HFE Mutations C282Y and H63D in Iranian Population With Type 2 Diabetes

authors:

avatar Neda Golchin 1 , avatar Hajie Bibi Shahbazian 2 , avatar Heshmatollah Shahbazian 2 , avatar Alireza Zare Bidoki 3 , avatar Javad Mohammadi Asl 2 , *

Cancer Research Center, Ahvaz Jundishapur University of Medical Sciences, Ahvaz, IR Iran
Diabetes Research Center, Ahvaz Jundishapur University of Medical Sciences, Ahvaz, IR Iran
Molecular Immunology Research Center, Tehran University of Medical Sciences, Tehran, IR Iran

how to cite: Golchin N, Shahbazian H B, Shahbazian H, Zare Bidoki A, Mohammadi Asl J. HFE Mutations C282Y and H63D in Iranian Population With Type 2 Diabetes. Jentashapir J Cell Mol Biol. 2015;6(2):e24659. https://doi.org/10.5812/jjhr.6(2)2015.24659.

Abstract

Background:

Type 2 diabetes (T2D) is a common metabolic disease caused by insulin secretion defects, which is associated with a variety of complications such as retinopathy, nephropathy, and neuropathy.

Objectives:

Regarding the relationship between type 2 diabetes and hereditary chromatists, we conducted a genetic analysis on two previously reported mutations C282Y and H63D related to the HFE gene in our population.

Patients and Methods:

Altogether, 145 patients with type 2 diabetes and 145 healthy controls were examined. A genotyping assay performed using electrophoresis of the DNA digestion products from MboI and RsaI for H63D and C282Y, respectively.

Results:

Results showed a significant difference between case and controls regarding C282Y (P value < 0.001) and H63D genotypes (P value = 0.013). We also found a relationship between both mutations and nephropathy. Moreover, the difference between C282Y genotypes of patients with retinopathy and healthy controls were statistically significant (P value = 0.020) while there was no association between H63D and retinopathy. In addition, observed differences of both mutations were significant when nephropathic patients compared to the controls.

Conclusions:

Our study showed a significant association between H63D and C282Y mutations and the risk of type 2 diabetes in Iranian population.

1. Background

Type 2 diabetes (T2D) is a common metabolic disease caused by insulin secretion defects, which is associated with a variety of complications such as retinopathy, nephropathy, and neuropathy (1, 2). Its worldwide prevalence is estimated to reach around 438 million people in 2030 (3) and a combination of genetic and environmental elements have been suggested in this regard (4). Several studies have found that higher body iron stores are related to the pathogenesis of T2D (5-7). Meanwhile, genome wide association studies suggested a number of genes associated with iron metabolism and the risk of monogenic and/or syndromic forms of T2D (8).

HFE, which is also responsible for hereditary hemochromatosis (HH), is one of those genes. It has revealed that there is a link between T2D and HH so that 20% to 50% of patients with HH also develop T2D (9-11). According to the Ensembl database (http://www.ensembl.org/index.html), HFE gene also called HLA-H and located on the small arm of chromosome 6 (6: 26,087,509-26,098,571), has 14 transcripts, 12 of them could be translated into protein. Two up-stream, missense variants of this gene known as C282Y, which converts 845G→A, and H63D exchanges 187 C→G have been under examination in different studies showing their clinical significance. Investigations on problems such as coronary heart disease (12, 13), cardiomyopathy (14), myocardial infarction (15), hepatic fibrosis (16, 17), colon cancer (18), atherothrombotic cerebral infarction (19), cumulative lead exposure (20), atherosclerosis (21), nonalcoholic fatty liver disease (22), gestational diabetes (23), type 1 diabetes (24), and type 2 (25) have found significant results.

2. Objectives

As studies on T2D showed inconsistent data, we aimed to investigate whether there is any relationship between HFE mutations and the risk of type 2 diabetes in Iranian population.

3. Patients and Methods

3.1. Study Population

Diabetic patients were selected according to standard international guidelines of diagnosis among those referring to the outpatient diabetes clinic of Golestan Hospital Ahvaz, Khuzestan, Iran. A total of 145 patients and 145 matched control subjects (with regard to ethnicity, sex, and age) were enrolled in our study. All the participants signed the informed consent forms prior to enrollment. The local Ethics Committee and appropriate institutional review board approved the study.

3.2. Sample Preparation

Three milliliter peripheral blood was collected in EDTA containing tubes and stored at -20°C. DNA extraction performed using a DNA extraction kit (Bioneer, Korea) under manufacturer protocol and DNA samples were stored at -70°C until the time of experiments.

3.3. Genotyping

Polymerase chain reaction (PCR) was performed using a PCR kit (Fermentas, Germany) via a thermocycler (Eppendorf, Germany). Restriction fragment length polymorphism (RFLP) was performed using RsaI and MboI restriction enzymes (Bioneer, Korea) (Table 1). Enzyme treated products were run on a 8% polyacrylamide gel visualized by a gel document (CAUTION-ST4, France) and sizes of DNA bands were distinguished using a 50 base pair ladder (GeneOn GmbH, Germany). Ten samples were sequenced by an ABI 3730 sequencer in order to confirm the RFLP data. The investigator responsible for genotype analysis was blinded to this study.

Table 1.

PCR-RFLP Test Information a

SNPPrimerProduct LengthRestriction EnzymeFragment Length
H63D (rs799945)294MboI138
F: ATGGTTAAGGCCTGTTGCTCTGTC99
R: CCCTTGCTGTGGTTGTGATTTTC57
C282Y (rs1800562)489RsaI245
F: TCCTCTTTCCTGTCAAGTGC244
R: GATGACTCCAATGACTAGGG

3.4. Statistical Analysis

Statistical analysis was accomplished using an SPSS (v16) software and OpenEpi online tool. Allele and genotype frequencies were determined by direct gene counting and analyzed by chi-square test. All of the alleles evaluated in this study complied with Hardy-Weinberg equilibrium. Odds ratios (OR) and 95% confidence intervals (CI) were estimated. All the tests were two-sided and the probability of less than 0.05 was considered as statistically significant.

4. Results

Genotyping results extracted from 145 Iranian patients (72 males and 73 females), with the mean age of 53.9 ± 9 years compared to healthy controls with the same sex ratio and the mean range of 51.3 ± 10 years are shown in Tables 2-5. 2-5. Differences between genotypes in diabetic patients and controls are illustrated in Table 2. As it has been shown, the frequency of GG genotype related to polymorphism C282Y is significantly higher, and the GA genotype is lower among patients in comparison with the controls (P < 0.001). Also, regarding the H63D, the differences between CG and CC genotypes are statistically significant in patients compared to healthy controls (P = 0.013). Tables 3 and 4 demonstrate patients who are classified in two groups with nephropathy and retinopathy complications, respectively; in 45 patients with nephropathy, C282Y genotypes were significantly different between patients with nephropathy and controls while the difference between H63D genotypes was not statistically significant. In addition, differences in genotypes of both C282Y and H63D were significant comparing our nephropathic patients and controls (P value < 0.001 and 0.006, respectively). Despite increased homozygote and heterozygote alleles at studied positions in patients with nephropathy compared to the retinopathic patients, it did not achieve significance (Table 5).

Table 2.

Genotype Frequencies of C282Y and H63D Genes in Patients With Type 2 Diabetes and Controls a

PositionGenotypesType 2 Diabetes, (n = 145)Controls, (n = 145)P Value bOdds Ratio (95% Confidence Interval)
C282Y0.000
GA97 (67)58 (40)3.01 (1.87 - 4.90)
GG48 (33)87 (60)0.33 (0.20 - 0.53)
H63D0.013
CG75 (52)54 (37)1.80 (1.12 - 2.89)
CC70 (48)91 (63)0.55 (0.34 - 0.88)
Table 3.

Genotype Frequencies of C282Y and H63D Genes in Diabetic Patients With Retinopathy and Controls a

PositionGenotypesRetinopathy (n = 45)Controls (n = 145)P Value bOdds Ratio (95% Confidence Interval)
C282Y0.020
GA27 (60)58 (40)2.24 (1.13 - 4.50)
GG18 (40)87 (60)0.44 (0.22 - 0.88)
H63D0.393
CG20 (44)54 (37)1.34 (0.67 - 2.66)
CC25 (56)91 (63)0.74 (0.37 - 1.47)
Table 4.

Genotype Frequencies of C282Y and H63D Genes in Diabetic Patients With Nephropathy and Controls a

PositionGenotypesNephropathy (n = 100)Controls (n = 45)P Value bOdds Ratio (95% Confidence Interval)
C282Y0.000
GA70 (70)58 (40)3.48 (2.03 - 6.04)
GG30 (30)87 (60)0.28 (0.16 - 0.49)
H63D0.006
CG55 (55)54 (37)2.05 (1.22 - 3.46)
CC45 (45)91 (63)0.48 (0.28 - 0.81)
Table 5.

Genotype Frequencies of C282Y and H63D Genes in Nephropathic and Retinopathic Patients a

PositionGenotypesNephropathy (n = 100)Retinopathy (n = 45)P ValueOdds Ratio (95% Confidence Interval)
C282Y0.245
GA70 (70)27 (60)1.55 (0.73 - 3.24)
GG30 (30)18 (40)0.64 (0.30 - 1.35)
H63D0.246
CG55 (55)20 (44)1.52 (0.74 - 3.12)
CC45 (45)25 (56)0.65 (0.31 - 1.33)

5. Discussion

This is the first study evaluated HFE gene polymorphisms among Iranian patients with type 2 diabetes. Our results suggested that both heterozygous and homozygous genotypes related to HFE gene polymorphisms are associated with T2D. We showed that only genotypes of C282Y are related to the retinopathy while there is an association between both studied SNPs and the nephropathy seen in T2D patients. The comparison between genotype frequency of T2D patients with retinopathy and those suffering from nephropathy indicated no significant outcome. These variations were subject of several studies.

Sampson et al. (26) found no excess of these HFE mutations in males with T2D. Also, study on 714 diabetic woman in the United States of America showed a significant relationship between higher body iron stores and HFE mutations while there were no significant difference in genotypes between case and controls (27). Similarly, Gomes et al. were unable to find a link between development of T2D and mentioned genotypes in Brazilian women (11). In addition, study on 167 diabetic African-American women declared no difference between frequency of C282Y and H63D between the case and control groups (28). Kankova et al. detected no difference between patients with type 2 diabetes mellitus and controls in Czech population. They found that ferritin levels were significantly higher in woman, but there is no relationship between these genotypes and ferritin levels (29). Study on Hellenic population showed similar results. The authors concluded that increased iron load in their patients linked to C282Y and H63D mutations while there is no difference in the frequency of genotypes between cases and controls (30). In another study, it is indicated that there is no association between distribution of these HFE mutations and T2D, but a relationship exists between H63D mutation and pathogenesis of late onset of disease in polish population (31).

These studies are in contrast with our data. Such discrepancies may have caused by the heterogeneity between different ethnic groups. In 2008, Sharifi et al. (32), which studied the association between mentioned genotypes with T2D in Iranian population could not find any significant correlation. In this case, while their study population was the same as ours, there are some statistical differences such as smaller population and different sex ratio, which caused such incongruity in results. Our data accord with Moczulski et al. (33) study on polish population, which demonstrated that hemochromatosis-causing mutations frequencies may play a role as T2D risk factors when 282Y and 63H mutations were greater in patients than healthy controls. A prospective cohort study, also, suggested an association between C282Y mutation and the incidence of the disease (34). In contrast, 63D allele, not C282Y related alleles, considered as a risk factor for Type 2 diabetes in a recent meta-analysis (25). By comparing genotypes of the two polymorphism with pathophysiological consequences of diabetes such as nephropathy and retinopathy (Tables 2 and 3), we found some significant data. Previously, it has been reported that 63D carriers have an elevated risk for nephropathy in type 2 diabetes (33), which is in consistent with our study and contrary to another study (35). Results from multivariate logistic regression in the Spanish population also approved the role of both C282Y alleles and H63D/H63D genotype in higher incidence of nephropathy (36). Retinopathy, which is another complication of type 2 diabetes, was also considered in some papers. We found that being a carrier of at least one C282Y allele (not H63D) boosts the risk of retinopathy, which is in agreement with previous study (36). It has been demonstrated that heterozygotes for C282Y might be under the risk for the development of proliferative diabetic retinopathy (PDR) (37). These findings showed that H63D plays no role in the incidence and development of retinopathy.

In conclusion, we observed that H63D and C282Y variants in HFE was in association with an increased risk of type 2 diabetes mellitus in Iranian patients. Future studies with a larger sample size and more rigorous study design are warranted.

References

  • 1.

    Andrulionyte L, Zacharova J, Chiasson JL, Laakso M, Stop-Niddm Study Group. Common polymorphisms of the PPAR-gamma2 (Pro12Ala) and PGC-1alpha (Gly482Ser) genes are associated with the conversion from impaired glucose tolerance to type 2 diabetes in the STOP-NIDDM trial. Diabetologia. 2004;47(12):2176-84. [PubMed ID: 15592662]. https://doi.org/10.1007/s00125-004-1577-2.

  • 2.

    Forbes JM, Cooper ME. Mechanisms of diabetic complications. Physiol Rev. 2013;93(1):137-88. [PubMed ID: 23303908]. https://doi.org/10.1152/physrev.00045.2011.

  • 3.

    Franks PW. Gene x environment interactions in type 2 diabetes. Curr Diab Rep. 2011;11(6):552-61. [PubMed ID: 21887612]. https://doi.org/10.1007/s11892-011-0224-9.

  • 4.

    Chauhan G, Spurgeon CJ, Tabassum R, Bhaskar S, Kulkarni SR, Mahajan A, et al. Impact of common variants of PPARG, KCNJ11, TCF7L2, SLC30A8, HHEX, CDKN2A, IGF2BP2, and CDKAL1 on the risk of type 2 diabetes in 5,164 Indians. Diabetes. 2010;59(8):2068-74. [PubMed ID: 20424228]. https://doi.org/10.2337/db09-1386.

  • 5.

    Jiang R, Manson JE, Meigs JB, Ma J, Rifai N, Hu FB. Body iron stores in relation to risk of type 2 diabetes in apparently healthy women. JAMA. 2004;291(6):711-7. [PubMed ID: 14871914]. https://doi.org/10.1001/jama.291.6.711.

  • 6.

    Rajpathak SN, Wylie-Rosett J, Gunter MJ, Negassa A, Kabat GC, Rohan TE, et al. Biomarkers of body iron stores and risk of developing type 2 diabetes. Diabetes Obes Metab. 2009;11(5):472-9. [PubMed ID: 19207293].

  • 7.

    Forouhi NG, Harding AH, Allison M, Sandhu MS, Welch A, Luben R, et al. Elevated serum ferritin levels predict new-onset type 2 diabetes: results from the EPIC-Norfolk prospective study. Diabetologia. 2007;50(5):949-56. [PubMed ID: 17333112]. https://doi.org/10.1007/s00125-007-0604-5.

  • 8.

    McCarthy MI. Progress in defining the molecular basis of type 2 diabetes mellitus through susceptibility-gene identification. Hum Mol Genet. 2004;13 Spec No 1:R33-41. [PubMed ID: 14722160]. https://doi.org/10.1093/hmg/ddh057.

  • 9.

    Njajou OT, Houwing-Duistermaat JJ, Osborne RH, Vaessen N, Vergeer J, Heeringa J, et al. A population-based study of the effect of the HFE C282Y and H63D mutations on iron metabolism. Eur J Hum Genet. 2003;11(3):225-31. [PubMed ID: 12673276]. https://doi.org/10.1038/sj.ejhg.5200955.

  • 10.

    Fernandez-Real JM, Lopez-Bermejo A, Ricart W. Cross-talk between iron metabolism and diabetes. Diabetes. 2002;51(8):2348-54. [PubMed ID: 12145144].

  • 11.

    Gomes KB, Carvalho MG, Coelho FF, Rodrigues IF, Soares AL, Guimaraes DA, et al. Lack of association of C282Y and H63D mutations in the hemochromatosis (HFE) gene with diabetes mellitus type 2 in a case-control study of women in Brazil. Genet Mol Res. 2009;8(4):1285-91. [PubMed ID: 19876870]. https://doi.org/10.4238/vol8-4gmr663.

  • 12.

    Rasmussen ML, Folsom AR, Catellier DJ, Tsai MY, Garg U, Eckfeldt JH. A prospective study of coronary heart disease and the hemochromatosis gene (HFE) C282Y mutation: the Atherosclerosis Risk in Communities (ARIC) study. Atherosclerosis. 2001;154(3):739-46. [PubMed ID: 11257277].

  • 13.

    Hannuksela J, Leppilampi M, Peuhkurinen K, Karkkainen S, Saastamoinen E, Helio T, et al. Hereditary hemochromatosis gene (HFE) mutations C282Y, H63D and S65C in patients with idiopathic dilated cardiomyopathy. Eur J Heart Fail. 2005;7(1):103-8. [PubMed ID: 15642540]. https://doi.org/10.1016/j.ejheart.2004.03.007.

  • 14.

    Surber R, Sigusch HH, Kuehnert H, Figulla HR. Haemochromatosis (HFE) gene C282Y mutation and the risk of coronary artery disease and myocardial infarction: a study in 1279 patients undergoing coronary angiography. J Med Genet. 2003;40(5). eee58. [PubMed ID: 12746412].

  • 15.

    Zorc M, Hruskovicova H, Petrovic MG, Milcic M, Peterlin B, Petrovic D. Haemochromatosis-causing mutations C282Y and H63D are not risk factors for coronary artery disease in Caucasians with type 2 diabetes. Folia Biol (Praha). 2004;50(2):69-70. [PubMed ID: 15222129].

  • 16.

    Chitturi S, Weltman M, Farrell GC, McDonald D, Kench J, Liddle C, et al. HFE mutations, hepatic iron, and fibrosis: ethnic-specific association of NASH with C282Y but not with fibrotic severity. Hepatology. 2002;36(1):142-9. [PubMed ID: 12085358]. https://doi.org/10.1053/jhep.2002.33892.

  • 17.

    Bugianesi E, Manzini P, D'Antico S, Vanni E, Longo F, Leone N, et al. Relative contribution of iron burden, HFE mutations, and insulin resistance to fibrosis in nonalcoholic fatty liver. Hepatology. 2004;39(1):179-87. [PubMed ID: 14752836]. https://doi.org/10.1002/hep.20023.

  • 18.

    Shaheen NJ, Silverman LM, Keku T, Lawrence LB, Rohlfs EM, Martin CF, et al. Association between hemochromatosis (HFE) gene mutation carrier status and the risk of colon cancer. J Natl Cancer Inst. 2003;95(2):154-9. [PubMed ID: 12529348].

  • 19.

    Hruskovicova H, Milanez T, Kobal J, Potisk KP, Petrovic D, Peterlin B. Hemochromatosis-causing mutations C282Y and H63D are not risk factors for atherothrombotic cerebral infarction. Med Sci Monit. 2005;11(7):BR248-52. [PubMed ID: 15990686].

  • 20.

    Zhang A, Park SK, Wright RO, Weisskopf MG, Mukherjee B, Nie H, et al. HFE H63D polymorphism as a modifier of the effect of cumulative lead exposure on pulse pressure: the Normative Aging Study. Environ Health Perspect. 2010;118(9):1261-6. [PubMed ID: 20478760]. https://doi.org/10.1289/ehp.1002251.

  • 21.

    Pankow JS, Boerwinkle E, Adams PC, Guallar E, Leiendecker-Foster C, Rogowski J, et al. HFE C282Y homozygotes have reduced low-density lipoprotein cholesterol: the Atherosclerosis Risk in Communities (ARIC) Study. Transl Res. 2008;152(1):3-10. [PubMed ID: 18593631]. https://doi.org/10.1016/j.trsl.2008.05.005.

  • 22.

    Valenti L, Dongiovanni P, Fracanzani AL, Santorelli G, Fatta E, Bertelli C, et al. Increased susceptibility to nonalcoholic fatty liver disease in heterozygotes for the mutation responsible for hereditary hemochromatosis. Dig Liver Dis. 2003;35(3):172-8. [PubMed ID: 12779071].

  • 23.

    Cauza E, Hanusch-Enserer U, Bischof M, Spak M, Kostner K, Tammaa A, et al. Increased C282Y heterozygosity in gestational diabetes. Fetal Diagn Ther. 2005;20(5):349-54. [PubMed ID: 16113552]. https://doi.org/10.1159/000086811.

  • 24.

    Ellervik C, Mandrup-Poulsen T, Nordestgaard BG, Larsen LE, Appleyard M, Frandsen M, et al. Prevalence of hereditary haemochromatosis in late-onset type 1 diabetes mellitus: a retrospective study. Lancet. 2001;358(9291):1405-9. [PubMed ID: 11705485]. https://doi.org/10.1016/S0140-6736(01)06526-6.

  • 25.

    Rong Y, Bao W, Rong S, Fang M, Wang D, Yao P, et al. Hemochromatosis gene (HFE) polymorphisms and risk of type 2 diabetes mellitus: a meta-analysis. Am J Epidemiol. 2012;176(6):461-72. [PubMed ID: 22908207]. https://doi.org/10.1093/aje/kws126.

  • 26.

    Sampson MJ, Williams T, Heyburn PJ, Greenwood RH, Temple RC, Wimperis JZ, et al. Prevalence of HFE (hemochromatosis gene) mutations in unselected male patients with type 2 diabetes. J Lab Clin Med. 2000;135(2):170-3. [PubMed ID: 10695662]. https://doi.org/10.1067/mlc.2000.104464.

  • 27.

    Qi L, Meigs J, Manson JE, Ma J, Hunter D, Rifai N, et al. HFE genetic variability, body iron stores, and the risk of type 2 diabetes in U.S. women. Diabetes. 2005;54(12):3567-72. [PubMed ID: 16306377].

  • 28.

    Acton RT, Barton JC, Bell DS, Go RC, Roseman JM. HFE mutations in African-American women with non-insulin-dependent diabetes mellitus. Ethn Dis. 2001;11(4):578-84. [PubMed ID: 11763282].

  • 29.

    Kankova K, Jansen EH, Marova I, Stejskalova A, Pacal L, Muzik J, et al. Relations among serum ferritin, C282Y and H63D mutations in the HFE gene and type 2 diabetes mellitus in the Czech population. Exp Clin Endocrinol Diabetes. 2002;110(5):223-9. [PubMed ID: 12148086]. https://doi.org/10.1055/s-2002-33071.

  • 30.

    Habeos IG, Psyrogiannis A, Kyriazopoulou V, Psilopanagiotou A, Papavassiliou AG, Vagenakis AG. The role of Hemochromatosis C282Y and H63D mutations in the development of type 2 diabetes mellitus in Greece. Hormones (Athens). 2003;2(1):55-60. [PubMed ID: 17003003].

  • 31.

    Malecki MT, Klupa T, Walus M, Czogala W, Greenlaw P, Sieradzki J. A search for association between hereditary hemochromatosis HFE gene mutations and type 2 diabetes mellitus in a Polish population. Med Sci Monit. 2003;9(2):BR91-5. [PubMed ID: 12601293].

  • 32.

    Sharifi F, Esmaeilzadeh A, Zali M. Hemochromatosis gene (HFE) mutations in patients with type 2 diabetes and their control group in an Iranian population. Saudi Med J. 2008;29(6):808-12. [PubMed ID: 18521456].

  • 33.

    Moczulski DK, Grzeszczak W, Gawlik B. Role of hemochromatosis C282Y and H63D mutations in HFE gene in development of type 2 diabetes and diabetic nephropathy. Diabetes Care. 2001;24(7):1187-91. [PubMed ID: 11423500].

  • 34.

    Salonen JT, Tuomainen TP, Kontula K. Role of C282Y mutation in haemochromatosis gene in development of type 2 diabetes in healthy men: prospective cohort study. BMJ. 2000;320(7251):1706-7. [PubMed ID: 10864547].

  • 35.

    Davis TM, Beilby J, Davis WA, Olynyk JK, Jeffrey GP, Rossi E, et al. Prevalence, characteristics, and prognostic significance of HFE gene mutations in type 2 diabetes: the Fremantle Diabetes Study. Diabetes Care. 2008;31(9):1795-801. [PubMed ID: 18566337]. https://doi.org/10.2337/dc08-0248.

  • 36.

    Oliva R, Novials A, Sanchez M, Villa M, Ingelmo M, Recasens M, et al. The HFE gene is associated to an earlier age of onset and to the presence of diabetic nephropathy in diabetes mellitus type 2. Endocrine. 2004;24(2):111-4. [PubMed ID: 15347835].

  • 37.

    Peterlin B, Globocnik Petrovic M, Makuc J, Hawlina M, Petrovic D. A hemochromatosis-causing mutation C282Y is a risk factor for proliferative diabetic retinopathy in Caucasians with type 2 diabetes. J Hum Genet. 2003;48(12):646-9. [PubMed ID: 14618419]. https://doi.org/10.1007/s10038-003-0094-3.