DEPARTAMENTO DE BIOLOGIA ANIMAL
ANALYSIS OF GENETIC
DISORDERS IN A PORT
FAMILIAL
Alexandra Paula dos Reis
Mestrado em Biologia Humana e Ambiente
UNIVERSIDADE DE LISBOA FACULDADE DE CIÊNCIAS
DEPARTAMENTO DE BIOLOGIA ANIMAL
ANALYSIS OF GENETIC MARKERS FOR CARDIOVA
DISORDERS IN A PORTUGUESE POPULATION WI
FAMILIAL HYPERCHOLESTEROLEMIA
Alexandra Paula dos Reis Gomes
Mestrado em Biologia Humana e Ambiente 2009
MARKERS FOR CARDIOVASCULAR
UGUESE POPULATION WITH
DEPARTAMENTO DE BIOLOGIA ANIMAL
ANALYSIS OF GENETIC
DISORDERS IN A PORTU
FAMILIAL
Professora ProfessoraAlexandra Paula dos Reis Gomes
Mestrado em Biologia Humana e Ambiente
UNIVERSIDADE DE LISBOA FACULDADE DE CIÊNCIAS
DEPARTAMENTO DE BIOLOGIA ANIMAL
ANALYSIS OF GENETIC MARKERS FOR CARDIOVA
DISORDERS IN A PORTUGUESE POPULATION WIT
FAMILIAL HYPERCHOLESTEROLEMIA
Dissertação orientada por:
Professora Doutora Ana Maria Viegas Crespo Professora Doutora Luciana Maria Gonçalves da Costa
Alexandra Paula dos Reis Gomes
Mestrado em Biologia Humana e Ambiente 2009
MARKERS FOR CARDIOVASCULAR
GUESE POPULATION WITH
AGRADECIMENTOS
Antes de mais, queria agradecer, talvez o mais importante, à Comissão de Fomento da Investigação em Cuidados de Saúde por ter apoiado financeiramente este projecto, sem o qual teria sido impossível a sua realização.
Em primeiro lugar gostaria de agradecer à Doutora Luciana Costa por me ter dado a portunidade de trabalhar neste projecto e orientado neste percurso mais longo que o previsto. Por todo o apoio, incentivo e compreensão quando as coisas não correram conforme o planeado. Pelo conhecimento transmitido, ideias sugeridas e também pelos “puxões de orelhas” quando mereci. Pelo carinho e amizade. Por tudo o que foi essencial para poder estar aqui agora a escrever este agradecimento.
Um agradecimento especial à Doutora Mafalda Bourbon, que apesar da impossibilidade de ser oficialmente minha orientadora, vestiu esse papel e como tal deu-me todo o apoio, orientação, força de vontade e conhecimento para que eu pudesse fazer este trabalho. Foi um grande pilar e sem ela não teria chegado até aqui com toda a certeza.
Agradeço à Doutora Ana Crespo primeiro por me ter dado a conhecer este projecto e por todo o apoio, preocupação e disponibilidade que demonstrou durante este percurso.
Às minhas colegas (Ana, Catarina, Liliana, Tânia, Vânia) e colega (João) de laboratório um enorme obrigado por me terem iniciado e integrado nesta vida de bata e luvas. Por me terem ensinado tudo o que precisei para realizar este trabalho, por todo o conhecimento transmitido, incentivo dado. Muito obrigada!
Um agradecimento especial à Ângela, minha companheira de tese, por ter estado presente e me ter ajudado neste percurso conturbado. Foi um grande apoio na integração desta nova etapa. Outro agradecimento especial, à Arminda por me ter ajudado neste trabalho e por ter sido também um grande pilar nos bons e maus momentos.
Às minhas amigas Bruna, Luciana, Vânia e Verónica que me acompanharam neste percurso universitário muitas vezes conturbado mas que foi o melhor período da minha vida até hoje. Sem vocês não teria conseguido chegar até aqui! Obrigada por fazerem parte da minha vida e pelo espírito coki! Obrigada por todas as noites mal dormidas, telefonemas, mensagens,
incentivos, ajuda. Simplesmente por estarem sempre presentes nos bons e maus momentos, incondicionalmente, e assim espero que continue.
Ao Yuri, obrigada por teres dado um outro significado à minha vida e por alargares a minha visão do mundo. Por todas as experiências incríveis vividas durante este período, por vezes difícil, foi sempre um grade incentivo e inspiração para continuar quando parecia ser impossível. Obrigada por seres tu. Valeu!
Aos meus pais por tudo e mais alguma coisa. Nunca vou conseguir agradecer o suficiente porque sem eles não teria conseguido concretizar os meus sonhos e desejos. Obrigada pela compreensão, confiança, amor, carinho, preocupação. Peço desculpa pela ausência, muita ausência que agora chega ao fim.
Ao meu irmão Pedro, que sem o perceber me dá força de vontade de continuar e vencer. Obrigada pelo amor, amizade, companheirismo, brincadeiras, risos. Sem ti tudo seria mais difícil.
INDEX ABSTRACT………...…i SUMÁRIO……….…ii ABBREVIATIONS………v INTRODUCTION...1 OBJECTIVES...8
MATHERIALS AND METHODS...9
1. Recruitment of the study population...9
2. Clinical characterization of the study population...9
3. Laboratorial characterization of the study population...10
3.1 Biochemical characterization...10
3.2 Hematological characterization...11
3.3 Evaluation of inflammatory status...11
3.4 Markers of Pro-oxidant /Antioxidant balance... 11
4. Genetic characterization of the study population... 12
4.1 Sample preparation... 12
4.2 Primer designing...13
4.3 DNA amplification...13
4.4 DNA Gel Electrophoresis...14
4.5 Purification of the PCR products...14
4.6 Sequencing of purified PCR products...14
4.7 Restriction Fragment Length Polymorphism...15
5. Stastistical Analysis...15
RESULTS...16
1. Biochemical data characterization...16
2. Immunological data characterization...21
3. Genetic data characterization...22
3.1 Biochemical data versus genotypes analysis………..…24
3.2 Immunological data versus genotypes analysis……….25
DISCUSSION………...26
REFERENCES……….34
APPENDIX A………..44
i ABSTRACT
Familial Hypercholesterolemia (FH) is a genetic disorder that leads to an increase in the levels of total and low density lipoprotein cholesterol promoting atherosclerosis (ATH) and premature cardiovascular disease (CVD). Inflammation has been considered to be involved in the pathogenesis of CVD namely the activity of pro-inflammatory cytokines and acute phase proteins. Also, there are other risk factors contributing to the development and progression of ATH and CVD as genetic and oxidative stress markers.
We intended to investigate the role of genetic, inflammatory and oxidative biomarkers in the clinical outcome of FH patients and study its putative correlation with CVD. There were selected 41 FH patients with CVD, 91 without CVD and 49 healthy individuals. All individuals were characterized through the determination of the lipid profile (high density lipoprotein, LDL and total cholesterol, triglycerides, apoA, apoB, lipoprotein(a)), measurement of the serum concentration of some inflammatory markers (Cp, haptoglobin and C reactive protein), pro-inflammatory cytokines (interleukin-6 and tumor necrosis factor-alpha), homocysteine and markers of antioxidant / pro-oxidant status (nitric oxid and oxLDL). The genetic characterization was achieved by studying polymorphisms in the genes encoding for LPL, APOAV, APOCIII, TNF-α, IL-6, MTHFR and NOS, which are thought to be involved in the inflammatory process and the predisposition to CVD.
The results showed that the group of FH patients with CVD presented increased total (p<0,001) and LDL cholesterol (p=0,001) and apoB (p<0,001) levels and decreased apoA1 (p=0,021) levels in relation to the FH group without CVD. In the FH group with CVD it was observed the highest oxLDL and the lowest NO concentrations. APOAV-1131C and APOCIII 3238G allele were associated with higher TG levels (p=0,013; p=0,042) in the FH group without CVD. MTHFR 677T allele was associated with high total cholesterol levels (p=0,006) in the FH group with CVD.
Markers of lipid metabolism are distinguishable between the groups analyzed however inflammatory and genetic markers need further studies to improve our knowledge of their role in CVD outcome.
Keywords: Familial Hypercholesterolemia, Atherosclerosis, Cardiovascular Disease,
ii SUMÁRIO
A Hipercolesterolemia Familiar (FH) é uma doença autossómica dominante causada por mutações nos genes codificantes para o receptor das lipoproteínas de baixa densidade (LDLR), para a apolipoproteína B (ApoB) ou para a pró-proteína convertase subtilisina/quexina 9 (PCSK9). Estas mutações traduzem-se fenotipicamente por níveis elevados de colesterol-LDL plasmático promovendo o processo aterosclerótico, que é um importante factor de risco para o desenvolvimento precoce da doença cardiovascular (DCV). Em Portugal, o Estudo Português de FH mostrou que apenas alguns dos indivíduos com FH sofrem de DCV sugerindo que, muito provavelmente, existem outros factores que condicionam o aparecimento desta doença. Vários estudos têm vindo a demonstrar a importância do processo inflamatório, no qual a mediação de citocinas pró-inflamatórias e proteínas de fase aguda parece conferir uma predisposição para a DCV. Para além destes existem outros factores de risco emergentes que contribuem para um estado de stress oxidativo que parece estar associado ao aparecimento de DCV, tais como a presença de níveis elevados de LDL oxidadas (oxLDL), de homocisteína (Hcy) assim como baixos níveis de óxido nítrico (NO).
O principal objectivo deste estudo foi estudar o papel de marcadores genéticos, inflamatórios e de stress oxidativo no aparecimento e desenvolvimento da DCV, em doentes com FH. Em particular, pretende-se investigar uma possível correlação entre alguns marcadores de inflamação e outros factores de risco associados à DCV. Para atingir este objectivo, foram seleccionados 132 indivíduos de ambos os sexos, geneticamente diagnosticados com FH, dos quais 41 com DCV e 91 sem DCV. Adicionalmete, foram recrutados indivíduos saudáveis como controlos. De cada indivíduo recolheram-se amostras de sangue periférico total, a partir das quais se realizaram todas as análises. A caracterização bioquímica das amostras obtidas foi feita com base na determinação do perfil lipídico (colesterol total, colesterol-LDL, colesterol-HDL, triglicéridos, apoAI, apoB e lipoproteína (a)). A caracterização hematológica incluiu a determinação da concentração da Hcy, mediante a utilização de testes laboratoriais standard. Adicionalmente, determinaram-se os níveis circulantes de marcadores inflamatórios (Cp, haptoglobina e proteína C reactiva de elevada sensibilidade (hsCRP)), por nefelometria. As concentrações de citocinas pró-inflamatórias nomeadamente de interleucina-6 (IL-6) e factor de necrose tumoral alfa (TNF-α) foram medidos no soro utilizando ensaios imunoenzimáticos. Com o objectivo de avaliar o balanço antioxidante / pró-oxidante mediram-se os níveis de NO e de oxLDL de forma a obter uma melhor caracterização da população em estudo. Finalmente, procedeu-se à pesquisa de polimorfismos genéticos dos genes que codificam para a LPL, ApoAV, ApoCIII, TNF-α, IL-6, MTHFR e NOS, os quais se pensa estarem directamente envolvidos com o processo inflamatório e a predisposição para a DCV.
iii
A comparação dos resultados obtidos com a quantificação dos pârametros bioquímicos entre o grupo controlo e o grupo da população total com FH revelou diferenças significativas nas concentrações de todos os parâmetros. O grupo FH apresentou valores mais elevados de colesterol total (p<0,001), colesterol LDL (p<0,001), triglicéridos (TG) (p<0,001), ApoB (p<0,001) e Lp(a) (p=0,022) e valores mais baixos de colesterol HDL (p=0,017) e ApoA1 (p=0,043) em relação ao grupo controlo. Quando se comparou os três grupos de estudo observaram-se diferenças significativas entre todos os grupos para o colesterol total (p<0,001), colesterol LDL (p<0,001) e apoB (p<0,001) sendo as suas concentrações progressivamente mais elevadas do grupo controlo para o grupo FH sem DCV e até ao grupo FH com DCV. O grupo FH com DCV apresentou concentrações significativamente mais baixas de apoA1 em relação ao grupo FH sem DCV (p=0,021). Quando se teve em conta a história familiar de DCV não houve diferenças significativas entre o grupo FH sem DCV e o grupo FH sem DCV mas com história familiar de DCV para os parâmetros medidos. A comparação dos resultados obtidos com a quantificação dos parâmetros imunológicos entre o grupo controlo e o grupo da população total com FH demonstrou que existem diferenças significativas nas concentrações de oxLDL (p=0,001) e de hsCRP (p=0,014). Na população com FH foram observadas concentrações mais elevadas de oxLDL e hsCRP em relação ao grupo controlo. Quando se comparou os três grupos só se observaram diferenças significativas nas concentrações de oxLDL entre o grupo controlo e o grupo de doentes FH sem DCV (p=0,017) embora o grupo de doentes FH com DCV apresente valores ainda mais elevados. Na análise da caracterização genética comparou-se os diferentes polimorfismos com os parâmetros bioquímicos medidos nos grupos de estudo. Na população com FH encontrou-se uma associação entre o alelo APOAV-1131C e concentrações elevadas de TG (p=0,029), entre o alelo TNF-α-308A e concentrações mais baixas de colesterol total (p=0,019) e colesterol LDL (p=0,038) e entre o alelo MTHFR 677T e concentrações elevadas de colesterol HDL (p=0,030) em relação ao genótipo homozigótico normal. Quando se fez a mesma análise no grupo de doentes FH sem DCV observou-se a mesma associação para o alelo APOAV-1131C e os TG (p=0,013) e para o alelo TNF-α-308A e o colesterol total (p=0,007). Uma nova associação foi encontrada neste grupo entre o alelo APOCIII 3238G e concentrações elevadas de TG (p=0,042) quando comparando com o genótipo homozigótico normal. No grupo de doentes FH com DCV apenas se encontrou uma associação que foi entre o alelo MTHFR 677T e concentrações elevadas de colesterol total (p=0,006). Em última análise comparou-se os polimorfismos com os parâmetros imunológicos medidos nos grupos de estudo. No grupo controlo encontrou-se associação entre o alelo TNF-α-308A e concentrações elevadas de Hcy (p=0,025) em comparação com o genótipo homozigótico normal. No grupo da população com FH encontrou-se associação entre o alelo MTHFR 677T e concentrações elevadas de TNF-α (p=0,049) em relação ao genótipo homozigótico normal.
Os resultados obtidos quanto ao perfil lipídico dos indivíduos parecem ser mais consistentes e fidedignos podendo ser utilizados como biomarcadores que distingam entre doentes
iv
com e sem DCV. Em relação à caracterização imunológica e genética da população de estudo, os resultados são ainda controversos e as associações obtidas pouco claras. Assim, o aumento do tamanho da amostra é um primeiro passo para poder confirmar estes resultados ou obter novos dados. Contudo, a DCV é uma doença complexa e existem outros factores que contribuem para o seu desenvolvimento e que futuramente serão contemplados quando se realizar um estudo em larga-escala.
Palavras-chave: Hipercolesterolemia Familiar, Aterosclerose, Doença Cardiovascular,
v ABBREVIATIONS
ANOVA Analysis of variance
Apo Apolipoprotein
APP Acute phase protein
APR Acute phase response
ATH Atherosclerosis
CABG Coronary Artery Bypass Graft
CHD Coronary heart disease
Cp Ceruloplasmin
CRP C-reactive protein
CVD Cardiovascular disease
DNA Deoxyribonucleic acid
dNTP deoxyribonucleotide
ddNTP dideoxyribonucleotide
EDTA Ethylenediamine tetraacetic acid
ELISA Enzyme linked immuno
sorbent assay
EPHF Portuguese Familial
Hypercholesterolemia study
Fe Iron
FH Familial
Hypercholesterolemia
Hcy Homocysteine
HDL High density lipoprotein
HHcy Hyperhomocysteinemia
Hp Haptoglobin
hsCRP high sensitive C-reactive protein
HSPG Heparin sulphate-proteoglycans
IL Interleukin
INSA Instituto Nacional de Saúde Doutor Ricardo Jorge
LCAT Lecithin-cholesterol acyltransferase
LDL Low density lipoprotein
LDLR Low density lipoprotein receptor
LPL Lipoprotein lipase
Mg2+ Magnesium
MI Myocardial infarction
mL mililiter
mRNA messenger ribonucleic acid
MTHFR Methylenetetrahydrofolate reductase
NADPH Nicotinamide adenine dinucleotide phosphate
ng nanogram
NH4 Ammonium
NO Nitric oxide
NOS Nitric oxide synthetase
ONOO- Peroxynitrite anion
oxLDL oxidized low density lipoprotein
PCR Polymerase chain reaction
PCSK9 Proprotein convertase subtilisin/kexin type 9
pmol picomol
PPAR Proliferator activated receptor
ROS Reactive oxygen species
TG Triglycerides
TMB Tetramethylbenzidine
TNF-α Tumor necrosis factor alpha
U/L Units per liter
µL microliter
µmol micromol
VLDL Very low density lipoprotein
VLDLR Very low density lipoprotein receptor
VSMC Vascular smooth muscle cells
1 INTRODUCTION
Familial Hypercholesterolemia
Cholesterol homeostasis is critical to human health as evidenced by the number of diseases that result from defects in cholesterol metabolism [1]. An example of one of these pathologies is familial hypercholesterolemia (FH), an autosomal dominant genetic disorder which is considered to be among the most common genetic disorders, since the heterozygous form has a prevalence of 1:500 in the majority European populations. Homozygoty is rare, estimated at one per million [2]. Based on the prevalence found in other European populations, it is estimated that there are around 20,000 cases of FH in Portugal. Despite these numbers, the disease is severely under-diagnosed in our country and before the Portuguese Familial Hypercholesterolemia Study (EPHF) has begun, under the supervision of our research unit, no clinical or genetic studies have been performed [3].
The FH phenotype usually results from mutations in the low density lipoprotein receptor (LDLR) gene, in the apolipoprotein B (APOB) gene or in the gene coding for proprotein convertase subtilisin/Kexin type 9 (PCSK9), recently discovered [4]. Thus, according to these data, at least three genes are involved in FH, wherein several mutations were described as being responsible for this disorder [5].
Clinically, FH manifests itself by high levels of total and LDL cholesterol, with normal levels of high density lipoprotein (HDL) cholesterol and triglycerides (TG), and a family history of hypercholesterolemia and premature cardiovascular disease (CVD) [5]. A lipid accumulation in tendons (xanthomas) and in the walls of major arteries (atheromas) is observed, leading to premature atherosclerosis (ATH) and increased risk of coronary heart disease (CHD) [4].
Familial Hypercholesterolemia and Cardiovascular Disease
CHD is a multifactorial disorder depending on both genetic and environmental factors [6]. Aging and male gender are correlated with increased CHD risk, since men usually developing the disease 10-15 years earlier than women, who are in general protected for CHD to a certain degree until menopause [7]. In fact, male FH patients usually develop CHD before the age of 55 years and female FH patients before the age of 65 years [3]. The most relevant cardiovascular risk factor is smoking, and environmental factor which increases the lifetime risk in twofold [8]. Lack of exercise and the associated adiposity, in addition to a high intake of saturated fats and a low intake of certain vitamins, are commonly associated with an increased CHD risk in the general population and are even more important in the context of FH pathophysiology [7]. The mechanism of action of these factors is, at least in part, thought to be associated and can influence differences in the plasma levels of lipids and atherogenic
2
lipoproteins. In fact, high levels of LDL cholesterol and low levels of HDL cholesterol have consistently been shown to be connected with CHD risk [9]. However, results obtained from the EPHF showed that only some individuals with clinically diagnosis of FH (20%), actually developed CHD [4]. Additionaly, a screening of mutations in clinical FH patients, was not able to identify a single mutation in any of the three genes analyzed in about 50% of the individuals, even though some of them presented a severe phenotype with very high total cholesterol values and premature CHD [4]. Therefore, it’s reasonable to hypothesize that there are other factors underlying the pathophysiology of this disorder.
Lipid metabolism, Atherosclerosis and Cardiovascular Disease
Increased lipid levels are known to be strongly associated with ATH and CHD [10]. Lipoprotein lipase (LPL) is an enzyme that is responsible for the hydrolysis of TG present in circulating lipoproteins such as chylomicrons and very low density lipoproteins (VLDL), releasing non-esterified fatty acids and 2-monoacylglycerol that are subsequently utilized by muscle and adipose tissue [11]. This enzyme is associated with the luminal side of capillaries, arteries and also macrophages where it exerts its function [12]. Besides its key role in lipid metabolism, LPL has been known to have additional functions. LPL can act like a non-catalytic bridging enabling it to bind at the same time to lipoproteins and various specific cell surface proteins including heparin sulphate-proteoglycans (HSPG), LDLR related protein, LDLR, VLDL receptor which consequently may cause an increase in lipoprotein accumulation and cellular uptake [11]. Also, LPL is involved in the interaction between monocyte surface and arterial endothelial cells, functioning this way as a monocyte adhesion protein [13]. Additionally, it has been demonstrated that LPL can promote vascular smooth muscle cells (VSMC) proliferation [14] and can directly promote the expression of the tumor necrosis factor alpha (TNF-α) gene, stimulate nitric oxide synthetase (NOS) expression by macrophages and activate endothelial nicotinamide adenine dinucleotide phosphate (NADPH) oxidase [15, 16, 17].
Alterations in LPL function have been implicated in several pathophysiological conditions such as hypertriglyceridaemia, including chylomicronemia, insulin resistance and diabetes, obesity, Alzheimer’s disease and ATH [18]. Herein, we will be focused on the pathogenesis of LPL in ATH and in CHD outcome.
Several polymorphisms within the coding region of LPL gene have been related to LPL activity which might contribute to the “pathophysiology of this enzyme” resulting from its altered function, namely D9N, N291S and S447X [19]. D9N (G A at position 280) induces an aspartate to asparagine change in the aminoacid sequence of the enzyme and this alteration promotes a deficiency in enzyme secretion [20]. N291S (A G at position 1127) alters an asparagine residue to serine and is capable to destabilize the homodimer complex formation of
3
the enzyme compromising its lipolytic activity. N291S variation was also associated with low levels of HDL cholesterol and a significant decline in LPL activity in males with ATH, contributing thus to the development of ATH [21]. Both polymorphisms are associated with a reduction in LPL enzymatic activity. The S447X (C G at position 1595) polymorphism generates a premature stop codon and consequently the loss of the terminal serine and glycine residues of the enzyme having contradictory effects on LPL and lipid levels to those observed in D9N and N291S polymorphisms [19, 22]. This last alteration has been correlated with low TG and high HDL cholesterol levels and protection against premature CHD, but only in men from the general population [23]. Hence, D9N and N291S are thought to have small deleterious effects on plasma HDL cholesterol and TG, and S447X is considered to have small beneficial effects [19].
Apolipoprotein AV (APOAV) is a newly identified lipoprotein that is thought to have an important role in TG metabolism [24]. There have been some association studies that link polymorphisms in APOAV and elevated TG levels particularly one in the promoter region, -1131T/C, with individuals carrying the rare C allele being at higher risk of developing severe CHD. There have been found an association between apoAV-1131C allele and increased TG levels but no relation with HDL cholesterol levels, which in general is normally inversely correlated to TG levels [25, 26]. APOAV gene is located near the apolipoprotein AI/CIII/AIV gene cluster where variations have been observed and associated with differences in lipid levels [27]. A great variety of alterations in this gene assembly have been described and linked to different genetic forms of dyslipidemia. The most broadly studied polymorphism is one in the untranslated region of apolipoprotein CIII (APOCIII) gene representing a change of a C to a G at position 3238 of the nucleotide. APOCIII is part of the protein portion of chylomicrons and VLDL and is thought to be involved in the regulation of TG rich lipoprotein (TRL) catabolism [28]. The rare G allele has been associated with elevated plasma TG levels [29].
Inflammation, Atherosclerosis and Cardiovascular Disease
Inflammation has long been considered to be involved with the occurrence of many disorders. The inflammatory process has been described as being caused by cytokines and chemokines secreted by various types of cells, including macrophages that ultimately activate systems such as the complement system and acute phase response (APR). When there is an impairment that avoids tissues to re-establish their own homeostasis, the responses became chronic and turn out in inflammatory disease. In fact, although immune responses to injury and infection are necessary to maintain homeostasis they can also cause problems, thus leading to chronic diseases, such as diabetes, ATH and CHD [30].
ATH is considered to be a multifactorial disease driven by inflammatory reactions [31]. The common cardiovascular risk factors (such as high lipid levels, hypertension, cigarette
4
smoking, stress, oxidative stress, diabetes mellitus, infectious agents, hyperhomocysteinemia (HHcy) and C-reactive protein (CRP)) result in injuries or insults to the arterial wall leading to endothelial dysfunction, to which the body adapts by continuously triggering a response throughout the years of exposure [30, 32]. During the inflammatory process there is an accumulation of monocytes ant T cells at the injury site where they are activated thus releasing proinflammatory cytokines such as interleukin-6 (IL-6), interleukin-1β (IL-1β) and TNF-α. These chemical mediators stimulate chemokine secretion resulting in selective recruitment of leukocytes to the injured tissue [30].
IL-6 is a pro-inflammatory cytokine that is synthesised by an enormous variety of cells like lymphocytes, monocytes, fibroblasts and endothelial cells. These molecule accomplishes many functions in inflammation, bone metabolism, immunity, endocrine functions and plays a key role in the regulation of the synthesis of acute phase proteins (APP) by the liver [33]. IL-6 has been associated with risk of CVD in a way that elevated concentrations of this cytokine alone predicted total and cardiovascular mortality over a five years follow-up, when taking in account the traditional CVD risk factors [34]. IL-6 mRNA was found at a 10- to 40-fold higher level in atherosclerotic arteries than in nonatherosclerotic vessels suggesting a pathogenic role of IL-6 in ATH [35]. The atherosclerotic process may be induced by IL-6, directly promoting the synthesis of coagulation factors or indirectly by inducing endothelial disfunction, monocyte/macrophage recruitment, differentiation, inflammation and matrix degradation [36]. Also, CHD risk was linked to high levels of IL-6 and this association was as powerful as those relating the already known main risk factors; however if it is a causal relation or not remains unclear because inflammatory markers synthesis may be induced during the atherosclerotic process [35, 37]. In 1998, Fishman and coworkers described a G/C polymorphism at position -174 in the promoter region of IL-6 gene which affected IL-6 transcriptional response to stimuli in vitro [38]. The IL-6 -174 C allele has been associated with elevated concentrations of IL-6 [39] and CRP [40] and with higher risk of CVD incidence and mortality [39, 41]. However, there is still controversy about the physiologic role of this polymorphism. In fact, some authors observed the association of the IL-6 -174G allele with high levels of IL-6 [38] while others found no association at all between IL-6 -174G/C polymorphism and IL-6 levels [36].
TNF-α is another pro-inflammatory cytokine produced by monocytes/macrophages and other cell types, that is involved in the pathogenesis of many diseases including infection, autoimmune disorders, cancer, neurodegenerative disease and even drug dependencies, among many others. Additionally, TNF-α has been related to susceptibility and development of CVD by its implication in vascular wound repair, insulin signalling and fat metabolism [42]. This cytokine has the ability to modify the expression of genes involved in lipid metabolism including LPL, proliferator activated receptors (PPARs), apoA-I, apoA-IV, apoE and lecithin-cholesterol acyltransferase (LCAT) [43]. TNF-α plays a key role in inflammation and is
5
associated with the pathogenesis of ATH. In fact, it has been shown that increased levels of TNF-α correlate with progression of early carotid ATH [44]. In 1992, Wilson and co-workers described a G to A polymorphism in the TNF-α gene at position -308 in the promoter region [45]. The rare TNF-α -308A allele is associated with increased transcription and production of TNF-α [46], and elevated concentrations of this cytokine are linked to increased myocardial infarction (MI) risk [47].
Interestingly, it was previously reported that there is an interaction between TNF-α and IL-6 where stimulation by TNF-α induces IL-6 production and IL-6 provides negative feedback to inhibit the production of TNF-α, on the inflammatory response in vivo [48, 49].
Along with other functions, cytokines are also able to initiate the APR, characterized by a modification in protein synthesis and consequently, an increase in the so called acute-phase proteins [50] like CRP, fibrinogen, ceruloplasmin (Cp) and haptoglobin (Hp), among others.
Oxidative Stress and Inflammation in Atherosclerosis and Cardiovascular Disease
CRP is an APP synthesized by the liver after cytokine stimulation [51] and is a marker of low grade inflammation [52]. It is well established that elevated CRP levels are associated with increased risk of CVD but it still remains uncertain if CRP is just a marker for CVD or if it has an active role in causing this disorder [53]. Nevertheless, CRP is thought to be the most reliable measured blood biomarker of vascular inflammation [54] and recently it has been shown that when compared with other inflammatory and lipid markers in predicting CVD, CRP overcame all of them, including LDL cholesterol [55]. During the inflammatory response the levels of PCR raise during the following 24-48 hours but long term variations are very stable over long periods of time which supports its strong predictive value [32].This protein has also been implicated in the pathogenesis of ATH in humans [52]. In fact, CRP has the ability to induce the expression of adhesion molecules, the synthesis of IL-6 and also to reduce the expression and bioavailability of endothelial NOS, in human endothelial cells. Moreover, CRP activates cytokine and tissue factor expression by macrophages and thus boost up LDL uptake by these cells [56]. In addition, CRP has been found in inflammatory fluids and deposited in atherosclerotic lesions so the circulating levels measured of this APP might not reflect correctly its tissue concentration, where its higher expression can induce the development and progression of ATH [57].
On the other hand, Cp is a multicopper oxidase protein, abundant in the plasma which accounts for 95% of total circulating copper in healthy adults. There are two isoforms of Cp, one secreted mainly by the liver and other anchored to the membrane of various cell types [58] by a glycosylphosphatidylinositol bound [59]. More recently, it was demonstrated by our research unit that Cp was also expressed in peripheral blood lymphocytes [60], reinforcing the
6
intimate connection between Cp and immune system. Previous studies showed an increase in Cp levels, during pregnancy and after several inflammatory conditions including CVD [61]. In fact, in 1956 Adelstain and co-workers were the first to correlate CVD and Cp by demonstrating that an increase in serum Cp was seen after MI [62]. Cp is a multifunctional protein which physiologic role still remains to be elucidated. In fact, Cp is able to oxidise different substrates such as biogenic amines, phenols and iron (Fe). However, Cp’s antioxidant activity has been reported for a long time and it seems to have a crucial importance in inflammatory process and APR [61]. This activity is evidenced not only by the capacity of Cp to act as a scavenger of superoxide anion radicals and other reactive oxygen species (ROS) [63] and by the inhibition of the Fenton reactions, but also by its ability to convert Fe2+ to Fe3+ (a non-toxic form) [61]. Although presenting a protective function, it is also known that Cp has the ability of oxidising lipids thus revealing a pro-oxidant activity. Actually, Ehrenwald and co-workers showed that Cp enhanced, rather than suppressed, the oxidation of LDL and that this activity was dependent on the integrity of its structure and its bound copper [64]. Oxidized LDL (oxLDL) can promote an inflammatory response in the artery wall contributing this way to the initiation and progression of the atherosclerotic process [65].
There are several studies that considered Cp itself a risk factor for CHD. In fact, Kok and co-workers observed that individuals that had the highest quantity of serum copper were those who had a four times higher risk of death from CHD [66]. It was also demonstrated that elevated serum LDL was associated with accelerated atherogenesis [67] and reduced serum HDL with higher risk of CHD, only in subjects with higher serum Cp [68].Consequently, it is believed that the contribution of Cp to the risk of CHD is not independent but rather depends on lipoprotein profile and perhaps other factors [69]. Oxidative modification of Cp may result in conformational changes that induce the release of free copper which in turn can generate ROS. These free radicals are involved in many pathological effects such as increased cell proliferation and apoptosis of endothelial cells and can as well react with nitric oxide (NO) to produce peroxynitrite anion (ONOO-)and other reactive nitrogen species.
NO is an important endothelium-derived relaxing factor synthesized from L-arginine by the NOS enzyme. Endothelial cells express constitutive endothelial NOS producing NO that diffuses from the endothelium to VSMC causing vascular relaxation and maintaining a vasodilator tone [70]. In the vascular endothelium, NO inhibits platelet and leukocyte adhesion, reduces VSMC migration and proliferation and limits the oxidation of atherogenic LDL [71]. All of these processes play a key role during atherogenesis and that is why NO is believed to be an important atheroprotective mediator. Impairments in NO production are associated with increases in cardiovascular risk factors [70]. The reduction of the bioavailability of NO might be, consequently, an important factor in CHD since NO is responsible for inhibiting proliferation of VSMC, adhesion molecules expression and lipid oxidation [69]. Hingorani and
7
co-workers [72] described a G to T polymorphism in the ecNOS gene at position 1917 resulting in the replacement of glutamic acid by aspartic acid at codon 298 (E298D). This polymorphism was identified as one of the most important risk factors for CHD, in individuals in the United Kingdom [73]. The presumed high risk E298D T allele is associated with high risk of developing CHD and MI, in individuals carriers of high-risk profiles [74]. Also, E298D T allele could be related to a reduction in NO bioavailability which can be a result of decreased enzymatic activity or cleavage of eNOS protein that may lead to the endothelium dysfunction [75].
As the result of a growing understanding of the atherosclerotic process, homocysteine (Hcy) is a newly emerging risk factor for CVD [32]. Hcy is an aminoacid present in the blood that turns out to be pathogenic when its concentration increases in plasma, a condition called HHcy, which is considered a major risk factor for vascular disease [76]. HHcy can be due to genetic or environmental factors such as nutrient-related disturbances that influence the Hcy metabolism [77]. The enzyme 5,10-Methylenetetrahydrofolate reductase (MTHFR) reduces 5,10-methylenetetrahydrofolate to 5-methyltetrahydrofolate, which is the primary circulatory form of folate and carbon donor for the re-methylation of Hcy to methionine [78]. Frosst P. and co-workers identified a common mutation in MTHFR gene due to a transition of a C to T at position 677 which led to an alteration in a highly conserved aminoacid alanine to valine in the MTHFR protein. This substitution was responsible for a reduction in the activity and increased thermolability of the enzyme. Moreover, a correlation between homozygous individuals for this alteration and elevated Hcy levels was previously reported [79]. Further studies revealed that the importance of this alteration depends on plasma levels of folate in such a way that homozygous individuals for the alteration that present low folate concentrations were those who had higher Hcy levels, when comparing with the non-homozygous ones [80]. MTHFR 677C/T polymorphism has thus been associated with ATH and with CVD risk [81].
8 OBJECTIVES
The general goal of this thesis was to investigate the role of the inflammatory process in the clinical characterization of patients with FH and study its putative correlation with CVD.
Specifically, we intended to search for genetic and biochemical/immunological biomarkers for CVD in a Portuguese FH population. In order to achieve this main objective we:
1. Collected blood samples from FH patients with and without CVD, and from healthy controls;
2. Measured inflammatory markers (PCR, Hp, Cp), pro-inflammatory cytokines (IL-6, TNF-α) and other risk factors associated with CVD (lipid profile, Hcy, NO) in all population under study;
3. Searched for genetic variation by analyzing some polymorphisms usually associated with CVD (LPL D9N, N291S, S447X, APOAV-1131T/C, APOCIII 3238C/G, TNF-α-308G/A, IL-6 -174G/C, MTHFR 677C/T, NOS E298D) in all population under study; 4. Correlated the levels of the biochemical and immunological parameters measured with
the polymorphisms analyzed;
5. Correlated the biochemical, immunological and genetic characterization with the presence/absence of CVD;
In summary, we looked for a better understanding into the interaction between inflammation, lipid metabolism and its connection to the CVD outcome, namely in FH patients.
9 MATHERIALS AND METHODS
1. Recruitment of the study population
A total number of 181 individuals of both sexes were recruited, with age ranging from 30 to 80 years old. The subjects consisted of 49 control individuals (34,7% male, 65,3% female, 43,7±11,6 years) and 132 individuals with a genetic diagnosis of FH (44,7% male, 55,3% female, 49,2±12,7 years). These FH individuals were selected from the EPHF [4] developed at the Grupo de Investigação Cardiovascular, Unidade de I&D Departamento de Promoção da Saúde e Doenças Crónicas, Instituto Nacional de Saúde Doutor Ricardo Jorge (INSA). The recruitment occurred between 1999 and 2009, with the cooperation of several clinicians from different Portuguese hospitals, which included a clinical questionnaire and a declaration of written consent where the subjects expressed their agreement with all the procedure including collection of biological samples and access to personal clinical information by the research team. The biological samples were obtained from peripheral blood by venous punction and collected to a serum gel tube and to an EDTA (Ethylenediamine tetraacetic acid) tube.
This project is based on the EPHF and has the approval from the ethic commission of INSA and from the National Commission for Data Protection.
2. Clinical characterization of the study population
The first group, consisting of 49 normolipidemic individuals, was assembled by recruiting healthy candidates. To verify the health status, common laboratory tests such as the determination of the lipid profile, hemogram and measurement of glycemia and CRP were performed. All these individuals presented normal values in all blood tests.
FH patients were selected to the EPHF following clinical criteria adapted from the “Simon Broome Heart Research Study” (Table I). According to these criteria, a person is considered a confirmed case of FH if presents total cholesterol serum concentrations over 290 mg/dL or LDL cholesterol over 190 mg/dL (for adults), and also has tendon xanthom or presents genetic evidence of a mutation in LDLR gene or APOB. All FH patients were submitted to a molecular analysis to search for a genetic alteration (mutation) in the LDLR or APOB genes in order to confirm the disease [4].
The subjects were then classified into two separated groups: patients with cardiovascular disease (41 individuals: 63,4% male, 36,6% female, 55,7±13,6 years) and patients without cardiovascular disease (91 individuals: 36,3% male, 63,7% female, 46,3±11,1
10
years). The first group included FH patients with CVD as Myocardial Infarction (MI), Coronary Artery Bypass Graft (CABG), Angina and Stroke, which had the first event before 55 years, in men and before 65 years, in women. The second group consisted of FH patients without CVD and included subjects that also had a family history of early CVD.
Table I
FH criteria adapted from the “Simon Broome Heart Research Trust” [4] Confirmed familial hypercholesterolemia is defined as:
(a) Index case: Child under 16 with total cholesterol over 200 mg/dl (5.20 mmol/l) or LDL cholesterol
over 120 mg/dl (3.10 mmol/l);
Index case: Adult with total cholesterol over 290 mg/dl (7.5 mmol/l) or LDL cholesterol over 190 mg/dl (4.9 mmol/l), and
(b) Tendon xanthoma in the index case or relative (parents, children, grandparents, siblings, aunts or
uncles), or
(c) Genetic evidence of a mutation in the LDL receptor gene or APOB Possible familial hypercholesterolemia is defined as:
(a) Index case: Child under 16 with total cholesterol over 200 mg/dl (5.20 mmol/l) or LDL cholesterol
over 120 mg/dl (3.10 mmol/l);
Index case: Adult with total cholesterol over 290 mg/dl (7.5 mmol/l) or LDL cholesterol over 190 mg/dl (4.9 mmol/l), and
(d) Family history of myocardial infarction before the age of 50 in grandparents or aunts or uncles, or
before the age of 60 in parents, siblings or children, and/or family history of elevated cholesterol levels (>290 mg/dl) in parents, siblings or children; or total cholesterol over 290 mg/dl (7.5 mmol/l) in grandparents and/or aunts or uncles.
3. Laboratorial characterization of the study population
3.1. Biochemical characterization
All the individuals from the three groups were biochemically characterized through the determination of the lipidic profile which included the measurement of total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides, ApoA1, ApoB and Lp(a), by an enzymatic colorimetric method following the manufacturer’s instructions in an automated chemistry analyzer (Hitachi 911, Boehringer Mannheim, Roche and Cobas Integra 400, Roche). Because of the impossibility to directly determinate LDL cholesterol values in Hitachi 911, the Friedewald formula was used to calculate an approximation of those values [82]. All measurements were performed in fresh serum of all individuals, obtained after the centrifugation of the blood sample collected to the serum gel tube. These analysis were performed at Unidade Laboratorial Integrada (INSA).
11
3.2. Hematological characterization
Hemogram was performed using Coulter Technology (only for the control group). Measurement of Hcy concentration was done by means of a fluorescence polarization immunoassay (Analyzer IMX), in serum samples (stored at -20ºC) of the individuals. These analysis were performed at Unidade Laboratorial Integrada (INSA).
3.3. Evaluation of inflammatory status
Individuals were immunological characterized through the measurement of circulating Cp, Hp and a high sensitivity method to circulating C-reactive protein (hsPCR) levels in serum samples (stored at -20ºC), performed by nephelometry, using a Beckman Array-TM protein system. These analysis were performed at Unidade Laboratorial Integrada (INSA).
Additionally, to further evaluate the inflammatory status, the levels of pro-inflammatory cytokines (IL-6 and TNF-α) produced at systemic level were quantified in serum samples (stored at -80ºC), by immunoenzymatic assays (ELISA), using available commercial kits (ELISA Kit for Human IL-6 / ELISA Kit for Human TNF-α – Uscn Life Science Inc.) following the manufacturer’s instructions in an automated analyzer (Dynex Technologies). The principle of the test consisted of a microtiter plate provided with this kit that had been pre-coated with an antibody specific to IL-6 / TNFα . Standards or samples were then added to the appropriate microtiter plate wells with a biotin conjugated polyclonal antibody preparation specific for IL-6 / TNFα. Next, Avidin conjugated to Horseradish Peroxidase was added to each microplate well and incubated. Then a 3,3’, 5,5’-tetramethylbenzidine (TMB) substrate solution was added to each well. Only those wells that contain IL-6 / TNFα, biotin-conjugated antibody and enzyme-conjugated Avidin exhibited a change in color. The enzyme-substrate reaction was terminated by the addition of a sulphuric acid solution and the color change was measured spectrophotometrically at a wavelength of 450 nm 2 nm. The concentration of IL-6 / TNF α in the samples was then determined by comparing the optical density of the samples to the standard curve.
3.4. Markers of Pro-oxidant /Antioxidant balance
The balance was evaluated by the measurement of the levels of NO and oxLDL through ELISAs, in serum samples (stored at -80ºC) of the individuals, using available commercial kits performed in an automated analyzer (Dynex Technologies). For the quantification of NO, total nitric oxide was measured (Total Nitric Oxide Assay – DRG International, Inc.). The principle of the test consisted of the enzymatic conversion of nitrate to nitrite, by the enzyme Nitrate
12
Reductase, followed by the colorimetric detection of nitrite as a colored azo dye product of the Griess reaction that absorbs visible light at 540 nm. The concentration of NO in serum samples was then determined by comparing the optical density of the samples to the standard curve. The quantification of oxLDL (Oxidized LDL ELISA – Mercodia) was based on the direct sandwich technique in which two monoclonal antibodies were directed against separate antigenic determinants on the oxidized apoB molecule. During incubation oxidized LDL in the sample reacted with anti-oxidized LDL antibodies bound to microtitration well. After washing, which removed non-reactive plasma components, a peroxidase conjugated anti-human apoB antibody recognized the oxidized LDL bound to the solid phase. After a second incubation and a simple washing step that removed unbound enzyme labeled antibody, the bound conjugate was detected by reaction with TMB. The reaction was stopped by adding acid to give a colorimetric endpoint, then read spectrophotometrically. The concentration of oxLDL in serum samples was then determined by comparing the optical density of the samples to the standard curve.
4. Genetic characterization of the study population
Genetic characterization of the study population was based on the study of 9 polymorphisms within 7 genes namely IL6 174G/C, MTHFR 677C/T, NOS E298D, TNFα -308G/A, APOAV -1131T/C, APOCIII 3238C/G, LPL D9N, LPL N291S and LPL S447X.
4.1. Sample preparation
Peripheral blood samples collected in an EDTA tube were maintained frozen until processing. Total genomic DNA samples were then obtained from all the individuals by DNA extraction directly from whole blood, using the commercial Wizard® Genomic DNA Purification
Kit (Promega Corp.). This kit is based on three main steps: first, red blood cells, white blood cells and nuclei lysis with cell lysis solution and nuclei lysis solution respectively; second, proteins removal by precipitation with protein precipitation solution; third, DNA precipitation with isopropanol. Specifically, for each individual it was added 9.0 mL of Cell Lysis Solution to a sterile 15 mL centrifuge tube and then transferred the blood to the centrifuge tube up to 13/14 mL. The tube was inverted 5–6 times to mix and then the mixture was incubated for 10 minutes at room temperature in a plate mixer to lyse the red blood cells before centrifugation at 2,000×g for 10 minutes at 18ºC (centrifuge 5810R, Eppendorf). Supernatant was removed and discarded as much as possible without disturbing the visible white pellet. The tube was vigorously mixed in a vortex until the white blood cells were resuspended. It was added 3 mL of Nuclei Lysis Solution to the tube containing the resuspended cells which was then inverted to lyse the white blood cells until the solution become very viscous. After that, it was added 1 mL of Protein
13
Precipitation Solution to the nuclear lysate, inverted 3-4 times and mixed vigorously in a vortex for 20 seconds. The tube was centrifuged at 2,000×g for 10 minutes at room temperature. The supernatant was transferred to a 15 mL centrifuge tube and it was added to the tube 3 mL of isopropanol at room-temperature. The solution was gently mixed by inversion until the white thread-like strands of DNA formed a visible mass. The white thread-like strands were removed with a P1000 to an eppendorf tube containing 1 mL of 70% ethanol at -20ºC. The tube was centrifuged at 2,000×g for 5 minutes at room temperature (microcentrifuga 5415D, Eppendorf). The supernatant was decanted and was added again 1 mL of 70% ethanol to the DNA. Again the tube was centrifuged at 2,000×g for 5 minutes at room temperature. The supernatant was decanted and the pellet was air-dried for 10–15 minutes. The DNA was rehydrated by adding 200 µL of sterile water and then incubating at room temperature in a thermomixer (Grant-bio, Alfagene) at (400-450 rpm) until total dissolution of DNA (12-24 h).
The DNA working solutions were prepared at concentration of 50 ng/µL and kept at 4ºC for short term storage while stock solutions were kept at -20ºC, for long term storage.
4.2. Primer designing
According to the target genomic regions of each polymorphism the primers were designed or adapted from previous studies.
Primers for MTHFR and TNF-α polymorphisms were designed in the program PrimerSelect (DNAstar) while primers for APOAV, APOCIII and LPL were designed in the program Primer3 [83]. Primers for IL-6 e NOS polymorphisms were ordered as previously described [36], [84] (Table AI).
4.3. DNA amplification
All DNA samples were amplified by Polymerase Chain Reaction (PCR) in a T3000 Thermocycler (Biometra®) for all of the nine target regions separately, using different primers and specific conditions (Table AI). The PCR reaction included 0.2 mM of each deoxyribonucleotide (dNTP) (Bioline), 10x NH4 of reaction buffer (Bioline), 1.5 mM of Mg2+ (Bioline), 10 pmol of each primer (Invitrogen), 1.25 U of Biotaq DNA Polymerase (Bioline), 100-200 ng of DNA and bidistilled water up to 25 µL (Table AII).
The PCR program used was as following: initial denaturation at 95ºC for 3 minutes, 35 cycles of denaturation at 94ºC for 45 seconds, annealing at specific temperature (Table AI) for 30 seconds and extension at 72ºC for 1 minute, and final extension at 72ºC for 7 minutes.
14
4.4. DNA Gel Electrophoresis
The DNA fragments were then detected in a 1.5% (m/v) concentrated agarose gel electrophoresis. The gel was prepared by dissolution of 1.5 g of agarose in 100 mL of Tris-Borate-Ethylenediamine tetraacetic acid (10xTBE buffer, Invitrogen) followed by staining with 2 µL of GelRed (Biotarget). After that, 6 µL of the PCR product were mixed with 3 µL of Bromophenol blue (deposition solution) and the mixture was then applied into de gel wells. The gel was subjected to an electric field with voltage of 90 volts during 35-40 minutes, then the DNA fragments were observed in a UV transilluminator and photographed with a standard camera.
4.5. Purification of the PCR products
The TNF-α, LPL, APOAV and APOCIII polymorphisms were identified by automated sequencing. First, the PCR products had to be cleaned-up from contaminating primers and remaining dNTPs with the purification kit ExoSAP-IT® (asb Corporation) which combines an Endonuclease I and a Shrimp alkaline phosphatase. In order to do that, it was added 1 µL de ExoSAP-IT® to 2.5 µL of PCR product followed by incubation at 37ºC for 15 minutes and posterior incubation at 80ºC for 15 minutes in a T3000 Thermocycler (Biometra), to inactivate the enzyme.
4.6. Sequencing of purified PCR products
The sequencing reaction was prepared using the commercial Big Dye® Terminator Cycle Sequencing Ready Reaction kit (version 2, Applied Biosystems) that uses dideoxyribonucleotides (ddNTPs) terminators labeled with fluorescent dyes. It was necessary to prepare the reaction mixture that included: 1 µL of purified PCR product, 2 µL of BigDye®, 1 µL of sequencing primer (20 µl of PCR primer and bidistilled water up to 100 µL) and bidistilled water up to 10 µL.
The reaction mixture was finally submitted to a specific reaction in a T3000 Thermocycler (Biometra®) with the following programme: 96ºC for 30 seconds, 25 cycles at 96ºC for 10 seconds followed by de 54-58ºC for 5 seconds, 60ºC for 4 minutes and cooling up to 4ºC. The direct sequencing of the PCR products was performed at Unidade Laboratorial de Uso Comum at INSA in a 3100 Genetic Analyzer with 16 capilars (Applied Biosystems) using the POP6 polymer, following manufacturer’s instructions. The polymorphic sites were detected by comparison of different DNA sequences in the program Sequence Scanner (Applied Biosystems) or BioEdit available in the Internet (http://www.mbio.ncsu.edu/BioEdit/BioEdit.html).
15
4.7. Restriction Fragment Length Polymorphism
IL-6, MTHFR and NOS polymorphisms detection was performed by restriction fragment length polymorphism (RFLP) analysis. The digestion reaction included: 10 µL of PCR reaction mixture, 7.8 µL nuclease-free water, 2 µL specific 10xbuffer, 0.2 µL restriction enzyme [IL-6 – LweI (Fermentas), MTHFR – TaqI (New England Biolabs) and NOS – MboI (New England Biolabs)
]
and the digestion occurred in a T3000 Thermocycler (Biometra®) during 4 hours (Table AIII). Then the different sized fragments were separated according to its length in a 3.5% (m/v) concentrated agarose gel electrophoresis. The gel was prepared by dissolution of 3.5 g of agarose in 100 mL of Tris-Borate-Ethylenediamine tetraacetic acid (10xTBE buffer, Invitrogen) followed by staining with 5 µL of Ethidium Bromide (Sigma). After that, 20 µL of the digestion product were mixed with 4 µL of Bromophenol blue (deposition solution) and the mixture was then applied into de gel wells. The gel was subjected to an electric field with voltage of 70 volts during 3 hours, then the DNA fragments were observed in a UV transilluminator and photographed with a standard camera.5. Statistical Analysis
All data obtained was analyzed with SPSS software (version 17.0).
To test for association between groups and sex and between groups and age brackets the Pearson Chi square test was used. In the cases of 2x2 tables the Fisher exact test was the option. When comparing two independent samples and the null hypothesis of normality was rejected, non parametric test Mann-Whitney. In the cases where normality of variables was verified the parametric Student’s t test was employed.
When comparing more than two independent samples and variances’ homogeneity was not verified, non parametric Kruskall-Wallis test was used. If the variables were normally distributed it was used the ANOVA test. In the cases where significant differences were detected, the Levene test was used to verify the homogeneity of the variances (p>0,05 indicates variances’ homogeneity). If the variances were similar the Tukey’s or LSD multiple comparisons were applied to distinguish between which groups there are differences.
If the variances were not similar then the Dunnett’s T3 test was utilized to compare groups in pairs.
16 RESULTS
1. Biochemical data characterization
A total of 181 subjects were enrolled for this study. However, only 163 individuals were biochemically characterized since there was no information available before medication, for the remaining individuals.
The study population was divided into two separate groups; the control group and the FH group constituted by all FH patients, with and without CVD. The biochemical data obtained from all these individuals is summarized in Table I. To test for normality, as all groups included more than 30 subjects, a Kolmogorov-Smirnov test was applied. All variables presented unequally distribution, with the exception of ApoB variable, so non parametric test Mann-Whitney was performed to test for association between the variables and the two study groups. ApoB association was performed using parametric Student’s t test. The results obtained from this analysis showed significant differences between these two groups in all the variables analyzed.
Table I – Clinical and biochemical characteristics of the individuals clustered by the presence or absence of FH.
Data are presented as mean ± SD. *t Student test
**Mann-Whitney test
Parameters Control FH patients p-value
n mean±SD n mean±SD Age (years) 49 43,69 ± 11,60 132 49,18 ± 12,68 Male (%) 17 34,7 59 44,7 Female (%) 32 65,3 73 55,3 Total cholesterol (mg/dL)** 49 178,78 ± 24,16 114 355,43 ± 80,09 <0,001 LDL cholesterol (mg/dL)** 48 105,08 ± 22,75 101 265,91 ± 81,41 <0,001 HDL cholesterol (mg/dL)** 48 58,13 ± 12,07 101 54,80 ± 20,19 0,017 Triglycerides (mg/dL)** 48 79,52 ± 33,68 99 162,70 ± 102,02 <0,001 ApoA1 (mg/dL)** 47 155,94 ± 24,58 103 146,52 ± 40,68 0,043 ApoB (mg/dL)* 47 75,36 ± 18,13 103 153,74 ± 53,89 <0,001 Lp(a) (mg/dL)** 45 33,80 ± 31,58 97 59,23 ± 68,87 0,022
17
In order to gain a better insight into FH pathophysiology, we further divided the study population into three groups; the control group, the FH without CVD group and the FH with CVD group. The biochemical data obtained from all these individuals is summarized in Table II / Figure 1. In order to search for the existence of significant differences between the parameters measured in all groups, it was necessary to verify first if all variables were normally distributed in order to choose the statistical test to apply. Normality was analyzed using the Shapiro-Wilk test since there was at least one group with less than 30 subjects. None of the variables had a normal distribution, except ApoB. To test for association between the variables and the three groups we had to use the non parametric test Kruskal-Wallis, even for ApoB since as it did not present homogeneity of variances we could not apply the parametric test ANOVA. The results obtained showed significant differences between the three groups in all parameters measured, with the exception of HDL cholesterol and Lp(a).
Table II – Biochemical characterization of the population under study divided into three groups: control, FH without CVD and FH with CVD.
Parameters Control FH without CVD FH with CVD p-value
n mean±SD n mean±SD n mean±SD
Age (years) 49 43,69 ± 11,60 91 46,25 ± 11,11 41 55,68 ± 13,63 Male (%) 17 34,7 33 36,3 26 63,4 Female (%) 32 65,3 58 63,7 15 36,6 Total cholesterol (mg/dL)* 49 178,78 ± 24,16 83 334,63 ± 62,80 31 411,13 ± 94,75 <0,001 LDL cholesterol (mg/dL)* 48 105,08 ± 22,75 72 243,97 ± 64,23 29 320,38 ± 94,33 <0,001 HDL cholesterol (mg/dL)* 48 58,13 ± 12,07 73 56,22 ± 22,14 28 51,11 ± 13,51 ns Triglycerides (mg/dL)* 48 79,52 ± 33,68 70 149,11 ± 91,66 29 195,48 ± 118,96 <0,001 ApoA1 (mg/dL)* 47 155,94 ± 24,58 75 151,77 ± 44,07 28 132,46 ± 25,45 0,003 ApoB (mg/dL)* 47 75,36 ± 18,13 75 139,81 ± 46,05 28 191,04 ± 56,38 <0,001 Lp(a) (mg/dL)* 45 33,80 ± 31,58 71 57,35 ± 68,27 26 64,35 ± 71,59 ns
Data are presented as percentage or mean ± SD. *Kruskal-Wallis Test
18
p=0,541
Figure 1 – Clinical and biochemical characterization in the groups under study: control, FH without CVD and FH with CVD. Only significant p values are presented.
p<0,001 p=0,001 p<0,001 p<0,001 p<0,001 p=0,021 P=0,001 p<0,001 p<0,001 P<0,001 p<0,001 p<0,001
19
Consequently, with the purpose of identifying between which groups there were differences in the parameters measured, multiple comparisons were made. The variables analyzed did not presented homogeneity of variances and therefore Dunnett’s T3 multiple comparisons were made (Table III). When comparing the controls with FH patients without CVD, we observed significant differences in total cholesterol, LDL cholesterol, triglycerides and ApoB. Also, a comparison between controls and FH patients with CVD, besides the parameters referred before there are also significant differences in ApoA1 values. Finally, comparison between the two FH groups (with CVD versus without CVD) showed significant differences in total cholesterol, LDL cholesterol, ApoA1 and ApoB values (Figure 1). As previously mentioned, HDL cholesterol and Lp(a) did not present differences between all groups compared.
Table III – Comparison of biochemical characteristics between the 3 groups.
Parameters Control FH without CVD FH with CVD
Control vs. FH
without DCV Control vs. FH with DCV FH without CVD vs. FH with CVD
n mean±SD n mean±SD n mean±SD p-value p-value p-value
Total cholesterol (mg/dL)* 49 178,78 ± 24,16 83 334,63 ± 62,80 31 411,13 ± 94,75 <0,001 <0,001 <0,001 LDL cholesterol (mg/dL)* 48 105,08 ± 22,75 72 243,97 ± 64,23 29 320,38 ± 94,33 <0,001 <0,001 0,001 HDL cholesterol (mg/dL) 48 58,13 ± 12,07 73 56,22 ± 22,14 28 51,11 ± 13,51 ns ns ns Triglycerides (mg/dL)* 48 79,52 ± 33,68 70 149,11 ± 91,66 29 195,48 ± 118,96 <0,001 <0,001 ns ApoA1 (mg/dL)* 47 155,94 ± 24,58 75 151,77 ± 44,07 28 132,46 ± 25,45 ns 0,001 0,021 ApoB (mg/dL)* 47 75,36 ± 18,13 75 139,81 ± 46,05 28 191,04 ± 56,38 <0,001 <0,001 <0,001 Lp(a) (mg/dL) 45 33,80 ± 31,58 71 57,35 ± 68,27 26 64,35 ± 71,59 ns ns ns
To try to identify the clinical value of family history in CVD outcome, the study population was reorganized. An additional group was created, including FH individuals without CVD but with a family history of early CVD (FH with fh of CVD) (Table IV), since this is a risk factor of early onset of the disease. By applying a Shapiro-Wilk test it was noticeable that only ApoB variable was normally distributed. Therefore, in order to test for association between the four groups and the biochemical parameters measured, a non parametric Kruskal-Wallis test was performed, also for ApoB as it did not present homogeneity of variances. Comparing these four groups, it was evident the differences in all parameters, except for HDL cholesterol (p=0,086) and Lp(a) (p=0,137).
Data presented as p-values (significance level: p<0,05) *Dunnett's T3 Multiple Comparisons
20 Table IV – Clinical and biochemical characteristics of the individuals between the three groups analyzed and an additionally group with FH individuals without CVD but with a family history of CVD, removed from the FH without CVD group.
Parameters Control FH without CVD FH with fh of CVD FH with CVD p-value n mean±SD n mean±SD n mean±SD n mean±SD
Age 49 43,69 ± 11,60 63 47,08 ± 12,05 28 44,39 ± 8,53 41 55,68 ± 13,63 Male (%) 17 34,7 25 39,7 8 28,6 26 63,4 Female (%) 32 65,3 38 60,3 20 71,4 15 36,6 Total cholesterol (mg/dL)* 49 178,78 ± 24,16 57 342,14 ± 64,07 26 318,15 ± 57,70 31 411,13 ± 94,75 <0,001 LDL cholesterol (mg/dL)* 48 105,08 ± 22,75 49 249,04 ± 69,19 23 233,17 ± 51,84 29 320,38 ± 94,33 <0,001 HDL cholesterol (mg/dL)* 48 58,13 ± 12,07 50 56,82 ± 24,49 23 54,91 ± 16,32 28 51,11 ± 13,51 ns Triglycerides (mg/dL)* 48 79,52 ± 33,68 48 164,17 ± 101,08 22 116,27 ± 55,57 29 195,48 ± 118,96 <0,001 ApoA1 (mg/dL)* 47 155,94 ± 24,58 53 150,58 ± 42,20 22 154,64 ± 49,21 28 132,46 ± 25,45 0,009 ApoB (mg/dL)* 47 75,36 ± 18,13 53 141,66 ± 47,82 22 135,36 ± 42,18 28 191,04 ± 56,38 <0,001 Lp(a) (mg/dL)* 45 33,80 ± 31,58 49 57,88 ± 77,11 22 56,18 ± 44,10 26 64,35 ± 71,59 ns
Table V – Comparison of clinical and biochemical characteristics between the 3 study groups and differences between them
Parameters Control FH without CVD FH with fh of CVD FH with CVD
FH without CVD vs. FH with fh of CVD FH without CVD vs. FH with CVD
n mean±SD n mean±SD n mean±SD n mean±SD p-value p-value Total cholesterol (mg/dL)** 49 178,78 ± 24,16 57 342,14 ± 64,07 26 318,15 ± 57,70 31 411,13 ± 94,75 ns 0,001 LDL cholesterol (mg/dL)** 48 105,08 ± 22,75 49 249,04 ± 69,19 23 233,17 ± 51,84 29 320,38 ± 94,33 ns 0,001 HDL cholesterol (mg/dL)** 48 58,13 ± 12,07 50 56,82 ± 24,49 23 54,91 ± 16,32 28 51,11 ± 13,51 ns ns Triglycerides (mg/dL)** 48 79,52 ± 33,68 48 164,17 ± 101,08 22 116,27 ± 55,57 29 195,48 ± 118,96 ns ns ApoA1 (mg/dL)** 47 155,94 ± 24,58 53 150,58 ± 42,20 22 154,64 ± 49,21 28 132,46 ± 25,45 ns 0,009 ApoB (mg/dL)* 47 75,36 ± 18,13 53 141,66 ± 47,82 22 135,36 ± 42,18 28 191,04 ± 56,38 ns <0,001 Lp(a) (mg/dL)** 45 33,80 ± 31,58 49 57,88 ± 77,11 22 56,18 ± 44,10 26 64,35 ± 71,59 ns ns
