combined analysis of iron SNPs and ORs evaluation stratified by APOE4 condition

To investigate possible cumulative effects or mutual interactions among iron SNPs, we calculated the risk resulting from the coexistence in patients of at least three polymorphic allele in the iron genes (multicarriers) stratified by the presence/absence of the APOE4 allele (S4 Table, panel A). The OR-values were firstly calculated comparing the whole group of cases and controls regardless to the APOE condition (OR1), this analysis did not show any significant protection towards cognitive impairment. Accordingly, when the risk was calculated considering only those cases and controls who did not carry the APOE4 (OR2), a weak risk protection was obtained in AD patients (OR = 0.65; 0.43–0.97; P = 0.035) and in the whole group of cases (OR = 0.74; 0.57–0.97; P = 0.030), whilst no significant risk reduction was observed in VaD and MCI.

Since coexistence of multiple iron SNPs in our cohort of patients provided evidence of a subtle protective role, we investigated if the opposite gene condition (i.e. no polymorphic alleles in iron genes) was associated with significant risk variations and, as done before, we further stratified by the presence/absence of the APOE4 allele. As shown in S4 Table, panel B, the OR-values were firstly calculated comparing the whole groups of cases and controls regardless to the APOE condition (OR1) showing significant risk association in all patient groups with the exception of MCI. By considering only those cases and controls positive for APOE4 allele, a massive increase in the risk was observed (OR3), reaching more than five-fold risk association among AD patients (OR = 5.11; 2.74–5.52; P<0.0001) which was higher than that obtained in the single analysis (OR = 4.49; 3.32–6.08; S2 Table). A similar improvement was observed in the rest of patients and in the whole group. Then the existence of an individual genetic background characterized by presence or absence of a substantial polymorphic iron-allele burden might differently (counter)act on other inherited established risk factors.

MMSE comparison according to iron-allele burden stratified by APOE4 condition

To explore whether APOE4 allele might differentially influence the MMSE score when surrounded by “opposite” genetic iron settings, we stratified patients according to the number of polymorphic iron genes (i.e. multicarriers versus no polymorphic alleles) and further by the presence/absence of APOE4 allele (S5 Table).

The number of patients carrying opposite genetic iron setting was equally distributed among the whole group of cases (OR = 1.05, 0.81–1.36) and in the different subgroups (AD: OR = 1.0, 0.65–1.51; VaD: OR = 1.11, 0.71–1.73; MCI: 1.06, 0.66–1.72). Interestingly, the median MMSE score did not significantly differ among the groups of patients stratified according to opposite iron-allele burden (S5 Table, panel A: multiple iron SNPs versus S5 Table, panel B: no iron SNPs), neither in the subgroups nor in the whole cohort. However, in the AD group those carrying multiple iron SNPs showed a lower median value than their counterpart negative for the iron SNPs (MMSE: 18.7; 15.3–22.5 versus 20.2; 17.6–22.9; P = 0.098) and this variation was more evident within the APOE4(+) subgroups of AD (MMSE: 17.4; 15.4–19.9 versus 19.0; 17.6–22.9; P = 0.038). Of interest, although a substantial polymorphic iron-allele burden seems do not affect the MMSE in the whole group (MMSE: 21.0, 17.4–24.4 versus 21.4, 18.5–23.7; P = 0.27) as well as in subgroups (AD: P = 0.098; VaD: P = 0.44; MCI: P = 0.45), the APOE4 significantly worsened the median MMSE of the whole group only in the presence of an elevated iron allele-burden (P = 0.006) with respect to the opposite genetic counterpart (P = 0.38).

Peripheral iron levels stratification by iron SNPs according to different genotypes

Although the poor relationship between dementia and circulating iron level, this is the most commonly assessed in the clinical routine among the iron parameters. As shown before, our cohort of patients did not have systemic iron overload or deficiency with MCI showing the highest mean values (MCI versus AD: P = 0.09; MCI versus VaD: P = 0.026). Normally, males have higher iron levels compared to females, accordingly in our dementia patients this was confirmed in VaD (76.80±35.89 μg/dL and 70.42 ±29.85 μg/dL in male and female respectively; P = 0.065) and in MCI subgroup (87.00±30.81 μg/dL and 80.23±26.64 μg/dL in male and female respectively; P = 0.037). Conversely, the two genders belonging to the AD subgroup showed completely overlapped iron levels (78.73±29.11 μg/dL and 78.78±31.56 μg/dL in males and females respectively; P = 0.495). Peripheral iron dyshomeostasis does not necessarily cause neurodegeneration or reflect brain iron deposit causing no genetic connection with neurodegeneration [70], and the common iron SNPs usually investigated are not considered strong determinants of iron accumulation [71, 72]. On the contrary, APOE4-allele has been widely confirmed as key determinant of brain iron deposit in dementia also contributing to cognitive decline [7377]. For this purpose, we stratified peripheral iron levels by SNPs and none of the SNPs in the iron hemostasis or APOE genes yielded significant differences in mean levels (Table 5). In an explorative way, we also investigated the presence of different genotype distributions between sexes within the different cognitive disease. No gender-related differences in genotype distributions were found in the whole group and in the different diseases subgroups (data not shown).


The iron hypothesis proposes Aβ pathogenicity via interaction with specific metals mainly unbalanced by suboptimal hepcidin-ferroportin axis [62, 7881] while the amyloid cascade hypothesis considers amyloidosis due to an abnormal Aβ deposit/clearance as the primary pathological event [11, 60, 61]. Amyloidosis, has been recently demonstrated differently tuned by the different APOE-alleles [79] in AD and in cognitively normal older adults [82] and the two hypotheses might share more than expected also in other neurodegenerative diseases with mutual causative interactions.

We found that the presence of the rare HFE 282Y-allele had a significant 3-fold risk reduction in all groups almost reaching a 5-fold reduction in MCI. These results are in accordance with previous reports ascribing to 282Y-allele a protective role against neurodegenerative disease [83, 84], though the recent metanalysis of Lin et al. did not confirm such a protection [37]. The HFE H63D polymorphism did not show any significant risk fluctuation in our study, though it has been reported a higher rate of the 63D-allele in neurodegenerative disease [85], also in combination with different APOE isoforms [8688], suggesting H63D a genetic risk factor for neurodegenerative diseases [89]. Conversely, the same metanalysis of Lin et al. supported a slight protective role of H63D [37], underlying the still conflicting results.

Brain iron load increases with age leading to oxidative stress that could progress towards neurodegeneration and cognitive decline in an unfavourable genetic milieu. Among younger individuals, brain iron levels sharply increase with low interindividual variance. Conversely, in midlife, brain iron concentration reaches a plateau showing higher interindividual variance [11]. In this scenario, gene-gene and gene-environment interactions could play a key role in the progression to cognitive impairment. Therefore, additional SNPs among key genes controlling iron homeostasis might be involved, as demonstrated by our group in a study focused on MS patients [28]. To gain insights on iron-dependent processes in cognitive impairment, we extended the analyses to further SNPs: FPN1 -8CG, HAMP -582AG and TF P570S.

We found that FPN1 -8G-allele was underrepresented in VaD patients, yielding a significant 1.5-fold risk reduction (P<0.0018), while a borderline reduction was observed among the whole cohort of patients (P = 0.047). In the same fashion, HAMP -582AG SNP was underrepresented among VaD subgroup (P = 0.04) as well in the whole cohort of cases (P = 0.025; computing together -582AG heterozygotes and -582GG homozygotes). Controversial results exist on the association between TF P570S and AD [51, 52], and a suggestive synergism has been reported between TF and HFE gene variants in AD patients [54]. In addition, a more recent study investigating genetic and biochemical markers in patients with AD associated several intronic variants in FPN1, TF and TFR genes, concluding that iron accumulation and oxidative damage might contribute to AD pathophysiology [43]. That study also found low circulating iron levels in the same subset of patients, suggesting a new iron-mediated mechanism in neurodegeneration. The reduced rate of iron excretion from cells could explain the intracellular iron overload and at the same time the increased production of toxic ROS together with the low circulating iron levels. This suggests that peripheral iron assessment could be a novel and effective early biomarker for treatment monitoring [43], although in our study we did not find any association.

How iron SNPs responsible for cellular iron retention could protect against cognitive diseases is not easy to clarify. Different brain regions could differently suffer from changes in iron homeostasis because of unequal “equipment” in iron receptors/transporters. Interestingly, iron surplus would enhance cell viability making cells more resistant to other toxic metals, through improved APP and ferritin translation, as demonstrated in vitro on Pb-induced damage and in vivo against Pb-toxicity [46].

Among the major genetic risk factors for AD, there is the APOE in the form of the APOE4 allele. It was not only involved in the amyloid cascade promotion but it has also been recently investigated as responsible for ROS production in AD and MCI patients [80] as well as for worsening memory performance in relation to Aβ-levels in cognitively normal adults [82].

Accordingly, we firstly investigated the risk-association for the presence of the APOE4 allele. As expected, in AD the associated risk was higher, reaching a quite 4.5-fold rising (P<0.0001), whilst among MCI and VaD the risk was less than 1.8-fold higher (P = 0.0003). Finally, APOE4 homozygotes in AD had risk values higher than 7.5-fold (P = 0.001). When we looked at the MMSE comparison stratifying patients according to the presence/absence of the APOE4 allele, the median MMSE was significantly lower (P = 0.038) only in AD patients who carried the APOE4 compared to non-carriers. A slight difference was also found in the whole cohort of cognitive impaired patients (P = 0.018).

Despite the unquestionable key role that iron genes have on neurodegenerative disease pathogenesis, iron SNPs have not been widely investigated, with the only exception of HFE and TF polymorphisms that are reported in dedicated meta-analyses (S6 Table). We hence explored if the coexistence of more than one gene variants could disclose possible cumulative interactions with significant effects on the susceptibility risk and on MMSE score. We firstly coupled each polymorphic allele in the iron genes with the APOE4 allele and recomputed the risk association. Given that the rate of the combined existence of any polymorphic iron allele with the APOE4 radically dropped off in patients and that the new combined risk association was inversely related to the strength of the iron allele, we can exclude that iron SNPs and APOE4 could synergize, as differently reported by other authors [8688]. On the contrary, we could speculate that a protective effect of iron SNPs might potentially counteract the APOE4 risk feature. As expected, the HFE 282Y allele, showing the highest protective effect within our patients, had the maximum effect on the coupled risk computation if compared to TF 570S allele that virtually had no consequence in similar risk calculation. As an example, the crude OR scored for the APOE4 allele in AD patients (OR(APOE4): 4.49; 3.32–6.08) was unmodified by the co-presence of the TF 570S allele (OR(APOE4+TF570S): 3.75; 2.40–5.85), while it completely extinguished when HFE 282Y allele was coinherited (OR(APOE4+HFE282Y): 0.98; 0.21–4.6). Also considering the extreme rare frequency of the combined coinheritance of the double-carrier status (i.e. HFE 282Y/APOE4), we could speculate an effect on the APOE4 allele risk association as follows: HFE C282Y>HFE H63D>HAMP -582AG>FPN1 -8CG>TF P570S.

Considering that iron-SNPs seem to account for risk reduction, we combined them together in a collective risk evaluation analysis. We firstly attempted at evaluating any possible interaction among the iron SNPs, providing that subjects carried at least three polymorphic alleles among the four iron genes. We then calculated the associated risk and the median MMSE scores. Concerning the combined risk association, no appreciable variations were recorded, though the exclusion of the APOE4 allele from the computation yielded a significant risk reduction in AD patients and in the whole group (P = 0.035 and P = 0.03 respectively). We therefore explored the “other side of the coin” and considered the absence of any polymorphic allele in the iron genes as a risk condition. This particular status yielded risk association ranging from 1.3 to more than 1.5-fold in all the subgroups, and unexpectedly the risk for APOE4 allele improved, reaching more than 5-fold in AD. As for the MMSE median scores, the APOE4 allele completely lost its detrimental effect on MMSE in the absence of a substantial polymorphic iron-allele burden. In fact, though the median total MMSE did not differ among the opposite iron-SNPs conditions (MMSE(3-iron SNPs): 21.0; 17.4–24.5 versus MMSE(No-iron SNPs) 21.4; 18.5–23.7; P = 0.27), the inheritance of the APOE4 allele had significant detrimental consequences on the MMSE score of the whole group only among those who co-carried a substantial iron allele burden (P = 0.006).

Overall, our results ascribe to iron genes variants a significant role in cognitive disease assigning from one hand a protective effect and on the other providing a suitable environment to maximize the APOE4 detrimental effects on cognitive capacity, being amplified in patients carrying a high iron-allele burden. Obviously, these observations deserve further confirmation in a larger cohort of patients. The poor strength of the results due to sub-sampling in the stratified analyses must be considered, even though the whole cohort completely resumes the single observations within the subgroups. Anyway, a possible interpretation of these opposite findings may be in part explained by the recently proposed role of APOE as allele-specific antioxidant molecule, with APOE4 less efficient in binding iron and reducing oxidative stress than APOE3, particularly when surrounded by a substantial toxic metal burden [73, 74]. Local iron overload goes together with increased ROS, and HFE variants fail in counteracting cellular iron uptake as confirmed by several biological/instrumental assessments correlating iron overload, severity of cognitive impairment and HFE gene condition [9092]. How iron-accumulating variants in HFE can protect against AD or other neurodegenerative disease still remains unclear. One possible explanation could rise from the observation that in AD iron is most dense in proximity of Aβ plaques [93], and promotes Aβ deposition [94]. Although the neurotoxicity of Aβ is enhanced by iron it also depends on the iron export action of APP itself [41]. Recently discovered, APP mRNA contains a functional 5’UTR-IRE, similar to that found in FPN1 and TF. Consequently, intracellular iron availability can increase APP, FPN1 and TF translation and, at high levels iron facilitates the cooperative APP-FPN1 action improving iron efflux (Fig 1), hampering the detrimental APP maturation and Aβ deposition [41]. This fine equilibrium is a recent therapeutic target for AD, aimed at generating optimal intracellular iron levels and optimizing brain iron homeostasis to contrast Aβ-genesis [41, 48, 95, 96]. APP upregulation, relevant for AD, is a key element also in general brain iron homeostasis since it actively proceeds during vascular accidents protecting in turn neurons from heme release in ischemic/haemorrhagic stroke, ascribing it a great value also for VaD patients [97, 98]. Another neurodegenerative disease in which iron SNPs play a crucial role is MS [28, 3033, 99]. In fact, although MS is a myelin-degeneration disease characterized by inappropriate immune responses, a strong relationship between iron and myelin exists [100]. Iron related genes and myelin related genes share indeed iron dysregulation as common pathogenic pathway. Myelin degeneration can occur early as consequence of brain iron dyshomeostasis. Subsequently, the lack of adequate counteracting mechanisms, in part genetically determined, may cause exacerbation/progression of the disease via oxidative stress mechanisms [28, 99]. In line with this, APP is crucial also for MS being upregulated in damaged axons and emerging as a promising therapeutic target [101]. Finally, a recent study evaluating results summarized from GWAS meta-analyses on circulating iron levels and common variants affecting iron homeostasis on AD risk did not identify any genetic relationship ascribing to peripheral iron a no causal role in initiation of AD, rather suggesting ferritin as leader mechanism by which APOE4 is considered a risk factor for AD [70].

Fig 1. Schematic model linking HFE, HAMP and TF to FPN1-APP complex in detoxifying action for balanced iron homeostasis.

The picture highlights the main pathways and molecular mediators involved in iron homeostasis maintenance by FPN1-APP cooperation. HAMP is Hepcidin; TF is Transferrin; TF-R is Transferrin receptor; HFE is Hemochromatosis; FPN is Ferroportin; IRP and IRE are Iron Regulatory Protein and Iron Responsive Element respectively.

In conclusion, although this study requires confirmation in larger cohorts of patients and in longitudinal analyses, our results strongly suggest that established genetic risk factors might be modulated by specific genetic backgrounds, making patients differently suited to manage iron overload thus resulting in different clinical phenotypes. In line with the recently recognized interconnections between iron and lipids, we have here extended to other iron gene variants the newly proposed influencing role of HFE gene on cholesterol metabolism with a strong translational potential, supporting pharmacogenetics studies aimed at optimally tuning intracellular iron levels to optimize brain iron homeostasis.

Supporting information

S1 Table. Primer sequences, restriction-product characteristics and PCR conditions.


S2 Table. Crude ORs evaluation in patients stratified by APOE allele combinations.


S3 Table. Mini mental state examination (MMSE) in patients stratified by type of dementia and SNPs genotypes.


S4 Table.

ORs evaluation comparing multicarriers (A) or no carriers (B) of polymorphic alleles in iron genes stratified by APOE4 condition.


S5 Table.

MMSE comparison in patients multicarrier (A) or no carriers (B) of polymorphic alleles in iron genes stratified by APOE4 condition.


S6 Table. Summary ORs from selected meta-analyses on AD and HFE and TF SNPs.



  1. 1. Fratiglioni L, Grut M, Forsell Y, Viitanen M, Grafstrom M, Holmen K, et al. Prevalence of Alzheimer's disease and other dementias in an elderly urban population: relationship with age, sex, and education. Neurology. 1991;41(12):1886–92. pmid:1745343
  2. 2. Ott A, Breteler MM, van Harskamp F, Claus JJ, van der Cammen TJ, Grobbee DE, et al. Prevalence of Alzheimer's disease and vascular dementia: association with education. The Rotterdam study. BMJ. 1995;310(6985):970–3. pmid:7728032
  3. 3. Gozzelino R, Arosio P. Iron Homeostasis in Health and Disease. Int J Mol Sci. 2016;17(1).
  4. 4. Zuliani G, Ranzini M, Guerra G, Rossi L, Munari MR, Zurlo A, et al. Plasma cytokines profile in older subjects with late onset Alzheimer's disease or vascular dementia. J Psychiatr Res. 2007;41(8):686–93. pmid:16600299
  5. 5. Ransohoff RM. How neuroinflammation contributes to neurodegeneration. Science. 2016;353(6301):777–83. pmid:27540165
  6. 6. Tisato V, Rimondi E, Brombo G, Volpato S, Zurlo A, Zauli G, et al. Serum Soluble Tumor Necrosis Factor-Related Apoptosis-Inducing Ligand Levels in Older Subjects with Dementia and Mild Cognitive Impairment. Dement Geriatr Cogn Disord. 2016;41(5–6):273–80. pmid:27304551
  7. 7. Sullivan JL. Is stored iron safe? J Lab Clin Med. 2004;144(6):280–4. pmid:15614249
  8. 8. Stankiewicz J, Panter SS, Neema M, Arora A, Batt CE, Bakshi R. Iron in chronic brain disorders: imaging and neurotherapeutic implications. Neurotherapeutics. 2007;4(3):371–86. pmid:17599703
  9. 9. Stankiewicz JM, Brass SD. Role of iron in neurotoxicity: a cause for concern in the elderly? Curr Opin Clin Nutr Metab Care. 2009;12(1):22–9. pmid:19057183
  10. 10. Singh AV, Subhashree L, Milani P, Gemmati D, Zamboni P. Interplay of iron metallobiology, metalloproteinases, and FXIII, and role of their gene variants in venous leg ulcer. Int J Low Extrem Wounds. 2010;9(4):166–79. pmid:21118859
  11. 11. Andersen HH, Johnsen KB, Moos T. Iron deposits in the chronically inflamed central nervous system and contributes to neurodegeneration. Cell Mol Life Sci. 2014;71(9):1607–22. pmid:24218010
  12. 12. Bou Khalil R, Khoury E, Koussa S. Linking multiple pathogenic pathways in Alzheimer's disease. World J Psychiatry. 2016;6(2):208–14. pmid:27354962
  13. 13. Casado A, Encarnacion Lopez-Fernandez M, Concepcion Casado M, de La Torre R. Lipid peroxidation and antioxidant enzyme activities in vascular and Alzheimer dementias. Neurochem Res. 2008;33(3):450–8. pmid:17721818
  14. 14. Padurariu M, Ciobica A, Hritcu L, Stoica B, Bild W, Stefanescu C. Changes of some oxidative stress markers in the serum of patients with mild cognitive impairment and Alzheimer's disease. Neurosci Lett. 2010;469:6–10. pmid:19914330
  15. 15. Puertas MC, Martinez-Martos JM, Cobo MP, Carrera MP, Mayas MD, Ramirez-Exposito MJ. Plasma oxidative stress parameters in men and women with early stage Alzheimer type dementia. Exp Gerontol. 2012;47(8):625–30. pmid:22664577
  16. 16. Li X, Jankovic J, Le W. Iron chelation and neuroprotection in neurodegenerative diseases. J Neural Transm (Vienna). 2011;118(3):473–7.
  17. 17. Guzior N, Wieckowska A, Panek D, Malawska B. Recent development of multifunctional agents as potential drug candidates for the treatment of Alzheimer's disease. Curr Med Chem. 2015;22(3):373–404. pmid:25386820
  18. 18. Zamboni P, De Mattei M, Ongaro A, Fogato L, Carandina S, De Palma M, et al. Factor XIII contrasts the effects of metalloproteinases in human dermal fibroblast cultured cells. Vasc Endovascular Surg. 2004;38:431–8. pmid:15490040
  19. 19. Zamboni P, Scapoli G, Lanzara V, Izzo M, Fortini P, Legnaro R, et al. Serum iron and matrix metalloproteinase-9 variations in limbs affected by chronic venous disease and venous leg ulcers. Dermatol Surg. 2005;31:644–9; discussion 9. pmid:15996413
  20. 20. De Mattei M, Ongaro A, Magaldi S, Gemmati D, Legnaro A, Palazzo A, et al. Time- and dose-dependent effects of chronic wound fluid on human adult dermal fibroblasts. Dermatol Surg. 2008;34(3):347–56. pmid:18177397
  21. 21. Singh AV, Vyas V, Montani E, Cartelli D, Parazzoli D, Oldani A, et al. Investigation of in vitro cytotoxicity of the redox state of ionic iron in neuroblastoma cells. J Neurosci Rural Pract. 2012;3(3):301–10. pmid:23188983
  22. 22. Gemmati D, Tognazzo S, Serino ML, Fogato L, Carandina S, De Palma M, et al. Factor XIII V34L polymorphism modulates the risk of chronic venous leg ulcer progression and extension. Wound Repair Regen. 2004;12(5):512–7. pmid:15453833
  23. 23. Zamboni P, Izzo M, Tognazzo S, Carandina S, De Palma M, Catozzi L, et al. The overlapping of local iron overload and HFE mutation in venous leg ulcer pathogenesis. Free Radic Biol Med. 2006;40(10):1869–73. pmid:16678024
  24. 24. Zamboni P, Tognazzo S, Izzo M, Pancaldi F, Scapoli GL, Liboni A, et al. Hemochromatosis C282Y gene mutation increases the risk of venous leg ulceration. J Vasc Surg. 2005;42(2):309–14. pmid:16102632
  25. 25. Gemmati D, Tognazzo S, Catozzi L, Federici F, De Palma M, Gianesini S, et al. Influence of gene polymorphisms in ulcer healing process after superficial venous surgery. J Vasc Surg. 2006;44(3):554–62. pmid:16950433
  26. 26. Tognazzo S, Gemmati D, Palazzo A, Catozzi L, Carandina S, Legnaro A, et al. Prognostic role of factor XIII gene variants in nonhealing venous leg ulcers. J Vasc Surg. 2006;44(4):815–9. pmid:16945500
  27. 27. Gemmati D, Federici F, Catozzi L, Gianesini S, Tacconi G, Scapoli GL, et al. DNA-array of gene variants in venous leg ulcers: detection of prognostic indicators. J Vasc Surg. 2009;50(6):1444–51. pmid:19958990
  28. 28. Gemmati D, Zeri G, Orioli E, De Gaetano FE, Salvi F, Bartolomei I, et al. Polymorphisms in the genes coding for iron binding and transporting proteins are associated with disability, severity, and early progression in multiple sclerosis. BMC Med Genet. 2012;13:70. pmid:22883388
  29. 29. Feder JN, Gnirke A, Thomas W, Tsuchihashi Z, Ruddy DA, Basava A, et al. A novel MHC class I-like gene is mutated in patients with hereditary haemochromatosis. Nat Genet. 1996;13(4):399–408. pmid:8696333
  30. 30. Zamboni P. The big idea: iron-dependent inflammation in venous disease and proposed parallels in multiple sclerosis. J R Soc Med. 2006;99(11):589–93. pmid:17082306
  31. 31. Rubio JP, Bahlo M, Tubridy N, Stankovich J, Burfoot R, Butzkueven H, et al. Extended haplotype analysis in the HLA complex reveals an increased frequency of the HFE-C282Y mutation in individuals with multiple sclerosis. Hum Genet. 2004;114(6):573–80. pmid:15014978
  32. 32. Ristic S, Lovrecic L, Brajenovic-Milic B, Starcevic-Cizmarevic N, Jazbec SS, Sepcic J, et al. Mutations in the hemochromatosis gene (HFE) and multiple sclerosis. Neurosci Lett. 2005;383(3):301–4. pmid:15955425
  33. 33. Ramagopalan SV, Cukjati M, Cernilec M, DeLuca GC, Dyment DA, Degenhardt A, et al. Mutations in the hemochromatosis gene and the clinical outcome of multiple sclerosis. J Neuroimmunol. 2008;203(1):104–7. pmid:18675463
  34. 34. Giambattistelli F, Bucossi S, Salustri C, Panetta V, Mariani S, Siotto M, et al. Effects of hemochromatosis and transferrin gene mutations on iron dyshomeostasis, liver dysfunction and on the risk of Alzheimer's disease. Neurobiol Aging. 2012;33(8):1633–41. pmid:21514009
  35. 35. Gazzina S, Premi E, Zanella I, Biasiotto G, Archetti S, Cosseddu M, et al. Iron in Frontotemporal Lobar Degeneration: A New Subcortical Pathological Pathway? Neurodegener Dis. 2016;16(3–4):172–8. pmid:26613252
  36. 36. Gebril OH, Kirby J, Savva G, Brayne C, Ince PG. HFE H63D, C282Y and AGTR1 A1166C polymorphisms and brain white matter lesions in the aging brain. J Neurogenet. 2011;25(1–2):7–14. pmid:21332426
  37. 37. Lin M, Zhao L, Fan J, Lian XG, Ye JX, Wu L, et al. Association between HFE polymorphisms and susceptibility to Alzheimer's disease: a meta-analysis of 22 studies including 4,365 cases and 8,652 controls. Mol Biol Rep. 2012;39(3):3089–95. pmid:21701828
  38. 38. Abboud S, Haile DJ. A novel mammalian iron-regulated protein involved in intracellular iron metabolism. J Biol Chem. 2000;275(26):19906–12. pmid:10747949
  39. 39. Ward DM, Kaplan J. Ferroportin-mediated iron transport: expression and regulation. Biochim Biophys Acta. 2012;1823(9):1426–33. pmid:22440327
  40. 40. Myhre O, Utkilen H, Duale N, Brunborg G, Hofer T. Metal dyshomeostasis and inflammation in Alzheimer's and Parkinson's diseases: possible impact of environmental exposures. Oxid Med Cell Longev. 2013;2013:726954. pmid:23710288
  41. 41. Bandyopadhyay S, Rogers JT. Alzheimer's disease therapeutics targeted to the control of amyloid precursor protein translation: maintenance of brain iron homeostasis. Biochem Pharmacol. 2014;88(4):486–94. pmid:24513321
  42. 42. Altes A, Bach V, Ruiz A, Esteve A, Remacha AF, Sarda MP, et al. Does the SLC40A1 gene modify HFE-related haemochromatosis phenotypes? Ann Hematol. 2009;88(4):341–5. pmid:18820912
  43. 43. Crespo AC, Silva B, Marques L, Marcelino E, Maruta C, Costa S, et al. Genetic and biochemical markers in patients with Alzheimer's disease support a concerted systemic iron homeostasis dysregulation. Neurobiol Aging. 2014;35:777–85. pmid:24199959
  44. 44. Leong WI, Lonnerdal B. Hepcidin, the recently identified peptide that appears to regulate iron absorption. J Nutr. 2004;134(1):1–4. pmid:14704284
  45. 45. Ding H, Yan CZ, Shi H, Zhao YS, Chang SY, Yu P, et al. Hepcidin is involved in iron regulation in the ischemic brain. PLoS One. 2011;6(9):e25324. pmid:21957487
  46. 46. Rogers JT, Venkataramani V, Washburn C, Liu Y, Tummala V, Jiang H, et al. A role for amyloid precursor protein translation to restore iron homeostasis and ameliorate lead (Pb) neurotoxicity. J Neurochem. 2016;138(3):479–94. pmid:27206843
  47. 47. Rogers JT, Randall JD, Cahill CM, Eder PS, Huang X, Gunshin H, et al. An iron-responsive element type II in the 5'-untranslated region of the Alzheimer's amyloid precursor protein transcript. J Biol Chem. 2002;277(47):45518–28. pmid:12198135
  48. 48. Duce JA, Tsatsanis A, Cater MA, James SA, Robb E, Wikhe K, et al. Iron-export ferroxidase activity of beta-amyloid precursor protein is inhibited by zinc in Alzheimer's disease. Cell. 2010;142(6):857–67. pmid:20817278
  49. 49. Wong CT, Saha N. Effects of transferrin genetic phenotypes on total iron-binding capacity. Acta Haematol. 1986;75(4):215–8. pmid:3096053
  50. 50. Lee PL, Ho NJ, Olson R, Beutler E. The effect of transferrin polymorphisms on iron metabolism. Blood Cells Mol Dis. 1999;25(5–6):374–9. pmid:10660486
  51. 51. Zambenedetti P, De Bellis G, Biunno I, Musicco M, Zatta P. Transferrin C2 variant does confer a risk for Alzheimer's disease in caucasians. J Alzheimers Dis. 2003;5(6):423–7. pmid:14757931
  52. 52. Zhang P, Yang Z, Zhang C, Lu Z, Shi X, Zheng W, et al. Association study between late-onset Alzheimer's disease and the transferrin gene polymorphisms in Chinese. Neurosci Lett. 2003;349(3):209–11. pmid:12951205
  53. 53. Wang Y, Xu S, Liu Z, Lai C, Xie Z, Zhao C, et al. Meta-analysis on the association between the TF gene rs1049296 and AD. Can J Neurol Sci. 2013;40(5):691–7. pmid:23968943
  54. 54. Kauwe JS, Bertelsen S, Mayo K, Cruchaga C, Abraham R, Hollingworth P, et al. Suggestive synergy between genetic variants in TF and HFE as risk factors for Alzheimer's disease. Am J Med Genet B Neuropsychiatr Genet. 2010;153B(4):955–9. pmid:20029940
  55. 55. Huang Y, Mahley RW. Apolipoprotein E: structure and function in lipid metabolism, neurobiology, and Alzheimer's diseases. Neurobiol Dis. 2014;72 Pt A:3–12.
  56. 56. Huang YA, Zhou B, Wernig M, Sudhof TC. ApoE2, ApoE3, and ApoE4 Differentially Stimulate APP Transcription and Abeta Secretion. Cell. 2017;168(3):427–41 e21. pmid:28111074
  57. 57. Morris JC, Roe CM, Xiong C, Fagan AM, Goate AM, Holtzman DM, et al. APOE predicts amyloid-beta but not tau Alzheimer pathology in cognitively normal aging. Ann Neurol. 2010;67(1):122–31. pmid:20186853
  58. 58. Verghese PB, Castellano JM, Garai K, Wang Y, Jiang H, Shah A, et al. ApoE influences amyloid-beta (Abeta) clearance despite minimal apoE/Abeta association in physiological conditions. Proc Natl Acad Sci U S A. 2013;110(19):E1807–16. pmid:23620513
  59. 59. Peters DG, Connor JR, Meadowcroft MD. The relationship between iron dyshomeostasis and amyloidogenesis in Alzheimer's disease: Two sides of the same coin. Neurobiol Dis. 2015;81:49–65. pmid:26303889
  60. 60. Hardy JA, Higgins GA. Alzheimer's disease: the amyloid cascade hypothesis. Science. 1992;256:184–5. pmid:1566067
  61. 61. Tanzi RE, Bertram L. Twenty years of the Alzheimer's disease amyloid hypothesis: a genetic perspective. Cell. 2005;120(4):545–55. pmid:15734686
  62. 62. Bush AI, Tanzi RE. Therapeutics for Alzheimer's disease based on the metal hypothesis. Neurotherapeutics. 2008;5(3):421–32. pmid:18625454
  63. 63. Lucotte G, Dieterlen F. A European allele map of the C282Y mutation of hemochromatosis: Celtic versus Viking origin of the mutation? Blood Cells Mol Dis. 2003;31(2):262–7. pmid:12972035
  64. 64. McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer's Disease. Neurology. 1984;34(7):939–44. pmid:6610841
  65. 65. Roman GC, Tatemichi TK, Erkinjuntti T, Cummings JL, Masdeu JC, Garcia JH, et al. Vascular dementia: diagnostic criteria for research studies. Report of the NINDS-AIREN International Workshop. Neurology. 1993;43(2):250–60. pmid:8094895
  66. 66. Petersen RC, Doody R, Kurz A, Mohs RC, Morris JC, Rabins PV, et al. Current concepts in mild cognitive impairment. Arch Neurol. 2001;58(12):1985–92. pmid:11735772
  67. 67. Gemmati D, Burini F, Talarico A, Fabbri M, Bertocco C, Vigliano M, et al. The Active Metabolite of Warfarin (3'-Hydroxywarfarin) and Correlation with INR, Warfarin and Drug Weekly Dosage in Patients under Oral Anticoagulant Therapy: A Pharmacogenetics Study. PLoS One. 2016;11(9):e0162084. pmid:27606428
  68. 68. Tombaugh TN, McIntyre NJ. The mini-mental state examination: a comprehensive review. J Am Geriatr Soc. 1992;40(9):922–35. pmid:1512391
  69. 69. Hixson JE, Vernier DT. Restriction isotyping of human apolipoprotein E by gene amplification and cleavage with HhaI. J Lipid Res. 1990;31(3):545–8. pmid:2341813
  70. 70. Lupton MK, Benyamin B, Proitsi P, Nyholt DR, Ferreira MA, Montgomery GW, et al. No Genetic Overlap Between Circulating Iron Levels and Alzheimer's Disease. J Alzheimers Dis. 2017;59(1):85–99. pmid:28582860
  71. 71. Hare D, Ayton S, Bush A, Lei P. A delicate balance: Iron metabolism and diseases of the brain. Front Aging Neurosci. 2013;5:34. pmid:23874300
  72. 72. Pirpamer L, Hofer E, Gesierich B, De Guio F, Freudenberger P, Seiler S, et al. Determinants of iron accumulation in the normal aging brain. Neurobiol Aging. 2016;43:149–55. pmid:27255824
  73. 73. Rottkamp CA, Raina AK, Zhu X, Gaier E, Bush AI, Atwood CS, et al. Redox-active iron mediates amyloid-beta toxicity. Free Radic Biol Med. 2001;30(4):447–50. pmid:11182300
  74. 74. Mutter J, Naumann J, Sadaghiani C, Schneider R, Walach H. Alzheimer disease: mercury as pathogenetic factor and apolipoprotein E as a moderator. Neuro Endocrinol Lett. 2004;25(5):331–9. pmid:15580166
  75. 75. Ayton S, Faux NG, Bush AI, Alzheimer's Disease Neuroimaging I. Ferritin levels in the cerebrospinal fluid predict Alzheimer's disease outcomes and are regulated by APOE. Nat Commun. 2015;6:6760. pmid:25988319
  76. 76. van Bergen JM, Li X, Hua J, Schreiner SJ, Steininger SC, Quevenco FC, et al. Colocalization of cerebral iron with Amyloid beta in Mild Cognitive Impairment. Sci Rep. 2016;6:35514. pmid:27748454
  77. 77. Ayton S, Faux NG, Bush AI. Association of Cerebrospinal Fluid Ferritin Level With Preclinical Cognitive Decline in APOE-epsilon4 Carriers. JAMA Neurol. 2017;74(1):122–5. pmid:27893873
  78. 78. Perry G, Cash AD, Smith MA. Alzheimer Disease and Oxidative Stress. J Biomed Biotechnol. 2002;2:120–3. pmid:12488575
  79. 79. Potter H, Wisniewski T. Apolipoprotein e: essential catalyst of the Alzheimer amyloid cascade. Int J Alzheimers Dis. 2012;2012:489428. pmid:22844635
  80. 80. Chico L, Simoncini C, Lo Gerfo A, Rocchi A, Petrozzi L, Carlesi C, et al. Oxidative stress and APO E polymorphisms in Alzheimer's disease and in mild cognitive impairment. Free Radic Res. 2013;47(8):569–76. pmid:23668794
  81. 81. Hofer T, Perry G. Nucleic acid oxidative damage in Alzheimer's disease-explained by the hepcidin-ferroportin neuronal iron overload hypothesis? J Trace Elem Med Biol. 2016;38:1–9. pmid:27329321
  82. 82. Lim YY, Williamson R, Laws SM, Villemagne VL, Bourgeat P, Fowler C, et al. Effect of APOE Genotype on Amyloid Deposition, Brain Volume, and Memory in Cognitively Normal Older Individuals. J Alzheimers Dis. 2017;58(4):1293–302. pmid:28550258
  83. 83. Buchanan DD, Silburn PA, Chalk JB, Le Couteur DG, Mellick GD. The Cys282Tyr polymorphism in the HFE gene in Australian Parkinson's disease patients. Neurosci Lett. 2002;327(2):91–4. pmid:12098643
  84. 84. Correia AP, Pinto JP, Dias V, Mascarenhas C, Almeida S, Porto G. CAT53 and HFE alleles in Alzheimer's disease: a putative protective role of the C282Y HFE mutation. Neurosci Lett. 2009;457(3):129–32. pmid:19429178
  85. 85. Sampietro M, Caputo L, Casatta A, Meregalli M, Pellagatti A, Tagliabue J, et al. The hemochromatosis gene affects the age of onset of sporadic Alzheimer's disease. Neurobiol Aging. 2001;22(4):563–8. pmid:11445256
  86. 86. Combarros O, Garcia-Roman M, Fontalba A, Fernandez-Luna JL, Llorca J, Infante J, et al. Interaction of the H63D mutation in the hemochromatosis gene with the apolipoprotein E epsilon 4 allele modulates age at onset of Alzheimer's disease. Dement Geriatr Cogn Disord. 2003;15(3):151–4. pmid:12584430
  87. 87. Pulliam JF, Jennings CD, Kryscio RJ, Davis DG, Wilson D, Montine TJ, et al. Association of HFE mutations with neurodegeneration and oxidative stress in Alzheimer's disease and correlation with APOE. Am J Med Genet B Neuropsychiatr Genet. 2003;119B(1):48–53. pmid:12707938
  88. 88. Connor JR, Lee SY. HFE mutations and Alzheimer's disease. J Alzheimers Dis. 2006;10(2–3):267–76. pmid:17119292
  89. 89. Ali-Rahmani F, Schengrund CL, Connor JR. HFE gene variants, iron, and lipids: a novel connection in Alzheimer's disease. Front Pharmacol. 2014;5:165. pmid:25071582
  90. 90. Nielsen JE, Jensen LN, Krabbe K. Hereditary haemochromatosis: a case of iron accumulation in the basal ganglia associated with a parkinsonian syndrome. J Neurol Neurosurg Psychiatry. 1995;59(3):318–21. pmid:7673967
  91. 91. Berg D, Hoggenmuller U, Hofmann E, Fischer R, Kraus M, Scheurlen M, et al. The basal ganglia in haemochromatosis. Neuroradiology. 2000;42(1):9–13. pmid:10663462
  92. 92. Zhu WZ, Zhong WD, Wang W, Zhan CJ, Wang CY, Qi JP, et al. Quantitative MR phase-corrected imaging to investigate increased brain iron deposition of patients with Alzheimer disease. Radiology. 2009;253:497–504. pmid:19709998
  93. 93. Connor JR, Menzies SL, St Martin SM, Mufson EJ. A histochemical study of iron, transferrin, and ferritin in Alzheimer's diseased brains. J Neurosci Res. 1992;31(1):75–83. pmid:1613823
  94. 94. Miyata M, Smith JD. Apolipoprotein E allele-specific antioxidant activity and effects on cytotoxicity by oxidative insults and beta-amyloid peptides. Nat Genet. 1996;14(1):55–61. pmid:8782820
  95. 95. Cho HH, Cahill CM, Vanderburg CR, Scherzer CR, Wang B, Huang X, et al. Selective translational control of the Alzheimer amyloid precursor protein transcript by iron regulatory protein-1. J Biol Chem. 2010;285(41):31217–32. pmid:20558735
  96. 96. Lambracht-Washington D, Qu BX, Fu M, Anderson LD Jr., Eagar TN, Stuve O, et al. A peptide prime-DNA boost immunization protocol provides significant benefits as a new generation Abeta42 DNA vaccine for Alzheimer disease. J Neuroimmunol. 2013;254(1–2):63–8. pmid:23036592
  97. 97. Komori N, Kittel A, Kang D, Shackelford D, Masliah E, Zivin JA, et al. Reversible ischemia increases levels of Alzheimer amyloid protein precursor without increasing levels of mRNA in the rabbit spinal cord. Brain Res Mol Brain Res. 1997;49(1–2):103–12. pmid:9387869
  98. 98. Fujioka M, Taoka T, Matsuo Y, Mishima K, Ogoshi K, Kondo Y, et al. Magnetic resonance imaging shows delayed ischemic striatal neurodegeneration. Ann Neurol. 2003;54(6):732–47. pmid:14681883
  99. 99. Sheykhansari S, Sitti M, Gemmati D, Zamboni P, Singh AV. Redox Metals Homeostasis in Multiple Sclerosis and Amyotrophic Lateral Sclerosis: A Review. Cell Death and Disease 2018. (in press).
  100. 100. Heidari M, Gerami SH, Bassett B, Graham RM, Chua AC, Aryal R, et al. Pathological relationships involving iron and myelin may constitute a shared mechanism linking various rare and common brain diseases. Rare Dis. 2016;4(1):e1198458. pmid:27500074
  101. 101. Matias-Guiu JA, Oreja-Guevara C, Cabrera-Martin MN, Moreno-Ramos T, Carreras JL, Matias-Guiu J. Amyloid Proteins and Their Role in Multiple Sclerosis. Considerations in the Use of Amyloid-PET Imaging. Front Neurol. 2016;7:53. pmid:27065425