A Search For Circulating Protein Markers in Neurological Disease Using 2-Dimensional SDS PAGE and Quantitative Densitometry

Abstract

Alzheimer's disease and multiple sclerosis are two prevalent neurological diseases that have no laboratory based tests to confirm their clinical diagnosis. In this study, protein abnormalities were searched for in the plasma and platelets of patients with Alzheimer's disease, and in the plasma and CSF of patients with multiple sclerosis. The aim of the search was to detect protein abnormalities that had the potential to be used as marker proteins for these diseases. Such marker proteins could be used as the basis of laboratory tests. The procedures used to perform the search were 2-dimensional SDS PAGE and quantitative densitometry. Before performing the search a study was made to find the optimum method for obtaining and analysing protein abundance data. It was found that quantitating only discrete protein regions, and converting the abundance data into compositions yielded the greatest reproducibility. Logratio transforms of the compositional data were analysed statistically to detect quantitative differences between control and affected individuals. In the analysis of plasma proteins, many proteins were seen to alter in relative abundance in multiple sclerosis and Alzheimer's disease. Almost all of the protein changes observed involved acute phase reactants, and so were of little diagnostic utility. In Alzheimer's disease, orosomucoid, haptoglobin α, haptoglobin β, apo D, RBP, and a group of unidentified proteins were significantly different (P < 0.01) between the Alzheimer's patients and a group of age and sex matched controls. In multiple sclerosis, plasma levels of orosomucoid, haptoglobin α, haptoglobin β, apo A1 and apo D, Ig G, Ig J, and a group of unidentified proteins were found to be significantly different (P <0-01) between a group of multiple sclerosis patients and their age and sex matched controls. Differing plasma levels of various acute phase reactants were seen for both diseases. It is possible that these differences are great enough to support clinical diagnosis. Changes in the acute phase reactant profile over time may also be useful in prognostic assessment. A larger study would be required before this could be assessed. A qualitative protein abnormality was also detected in a group of females affected with multiple sclerosis. This protein was isolated and a partial amino acid sequence was obtained for it. Based on sequence homology, the protein was identified as haptoglobin α. However, as the haptoglobins are a class of acute phase reactants, this protein would be of limited diagnostic utility. Several quantitative differences were also detected between the CSF proteins of multiple sclerosis patients and their controls. As in the plasma studies, all of the protein changes detected were acute phase reactants, and so were of limited diagnostic utility. The proteins significantly different (P < 0.05) between the controls and multiple sclerosis patients were Apo A1, haptoglobin β, fibrinogen β, α1 AT, and actin. A study was also made to see if any disease related pattern in the distribution of proteins in the plasma and CSF could be detected. Such a pattern could suggest differences in protein processing in the disease state, or may indicate a breach in the blood-brain barrier or choriod plexus. Two separate studies were done. In the first a search was made for proteins in the plasma and CSF that appeared to correlate in terms of relative abundance. For this study, correlation analysis was performed on the log10 ratio values of all of the plasma and CSF proteins. Those protein combinations found to correlate significantly ( P<0.05) were subjected to further analysis, where compositional data for plasma and CSF proteins for every individual were expressed as a series of CSF:Plasma ratios. These values were analysed statistically to detect any disease related changes. There were correlations between nineteen potential multiple sclerosis related proteins, although almost all involved acute phase reactants. Most of the correlations were positive; that is, increasing CSF concentration was associated with increasing plasma concentration. Based on a knowledge of the biochemistry of the proteins studied, no obvious pattern to the correlations was seen. The complexity of the correlations may reflect the complexity of the acute phase response to chronic inflammation. In the second study, proteins that could be identified in both the plasma and CSF were examined. Eleven proteins were studied. The proteins were orosomucoid, ceruloplasmin, α-1 antichymotrypsin, α2 HS glycoprotein, haptoglobin β(2), haptoglobin α(3), Apo D, and Apo A1. No pattern emerged between the relative plasma and CSF concentrations of these proteins. The observations did suggest a complex pattern of protein synthesis in response to multiple sclerosis. The pattern may have emerged as the result of protein movement between the CSF and plasma and/or co-regulated hepatic and CNS protein synthesis. A quantitative study of platelet proteins in Alzheimer's disease was also carried out, where platelets prepared in the presence of a protease inhibitor cocktail (1 mM phenylmethyl sulphonylfluoride, 100 µM N-ethylmaleimide, 10 mM EDTA, 100 µM iodoacetamide, I µM pepstatin, and 150 Kallikrein inhibitor units/ml of aprotinin) were compared to platelets prepared in the absence of the inhibitor cocktail. By comparing platelets prepared in the presence of protease inhibitors for controls and Alzheimer's patients, an assessment was made of changes in protein expression and/or abundance in Alzheimer's disease. An assessment of the effect of proteolysis in disease was made by comparing platelet proteins prepared in the absence of protease inhibitors from Alzheimer's patients and controls, and also by comparing platelets prepared in the absence of protease inhibitors to those prepared in the presence of protease inhibitors for both controls and patients. Proteins that were seen to be proteolytically processed differently, but which were not observed to alter in abundance between the platelets of patients and controls prepared in the presence of protease inhibitors, were considered more likely to represent disease related proteolytic events, as altered rates of expression could be ruled out. The major assumption made was that proteolysis ums inhibited by the protease inhibitor cocktail used. When platelet proteins prepared in the absence of protease inhibitors were examined, several significant differences were observed. The proteins significantly different (P <0.05) between Alzheimer's patients and controls were ER-60 protease inhibitor, serotransferrin, and three unidentified proteins. When platelet proteins prepared in the presence of protease inhibitors were examined, several differences were also seen. Those proteins found to differ significantly (P <0.05) between controls and Alzheimer's patients were ER-60 protease inhibitor and five unidentified proteins. It is difficult to assess the significance of the other platelet protein changes observed in this study due to the limited number of platelet proteins that are currently identified in the Swiss-Prot database. The results obtained however, do add further support to the suggestion of Inestros et al. (1993), that platelets may provide a systemic marker for Alzheimer's disease. Extension of the results obtained in this project could yield useful information regarding the aetiology and pathogenesis of both disorders. Further study may also yield information that may be used in the detection and diagnosis of the diseases.