Final Activity Report Summary - PLATELET-CHIP (Protein Biochip Array Technology for Diagnosis of Alzheimer's disease in Blood Platelets)
A correct clinical diagnosis in the early stage of Alzheimer's disease (AD) is mandatory to have the possibility for an appropriate treatment of this neurodegenerative disorder in future. Therefore, it is very helpful to identify relevant drug targets of the brain also in the blood. The aim of this project was to develop a multiparameter platelet protein chip panel, for the diagnosis of AD out of blood. The interesting aspect of this approach is that platelets can be used as peripheral mirrow for biochemical abnormalities in the brain cells. Thereby, they can indicate AD treatment targets. Since in numerous previous clinical studies many different phenotypic changes were described in platelets of AD patients, our academic working group made a gel-based proteome analysis for the identification of AD-related protein profiles in platelets.
Thereby, 17 different AD-related protein expression changes was identified in the platelet proteome. This was the initial situation of this Marie Curie project to develop a diagnostic multiparameter routine assay for the quantification of these AD biomarkers. After an intensive research and discussion with the assay design department of the diagnostic company Randox about the feasibility of these targets we focussed on following AD-related proteins for the platelet chip.
- Apolipoprotein E4 allele: frequency in AD 60 %, in matched controls 10 %.
wt Glutathione- S transferas omega 1 (GSTO-1) allele: frequency in non-ApoE4 AD 95 %, in matched controls: 44 % monoamine oxidase (Mao-B) protein level increase in AD by 31 %, P<0.001 tropomyosin protein level increase in AD by 25 %, P = 0.002.
Coagulation factor XIIIa protein level increase in AD by 26 %, P = 0.005.
In parallel, the academic group characterised in the platelet proteome database of this study proteins with a very low biological variation (e.g. 14-3-3 gamma, ERK2) which would be a very promising tool as loading control like GPDH for an immunological assay. These proteins should enable a normalisation step on the platelet protein chip and thereby allow a one-step quantification for AD biomarkers in the platelet lysate.
In the last two years, a statistical evaluation of the clinical proteome AD data revealed that the combination of these AD biomarkers by different algorithms increases the area under (AUC) of the ROC curve from the most significant AD marker Mao-B with AUC = 0.837 to: ROC curves (calculated from genotyping and 2D-DIGE): AUC = 0.939 sign. < 0.001 (95 % CI: 0.885 - 0.993).
Haptoglobin, glycerol phosphate dehydrogenase and integrin linked kinase 1 which was also in the production pipeline were stopped after the first year since they could not be verified in the verification phase with new clinical samples in the second year.
Randox protein biochip technology: Therefore, an antibody sandwich-pair is required. The capture antibody is spotted on the biochip surface the detection antibody is HRP-conjugated prior use and serves as target specific secondary antibody. Neat protein is used as calibrator material, lysed gel-filtrated platelets (GFP) are analysed for expressed protein concentration. Calibration curves have to be prepared prior sample analysis and suitable assay ranges for all targets defined accordingly to expected sample concentrations.
One important technical finding in this project was that polyclonal antibodies were rather unsuited for this chip technology. Accordingly, monoclonal sheep antibodies had to be developed against all these targets, except XIIIa and ERK2 which was commercially available. As most of the immunogens are expressed in recombinant (bacterial) systems and monoclonal antibodies raised against them, problem with recognition of human, endogenous target proteins can appear.
All working assays have to be analysed for cross reactivity with related proteins, e.g. isoforms. A new ApoE4 antibody was developed from the Medical University of Vienna. The ROC curve of the developed antibody chip panel Apo E4, XIIIa and GSTO-1 should be 0.896 calculated from the clinical samples of the clinical study.
Thereby, 17 different AD-related protein expression changes was identified in the platelet proteome. This was the initial situation of this Marie Curie project to develop a diagnostic multiparameter routine assay for the quantification of these AD biomarkers. After an intensive research and discussion with the assay design department of the diagnostic company Randox about the feasibility of these targets we focussed on following AD-related proteins for the platelet chip.
- Apolipoprotein E4 allele: frequency in AD 60 %, in matched controls 10 %.
wt Glutathione- S transferas omega 1 (GSTO-1) allele: frequency in non-ApoE4 AD 95 %, in matched controls: 44 % monoamine oxidase (Mao-B) protein level increase in AD by 31 %, P<0.001 tropomyosin protein level increase in AD by 25 %, P = 0.002.
Coagulation factor XIIIa protein level increase in AD by 26 %, P = 0.005.
In parallel, the academic group characterised in the platelet proteome database of this study proteins with a very low biological variation (e.g. 14-3-3 gamma, ERK2) which would be a very promising tool as loading control like GPDH for an immunological assay. These proteins should enable a normalisation step on the platelet protein chip and thereby allow a one-step quantification for AD biomarkers in the platelet lysate.
In the last two years, a statistical evaluation of the clinical proteome AD data revealed that the combination of these AD biomarkers by different algorithms increases the area under (AUC) of the ROC curve from the most significant AD marker Mao-B with AUC = 0.837 to: ROC curves (calculated from genotyping and 2D-DIGE): AUC = 0.939 sign. < 0.001 (95 % CI: 0.885 - 0.993).
Haptoglobin, glycerol phosphate dehydrogenase and integrin linked kinase 1 which was also in the production pipeline were stopped after the first year since they could not be verified in the verification phase with new clinical samples in the second year.
Randox protein biochip technology: Therefore, an antibody sandwich-pair is required. The capture antibody is spotted on the biochip surface the detection antibody is HRP-conjugated prior use and serves as target specific secondary antibody. Neat protein is used as calibrator material, lysed gel-filtrated platelets (GFP) are analysed for expressed protein concentration. Calibration curves have to be prepared prior sample analysis and suitable assay ranges for all targets defined accordingly to expected sample concentrations.
One important technical finding in this project was that polyclonal antibodies were rather unsuited for this chip technology. Accordingly, monoclonal sheep antibodies had to be developed against all these targets, except XIIIa and ERK2 which was commercially available. As most of the immunogens are expressed in recombinant (bacterial) systems and monoclonal antibodies raised against them, problem with recognition of human, endogenous target proteins can appear.
All working assays have to be analysed for cross reactivity with related proteins, e.g. isoforms. A new ApoE4 antibody was developed from the Medical University of Vienna. The ROC curve of the developed antibody chip panel Apo E4, XIIIa and GSTO-1 should be 0.896 calculated from the clinical samples of the clinical study.