3-Hydroxy-3-methylglutaryl-CoA reductase inhibitors, known as statins, are the most prescribed and effective pharmacological therapy for treating hypercholesterolemia and prevent cardiovascular events(Reference Thompson, Panza and Zaleski1–Reference Tomaszewski, Stêpieñ and Tomaszewska3). Despite their favourable overall safety profile, significant adverse effects of statin treatment have been reported, especially statin-associated muscle symptoms, defined as myalgia or muscle weakness with or without an elevation in serum creatine kinase (CK) levels(Reference Thompson, Panza and Zaleski1,Reference Fuhrmeister, Tews and Kromer2,Reference Ahmadi, Ghorbanihaghjo and Naghi-Zadeh4–Reference Selva-O’Callaghan, Alvarado-Cardenas and Pinal-Fernández7) . It has been suggested that the statin-associated muscle symptoms are associated with oxidative stress, in which the imbalance between antioxidants and the over-generation of free radicals, including reactive oxygen species, could induce apoptosis in the skeletal muscles(Reference Ahmadi, Ghorbanihaghjo and Naghi-Zadeh4). The inclusion of foods with antioxidant characteristics may be a worthwhile strategy to improve antioxidant capacity(Reference Cardoso, Busse and Hare8) and reduce statin-associated muscle symptoms risk in patients using statins. The consumption of Brazil nuts (Bertholletia excelsa, family Lecythidaceae) has become popular among recent studies due to its significant amount of Se in the main form of selenomethionine, being a viable strategy to increase dietary intake of Se in different conditions and populations(Reference Donadio, Rogero and Guerra-Shinohara9,Reference Cardoso, Duarte and Reis10) . Moreover, the increase of Se intake due to the regular consumption of Brazil nuts has been associated with antioxidant and anti-inflammatory effects, modulation of blood lipids, decreased atherogenic risk(Reference Cardoso, Duarte and Reis10) and considered as coadjutant therapy in the prevention/treatment of CVD(Reference Silva, Cardozo and Cruz11).
In the human population, the individual variation in response to Se consumption, irrespective of baseline Se status, may indicate gene variants’ influence(Reference Reszka, Jablonska and Gromadzinska12). Genetic polymorphisms have been recognised as an important source of interindividual variation in response to nutritional supplementation. Several SNP in selenoprotein genes could give rise to interindividual variations in Se metabolism and response to Se supplementation(Reference Donadio, Rogero and Guerra-Shinohara9,Reference Donadio, Rogero and Cockell13,Reference Méplan, Crosley and Nicol14) . In particular, SNP in glutathione peroxidase (GPX1) (encoding GPX1) and selenoprotein P (SELENOP) (encoding SELENOP) have been shown to affect blood Se or selenoprotein levels after supplementation. The GPX1 rs1050450 polymorphism is a proline to leucine substitution at codon 197, resulting in reduced enzyme activity and higher DNA damage levels(Reference Kopp, Outzen and Olsen15). The impact of SELENOP variation in codon 234, associated with alanine to threonine change (rs3877899) and guanine to adenine transition within 3' untranslated region (UTR) of SELENOP mRNA (rs7579), may result in changes in Se metabolism, but the involved mechanisms are not entirely elucidated(Reference Méplan, Crosley and Nicol14). Moreover, polymorphisms or mutations in selenoproteins’ genes and synthesis cofactors may negatively influence the natural protection against oxidative stress and could be potential determinants of blood lipid levels, increasing the risk for many diseases(Reference Cardoso, Busse and Hare8,Reference Zoidis, Seremelis and Kontopoulos16,Reference Galan-Chilet, Tellez-Plaza and Guallar17) .
In a previous dietary intervention with patients using statins regularly(Reference Watanabe, Fernandes de Lima and Ferraz-Bannitz18), we found that one Brazil nut intake for 3 months modulated the oxidative stress by increasing erythrocyte GPX activity and blood Se levels while influencing the decreased levels of malondialdehyde. We also found that Se status post-Brazil nut consumption did not enhance selenoproteins gene expression(Reference Watanabe, Fernandes de Lima and Ferraz-Bannitz18). A hypothesis that may enrich the discussion presented in our previous study is that the response to Brazil nut consumption is influenced by specific polymorphisms, especially those involved in Se transportation and distribution, affecting biomarkers of metabolic pathways Se acts. Thus, in the present study, we expanded our work on the same thirty-two patients from the previous study to investigate whether SNP in GPX1 (rs1050450) and SELENOP (rs3877899 and rs7579) modify the effect of increased dietary Se intake after Brazil nut consumption on biomarkers of Se status, antioxidant protection, muscle homoeostasis, lipid profile and selenoproteins mRNA expression in the same study population.
Subjects and methods
This was an open, non-randomised, single-centre study performed at the Ribeirão Preto Medical School University Hospital, University of São Paulo, Brazil, from January 2017 to September 2018. Thirty-two patients fulfilled the inclusion criteria: age between 18 and 60 years, both sexes and using statin continuously. We did not include patients under at least one of the following conditions: nuts allergy, taking multivitamins or mineral supplements, using antibiotics or other medications that are also metabolised by cytochrome P450 3A4 (CYP3A4), in current tobacco or alcohol consumption, engaging in intense physical activity, severe cardiac complications, thyroid disorders, liver disease, kidney failure or neoplasia.
In order to investigate the influence of SNP in response to daily Brazil nut consumption, the volunteers were genotyped for SNP in GPX1 (rs1050450) and SELENOP (rs3877899 and rs7579) and distributed as homozygous wild type, heterozygous and homozygous variant.
All participants received one unit of Brazil nut daily for 3 months. The composition of Brazil nuts was described previously(Reference Watanabe, Fernandes de Lima and Ferraz-Bannitz18). The Se content in Brazil nuts was 58·1 (sem 2·1) µg/g, and the average weight was 5 g; therefore, each nut provided approximately 290 µg of Se(Reference Watanabe, Fernandes de Lima and Ferraz-Bannitz18). Clinical evaluation and interviews were conducted at the beginning of the study to obtain general information (age, sex) and personal habits (co-morbidities, presence of allergies, use of medications, vitamins and mineral supplements consumption, and physical activity). Additionally, we collected venous blood samples for biochemical assays. The blood sample collection occurred before (pre-intervention period) and after (post-intervention period) 12 weeks of Brazil nuts consumption. The blood collection was performed in the University Hospital Clinical Research Unit (UPC) after 8 h of fasting. For blood plasma collection, blue-capped trace metals-free tubes containing K2EDTA anticoagulant (BD Vacutainer®) were used. Separation of whole blood to obtain plasma and erythrocytes occurred immediately after collection. An aliquot of 500 μl of whole blood was stored in 1·5 ml sterile plastic tubes for RNA extraction and subsequent gene expression. Samples were stored at –80° until the time of analysis. Data of oxidative stress, CK activity and selenoproteins mRNA expression were previously reported in(Reference Watanabe, Fernandes de Lima and Ferraz-Bannitz18) but were re-examined under a different stratification in the present study, according to the genotype for SNP in GPX1 (rs1050450) and SELENOP (rs3877899 and rs7579).
The Ethics Committee of Ribeirão Preto Medical School at the University of São Paulo, Brazil (protocol number CAAE: 56221916.5.0000.5440) approved the study protocol. Informed consent was obtained from all individual participants. The trial was registered at the Brazilian Clinical Trial Registry under identification number (RBR-7rwgzt).
Weight was measured with an electronic platform (Filizola) scale with a precision of 0·1 kg and a maximum capacity of 300 kg. Height was measured with a vertical shaft with 0·5 cm graduation. BMI was calculated by dividing the body mass by the square of the body height, universally expressed in units of kg/m2.
Evaluation of dietary intake
Dietary intake was assessed by food record over three non-consecutive days. Patients were oriented to self-record the type and amount of food and beverages consumed. The information was processed in the nutritional analysis programme Dietpro®, 5i version. Brazilian databases were used to determine the food chemical composition: the food chemical composition table developed by Sonia Tucunduva Philippi(Reference Philippi19) and the Brazilian food chemical composition table, TACO(20).
The Clinical Analysis Laboratories of HCFMRP-USP take part in rigorous external quality control programmes nationally and internationally, among them: Laboratory Testing Proficiency, Brazilian Society of Clinical Pathology/Laboratory Medicine (SBPC/ML), College of American Pathologists (CAP) and Oneworld Accuracy.
Commercially available kits by Labtest (Minas Gerais) were used following the manufacturer’s protocol. Serum CK activity was measured by UV kinetic increasing reaction (Cat. No. 117). Total cholesterol in serum was measured by quantitative endpoint colorimetric assay (Cat. No. 76). LDL (Cat. No. 111) and HDL (Cat. No. 145) were measured by selective surfactant. TAG (Cat. No. 87) were measured by the Enzyme-Trinder method. The intra-assay CV was less than 10 % for all of the serum markers. As each biomarker was made within the same assay, we did not consider the inter-assay variability. GPX activity was measured in erythrocytes according to the method of Paglia and Valentine(Reference Paglia and Valentine21), as described previously(Reference Watanabe, Fernandes de Lima and Ferraz-Bannitz18). The determination of total Se concentration in plasma and erythrocytes was performed according to Batista et al. (Reference Batista, Rodrigues and Nunes22) by an Inductively Coupled Plasma MS (NexIon 2000B, Perkin Elmer). Samples were diluted in the ratio 1:50 with a solution containing Triton X-100 0·01 % (v/v), HNO3 0·05 % (v/v) and 10 mg/l-1 rhodium (Rh) as an internal standard. The concentration of the analytical calibration standards ranged from 0 to 50 μg/l(Reference Batista, Rodrigues and Nunes22). To verify the accuracy and precision of data, the analysis of reference materials – QMEQAS07B06 and QMEQAS07B03 human blood from the L’Institut National de Santé Publique du Quebec (Canada) – was performed at the beginning of the assay and after running ten samples. Originated values were constantly in agreement with target values (recoveries > 90 % and t test, 99 %). Repeatability (intra-assay precision) was < 8 %. The method detection limit is 0·2 μg/l.
Genotyping of the rs1050450, rs7579 and rs3877899 SNP
Isolation of DNA from whole blood was carried out using a PureLink Genomic DNA kit (Invitrogen, Life Technologies Inc.), and the concentration was measured using a NanoDrop ND 1000 spectrophotometer (Thermo Fisher Scientific). Polymorphisms in GPX1 (rs1050450) and SELENOP (rs7579 and rs3877899) genes were determined by real-time quantitative PCR with TaqMan SNP Genotyping assays (Life Technologies). The assays were obtained as pre-designed from Applied Biosystems for rs7579 and rs3877899 (ID Assays C___8806056_10 and C___2841533_10, respectively) and custom-made through Custom TaqMan® Genomic Assays service for rs1050450 (GPX1/rs1050450: primers: F: 5′-TGT GCC CCC TAC GCA GGT ACA-3′, R: 5′-CCC CCG AGA CAG CAG CA-3′, T-allele: 5′-FAM-CTG TCT CAA GGG CTC AGC TGT-MGB-3′, C-allele: 5′-VIC-CTG TCT CAA GGG CCC AGC TGT-MGB-3′). The quantitative PCR reaction contained 3·125 μl of 1Å˜ GoTaq® Master Mix (Promega), 0·156 μl of the 20Å˜ SNP Genotyping Assay (Life Technologies) and 20 ng of genomic DNA in a 6·741 μl total reaction volume. Samples were run on Applied Biosystems 7500 real-time PCR system (Applied Biosciences/Thermo Fisher Scientific) under standard conditions. The validity of GPX1 rs1050450 quantitative PCR results was verified by direct sequencing 5 % of the samples using the BigDye FN Sequencing Kit (PE Applied Biosystems) and the following primers: primer forward 5′-CATCGAAGCCCTGCTGTCT-3′; primer reverse 5′-CACTGCAACTGCCAAGCA-3′. The agreement of the genotypes determined for the blinded quality control samples was 100 %
GPX1 and SELENOP gene expression
We previously described the procedures of RNA extraction and cDNA synthesis(Reference Watanabe, Fernandes de Lima and Ferraz-Bannitz18). The analysis of gene expression was performed by quantitative PCR using TaqMan Gene expression Assays for GPX1 (Hs00829989_gH) and SELENOP (Hs01032845_m1). β-actin (4352935E) mRNA expression was used as a reference gene. The relative expressions were calculated using the 2–ΔΔCt method(Reference Livak and Schmittgen23).
For all statistical analyses, homozygous and heterozygous individuals for the variant alleles were combined in one category, leaving the homozygous wild type in another category to increase the statistical power. The χ 2 test was used to determine whether genotype frequencies followed the Hardy–Weinberg equilibrium. A repeated-measures ANOVA adjusted for multiple comparisons with the Bonferroni test was used to determine the intergroup effect between genotypes. According to genotype, the post-intervention data (after Brazil nut consumption) were compared with pre-intervention data (before Brazil nut consumption) and were assessed by the paired Student’s t test or Wilcoxon test. Pearson’s test was applied to determine the correlations between genotype, GPX enzyme activity and erythrocyte Se during the intervention period. Data were plotted in Statistical Package for the Social Sciences software version 14·0 (SPSS) and GraphPad Prism software (GraphPad Software Inc.). Differences were considered significant if P < 0·05.
Thirty-two participants were included and completed the study protocol. Among them, 59·4 % were men, and 40·6 % were women. Participants’ mean age was 50·1 (sem 7·6) years. According to BMI, patients were classified as obese or overweight (31·1 (sem 3·8) kg/m2). According to the analysis of non-consecutive dietary food records, at baseline, the participants’ energy intake was 8487·9 (sem 1266·5) kJ/d and, all participants maintained their regular diet during the intervention (data not shown). At the beginning of the study, as part of the questionnaire of personal habits, participants were asked about the type, intensity and frequency of physical activity. Most of the participants were not practicing regular physical activity (84·4 %), and they did not change this habit during the study. Regarding the type of cholesterol-lowering medication, 71·9 % were using simvastatin (40 mg) and 28·1 % atorvastatin (40 mg). The pre-supplementation variables of the study population were not different among genotype groups (data not shown).
Genotype distribution and variant allele frequencies for the GPX1 rs1050450, SELENOP rs7579 and SELENOP rs3877899 polymorphisms are shown in Table 1. One participant has excluded from SELENOP rs7579 evaluation due to the inconclusive genotype. The genotype distribution of the studied polymorphisms was in Hardy–Weinberg equilibrium (data not shown).
C, cytosine; T, thymine; A, adenine; G, guanine; GPX, glutathione peroxidase; SELENOP, selenoprotein P; MAF, minor allele frequency.
The pre-and post-Brazil nut consumption and genotype effect on erythrocyte GPX activity, plasma and erythrocyte Se, and CK activity for each SNP are shown in Table 2. In the present study, the significant increase in plasma and erythrocyte Se after Brazil nut consumption occurred regardless of the genotypes. Also, the increase in both erythrocyte GPX activity and CK activity was not significant post-intervention in the presence of variant alleles in SELENOP.
C, cytosine; T, thymine; A, adenine; G, guanine; GPX, glutathione peroxidase.
† Repeated-measures two-way ANOVA adjusted for multiple comparison with Bonferroni test. Differences in genotype effect are represented as *P < 0·05.
Total cholesterol, TAG, LDL and HDL levels were stratified according to SELENOP rs3877899 genotypes (Table 3). After the intervention, TAG and LDL levels decreased significantly in wild-type individuals. However, in allele T carriers’ presence, we did not find significant differences between pre- and post-intervention for the same parameters, and the intergroup comparison indicated higher levels of TAG and LDL in these individuals than the wild-type carriers. No differences were observed in the lipid profile in the presence of rs7579 and rs1050450 polymorphisms (data not shown).
C, cytosine; T, thymine; A, adenine; G, guanine.
† Wilcoxon matched-pairs signed-rank test. Differences in genotype effect are represented as *P < 0·05.
We also investigated the effects of genotype and Brazil nut consumption in selenoprotein genes’ mRNA expression (Fig. 1). After Brazil nut consumption, GPX1 expression’s up-regulation occurred only in individuals with CC genotype for rs1050450 (P = 0·0361) but not in CT + TT (P = 0·4375). (Fig. 1(a)). In the presence of the variant allele A for rs7579, we observed an up-regulation of SELENOP mRNA expression when compared with the wild type despite the intervention (P Genotype = 0·05) (Fig. 1(c)). SELENOP mRNA expression was not influenced by rs3877899 polymorphism in response to Brazil nut consumption (Fig. 1(b)).
Individual variations in Se supplementation response may partly be reflected by the selenoprotein gene variant(Reference Reszka, Jablonska and Gromadzinska12,Reference Kopp, Outzen and Olsen15,Reference Duntas and Benvenga24) . Recent studies have demonstrated that selenoprotein SNP, including GPX1 rs1050450, SELENOP rs3877899 and SELENOP rs7579, may interfere with Se utilisation and effectiveness(Reference Reszka, Jablonska and Gromadzinska12,Reference Duntas and Benvenga24) . The study of Méplan et al. provided 100 µg of sodium selenite/d and observed that SNP in selenoproteins might predict the behaviour of biomarkers of Se status and response to supplementation, influencing interindividual differences in Se metabolism(Reference Méplan, Crosley and Nicol14). In the present study, participants received about three times more Se, approximately 290 µg of Se/d, which is higher than the RDA for adults (55 mg/d) but less than the upper limit (400 mg/d) established by the Institute of Medicine (IOM), 2000(Reference Watanabe, Fernandes de Lima and Ferraz-Bannitz18). Nevertheless, we did not find significant differences in blood Se levels between selenoprotein SNP genotypes, pre- and post-intervention. Thus, the current analysis indicated that the variations in Se biomarkers assessed were strongly dependent on Se status, and at saturated Se condition, some functional differences of gene polymorphisms could be masked.
Considering the role of GPX in the antioxidant defense system, it appears that subjects with the variant allele in SELENOP rs7579 and rs3877899 had lower enzyme activity, indicating that these SNP could affect the antioxidant protection. Studies have demonstrated that the presence of SELENOP variant alleles could modify the efficiency of selenocysteine incorporation into SELENOP, the stability of SELENOP and cellular Se uptake, affecting the Se metabolism, blood concentrations and the synthesis of other selenoproteins, including antioxidant enzymes(Reference Cardoso, Busse and Hare8,Reference Donadio, Rogero and Guerra-Shinohara9,Reference Méplan, Crosley and Nicol14,Reference Kopp, Outzen and Olsen15,Reference Méplan, Nicol and Burtle25) . Previous studies have also correlated the rs7579 polymorphism with several clinical conditions characterised by oxidative stress, such as diabetes and cancer(Reference Zhou, Luo and Yin26,Reference Gharipour, Ouguerram and Nazih27) . Moreover, Gharipour et al. observed that the existence of variant alleles in rs3877899 decreased the antioxidant activity in Iranian patients with CVD, confirming the importance of SELENOP in antioxidant defense(Reference Gharipour, Ouguerram and Nazih27).
The muscular side effects associated with statin treatment could resemble the symptoms of Se deficiency(Reference Gröber, Schmidt and Kisters28). Thus, Se supplementation could mitigate statin’s side effects, mainly due to its antioxidant defense role(Reference Watanabe, Fernandes de Lima and Ferraz-Bannitz18). Nonetheless, in the presence of SELENOP rs7579 and rs3877899 variant alleles, the CK activity, a biomarker to muscle damage, did not change after Brazil nut consumption, indicating a possible impairment in the Se carrying capacity to extra-hepatic tissues via SELENOP(Reference Gharipour, Ouguerram and Nazih27).
Previous studies have demonstrated that the intake of Brazil nuts improves blood lipid profile in adults(Reference Donadio, Rogero and Guerra-Shinohara29,Reference Carvalho, Huguenin and Luiz30) , which could be an auxiliary strategy in treating dyslipidaemia, resulting in a reduction in the statin dosage and avoiding muscular side effects dose dependent. However, a genetic variation in SELENOP rs3877899 altered the response to Brazil nut consumption regarding LDL and TAG levels, suggesting that this polymorphism influenced Se’s ability to improve biomarkers of lipid profile.
Alterations in selenoprotein activity and concentration during Se depletion and repletion are accompanied by changes in the mRNA levels(Reference Reszka, Jablonska and Gromadzinska12). In the present study, Brazil nut consumption effectively increased GPX1 mRNA expression in circulating blood leukocytes only in individuals of the CC genotype at rs1050450, corroborating with Cardoso et al. and Donadio et al. (Reference Cardoso, Busse and Hare8,Reference Donadio, Rogero and Guerra-Shinohara9,Reference Donadio, Rogero and Cockell13) . Nevertheless, other studies did not find an association between Se supplementation and selenoprotein transcript levels in circulating leukocytes and whole blood(Reference Reszka, Jablonska and Gromadzinska12,Reference Reszka, Gromadzinska and Jablonska31,Reference Sunde, Paterson and Evenson32) . A possible explanation for these conflicting findings could be the excess levels of circulating Se after Brazil nut consumption affecting selenoproteins’ expression and activity(Reference Watanabe, Fernandes de Lima and Ferraz-Bannitz18,Reference Misra, Boylan and Selvam33,Reference Duarte, Reis and Rogero34) . Jablonska et al. (Reference Jablonska, Gromadzinska and Reszka35) also found a functional significance of rs1050450 polymorphism resulting in a different response of GPX1 activity to Se supplementation. They pointed to the importance of the genetic background in assessing the Se status, especially in the supplementation trials(Reference Jablonska, Gromadzinska and Reszka35). Another possibility could be Se’s chemical form since the Se compounds with singular chemical characteristics are metabolised by distinct pathways, varying in their abilities to produce distinct Se metabolites(Reference Weekley and Harris36,Reference Burk and Hill37) . Interestingly, Se in Brazil nuts is found in the main form of selenomethionine (concentration ranges from 75 to 90 %), which has a high bioavailability and low toxicity(Reference Watanabe, Fernandes de Lima and Ferraz-Bannitz18,Reference Rita Cardoso, Apolinário and da Silva Bandeira38) , being a valuable alternative for Se supplementation(Reference Cardoso, Duarte and Reis10).
The variant allele for rs7579 was associated with increased SELENOP expression pre- and post-Brazil nut supplementation. One possibility is the influence of SELENOP polymorphisms in the ratio of SELENOP isoforms in plasma, affecting the SELENOP mRNA expression, especially the 60 kDa isoform, found in higher amounts in the presence of the variant allele A for rs7579(Reference Cardoso, Busse and Hare8,Reference Donadio, Rogero and Guerra-Shinohara9,Reference Reszka, Jablonska and Gromadzinska12,Reference Méplan, Nicol and Burtle25) .
In this study, due to the small size sample, it was not possible to detect the effect of sex on biochemical parameters. The other limitation was the absence of a control group. However, since this study’s primary purpose was to investigate the effect of the genotypes on the dietary intervention response, we decided that the wild-type genotype for each SNP would be the best control.
In conclusion, our results indicated that SNP in SELENOP were associated with interindividual differences in Se utilisation after Brazil nut consumption in patients using statins, indicating that the antioxidant protection, muscle homoeostasis and lipid profile are depending upon the presence of genetic polymorphisms. We also demonstrated the influence of selenoprotein polymorphisms rs1050450 and rs7579 on the mRNA expression of the selenoproteins GPX1 and SELENOP. However, more studies are required to understand the specific molecular mechanisms by which the genetic components interact with other interindividual variations, affecting the response of Se supplementation and, consequently, the Se metabolism. This knowledge could assist in a more individual approach to Se supplementation in clinical practice.
The authors are grateful for the assistance from Renata Danielle Scchieri Pugliesi (Laboratory Technician) and Margaret de Castro (Coordinator) of the Molecular Endocrinology Laboratory of the Ribeirão Preto Medical School of the University of São Paulo (USP), and Lucia A Seale (Assistant researcher) of the Cell and Molecular Biology Laboratory of the University of Hawaii at Manoa, HI, USA.
This work was supported by Brazilian grants from Fundação de Amparo a Pesquisa do Estado de São Paulo (FAPESP) to L. M. W. under grant 2016/05677-7 and to F. B. J. and A. M. N. under grant 2018/24069-3. Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) also supported L. M. W. under grant 1591486. Funding sources had no involvement in the study design, collection, analysis and interpretation of data from the present study.
Statement of authors’ contributions to the manuscript: conceptualization - L. M. W., F. B. J. and A. M. N. Investigation - L. M. W. and L. F. L. Visualization - L.M.W., R. F. B., R. D., M. P. G., A. C. B. and M. C. F. F. Formal analysis, validation and writing - L. M. W. and A. C. B.Project administration - L. M. W. All authors read and approved the final manuscript.
The authors declare that they have no financial conflict of interest.