Development of a new biochip array for APOE4 classification from plasma samples using immunoassay-based methods.

BACKGROUND: Apolipoprotein E (APOE) is a key player in lipid transport and metabolism and exists in three common isoforms: APOE2, APOE3 and APOE4. The presence of the E4 allelic variant is recognized as a major genetic risk factor for dementia and other chronic (neuro)degenerative diseases. The availability of a validated assay for rapid and reliable APOE4 classification is therefore advantageous. METHODS: Biochip array technology (BAT) was successfully applied to identify directly the APOE4 status from plasma within 3 h, through simultaneous immunoassay-based detection of both specific APOE4 and total APOE levels. RESULTS: Samples (n=432) were first genotyped by polymerase chain reaction (PCR), and thereafter, using BAT, the corresponding plasma was identified as null, heterozygous or homozygous for the E4 allele by calculating the ratio of APOE4 to total APOE protein. Two centers based in Austria and Ireland correctly classified 170 and 262 samples, respectively, and achieved 100% sensitivity and specificity. CONCLUSIONS: This chemiluminescent biochip-based sandwich immunoarray provides a novel platform to detect rapidly and accurately an individual's APOE4 status directly from plasma. The E4 genotype of individuals has been shown previously to affect presymptomatic risk, prognosis and treatment response for a variety of diseases, including Alzheimer's disease. The biochip's potential for being incorporated in quantitative protein biomarker arrays capable of analyzing disease stages makes it a superior alternative to PCR-based APOE genotyping and may deliver additional protein-specific information on a variety of diseases in the future.

A platelet protein biochip rapidly detects an Alzheimer's disease-specific phenotype.

Alzheimer's disease (AD), a multifactorial neurodegenerative condition caused by genetic and environmental factors, is diagnosed using neuropsychological tests and brain imaging; molecular diagnostics are not routinely applied. Studies have identified AD-specific cerebrospinal fluid (CSF) biomarkers but sample collection requires invasive lumbar puncture. To identify AD-modulated proteins in easily accessible blood platelets, which share biochemical signatures with neurons, we compared platelet lysates from 62 AD, 24 amnestic mild cognitive impairment (aMCI), 13 vascular dementia (VaD), and 12 Parkinson's disease (PD) patients with those of 112 matched controls by fluorescence two-dimensional differential gel electrophoresis in independent discovery and verification sets. The optimal sum score of four mass spectrometry (MS)-identified proteins yielded a sensitivity of 94 % and a specificity of 89 % (AUC = 0.969, 95 % CI = 0.944-0.994) to differentiate AD patients from healthy controls. To bridge the gap between bench and bedside, we developed a high-throughput multiplex protein biochip with great potential for routine AD screening. For convenience and speed of application, this array combines loading control-assisted protein quantification of monoamine oxidase B and tropomyosin 1 with protein-based genotyping for single nucleotide polymorphisms (SNPs) in the apolipoprotein E and glutathione S-transferase omega 1 genes. Based on minimally invasive blood drawing, this innovative protein biochip enables identification of AD patients with an accuracy of 92 % in a single analytical step in less than 4 h.

Platelets, a reliable source for peripheral Alzheimer's disease biomarkers?

Peripheral biomarkers play an indispensable role in quick and reliable diagnoses of any kind of disease. With the population ageing, the number of people suffering from age-related diseases is expected to rise dramatically over the coming decades. In particular, all types of cognitive deficits, such as Alzheimer's disease, will increase. Alzheimer's disease is characterised mainly by coexistence of amyloid plaques and neurofibrillary tangles in brain. Reliable identification of such molecular characteristics antemortem, however, is problematic due to restricted availability of appropriate sample material and definitive diagnosis is only possible postmortem. Currently, the best molecular biomarkers available for antemortem diagnosis originate from cerebrospinal fluid. Though, this is not convenient for routine diagnosis because of the required invasive lumbar puncture. As a consequence, there is a growing demand for additional peripheral biomarkers in a more readily accessible sample material. Blood platelets, due to shared biochemical properties with neurons, can constitute an attractive alternative as discussed here. This review summarises potential platelet Alzheimer's disease biomarkers, their role, implication, and alteration in the disease. For easy comparison of their performance, the Hedge effect size was calculated whenever data were available.

Identification and validation of platelet low biological variation proteins, superior to GAPDH, actin and tubulin, as tools in clinical proteomics.

Accurate biomarker quantification requires carefully chosen normalisation procedures. When single proteins are used as loading controls (LCs), it is crucial that their expressional stability must be known. Platelets are an important biomarker source, especially for neurological diseases. We performed a systematical analysis of the platelet proteome to identify proteins suitable as LCs, using the 2-D DIGE system. We first screened a healthy population (n=137), aged between 18 and 104years, to find proteins with small coefficients of total variation (CVtot), herein termed low biological variation proteins (LBVP). Thereafter, expressional stability was verified in 101 patients suffering from Alzheimer's- (AD), Parkinson's- disease, vascular dementia or schizophrenia. Interestingly, traditional LCs such as tubulin beta-1 and GAPDH, were not found amongst LBVP. The least variable protein, calculated over all 238 individuals, was 14-3-3 gamma, with a CVtot of 9.3%, showing no gender, age or disease dependency. The normalisation capability of 14-3-3 gamma was superior to traditional LC in quantifying Western blot signals of the platelet AD-biomarker Monoamine Oxidase B of patient versus controls. Similar results were obtained with HepG2 cells, treated in vitro with DNA-methyltransferase inhibitor 5-aza-2'deoxicytidine. Finally, we provide a list of alternative normalisation candidates for accurate biomarker quantification. BIOLOGICAL SIGNIFICANCE: This paper suggests a considerable list of platelet proteins with a lower biological variation than well known "housekeeping" proteins like GAPDH and tubulin. Spot abundances of found proteins are middle ranged and unaffected by gender, age and certain diseases. Hence, listed proteins might be valuable normalisation candidates used additionally or alternatively. Platelet's least variable protein 14-3-3 gamma is validated as normalisation protein in platelet biomarker quantification. Furthermore 14-3-3 gamma is demonstrated to be also stable expressed by in HepG2, cells others than platelets, when treated by DNA methylation inhibitor.

A combined proteomic and genetic analysis of the highly variable platelet proteome: from plasmatic proteins and SNPs.

High biological variation in protein expression represents a major challenge in clinical proteomics. In a study based on 2D-DIGE, we found that the standardised abundance of only a few proteins varied by more than 50%. While some of the highest variable proteins in platelets of 52 healthy elderly were of plasmatic origin, such as albumin or haptoglobin, absence of several other high-abundant plasma proteins strongly suggests that plasma-derived proteins represent an integral part of the platelet proteome. Amongst the highly variable platelet-derived proteins, two spots were both identified as GSTO1 and assigned to either the wild-type or mutant isoform of SNP A140D. Remarkably, when the spots were considered within the respective genotype groups, their CV decreased to about the median variation. Albeit 2D-DIGE allowed correct genotyping, two individuals seemed to be GSTO1*A140 deficient. Probing 2D-Western blots with novel mAb, however, detected A140 protein as additional spot at pH 8.1, caused by the SNPs E155del and E208K. In contrast to previous studies, we show that GSTO1 protein is expressed in vivo, despite the deletion E155. Our data indicate that incorporation of exogenous proteins and genetic polymorphisms of endogenous proteins represent the main source of extreme biological variation in the platelet proteome.

Identification of Non-HLA antigens targeted by alloreactive antibodies in patients undergoing chronic hemodialysis.

The only treatment of end-stage renal disease patients undergoing chronic dialysis is kidney transplantation. However, about half of graft recipients encounter organ loss within ten years after renal transplantation. There is emerging evidence that the presence of alloreactive antibodies against non-HLA antigens in the serum of the recipient prior transplantation is associated with higher incidence of chronic rejection. However, the molecular identity of these antigens is largely unknown. To determine the most common non-HLA antigens, we tested lymphocytic extracts from 20 healthy volunteers with sera of 28 patients on the transplantation waiting list by Western blotting. There was a group of five proteins that was recognized by most sera. Using patient's own lymphocytes revealed that autoimmunity plays a minor role in this recognition. Two-dimensional Western blotting experiments followed by mass spectrometry identified the antigens as tubulin beta chain, vimentin, lamin-B1, and Rho GDP-dissociation inhibitor 2. A detailed analysis of vimentin expression revealed that the antigenic 60 kDa isoform is underrepresented in patient's lymphocytes in comparison to those of healthy volunteers. The study revealed that preformed alloreactive antibodies are directed against a small number of specific protein isoforms. Our findings could provide a basis for future improvement of donor-recipient matching.

The role of proteomics in dementia and Alzheimer's disease.

Proteomic analysis enables us to identify dementia-related protein profiles of both genetic and environmental origins. In this review, current proteomics technologies are described including many examples of clinical proteomics studies. Many of these studies present only results of the discovery phase. Progression to the validation phase was achieved by developing more advanced proteomics technologies such as fluorescence two-dimensional differential gel electrophoresis or isobaric tagging for relative and absolute protein quantification. These technologies will lead to the design of several new Alzheimer's disease-related protein panels for the analysis of CSF. On these new panels, established markers such as tau and Abeta42 will be used in combination with novel markers, for example beta-2-microglobulin, brain-derived neurotrophic factor 1 and fragments of cystatin C. However, there are still limitations to using proteomic assays. The preparation of homogeneous sample material is difficult due the complexity of brain tissue. Laser capture microdissection and recently developed more sensitive proteomics methods, for example fluorescence saturation labelling, will overcome these limitations. Combining proteomics with approaches at the level of the genome and transcriptome followed by interpretation by systems biology will soon shed further light on dementia-related pathogenesis.