台灣人族群阿茲海默氏症及血管型失智症的生物標記評估

全文

(1)國立臺灣師範大學生命科學系博士論文. 台灣人族群阿茲海默氏症及血管型失智症的生物標記評估 Estimation of biomarker candidates for Alzheimer’s disease and/or vascular dementia in Taiwanese population. 研究生：王. 秀. 觀. Hsiu-Kuan Wang. 指導教授：李. 桂. 楨博士. Guey-Jen Lee-Chen 中華民國九十八年七月.

(2) ACKNOWLEDGMENTS Special thanks to: Dr. Guey-Jen Lee-Chen, Adviser, who provided kind and strict direction from my Master’s to Doctor’s stages. Over the past years, she has wholeheartedly shared her scientific knowledge, technique, experience and wisdom in my life and study. I would also like to thank the following committee: Dr. Chiung-Mei Chen, Dr. Yijuang Chern, Dr. Yung-Feng Liao, Dr. Hsiu-Mei Hsieh and Dr. Ming-Tsan Su, who shared good study experience and corrected the manuscript with clarity and care. Thanks for their guidance and enthusiasm support. I appreciate to all research colleagues in D311 laboratory, Department of Life Science, National Taiwan Normal University. With your help, I could enjoy myself in the arduous research. Thank you, every partner. And the most special thanks to my dear father, mother and my babies Jolin and Andy, who supported and helped me toward the completion of this thesis. I love you so much!.

(3) Index Index ........................................................................................................... I 摘要.......................................................................................................... IV Abstract ......................................................................................................V List of tables and figures.......................................................................... VI Introduction.................................................................................................1 Alzheimer’s disease and vascular dementia ........................................1 AD/VaD diagnostic biomarkers ...........................................................1 Apolipoprotein E..................................................................................2 Angiotensin I-converting enzyme........................................................3 Tissue kallikrein 1................................................................................4 Interleukin-1α and interleukin-1β ........................................................4 HSPA5 chaperone ................................................................................5 Protein candidate markers for Alzheimer’s disease.............................7 Amyloid precursor protein forms ratio as biomarker for AD...............7 Oxidation-sensitive protein as biomarker for AD................................9 Specific Aims ............................................................................................12 Aim 1. To identify probable genetic factors for AD and VaD ...........12 Aim 2. To identify protein biomarkers from transformed lymphoblasts from AD patients and controls.......................12 Materials and methods ..............................................................................13 I. Case-control study to identify genetic risk markers for dementia .13 Subjects .......................................................................................13 Genomic DNA extraction ............................................................14 Polymerase chain reaction (PCR) and genotyping ....................14 Statistical analysis.......................................................................15 II. ACE promoter -240 A/T and Alu I/D reporter functional assay ..16 Promoter constructs ....................................................................16 I.

(4) (A) ACE-firefly/TK-Renilla dual luciferase reporter constructs .......................................................................16 (B) ACE-firefly luciferase - Alu I/D reporter constructs (pGL3-ACE-A/T-I/D) .....................................................17 Preparation of electro-competent cells .......................................17 Ligation .......................................................................................18 Electroporation ...........................................................................18 Minipreparation of plasmid DNA ...............................................18 Midipreparation of plasmid DNA ...............................................19 Cell cultivation ............................................................................20 Promoter functional assay and statistical analysis ....................20 III. HSPA5 promoter SNPs and ER stress functional assay ..............21 Lymphoblastoid cell cultivation ..................................................21 RNA isolation ..............................................................................21 Real-time quantitative RT-PCR analysis of HSPA5 expression ..22 IV. Western blot analysis of APP forms ratio.....................................22 Cell lysate preparation................................................................22 Western blotting and statistical analysis.....................................23 V. Western blot analysis of oxidation-sensitive protein.....................24 Preparation of DNPH-derivatized proteins ................................24 Carbonyl immunreactivity assay and statistical analysis...........24 Results.......................................................................................................26 I. Case-control study to identify genetic risk markers for dementia .26 APOE, ACE and KLK1 gene polymorphisms and the risk of dementia ......................................................................................26 IL-1α and IL-1β gene promoter polymorphism and the risk of dementia ......................................................................................27 HSPA5 gene promoter polymorphism and the risk of dementia.28 II. ACE promoter functional study ....................................................30 II.

(5) ACE-firefly/TK-Renilla dual luciferase reporter constructs and assay ............................................................................................30 ACE-firefly luciferase - Alu I/D reporter constructs and assay .30 III. ER stress and HSPA5 expression study.......................................31 IV. APP forms ratio study..................................................................31 V. Oxidation-modified protein study...............................................32 Discussion .................................................................................................33 APOE, ACE and KLK1 gene polymorphisms and the risk of dementia ............................................................................................................33 IL-1α and IL-1β gene promoter polymorphisms and the risk of dementia .............................................................................................36 HSPA5gene promoter polymorphism and the risk of dementia..........38 APP forms ratio study ........................................................................40 Oxidation-modified protein study ......................................................42 Conclusions...............................................................................................45 References.................................................................................................46 Legends for tables and figures ..................................................................72. III.

(6) 摘要阿茲海默症與血管型失智症是最普遍的兩類失智症。目前有關阿茲海默症的疾病診斷生物標記分子研究中，已知的遺傳標記分子有脂蛋白 E 基因ε4 對偶基因、類澱粉蛋白前驅蛋白基因突變、早老素 1 及早老素 2 基因突變等。腦部及腦脊髓液中 tau 蛋白與類澱粉蛋白含量則為已確認的阿茲海默症疾病診斷蛋白標記分子。本研究利用台灣阿茲海默症、血管型失智症與正常人族群基因分型技術，分析候選基因多型性變異與疾病的相關性。結果發現脂蛋白 E 基因ε4 對偶基因是阿茲海默症、而非血管型失智症的風險因子；血管收縮素轉換酶基因 DD 基因型、D 對偶基因和-240 T – Alu D 單套型是阿茲海默症與血管型失智症的風險因子；此外，實驗結果顯示介白素 1α基因-889 CT 基因型對於 70 歲以上的血管型失智症罹病感受方面具有潛在的保護功能；同時，熱休克 A5 基因-415 AA/-180 GG 基因型、-415 A/-180 G 對偶基因會減低罹患阿茲海默症的風險。基因表現分析結果亦顯示，具有-415 A/-180 G 對偶基因的細胞在遭受內質網壓力之後，其熱休克 A5 蛋白被誘發產生的程度明顯增加。由於尋找阿茲海默症淋巴細胞的生物標記的可行性，本論文亦分析了 8 個阿茲海默症及 4 個年齡、性別配合之正常人的 APP 蛋白型式及氧化蛋白，但未找到明顯的蛋白標記。以上實驗結果顯示，上述與阿茲海默症及/或血管型失智症的罹病相關的基因，可作為協助疾病診斷的遺傳標記分子。. 關鍵字：阿茲海默症，血管型失智症，遺傳變異，疾病相關性，基因表現分析. IV.

(7) Abstract Alzheimer’s disease (AD) and vascular dementia (VaD) are the most prevalent forms of dementia. To date, the allelic variants of apolipoprotein E (APOE), genetic mutations in the amyloid precursor protein (APP), presenilin 1 (PS1), and presenilin 2 (PS2) genes, the levels of tau proteins and amyloid beta-peptides in brain and cerebrospinal fluid (CSF) are the well-documented biomarkers for AD. In the study, AD patients, VaD patients and ethnic-matched nondemented controls were analyzed by means of genotype association method. APOE ε4 allele acts as risk factors only for Taiwanese AD, not for VaD. Angiotensin I-converting enzyme (ACE) DD genotype, ACE D allele and ACE -240 T – Alu D haplotype are risk factors for both AD and VaD. Besides, the results suggest a potential protective role of IL-1α -889 CT genotype in VaD susceptibility among Taiwanese over 70 years. Also a decrease in risk of developing AD for HSPA5 -415 AA/-180 GG genotype and HSPA5-415 A/-180 G allele was found. Expression analysis revealed that induction of HSPA5 expression after ER stress was markedly increased in the cells with the -415 A/-180 G allele. Since finding biomarkers in lymphocytes for AD is feasible, lymphoblastoid cells from 8 AD patients and 4 age-gender matched controls were used to study different APP forms and oxidized proteins associated with AD. However, no apparent protein marker was identified. Thus, the data suggested that genes involved in the prevalence of AD and/or VaD and may act as the genetic biomarkers for dementia diagnosis. Key words: Alzheimer’s disease, vascular dementia, genetic variation, disease association, expression analysis. V.

(8) List of tables and figures Table 1. Primers and conditions for PCR amplification ...........................73 Table 2. Genotype and allele distributions of APOE polymorphisms in cases and controls .......................................................................74 Table 3. Association analysis of APOE with AD and VaD ......................75 Table 4. Genotype, allele and haplotype distributions of ACE polymorphisms in cases and controls .........................................76 Table 5. Association analysis of ACE with AD and VaD .........................77 Table 6. Genotype and allele distributions of KLK1 polymorphisms in cases and controls .......................................................................78 Table 7. Genotype and allele distributions of IL-1α polymorphism in cases and controls .......................................................................79 Table 8. Genotype and allele distributions of IL-1β polymorphism in cases and controls .......................................................................80 Table 9. Distributions of IL-1α and IL-1β pairwise genotypes in cases and controls........................................................................................81 Table 10.Association analysis of IL-1α -889 and IL-1β -511 with AD and VaD .............................................................................................82 Table 11.Pairwise linkage disequilibrium measures for HSPA5 SNPs ....83 Table 12.Genotype, allele and haplotype distributions of HSPA5 in cases and controls and association analysis with AD and/or VaD .......84 Table 13.Possible biomarkers for Alzheimer’s disease identified in cerebrospinal fluid (CSF), blood (serum/plasma) and brain tissues of AD patients and controls.............................................85 Table 14.Demographic and clinical characteristicsa of the lymphoblastoid cell lines ......................................................................................86 Figure 1.Genotype analysis of APOE polymorphism. .............................87 VI.

(9) Figure 2.Genotype analysis of ACE polymorphisms. ..............................88 Figure 3.Genotype analysis of tissue kallikrein gene promoter polymorphism. ............................................................................89 Figure 4.Genotype analysis of IL-1 promoter polymorphisms. ...............90 Figure 5.Genotype analysis of HSPA5 promoter polymorphisms. ..........91 Figure 6.ACE promoter dual luciferase reporter assay.............................92 Figure 7.ACE Alu I/D functional assay. ...................................................94 Figure 8.Effect of HSPA5 genotype on mRNA expression......................95 Figure 9.Estimation of APPr from lymphobalstoid cell lysates. ..............96 Figure 10.Estimation of oxidated protein from lymphobalstoid cell lysates. .....................................................................................................97. VII.

(10) Introduction Alzheimer’s disease and vascular dementia Dementia is a heterogeneous condition caused by a number of different disorders. Clinically patients with dementia display progressive decline in memory and other cognitive functions sufficient to interfere with performance of usual activities. The prevalence of dementia increases as the population ages, with 5-10% in people over 65 years and 20% in people over 80 (Saunders et al., 1993). Alzheimer’s disease (AD) and vascular dementia (VaD) are the most prevalent forms of dementia. AD is neuropathologically characterized by cortical atrophy and the accumulation of amyloid plaques and neurofibrillary tangles in the brain. VaD is a clinical syndrome causing cognitive impairment as a result of cerebrovascular insults arising from a variety of pathologies. Despite being different disease, AD and VaD share many similar risk factors and overlapping clinical symptoms. (De La Torre, 2002). Even though, both clinical entities differ in their major phenotypic and genotypic profiles, as revealed by clinical and genomics studies (Cacabelos, 2004). AD/VaD diagnostic biomarkers To date, the diagnosis of AD is made primarily on clinical grounds but there is no reliable diagnostic test. It was reported that accuracy of the clinical diagnosis of AD is quite variable, causing a significant number of AD patients undiagnosed or falsely positive diagnosed (Grundke-Iqubal et al., 2006). Therefore, it is necessary to search for diagnostic markers to ensure that disease therapies are targeted at the correct patient population, to initiate early treatment and to monitor disease course (Hye et al., 2006). There were a few identified AD diagnostic biomarkers that can be divided in two groups: genetic markers and protein markers. Among genetic AD 1.

(11) markers, genetic mutations in the amyloid precursor protein (APP) (Goate et al., 1991), presenilin 1 (PS1) (Sherrington et al., 1995), and presenilin 2 (PS2) (Levy-Lahad et al., 1995) genes can lead to early-onset familial AD (reviewed in Hardy, 1997). Although various mutations in the amyloid precursor protein, presenilin 1, and presenilin 2 genes can lead to familial AD, the etiologies of sporadic cases remain unclear. The presence of the apolipoprotein E (APOE) ε4 allele represents a well-documented risk factor for late-onset AD (Corder et al., 1993; Nalbantoglu et al., 1994; Duara et al., 1996; Farrer et al., 1997). It has also been proposed as a risk factor for VaD (Slooter et al., 1997; Kalman et al., 1998; Marin et al., 1998). On the other hand, the most significant protein biomarkers are levels of tau proteins and amyloid beta (Aβ)-peptides (Zhang et al., 2005; Abdi et al., 2006; Simonsen et al., 2008; Zetterberg et al., 2008) in brain and cerebrospinal fluid (CSF) and plasma. Nevertheless, as dementia disease is multifactorial and heterogeneous, neither of recent biomarkers allows the ultimate disease diagnosis. Thus, identification of various disease biomarkers will help improvement in diagnoses and development of potent therapeutic drugs. Apolipoprotein E Human apolipoprotein E (apo E) is a glycoprotein that contains 299 amino acids (Rall et al., 1982). Apo E is synthesized primary by the liver and the brain but also by other tissues including astrocytes, macrophages and monocytes (Basu et al., 1982). There are three major apo E protein isoforms (E2, E3 and E4) that are the products of three allelic forms (ε2, ε3 and ε4) of APOE gene located on the chromosome 19q13.2 (Utermann et al., 1980; Strittmatter et al., 1993). These alleles give rise to six different genotypes and the most common is ε3ε3. The three protein isoforms differ by the interchange of cysteine (Cys) and arginine (Arg) 2.

(12) residues at positions 112 and 158 of the mature apo E protein. That is, the common wild type apo E3 is Cys112/Arg158, the apo E2 is Cys112/Cys158 and the apo E4 is Arg112/Arg158 (Siest et al., 1995; Weisgraber et al., 1981). There is an increased ε4 allele frequency in multi-infarct dementia patients compared to normal controls (Shimano et al., 1989). Several studies have reported increased ε4 allele frequencies in early-onset AD, late-onset AD, familial AD and sporadic AD (Corder et al., 1993; Saunders et al., 1993; Strittmatter et al., 1993; van Duijn et al., 1994). Individuals carrying two ε4 alleles are at higher risk and have an earlier onset age of disease than those with one or no ε4 allele (Corder et al., 1993), indicating that the risk of developing AD seems to be ε4 allele dose dependent. By the way, studies have reported that the risk associated with ε4 allele in AD would seem to be age- and gender- dependent (Rebeck et al., 1994; Sobel et al., 1995). Several studies have reported apo E protein binds to synthetic and soluble Aβ peptide in CSF (Wisniewski et al., 1993). In particular, apo E4 binds Aβ peptide more efficiently and causes novel monofibrils formation that precipitated to form dense structures (Sanan et al., 1994). It is suggested that the interaction between apo E2 / E3 and tau protein serves as protection against tau phosporylation and tangle formation (Strittmatter et al., 1994). Angiotensin I-converting enzyme Vascular risk factors such as elevated serum cholesterol levels and high blood pressure have been shown to be associated with the development of AD in large cohort studies (Kivipelto et al., 2001). Angiotensin I-converting enzyme (ACE) is a key factor in the production of angiotensin II and the degradation of bradykinin, which are involved in vascular physiology. An Alu insertion/deletion (I/D) polymorphism of the ACE gene determines 3.

(13) most of the plasma ACE activity genetically (Rigat et al., 1990). Although inconsistent, several results of case-control studies suggest that the ACE Alu I/D polymorphism may confer a genetic susceptibility to dementia (Cheng et al., 2002; Elkins et al., 2004; Farrer et al., 2000; Kehoe et al., 1999; Narain et al., 2000; Richard et al., 2001), suggesting a major implication of vascular risk factors in the occurrence of dementia. Tissue kallikrein 1 Tissue kallikrein processing low molecular weight kininogen to produce the potent vasoactive kinin peptide (Bhoola et al., 1992) has been examined for its potential role in human essential hypertension and associated complications. The tissue kallikrein gene, KLK1, is uniquely polymorphic with a poly-G length polymorphism coupled with multiple single base substitutions in the -121 to -131 promoter region. The functional analysis suggests that different haplotype alleles contribute significantly to the level of kallikrein expression (Song et al., 1997). Interleukin-1α and interleukin-1β The interleukin 1 (IL-1) cytokine family is thought to participate in processes leading to neurodegeneration. IL-1 has been demonstrated to enhance both the synthesis and the translation of amyloid-β protein precursor mRNA in human endothelial cells or astrocytes (Goldgaber et al., 1989; Rogers et al., 1999), suggesting a direct role for this cytokine in the formation of senile plaque. Decreased production of the intrathecal IL-1 receptor antagonist in patients with AD also suggests a propensity towards inflammation in AD (Tarkowski et al., 2001). On the other hand, IL-1 may stimulate reactive astrocytes and display neuroprotective properties (Giulian and Lachman, 1985). Thus, IL-1 may exert both toxic and protective effects on AD. The IL-1 gene family encodes three structurally and functionally 4.

(14) related polypeptides, IL-1α, IL-1β and IL-1 receptor antagonist (Nothwang et al., 1997). IL-1α and IL-1β act through the same cell surface receptors and have similar biological activities (March et al., 1985). They have been implicated in the pathophysiology of several chronic inflammatory diseases, including. rheumatoid. arthritis,. inflammatory. bowel. diseases. and. insulin-dependent diabetes mellitus (Dinarello and Wolff, 1993; McDowell et al., 1995; Nemetz et al., 1999). The IL-1α -889 C/T and IL-1β -511 C/T single nucleotide polymorphisms (SNPs) were reported to be related to the amount of cytokine produced (Dominici et al., 2002; Hall et al., 2004). Thus these two genetic variations may account for individual susceptibility to inflammatory-associated conditions. Association of IL-1α -889 C/T SNP with risk and age of onset of AD has been shown (Du et al., 2000; Grimaldi et al., 2000; Hedley et al., 2002; Kolsch et al., 2001; Nicoll et al., 2000; Rebeck, 2000), although not all association studies have been able to replicate these findings (Fidani et al., 2002; Ki et al., 2001; Kuo et al., 2003; Tsai et al., 2003). IL-1β -511 C/T SNP alone (Wang et al., 2005; Yucesoy et al., 2006) or in combination with the IL-1α -889 C/T (Grimaldi et al., 2000; Nicoll et al., 2000) has been shown to be associated with AD, but such finding was not observed in another study (Ehl et al., 2003). For VaD, only one study showing lack of association of IL-1α -889 C/T SNP (Kuo et al., 2003) and one showing association of IL-1β -511 C/T SNP (Yucesoy et al., 2006) were reported. Thus, further evaluation of the association of IL-1 gene polymorphisms with AD and VaD and their roles in pathogenesis is needed. HSPA5 chaperone HSPA5 (heat shock 70 kDa protein 5) gene encodes the HSPA5 (GRP78/BiP) protein which functions as endoplasmic reticulum (ER) chaperone essential for regulating ER stress response. When unfolded proteins accumulate in ER by some reasons, ER stress response consisting 5.

(15) the induction of ER chaperone such as HSPA5 begins for promoting the folding of unfolded proteins (Schroder and Kaufman, 2005). Modification of amyloid precursor protein (APP) in the ER is essential for the generation of Aβ (Selkoe, 1999) and it is known that ER is proposed to be important for Aβ-induced apoptosis of neuronal cells and there exists a close relationship between the ER stress response and Aβ (Yan et al., 1997; Sato et al., 2001; Katayama et al., 2001, 2004). By the way, several studies have shown that the ER chaperone HSPA5 may play a possible protective role against Aβ and AD (Yang et al., 1998; Yoo et al., 2001; Kakimura et al., 2002; Vattemi et al., 2004; Hoozemans et al., 2005; Hoshino et al., 2007). Thus, the ER cheperone such as HSPA5 might involve in the pathogenesis of AD. There are four single nucleotide polymorphism (SNP) sites in the promoter region of HSPA5 gene: -415G/A, -378C/T, -370C/T and –180 del/G (Kakiuchi et al., 2005). A promoter assay revealed that these polymorphisms affect the promoter activity and the haplotype of HSPA5 gene was nominally associated with bipolar disorder in Japanese population (Kakiuchi et al., 2005). However, the genetic association analysis of HSPA5 SNP polymorphisms with AD and/or VaD is rare. Heat shock 70 kDa protein 5 (HSPA5), also referred to as glucose-regulated protein (GRP78) 78 kDa or Bip (immunoglobulin heavy chain-binding protein), is involved in folding and assembly of proteins in the endoplasmic reticulum (ER). The synthesis of HSPA5 is not significantly affected by heat shock. Instead, its synthesis can be stimulated by a variety of environmental and physiological stress conditions that perturb ER function and calcium homeostasis (Lee, 2001). As an ER stress signal regulator as well as its ability to block the apoptotic process, HSPA5 plays an important role in the ER stress and unfolded protein response (UPR) (Lee, 2005). HSPA5 can bind to APP and decrease Aβ peptides secretion (Yang et al., 1998). Both familial AD-linked PS1 mutation and sporadic AD-linked 6.

(16) aberrant spliced PS2 isoform impair UPR signaling and lead to vulnerability to ER stress (Katayama et al., 2001; Sato et al., 2001). HSPA5 promoter polymorphisms [-415 G/A (rs391957), -370 C/T (rs17840761) and -180 del/G (rs3216733) polymorphisms] have been suggested as a risk factor for bipolar disorder in Japanese population (Kakiuchi et al., 2005). Protein candidate markers for Alzheimer’s disease Several studies have investigated protein biomarker candidates for AD in various tissues including CSF (Blum-Degen et al., 1995; Hampel et al., 2004; Ida et al., 1996; Jensen et al., 1995; Kanai et al., 1998; Kurz et al., 1998; Motter et al., 1995; Simonsen et al., 2008; Tapiola et al., 1997; Zhang et al., 2005), blood (Hye et al., 2006; Kuo et al., 1999; Matsubara et al., 1999; Zhang et al., 2004) and brain tissues (Shiozaki et al., 2004) (Table 13). The most reproduced findings by these approaches are that there exists decreased Aβ42 and increased total tau and phospho-tau in CSF in AD (Blennow and Hampel, 2003; Sunderland et al., 2005). However, the estimation of AD biomarker using human lymphoblast cells is rare reported. Previously, several studies have shown substantial deficits in peripheral tissues including lymphocytes that parallel neuronal dysfunction in several polyglutamine diseases (Borovecki et al., 2005; Colin et al., 2005; Maglione et al., 2005; Nagata et al., 2004). These studies suggest that finding biomarkers in lymphocytes or lymphoblastoid cells for AD is feasible. Thus, in the study, we aim to develop biomarkers for disease diagnosis and/or early detection from 8 AD patients and 4 age-gender matched control human lymphoblastoid cells. Amyloid precursor protein forms ratio as biomarker for AD The proteolysis of the amyloid precursor protein (APP) leads to the extracellular and intracellular accumulation of a 40- or 42-residue β-amyloid 7.

(17) protein (Aβ40 or Aβ42) (Beyreuther and Masters, 1991; Kosik, 1992) in Alzheimer’s disease. The APP gene localized on human chromosome 21 has 19 exons (Selkoe et al., 1988; Yoshikai et al., 1990). The family of APP transmembrane glycoproteins now extends to ten named isoforms (http://us.expasy.org/uniprot/P05067) generated by alternatively splicing of the APP gene, principally of exons 7, 8, and 15. APP695, APP751, and APP770 are the three isoforms most relevant to AD. APP695 is restricted to the central nervous system (CNS), and APP751 and APP770 are both expressed in peripheral and CNS tissues (Kang and Muller-Hill, 1990; Kitaguchi et al, 1988; Tanzi et al., 1988). The presence of APP in peripheral cells, such as endothelial and blood cells, may contribute to Aβ deposition (Ghilardi et al., 1996; Shayo et al., 1997). The circulating forms of APP in blood may derive from platelets (Bush et al., 1990) or activated lymphocytes/monocytes (Monning et al., 1990). It has demonstrated that AD patients show a specific alternation in levels of platelet APP forms (Davies et al., 1997; Rosenberg et al., 1997). A marked decrease in the ratio between the 120 to 130-kDa APP and 106 to 110-kDa APP forms was found in platelets of AD patients compared with control individuals. (Rosenberg et al., 1997; Luca et al., 1998; Baskin et al., 2000; Borroni et al., 2001). Besides, the ratio of platelet APP isoforms varied according to Clinical Dementia Rating score, there existed a positive correlation between the platelet APP isoforms ratio and individual Mini-Mental State Examination (MMSE) score (Luca et al., 1998) and the platelet APP isoforms ratio declines in a linear manner with the severity of dementia (Tang et al., 2006). In the meanwhile, administration of actylcholinesterase (AChE) inhibitors increases the ratio of APP forms in platelets of AD patients, suggesting that the peripheral marker might be useful to monitor not only disease progression (Baskin et al., 2000) but also pharmacologic manipulations (Borroni et al., 2001). The platelets of AD patients and control individuals are not conveniently available in our study. 8.

(18) Monning et al. had demonstrated that mitogen-activated peripheral mononuclear blood leucocytes (PMBL), like activated platelets, secrete APP (Monning et al., 1990). Thus, the study try to test whether the APP ratio also could be observed in lymphoblastoid cell lines of AD patients and control individuals and verify whether the lymphoblastoid cell lines could act as the peripheral marker of pharmacologic manipulations. Oxidation-sensitive protein as biomarker for AD Oxidative damage has been implicated acting as an important role in aging, cancer, chronic inflammatory diseases, cardiovascular diseases, and neurodegenerative diseases such as Alzheimer’s disease (Oliver et al., 1987; Halliwell and Gutteridge, 1990; Weitzman and Gordon, 1990; Smith et al., 1991; Stadtman, 1992). Not only the environmental factors such as ionizing radiation and chemical substances, but also the normal cellular metabolism by mitochondrial electron transport and cellular redox system and immune responses could generate oxidative stress (Cance et al., 1979; Weitzman and Gordon, 1990). Activated phagocytes (Klebanoff and Waltersdorph, 1988), pro-oxidant enzymes such as xanthine oxidase (Granger, 1988) and mixed function oxidase (Stadtman, 1986) and normal cellular metabolism (Adelman et al., 1988; Cance et al., 1979) are the sources of endogenous oxidants. The reactive oxygen compounds associated pathology of diseases derives from their ability to modify intra- and extracellular macromolecules. —. nucleic. acids,. lipids. and. proteins. through. metal-catalyzed oxidation reaction (Cance et al., 1979; Harman, 1981; Davies, 1987a; Shigenaga et al., 1989; Halliwell and Gutteridge, 1990; Stadtman, 1990; Smith et al., 1991; Beckman and Ames, 1997). Proteins are the major targets for oxidative modification (Davies, 1987a; Stadtman, 1990). Oxidative modification of proteins modifies and/or converses the side chains of some amino acids (Davies, 1987b; Amici et al., 1989; Stadtman 9.

(19) and Berlett, 1991; Heinecke et al., 1993). The biochemical characteristics of the oxidated-modified proteins, such as enzymatic activity (Levine, 1983; Oliver et al., 1987), DNA binding activities of transcription factors (Wolff and Dean, 1986; Pognonec et al., 1992) and the susceptibility to proteolytic degradation (Davies, 1987a) could be modulated. Oxidative damages to proteins, lipids, or DNA may all be seriously deleterious and may be concomitant (reviewed in Dalle-Donne et al., 2003). Thus, study for the oxidation-sensitive proteins as biomarkers could be meaningful for AD research. Carbonyl (CO) groups (aldehydes and ketones) are produced on site-specific side chains of amino acids (especially Pro, Arg, Lys, Thr, Cys and His) when they are oxidized (reviewed in Berlett and Stadtman, 1997; Dalle-Donne et al., 2003). Carbonyl formation is an important detectable marker of protein oxidation (Aksenov et al., 2001). Oxidation-sensitive proteins have been considered as potential biomarkers for AD (Choi et al., 2002). Recently, the usage of protein carbonyl groups as biomarkers of oxidative stress is widespread. Statistically significant increase of a 78 kDa carbonylated protein band intensity was observed in AD subjects plasma compared with controls by Western blotting (Conrad et al., 2000). The authors further identified the presence of protein carbonyl immunoreactivity in fibrinogen γ-chain precursor protein and α-1-antitrypsin precursor in AD and control plasma by proteomics analysis (Choi et al., 2002). Besides, the carbonyl immunoreactivity in also found in β-tubulin, β-actin, creatine kinase BB (CKBB), glutamine synthetase (GS), ubiquitin carboxy-terminal hydrolase L-1 (UCH L-1), glial fibrillary acidic protein (GFAP), dihydropyrimidinase-related protein-2 (DRP-2) and DJ-1 in Alzheimer’s disease and/or control brain extracts by proteomics analysis (Aksenov et al., 2001; Castegna et al., 2002; Boyd-Kimball et al., 2005; Choi et al., 2006). In this study, we test the protein carbonyl immunoreactivity in Alzheimer’s 10.

(20) disease and control lymphoblastoid cells and try to identify the oxidation-sensitive protein biomarkers for AD from peripheral tissues.. 11.

(21) Specific Aims Aim 1. To identify probable genetic factors for AD and VaD Since the established known genetic variations, APP, PS1, PS2 gene and APOE ε4 allele, only account for a small portion of AD and VaD, more genetic factors remain to be identified. Therefore, the main purpose of this thesis is to identify probable genetic risk factors for AD and VaD. Using case-control study, the APOE, ACE, KLK1, IL-1α, IL-1β and HSPA5 genes as genetic candidate markers for two type of dementia in Taiwanese: AD and VaD were investigated. Aim 2. To identify protein biomarkers from transformed lymphoblasts from AD patients and controls In the AD study so far, the effective treatments and good indicators for disease progression are not yet available. Since finding biomarkers in lymphocytes or lymphoblast cells for AD is feasible, Epstein-Barr virus (EBV) transformed lymphoblasts from 8 AD patients and 4 controls were used for biomarkers investigation in the thesis. These transformed lymphoblastoid cell lines can provide a cellular model for protein biomarker investigations for AD by using Western blot analysis.. 12.

(22) Materials and methods I. Case-control study to identify genetic risk markers for dementia Subjects Patients were recruited from the dementia outpatient clinic of Chang Gung Medical Center. All examinations were performed by Drs W.-C. Hsu, H.-C. Fung, J.-C. Lin, H.-P. Hsu, C.-M. Chen, Y.-R. Wu, K.-H. Chang and L.-S. Ro (Department of Neurology, Chang Gung Memorial Hospital) after obtaining informed consent from patients and control individuals. Patients with a previous clinical history of neurological, psychiatric, somatic, or toxic causes for dementia were excluded. Evaluation included general physical and neurological assessment, the Mini-Mental State Examination (MMSE) and the Hachinski ischemia score (HIS). Laboratory studies included complete blood cell count, biochemistry analysis, erythrocyte sedimentation rate, vitamin B12 and folic acid levels, thyroid-stimulating hormone level and syphilis serological testing. Each patient underwent a brain computerized tomography scan (CT) or magnetic resonance imaging (MRI). At least 2 neurologists examined all the patients, and confirmed that they fulfilled the DSM-IV criteria for dementia. The diagnosis of AD was made by consensus, according to the criteria of the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer’s Disease. and. Related. Disorders. Association. for. probable. AD. (NINCDS-ADRDA) (McKhann et al., 1984). VaD was diagnosed using the criteria of the National Institute of Neurological Disorders and Stroke and the Association Internationale pour la Recherche et l’Enseignement en Neurosciences (NINDS-AIREN) (Roman et al., 1993). All VaD patients in this study had HIS of > 7. They belonged to the multi-infarct dementia subtype of VaD. The brain CT or magnetic resonance imaging of those patients showed a good clinical-radiographic correlation with at least one 13.

(23) apparent infarction (Moroney et al., 1997). Patients whose HIS were between 4 and 7, with dementia caused by a possible combination of VD and AD and those whose brain CT study only showed peri-ventricular white matter changes were excluded. Unrelated controls without stroke and cognitive impairment were recruited from the same community; similar ethnic origin and area of residence of the controls and the patients limit the possible confounding effect of population stratification. Genomic DNA extraction Genomic DNA extraction from whole blood (peripheral leukocytes) or lymphoblastoid cells is carried out using DNA Extraction Kit (Cat. No. 200600, Stratagene). Three volume of 1× solution 1 was mixed well with the blood sample and incubated on ice for 5 minutes. After centrifugation at 2,000 rpm for 10 minutes, the supernatant was removed and the pellet was resuspended in 2 ml of solution 2. Then 10 μl of pronase is added to the suspension and incubated at 60°C for several hours, or at 37°C overnight for several days. The suspension was placed on ice for 10 minutes, followed by mixing with 0.8 ml of solution 3 and stood on ice for 10 minutes. The protein precipitate was removed by centrifugation at 3,400 rpm for 15 minutes at 4°C, and the supernatant containing nucleic acid was treated with 6 μl of RNase at 37°C for 15 minutes. DNA precipitate was separated out by mixing with 2.5 ml of isopropanol, transferred to a fresh tube, and centrifuged at 14,000 rpm for 1 minute. The pellet was rinsed once with 70% ethanol and air-dried. DNA was dissolved in adequate volume of ddH2O and the concentration was determined using the spectrophotometer. Polymerase chain reaction (PCR) and genotyping Genotypes for APOE ε2/ε3/ε4, ACE -240 A/T and Alu I/D, IL-1α -889 14.

(24) C/T, IL-1β -511 C/T, as well as HSPA5 -415 G/A, -370 C/T and -180 del/G were determined by polymerase chain reaction (PCR)-based restriction and/or electrophoresis assay (Table 1). Briefly, 100 ng of genomic DNA, 0.4 μM of each primer, 100-200 μM dNTPs, 0.8-3.0 mM MgCl2, 10 mM of Tris pH 8.3, 50 mM KCl, 0.5 U Taq polymerase were prepared in a final volume of 25 μl. Genotypes for ACE Alu I/D were determined by direct electrophoresis of PCR products. Genotypes for APOE ε2/ε3/ε4, ACE -240 A/T, IL-1α -889 C/T, IL-1β -511 C/T, and HSPA5 -415 G/A and -370 C/T were determined by electrophoresis after restriction analysis. Genotypes for HSPA5 -180 del/G were determined by analyzing PCR-amplified products in a linear polyacrylamide gel on an automated MegaBACE Analyzer (Molecular Dynamics, Division of Amersham Pharmacia Biotech). Alleles del and G were confirmed by DNA sequencing. Genotypes for KLK1-130 GN were determined as previously described (Lee-Chen et al., 2004). PCR-amplified products were analyzed in a linear polyacrylamide gel and allele sizes were determined by comparing migration relative to molecular weight standards. In addition, aliquots of the amplified products were mixed with an equal volumn of 95% formamide buffer and subjected to single strand conformation polymorphism (SSCP) analysis using GeneGel Excel gels as recommended by the manufacture (Pharmacia Biotech). Statistical analysis Differences in genotype frequencies between groups were assessed by the Chi-square test. The expected genotypic frequency under random mating was computed using the algorithm described by H. Levene (Levene, 1949), while Chi-square analysis was used to test for the Hardy-Weinberg equilibrium (Yeh and Boyle, 1997). The pair-wise haplotype frequencies were computed by gene counting (Hill, 1974) and tested by Chi-square test. 15.

(25) for significance. A P-value of statistical significance was adjusted by Bonferroni correction for each set of comparison. For HSPA5 gene, the SNPSpD method (Nyholt, 2004) was used to generate an adjusted significance threshold for correction of multiple SNP testing. The experiment-wide significance threshold of 0.030 was required to keep the type I error rate at 5%. Measures of pairwise linkage disequilibrium (LD) between SNPs including Lewontin’s standardized disequilibrium coefficients (D’), The squared pairwise correlations (Δ2), and eigenvalues (λs) were computed with the LDMAX software-part of the GOLD Command Line Tool package (Abecasis and Cookson 2000). PHASE version 2.1 was used to infer the HSPA5 gene haplotypes (Stephens et al., 2001). The HSPA5 pairwise haplotype frequencies were computed and Chi-square tested for significance. Odds ratios (ORs) with 95% confidence intervals (95% CI) were calculated to test the association between genotype/allele/haplotype and disease. II. ACE promoter -240 A/T and Alu I/D reporter functional assay Promoter constructs (A) ACE-firefly/TK-Renilla dual luciferase reporter constructs The ACE promoter fragment (-1204 ~ +4, where +1 represents the first nucleotide to be transcribed) containing -240 A or T allele were cloned into the pGEM-T Easy vector (Promega) and sequenced ( 劉 , 2003). The SacI-SalI fragment containing HSV-TK promoter and Renilla luciferase gene from phRL-TK plasmid and the XhoI-HindIII fragment containing EcoRI site from the polylinker region of pcDNA3 were placed between the SacI/HindIII sites of the pGL3-basic vector to generate dual luciferase reporter vector (侯, 2005). The cloned ACE promoter fragments were excised with EcoRI (in the pGEM-T Easy vector) and inserted upstream of 16.

(26) the firefly luciferase reporter gene at the EcoRI site (in the added polylinker region) of the dual luciferase reporter vector. The orientation of the ACE promoter in dual luciferase reporter plasmid was confirmed by restriction analysis. (B). ACE-firefly. luciferase. -. Alu. I/D. reporter. constructs. (pGL3-ACE-A/T-I/D) The ACE promoter fragments containing -240 A or T allele were inserted upstream of the firefly luciferase reporter gene at the EcoRI site added in the polylinker region of the pGL3-basic vector to generate pGL3-ACE-A or -T plasmids ( 劉 , 2003). The ACE Alu I and D allele-containing fragments were cloned into the pGEM-T Easy vector and sequenced. The Alu I and D allele-containing fragments were excised with AccI and HpaII (in the pGEM-T Easy vector) and inserted downstream of the firefly luciferase reporter gene at the AccI site of the pGL3-ACE plasmid. The insertion and orientation of the Alu I/D in pGL3-ACE reporter plasmid was confirmed by EcoRI and AatII-AccI restriction analysis, respectively. Preparation of electro-competent cells The day before preparation, bacteria (TOP 10F’, Cat. No.50-0059, Invitrogen) were cultured overnight in 100 ml of LB broth containing tetracycline. On the day, bacterial culture was aliquot into four 1-liter flasks, each containing 125 ml of LB broth, and inoculated with shaking at 37°C until the OD600 reached 0.7~0.8. Cells were chilled on ice for 10 minutes and pelleted by centrifugation at 4,000 rpm at 4°C for 5 minutes. The supernatant was removed and resuspended, pooled the cell pellet from 500 ml culture with 250 ml of cold ddH2O. Cells were centrifuged at 4,500 rpm at 4°C for 5 minutes and washed with 250 ml of cold ddH2O twice. The supernatant was poured off and cells were resuspended in 1 ml of ddH2O 17.

(27) containing 20% glycerol. Finally, cells were aliquot into eppendorf tubes in 40 μl amounts and stored at -80°C. Ligation The adequate amount of insert DNA fragment was mixed with vector, 0.5 μl of 10× ligation buffer and 1.25 U of T4 DNA ligase (Cat. No. EL0331, Fermentas) in a volume of 5 μl. The ligation reaction was incubated at room temperature for 1 hour or 16°C overnight. Electroporation The ligation reaction was inactivated at 65°C for 10 minutes and kept on ice after spinning down. Immediately prior electroporation, 1 μl ligation sample was mixed with 20 μl aliquot of electro-competent cells and then added to the 1 mm-gap cuvatte (Cat. No. 4307-000-569, Eppendorf). The electroporation was carried out using MicroPulser Electroportor (BIO-RAD) with 1.25 kV under the manual program. Chilled LB broth (500 μl) was added immediately to the cuvatte. The content of cuvatte was then transferred to the tube, incubated at 37°C for at least 30 minutes, and plated on LB medium containing antibiotic for incubation at 37°C for 16~18 hours. Minipreparation of plasmid DNA A. single. bacterial. colony. was. transferred. into. 1. ml. of. antibiotic-containing LB broth. After overnight incubation at 37°C with vigorous shaking, the bacteria were harvested by 14,000 rpm centrifugation for 1 minute. The bacterial pellet was resuspended in 70 μl of solution I (50 mM glucose - 25 mM Tris-HCl pH 8.0 - 10 mM EDTA pH 8.0) by vigorous vortexing. Then 140 μl fresh solution II (2.0 N NaOH - 1% SDS) was added into the bacterial suspension and the contents was mixed by inverting 18.

(28) several times followed mixing with 105 μl solution III (3 M potassium acetate). The cell debris and genomic DNA were precipitated at 14,000 rpm for 5 minutes, and the supernatant containing plasmid DNA was transferred to a fresh tube. The plasmid DNA pellet was precipitated after mixing with 0.8 volume of isopropanol and centrifuged at 14,000 rpm for 5 minutes. The DNA pellet was washed in 70% ethanol and dried in the air. The plasmid DNA was dissolved in 40 μl ddH2O containing 10 μg/ml RNnase and checked by 0.8% agarose gel electrophoresis. Midipreparation of plasmid DNA Large-scale isolation of plasmid DNA was performed with Midi-V100TM Ultrapure Plasmid Extraction System (Viogene). A single bacterial colony was incubated in 65 ml broth at 37°C for 20 hours and then the bacteria cells were harvested by centrifugation at 8,000 g for 5 minutes. The cell pellet was resuspended in 4 ml of VP1 buffer containing RNase. The cell suspension was mixed with 4 ml of VP2 buffer and with 4 ml of VP3 buffer, sequentially. The cell debris was centrifuged at 12,000 g for 15 minutes and the supernatant was applied to a Midi-V100TM column, which was pre-equilibrated with 10 ml of VP4 buffer. The column was washed with 15 ml of VP5 buffer and the plasmid DNA was eluted with 5 ml of VP6 buffer. Then DNA was precipitated by mixing 3.75 ml of isopropanol, allotted to 1.5 ml tube, and centrifuged at 14,000 rpm for 10 minutes. After removing the supernatant, DNA pellet was dissolved in 63 μl of ddH2O for each tube and combined to one tube. To eliminate residual salt, DNA was precipitated by mixing with 20 μl of 5 M NaCl and 1 ml of pure ethanol, followed by centrifugation at 14,000 rpm for 5 minutes. The supernatant was removed and DNA was allowed to dry in the air. Finally, DNA was dissolved in 400 μl of ddH2O and the concentration was determined using the 19.

(29) spectrophotometer. Cell cultivation Human embryonic kidney 293 (HEK-293) (ATCC No. CRL-1573) and neuroblastoma IMR-32 (ATCC No. CCL-127) cell lines were maintained in DMEM supplement with 10% fetal bovine serum (FBS), 1.0 mM sodium pyruvate, 1.5 g/L sodium bicarbonate, 100 U/ml penicillin, and 100 U/ml streptomycin, at 37°C in an atmosphere containing 5% CO2. Promoter functional assay and statistical analysis Cells were plated into 96-well dishes (105 cells/well) and transfected with ACE-firefly/TK-Renilla dual luciferase reporter plasmid (0.3 μg) or co-transfected. with. ACE-firefly. luciferase. reporter. plasmid. (pGL3-ACE-A/T-I/D, 0.294 μg) and internal Renilla luciferase control plasmid (phRL-TK, 0.006 μg) by the lipofection method (GibcoBRL). Briefly, 0.3 μg of DNA and 1 μl of LipofectamineTM 2000 reagent (LF2000) (Cat. No. 11668-019, Invitrogen) was first diluted into 100 μl Opti-MEM I ®. reduced serum medium (Cat. No. 31985-062, Invitrogen), respectively. The dilution solution was incubated for 5 minutes at room temperature. Then DNA and LF2000 were combined in the well and incubated for 20 minutes at room temperature. At the meanwhile of DNA-LF2000 complexes formation, the cell suspension was prepared in 10% FBS-DMEM without antibiotics. The cell suspension was added to the DNA-LF2000, mixed gently by rocking back and forth, and incubated at 37°C in an atmosphere containing 5% CO2. After 48-h incubation, cell lysates were prepared and the activity of each promoter was directly measured by the ratio of the firefly luciferase level to the Renilla luciferase level using a dual luciferase assay system (Promega). For each reporter construct, three independent 20.

(30) transfection experiments were performed. An allele-specific difference in luciferase activity was tested using the two-tailed Student’s t test. III. HSPA5 promoter SNPs and ER stress functional assay Lymphoblastoid cell cultivation Lymphoblastoid cell lines were established (Food Industry Research and Development Institute, Taiwan) after obtaining informed consent from patients. Cells were matained in RPMI 1640 medium (GIBCO) contaning 10% FBS, 1.0 mM sodium pyruvate, 1.5 g/L sodium bicarbonate, 100 U/ml penicillin, and 100 U/ml streptomycin, at 37°C in an atmosphere containing 5% CO2. Cells were treated with thapsigargin (300 nM) for 3 hours for ER stress assay. The lymphoblastoid cells DNA were extracted to verify the APOE ε2/ε3/ε4 and HSPA5 promoter -415G/A and -180del/G genotypes. RNA isolation Total RNA of lymphoblastoid cells with or without thapsigargin treatment was extracted using the Trizol reagent (Invitrogen) (Cat. No. 15596-018, Invitrogen) according to the manufacturer’s specifications. Cells were harvested by centrifugation at 1,000 rpm for 5 minutes followed with cold PBS washing twice. After PBS washing, 500 μl of Trizol reagent was added to the cell pellet and cells were dispersed into Trizol reagent by gently pipetting. The cell suspension was incubated on ice for 5 minutes, mixed well with 1/5 volumn of chloroform and incubated on ice for another 5 minutes. RNA was separated from DNA and proteins by centrifugation at 4°C for 15 minutes. The colorless, upper aqueous phase was carefully removed to a fresh tube avoiding the material that collected at the interface, and mixed with 0.8 volume of isopropanol. The mixture was sat at -20°C for at least 1 hour and centrifuged at 4°C for 15 minutes to precipitate RNA. 21.

(31) The supernatant was discarded and RNA pellet was rinsed with 70% DEPC-ddH2O. RNA was air dried and dissolved in adequate volume of DEPC-ddH2O. The quality and quantity of RNA samples were determined by the agarose electrophoresis and the absorbance at 260 nm, respectively. Real-time quantitative RT-PCR analysis of HSPA5 expression RNA (6.25 μg) was treated with RNase-free DNase (Stratagene) in 25 μl of reaction volume and then 2.5 μg (10 μl) of DNase-treated RNA was reverse-transcribed to cDNA using High Capacity cDNA Reverse Transcription Kit (Cat. No. 4368814, Applied Biosystems) with random primers. Real-time quantitative PCR experiments were performed in the ABI PRISM® 7000 Sequence Detection System (Applied Biosystems). The transcribed cDNA (12.5 ng) were used for real-time PCR with 1× TaqMan® universal PCR Master mix (Cat. No. 4304437, Applied Biosystems) and 1× TaqMan fluorogrnic probes Hs99999174_ml for HSPA5 and 4326321E for HPRT1 (endogenous control) (Applied Biosystems). The PCR condition was: initiation step at 50°C for 2 minutes, denaturation at 95°C for 10 minutes, then 40 cycles of denaturing at 95°C for 15 seconds and combined annealing and extension at 60°C for 1 minute. Experimental samples and no-template controls were all run in duplicate. Results from duplicate reactions were averaged and the fold △CT. difference in the HSPA5 expression was calculated using the formula 2 where △CT. ,. = CT (HPRT1) – CT (HSPA5). For each cell line, three. independent experiments each performed in duplicate were carried out. Statistical analysis of differences between the groups was carried out using one-way analysis of variance. IV. Western blot analysis of APP forms ratio Cell lysate preparation 22.

(32) Lymphoblastoid cells from AD patients as well as age- and gender-match controls were harvested by centrifugation at 1,000 rpm for 5 minutes followed with cold PBS washing twice. The cell pellet was resuspended in 200 μl lysis buffer (50 mM Tris-HCl pH 7.5 - 150 mM NaCl - 2 mM EDTA pH 8.0 - 2% Triton-X100 - 0.2% NP-40) containing the protease inhibitor mixture. Cells were kept on ice and sonicated until the cell suspensions were clear. Then the cell suspension was incubated on ice for 20 minutes. Protein extracts were centrifuged at maximum speed at 4°C for 30 minutes and the supernatant was transferred to a fresh tube. The protein concentration was determined using the Bio-Rad Protein Assay (Cat. No. 500-0006, Bio-Rad). Western blotting and statistical analysis Equal amount of cell lysates were denatured in 1× sample buffer (50 mM Tris pH 6.8 - 2% SDS - 10% Glycerol - 2.5% -mercaptoethanol 0.005% bromophenolblue) at 95°C for 10 minutes and separated on 8% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and electrophoretic transferred to nitrocellulose membrane (Cat. No. EI9051, Invitrogen) in transfer buffer (25 mM Tris - 0.2 M glycin - 20% methanol) at 45 V for 1 hour. The membrane was blocked in PBS containing 10% non-fat milk for at least 1 hour at room temperature or overnight at 4°C. The membrane was rinsed with washing buffer (10 mM Tris pH 8.0 - 0.05% tween-20) and incubated with primary antidody 22C11 (mouse monoclonal antibody, 1:250 dilution, Cat. No. MAB348, Chemicon) at room temperature for at least 1 hour. The membrane was washed three times with washing buffer for 15 minutes each. Then the membrane was incubated with horse-radish peroxidase (HRP)-conjugated goat anti-mouse IgG antibody (1:10,000 dilution, Jackson ImmunoResearch) for 1 hour and washed three times with 23.

(33) washing buffer for 15 minutes each. The immune complexes were detected with enhanced-chemiluminesent substrate (Cat. No. WBKLS0500, Millipore) by ImageReader LAS-3000 software (LAS-3000, Fujifilm). Protein bands were quantified using Alpha1220 densitometry software and expressed as the Integrated Density Values (IDV). The APP isoform values were calculated and tested by two-tailed Student’s t test for significance. V. Western blot analysis of oxidation-sensitive protein Preparation of DNPH-derivatized proteins Denature 5 μl (30 μg) protein from each lymphoblastoid cell sample by adding 5 μl 12% SDS for a final concentration of 6% SDS. Protein denatured in 6% SDS was then mixed with an equal volume (10 μl) of the 1X DNPH (2,4-dinitrophenylhydrazine) solution (Cat. No. 90448, OxyBlotTM Protein Oxidation Detection Kit, S7150, Chemicon) and incubated at room temperature for 15 minutes. The solution was neutralized by adding 7.5 μl neutralization solution (Cat. No. 90449, OxyBlotTM Protein Oxidation Detection Kit, S7150, Chemicon) and prepared for loading onto SDS-polyacrylamide gel by adding 2-mercaptoethanol (reducing reagent) to the sample mixture to achieve a final concentration of 0.74 M solution. Carbonyl immunreactivity assay and statistical analysis The carbonyl groups in protein side chains were derivatized to 2,4-dinitrophenylhydrazone (DNP-hydrazone) by reaction with DNPH. The DNP-derivatized protein samples were separated on 10% SDS-PAGE and electrophoretic transferred to nitrocellulose membrane as described. The membrane was blocked in PBS-T containing 1% bovine serum albumin (BSA, Lot. No. 61806, Bionovas) for at least 1 hour at room temperature or overnight at 4°C. The membrane was rinsed with washing buffer (1X PBS 24.

(34) 0.05% tween-20) and incubated with primary anti-DNP antibody (rabbit monoclonal antibody, 1:150 dilution, Cat. No. 90451, OxyBlotTM Protein Oxidation Detection Kit, S7150, Chemicon) at room temperature for at least 1 hour. The membrane was washed three times with PBS-T washing buffer for 15 minutes per wash. Then the membrane was incubated with horse-radish peroxidase (HRP)-conjugated goat anti-rabbit IgG antibody (1:300 dilution, Cat. No. 90452, OxyBlotTM Protein Oxidation Detection Kit, S7150, Chemicon) for 1 hour and washed three times with PBS-T washing buffer for 15 minutes per wash. Bands containing DNP proteins were visualized by enhanced-chemiluminesent substrate and quantified as described. After removal of the membrane-attached antibodies by stripping buffer (0.4 M glycine - 0.1% SDS - 1% Tween 20 - pH 2.2), the membrane was ready for blocking stage again. After blocking, the membrane was incubated with primary anti-actin antibody (mouse monoclonal antibody, 1:10000 dilution) for 1 hour, secondary HRP-conjugated goat anti-mouse IgG antibody (1:10000 dilution) for another 1 hour. Actin bands were visualized as described and used as loading controls to normalize the images of the carbonyl protein bands. Alternatively, the membrane after electrophoretic transfer was first stained with Ponceau S solution (0.1% (w/v) Ponceau S 5% (v/v) acetic acid) for total proteins as loading controls. Incubate the membrane for 30~60 minutes in Ponceau S solution with gentle shaking. Rinse the membrane in distilled water until the background is clean and the stained bands were visulized as described. The membrane was ready for blocking stage after the stain was removed from the protein bands by continued washing. Blocking the mambrane before proceeding to the antibody incubation as described above. Finally, the normalized carbonyl values of oxidative proteins were analyzed statistically by two-tailed Student’s t test for significance. 25.

(35) Results I. Case-control study to identify genetic risk markers for dementia For the case-control study, genotype distributions in AD and controls did not deviate significantly from Hardy–Weinberg equilibrium for all the polymorphisms examined (data not shown). APOE, ACE and KLK1 gene polymorphisms and the risk of dementia The genotype of the APOE polymorphism was analyzed using PCR-RFLP method. HhaI restriction analysis of the PCR products produced 144 + 96, 144 + 48, and 72 + 48 bp fragements for ε2, ε3 and ε4 allele, respectively (Table 1, Figure 1). The genotype and allele distributions of the APOE polymorphism in patients and controls are displayed in Table 2. The association analysis of APOE with AD and VaD is displayed in Table 3. The risk of AD was significantly increased for individuals with APOE ε3ε4 genotype (odds ratio 2.73, 95% CI 1.61-4.72, P < 0.001) and APOE ε4 allele (odds ratio 3.73, 95% CI 2.38-5.98, P < 0.001) (Table 2, 3). The analysis of the ACE -240 A/T polymorphism was also carried out using PCR-RFLP method. XbaI restriction analysis of the PCR products could obtain 137 bp fragement (A allele) versus 114 bp fragement (T allele) (Table 1, Figure 2A). The analysis of the ACE Alu I/D polymorphism was performed using agarose gel electrophoresis of PCR products: 490 bp (I allele) versus 193 bp (D allele) (Table 1, Figure 2B). The genotype, allele and haplotype distributions of the ACE polymorphisms in patients and controls are displayed in Table 4. The association analysis of ACE with AD and VaD is displayed in Table 5. There was no statistically significant difference between patients and controls for the ACE -240 A/T polymorphism. However, the risk of both AD and VaD were significant for people with ACE DD genotype (odds ratio 4.29 ~ 4.96), D allele (odds ratio 26.

(36) 1.90 ~ 1.95), or T-D haplotype (odds ratio 2.91 ~ 3.18) (Table 5, P = 0.000 ~ 0.005). The analysis of KLK1 -130 GN polymorphism was first carried out by electroforesis of fluorescenced PCR products (Table 1, Figure 3A). The resulting data were confirmed with electrophoresis of denatured PCR products for SSCP/heteroduplex analyses (Figure 3B). The genotype and allele distributions of the KLK1 polymorphism in patients and controls are displayed in Table 6. There was no statistically significant difference between patients and controls. IL-1α and IL-1β gene promoter polymorphism and the risk of dementia Both IL-1α -889 C/T and IL-1β -511 C/T polymorphisms were analyzed using PCR-RFLP method. NcoI restriction analysis of the IL-1α -889 C/T PCR products could obtain 79 bp fragement (C allele) versus 99 bp fragement (T allele) (Table 1, Figure 4A). AvaI restriction analysis of the IL-1β -511 C/T PCR product could obtain 190 and 115 bp fragements (C allele) versus 305 bp fragement (T allele) (Table 1, Figure 4B). The genotype/allele distributions of the IL-1α -889 C/T and IL-1β -511 C/T polymorphisms in patients and controls are displayed in Table 7 and Table 8. There was no significant difference in genotype distribution between AD or VaD patients and controls for the two SNPs examined. Since age is a major risk factor for dementia, the studied groups were further analyzed for those over 70 years of age at study recruitment. The IL-1α -889 CT genotype frequency was notably lower in the VaD patients than in the controls in this subgroup aged over 70 years (Table 7, 9.1 versus 22.9%, P = 0.036). When the allele distributions of the IL-1α -889 and IL-1β -511 SNPs were analyzed, significantly lower frequency of the IL-1α -889 T allele in this VaD subgroup than that of the age-matched controls was observed (Table 7, 4.5 versus 13.3%, P = 0.012). Analysis of pairwise IL-1α -889 and IL-1β 27.

(37) -511 haplotype distribution showed random association between IL-1α -889 and IL-1β -511 sites (P = 0.819). However, T-carrying genotypes in both sites were also notably lower, although not statistically significant, in this VaD patient subgroup than in the age-matched controls (Table 9, 7.6 versus 20.0%, P = 0.043). Odds ratios of the at-risk genotype and pairwise genotypes were calculated by comparing each value to the common CC genotype. As shown in Table 10, the IL-1α -889 CT genotype alone showed an odds ratio of 0.34, 95% CI: 0.12-0.83 (P = 0.026) for VaD among individuals over 70 years of age. The combined IL-1α -889 and IL-1β -511 T-carrying genotypes showed an odds ratio of 0.26, 95% CI: 0.08-0.79 (P = 0.024) for VaD among individuals over 70 years. When the APOE genotype was evaluated in combination with the IL-1α -889 genotype, APOE non-ε4-carrying genotypes also strengthened the negative association of the IL-1α -889 CT genotype with VaD (Table 10, odds ratio: 0.23; 95% CI: 0.07-0.65, P = 0.011). No significant difference was found for both the AD (Table 10, P = 0.036) and VaD subgroups (Table 10, P = 0.671) when sex dependency was examined. HSPA5 gene promoter polymorphism and the risk of dementia HSPA5 -415 G/A and -370 C/T polymorphisms were analyzed using PCR-RFLP method. XmnI restriction analysis of the HSPA5 -415 G/A PCR product could obtain 212 and 92 bp fragements (G allele) versus 304 bp fragement (A allele). BstYI restriction analysis of the HSPA5 -370 C/T PCR product could obtain 206 and 98 bp fragements (C allele) versus 174 and 98 bp fragements (T allele) (Table 1, Figure 5A). The analysis of HSPA5 -180 del/G polymorphism was performed by electrophoresis of fluorescenced PCR products (Table 1, Figure 5B). The genotyping data were confirmed with electrophoresis of denatured PCR products for SSCP analyses (Figure 5C). The SNPSpD method was employed for correction of multiple SNP 28.

(38) testing. SNPSpD output of three λs was shown in Table 11. As described by Cheverud (2001), high correlation among variables leads to high λs. In this case, the first λ (1.59) is less than 2 (the number of variables in the correlation matrix), suggesting that not all variables are completely correlated. The magnitude of pair-wise LD was quantified by the metrics D’ and Δ2. The D’ coefficient of -415 G/A and -180 del/G was equal to 1 (D’ = 1.0), strongly suggesting that there has been no recombination in the region over time, and a very strong LD was observed between -415 G/A and -180 del/G sites (Δ2 = 1.0). In fact, SNPs -415 G/A and -180 del/G were completely linked in our study groups. The genotype, allele and haplotype frequency distributions of the HSPA5 -415 G/A, -370 C/T and -180 del/G polymorphisms in patients and controls are displayed in Table 12. The overall -415 G/A (also -180 del/G) genotype distribution was significantly different between the AD cases and controls (P = 0.021), with -415 AA (also -180 GG) genotype being less frequent among AD cases (Table 12, 5.5% versus 12.2%, P = 0.014). The allele frequency distribution at -415 (also -180) site was also significantly different between the AD cases and controls, with -415 A (also -180 G) allele being less frequent among AD cases (Table 16, 24.3% versus 31.9%, P = 0.027). The haplotypes of the HSPA5 gene were estimated using the PHASE 2.1 program. As summarized in Table 12, the distribution of haplotype between AD patients and controls was significantly different (P = 0.027), with A–C–G haplotype (-415,-370 and -180 sites) being less frequent among AD cases (24.3% versus 31.9%, P = 0.027). Odds ratios of the at-risk genotype, allele and haplotype were calculated by comparing each value to the common genotype, allele or haplotype. As shown in Table 12, the -415 AA (also -180 GG) genotype, -415 A (also -180 G) allele and -415-370-180 A–C–G haplotype showed odds ratios of 0.38 (95% CI 29.

(39) 0.18-0.75, P = 0.007), 0.69 (95% CI 0.51-0.91, P = 0.009) and 0.71 (95% CI 0.52-0.97, P = 0.029), respectively, for AD. II. ACE promoter functional study ACE-firefly/TK-Renilla dual luciferase reporter constructs and assay Previously, the Alu D allele was found to be associated with a higher ACE activity (Rigat et al., 1990) and there was a complete or strong non-random association for the polymorphic alleles at -5491, -5466, -3892, -3692, -240, and Alu D allele (Keavney et al., 1998; Liu et al., 2004). In this study, the effect of ACE -240 A/T on gene expression was determined by the promoter functional assay. The ACE promoter fragments were inserted upstream of the firefly luciferase reporter gene at the EcoRI site of the dual luciferase reporter vector (侯, 2005) to generate ACE-firefly/TK-Renilla dual luciferase reporter plasmids (Figure 6A). Dual luciferase reporter constructs containing the -240 A or T allele displayed similar transcriptional activity in both IMR-32 and HEK-293 cells (Figure 6B). ACE-firefly luciferase - Alu I/D reporter constructs and assay The ACE Alu I and D allele-containing fragments were inserted downstream of the firefly luciferase reporter gene at the AccI site of the pGL3-ACE plasmid ( 劉 , 2003) to generate pGL3-ACE-A-I and pGL3-ACE-T-D plasmids (Figure 7A). As shown in Figure 7B, EcoRI restriction mapping produced fragments with predicted sizes: 4.9 kb for pGL3-basic (lane 1), 4.9 and 1.2 kb for pGL3-ACE-A (lane 2) and pGL3-ACE-T (lane 3), 2.9, 2.0, 1.2 and 0.5 kb for pGL3-ACE-A-I (lanes 4 and 5), 2.9, 2.0, 1.2 and 0.2 kb for pGL3-ACE-T-D (lanes 6 and 7). The orientation of the Alu I and D was confirmed by AatII-AccI restriction analysis, with forward insertion for pGL3-ACE-A-If (lane 8, 4.1 and 2.5 kb) 30.

(40) and pGL3-ACE-T-Df (lane 9, 3.8 and 2.5 kb), and the reverse insertion for pGL3-ACE-A-Ir (lane 10, 3.6 and 3.0 kb) and pGL3-ACE-T-Dr (lane 11, 3.6 and 2.7 kb). The effect of ACE Alu I/D on gene expression was examined by the promoter functional assay. As shown in Figure 7C, reporter constructs containing the -240 A + Alu I or -240 T + Alu D did not display significantly different activity in both IMR-32 and HEK-293 cells. III. ER stress and HSPA5 expression study To test the relevance of the HSPA5 promoter SNP and ER stress response, HSPA5 mRNA expressions in lymphoblastoid cell lines from three patients with -415 GG/-180 deldel, four patients with -415 GA/-180 delG and one patient with -415 AA/-180 GG were examined. While no significant difference was observed between the median (range) fold for -415 GG/-180 deldel and -415 GA/-180 delG: 5.91 (4.78 to 6.59) and 5.84 (3.47 to 7.00), the fold for -415 AA/-180 GG was notably low (3.42) (Figure 8A). The response to ER stress in lymphoblastoid cells carrying -415 AA/-180 GG and one representative lymphoblastoid cell line with -415 GG/-180 deldel was further examined by quantitative RT-PCR. The HSPA5 mRNA of the cells carrying -415 AA/-180 GG treated with 300 nM thapsigargin was significantly higher than that of the untreated -415 AA/-180 GG cells (Figure 8B, P = 0.001) or the treated -415 GG/-180 deldel cells (P = 0.004). In the untreated situation, the HSPA5 mRNA level in -415 AA/-180 GG cells was significantly lower than that of -415 GG/-180 deldel cells (Figure 8B, P = 0.009). IV. APP forms ratio study Lymphoblastoid cells from AD patients (P1 ~ P8) and age- and gender-matched controls (C1, C4, C6 and C7) (Table 14) were used for APP forms ratio study. Total proteins from each lymphoblastoid cell line were 31.

(41) separated on 8% SDS-PAGE and processed for Western blot analysis using monoclonal antibody 22C11 raised against the N-terminal domain of APP, thereby all APP forms present in the samples should be recognized. A representative Western blot analysis was shown in Figure 9A. The optical densities of the bands at 110, 120 and 130 kDa were measured by image analysis. While 130-kDa band can only be detected in P5 and C7, the 120-, 110- and 106-kDa 22C11 immunoreactive bands are clearly visible in all lymphoblastoid lysates examined. The ratio between the upper (130 and 120 kDa) and lower (110 kDa) APP forms was measured. As shown in Figure 9B, mean (range) APP forms ratios for AD patients (n = 8) and controls (n = 4) were 0.77 (0.44 ~ 1.11) and 0.95 (0.25 ~ 1.65), respectively. No significant different was observed (P = 0.66). Mean APP forms ratios for ε4-carriers (n = 4) and non-ε4 carriers (n = 8) were 0.89 (0.38 ~ 1.40) and 0.71 (0.33 ~ 1.09), respectively. Again no significant different was observed (P = 0.50). V. Oxidation-modified protein study The oxidation-modified proteins in lymphoblastoid lysates from AD patients (P1 ~ P8) and age- and gender-matched controls (C1, C4, C6 and C7) were examined by immunostained with anti-DNP antibody (Figure 10A). The image of the oxidized protein bands at molecular weight 43 kDa for β-actin and 51 kDa for unknown protein were quantitated with β-actin stain or Ponceau S stain as loading controls. As shown in Figure 10B, when the oxidative level of each lane was normalized with the average level of all controls in each blot, 51-kDa band median (range) for patients and controls were 1.02 (0.46 ~ 1.58) and 1.00 (0.86 ~ 1.14), respectively. The difference in the oxidative level of 51-kDa band between patients and controls was not significant (P = 0.93).. 32.

(42) Discussion APOE, ACE and KLK1 gene polymorphisms and the risk of dementia The etiology and pathogenesis of AD and VaD are largely unknown. In our AD group, the APOE ε4 frequency is 27.2% (Table 2) and the odds ratio of carrying at least one allele of APOE ε4 is 3.73 (Table 3), which is parallel to those reported in Taiwanese (Huang et al., 2002; Lai et al., 2003). Our study supports previous results in the medical literatures that demonstrated a strong association between APOE ε4 and AD with diverse ethnic backgrounds (Corder et al., 1993; Duara et al., 1996; Farrer et al., 1997). The APOE ε4 allele was identified as a genetic risk factor for VaD among the Caucasians (Slooter et al., 1997; Kalman et al., 1998; Marin et al., 1998), but not among the Japanese (Kawamata et al., 1994), Latin Americans (Molero et al., 2001), Koreans (Bang et al., 2003) and Taiwanese populations (Huang et al., 2002; Lai et al., 2003). Our study did not show an association between APOE ε4 and VaD. The negative results may be explained by the fact that the gene(s) responsible for the disease may be different in different populations and that the epistatic gene-gene interactions or specific multi-loci haplotype may be involved in determining the association. Tissue. kallikrein. (kallikrein-kinin. system). and. angiotensin. (renin-angiotensin system) regulate blood circulation. Compared to the initially found A, B, H, and K alleles in a small Asian populations (Song et al., 1997), two additional rare alleles (M and I) were seen in our study (Table 6). Low urinary kallikrein excretion was associated with hypertension and renal disease (Margolis et al., 1971) and the association of the -130 GN polymorphism with renal dysfunction and hypertension has. 33.

(43) been shown (Yu et al., 2002; Lee-Chen et al., 2004). Our study did not show an association of the KLK1 -130 GN polymorphism with either AD or VaD. Both the D and I alleles of ACE I/D polymorphism have been inconsistently associated with AD (Kehoe et al., 1999; Farrer et al., 2000; Narain et al., 2000; Yang et al., 2000; Richard et al., 2001; Cheng et al., 2002; Elkins et al., 2004). Different from most of the previous studies focusing exclusively upon a single ACE I/D polymorphism, we used a haplotype-based approach for ACE polymorphisms that comprised the Alu I/D allele in intron 16 and -240 A/T allele in the promoter region. Our results are compatible with the studies reported by Farrer and colleagues (2000) and Richard and colleagues (2001) which showed that the ACE D allele increased the risk for AD. However, our study differs from those of the meta-analysis (Kehoe et al., 1999; Elkins et al., 2004) and previous studies on Taiwanese and Chinese populations (Yang et al., 2000; Cheng et al., 2002), in which the I allele was shown to be associated with AD. The heterogeneity of risk for AD associated with the I allele or D alleles appeared to be related, at least in part, to the sampling bias, the effects of race and age of populations (Barley et al., 1994; Elkins et al., 2004). Previous studies have found that the ACE D allele is associated with essential hypertension (Zee et al., 1992) and coronary heart disease (Cambien et al., 1992). The differential results of association studies of the Alu I/D allele with AD may occur, if the cohorts recruit different proportions of individuals with hypertension or coronary heart disease who tend to carry the D allele. The ACE Alu D allele was shown to be a genetic risk factor for myocardial infarcts, cardio-vascular disease, coronary heart disease, hypertension, diabetes mellitus, and stroke, all of which are risk factors for VaD (Cambien et al., 1992; Zee et al., 1992; Morris et al., 1994; Sharma et al., 1994; Markus et al., 1995; Catto et al., 1996; Kario et al., 1996; Hsieh et 34.