Current Alzheimer Research, 2012, 9, 000-000
1
Can Apolipoproteins and Complement Factors be Biomarkers of Alzheimer’s Disease? Pallavi Manrala,#, Pratibha Sharmaa,#, Gururao Hariprasada, Chandralekhab, Manjari Tripathic and Alagiri Srinivasana,* a
Department of Biophysics; bDepartment of Anaesthesiology; cDepartment of Neurology, All India Institute of Medical Sciences, New Delhi, India (# these authors have contributed equally) Abstract: Alzheimer’s disease is the most common cause of dementia in elderly persons. Quick diagnosis of Alzheimer's disease will allow treatments that may help slow its progression. The correlation between cerebrospinal fluid (CSF) parameters and progression of Alzheimer's disease is higher than and independent of other risk factors. We have compared sixteen CSF samples of clinically diagnosed Alzheimer’s disease patients with non demented subjects using proteomics approach. Apolipoprotein E, apolipoprotein J, complement C4b, hemopexin and complement factor B were identified as differentially expressed proteins. Pathway analyses show that these proteins have interacting partners in Alzheimer’s and apoptotic pathways. The possible roles of these proteins in relation to the disease are discussed.
Keywords: Biomarkers, CSF, diagnosis, pathways. INTRODUCTION Alzheimer’s disease (AD) is a progressive and debilitating brain disease. The pathological brain changes occur 20-30 years prior to the onset of clinical symptoms. Therapeutic intervention during the early course of disease may have efficacious outcomes. Considering the intensive ongoing research towards developing disease modifying therapies, there is an urgent need for timely diagnosis [1, 2]. Present-day diagnosis of AD relies on patient history, brain imaging and neuropsychological assessment by neurologists. Diagnosis of probable AD is made when the disease has progressed far enough that dementia sets in [3]. Diagnosis is confirmed by histopathological examination of the brain for amyloid plaques and neurofibrillary tangles in brain autopsy samples. Obtaining a brain biopsy is a highly invasive procedure and cannot be used for routine diagnosis. A simple objective biochemical test which could supplement existing diagnostic tests for diagnosis of AD is highly desired. Pathological changes in brain occurring in the course of AD will result in altered CSF chemistry. AD has multifactorial etiology and a heterogeneous neuropathological presentation. A single biomarker probably cannot reach a desired diagnostic specificity and sensitivity. A panel of proteins showing specific changes in expression pattern will most accurately reflect the presence of disease. Proteomics is an unbiased approach for the simultaneous display of all proteins expressed in a biological sample at a particular time. A number of previous studies have carried out CSF analyses for AD [4, 5] and other neurological disorders [6-8] using various combinations of proteomic tools. In the present study, we have compared CSF samples *Address correspondence to this author at the Department of Biophysics, All India Institute of Medical Sciences, New Delhi-110 608 India; Tel: +91 11 2659 4240; Fax: +91 11 2658 8663; E-mail:
[email protected]
1567-2050/12 $58.00+.00
of AD and non demented elderly patients using twodimensional difference-in-gel electrophoresis and mass spectrometry. This is a first ever proteomic study of Indian AD patients. MATERIALS AND METHODS CSF Collection Ante mortem lumbar CSF samples (~ 4 ml) were obtained from 8 patients diagnosed clinically for Alzheimer’s disease in cognitive decline clinic and admitted in Department of Neurology. These patients were diagnosed using NINCDS-ADRDA criteria. MMSE score and MRI findings were also used to rule out other causes of dementia. Three patients had a MMSE score of 15, two had a score of 20 and the last three had a score of 22. CSF samples were also obtained from 8 gender matched non-demented patients of similar age group undergoing spinal anaesthesia for lower limb surgery. These subjects were cognitively normal with a MMSE score of 30. Informed consent was taken before drawing CSF sample. The study was approved by the institutional ethics committee. Samples were centrifuged at 4˚C at 2000 rpm for 10 min to separate cells. Samples contaminated with blood were discarded. The supernatant was aliquoted (~500 l) in polypropylene tubes before freezing at -80˚C. Removal of High Abundant Proteins High-abundant proteins (albumin, IgG, IgA, alpha-1antitrypsin, transferrin, haptoglobin and fibrinogen) were selectively removed by using multi affinity removal system (MARS Hu7 kit, Agilent Technologies, Santa Clara, CA). Briefly 1 ml CSF sample was concentrated (4x) using centrifugal filter units (5 kDa cut-off) and mixed with buffer A (Equilibration/ Loading buffer) to a final volume of 600 l and filtered through microcentrifuge filter (0.22 m, Agilent © 2012 Bentham Science Publishers
2 Current Alzheimer Research, 2012, Vol. 9, No. 5
Manral et al.
Technologies, Santa Clara, CA). Filtrate was passed through the immunoaffinity column at a flow rate of 200 l/ min. CSF samples devoid of high abundant proteins (flow through and wash) were concentrated using centrifugal filter units. Depleted samples were mixed with four times volume of cold acetone and kept in -20˚C for protein-precipitation. Protein pellets were air dried and dissolved in lysis buffer (8M urea, 2M thiourea, 4% CHAPS). Protein concentration was determined by 2D Quant Kit (GE Healthcare, Piscataway, NJ). Bound proteins were eluted with 2 ml of buffer B (Elution buffer). Column was regenerated with 4 ml of buffer A for the next cycle. Experimental Design Protein samples (25 g) were labelled with 200 pmole of fluorescent cyanine dyes (GE Healthcare, Piscataway, NJ). Internal standard was constituted by pooling equal amount of protein from all 16 CSF samples. 25 g protein of the internal standard mixture was labelled with fluorescent dye Cy2. AD and cognitively normal samples were labelled with Cy3 or Cy5 as indicated in Supplementary (Table 1). Labelling reaction was carried out at 4˚C in dark for 30 min and the reaction was quenched with 10nmole lysine. Each gel was run with three different protein samples labelled with three different dyes. Dye labelling was reversed for some samples to prevent any bias in protein labelling. Table 1.
Characteristics of AD and Normal Subjects
Demographic Variables
AD
Control
Mean age (SD)
62.6(9.5)
60.7(12.2)
Number of subjects (F/M)
8 (3/5)
8 (3/5)
MMSE (Number of subjects)
Disease duration, y
15 (3) 20 (2) 22 (3) 3-4
30 (8)
N/A
AD = Alzheimer’s disease; SD = Standard Deviation; F = female; M = males; MMSE = Mini-mental state Examination; N/A = not applicable; y = year.
Protein Separation by 2D-Electrophoresis Samples labelled with Cy dyes were mixed and reconstituted with rehydration buffer containing 8 M urea, 2 M thiourea, 4 % CHAPS, 0.3 % DTT, 0.002 % bromophenol blue and 0.5 % IPG buffer (pH 3-10 NL). IPG strips (13 cm, pH 3-10 NL; GE Healthcare, Piscataway, NJ) were rehydrated with Cy dye labelled protein mixture for 15h at room temperature. Isoelectric focussing was performed to a total of 26 kVh at 20˚C. The cysteine sulfhydryls were reduced and carbamidomethylated by incubating the focussed strip in 1% DTT for 15 min followed by second incubation with 2.5% iodoacetamide in loading buffer (6 M urea, 50 mM Tris-HCl (pH 8.8), 2% SDS, 30% glycerol, 0.002% bromophenol blue). The IPG strips were loaded onto a homogenous 11% acrylamide gel. SDS PAGE was performed at 15 mA/gel for 30 minutes and then at 30mA/
gel for 3 hours at 20˚C. Eight analytical and one preparative gels were run. Gel Imaging, Spot Detection and Statistical Analysis Gels were scanned for Cy2, Cy3 and Cy5 using Typhoon trio (GE Healthcare, Piscataway, NJ) at 200 m resolution. Imaging for Cy2 at 488-nm laser/ 520-nm band pass (BP) 40; Cy3, 532-nm laser/ 580-nm BP 30 and for Cy5, 633-nm laser/ 670-nm BP 30 was carried out. Three images were obtained from each gel corresponding to cognitively normal, AD and internal standard. The spots were detected and quantified in Differential In-gel Analysis (DIA) mode of DeCyder software 7.0 (GE Healthcare, Piscataway, NJ). The estimated number of spots for the detection procedure was set to 1250. Dust particles and protein streaks were manually removed. The intensity of spots in the Cy3 and Cy5 images were normalised to that of Cy2 image in the same gel. Biological Variation Analysis (BVA) mode in DeCyder software was used to simultaneously match all 24 proteinspot maps from eight gels. Using Cy3:Cy2 and Cy5:Cy2 DIA ratios, average abundance changes and student’s t-test p values were calculated. Only those spots with 1.5 fold difference in volume after normalization between Cy3 and Cy5 and a p-value < 0.05 were defined as spots of interest. Preparative gel was stained with colloidal coomassie blue G250 stain. Spots of interest were then excised manually from preparative gel. In-Gel Digestion All reagents and solvent used for digestion were of LCMS grade (Sigma-Aldrich, St.Louis, MO). Coomassiestained gel-pieces were destained with 100 l of 1:1 solution of acetonitrile (ACN) and 40 mM ammonium bicarbonate (ABc). The gels were then dehydrated with 100 l of 100%ACN for 10 min twice and dried. Mass spectrometry grade trypsin (GE Healthcare, Piscataway, NJ) 500 ng in 25 l of 40 mM ABc containing 10% ACN was added to the gel pieces for 30 min in ice. Additional 25 l of 40 mM ABc containing 10% ACN was added to gel pieces and incubated at 37˚C for 20h. The supernatant was collected. The gel pieces were re-extracted twice with 50l of 50% ACN containing 0.1 % formic acid (FA). Supernatants were pooled and vacuum dried. Dried peptides were resuspended in 10 l of 2% ACN containing 0.1% FA (solvent 1). Protein Identification by Tandem Mass Spectrometry The peptides were analysed in NanoLC-ESI-Q-TOF MS/MS system (Applied Biosystems, Foster City, CA). The LC solvents used for gradient elution were solvent 1 and 98% ACN containing 0.1% FA (solvent 2). LC method of 70 min gradient reaching 100% solvent 2 at a flow rate of 400 nl/min was used for column elution. The run was monitored with the total ion count in the positive ion mode. The product ions were monitored in the m/z range of 400-1600 for MS and 140-1600 for MS/MS. The run was controlled by Analyst QS1.1 software. Electro-spray ionisation was carried out at an ion spray voltage of 2300 V. Mass spectra were acquired for 70 min setting the parameters for information dependent acquisition (IDA) method. The data were used to search sequences in Uniprot database using Protein Pilot 3.0
Markers of Alzheimer’s Disease
Current Alzheimer Research, 2012, Vol. 9, No. 5
3
software (Applied Biosystems, Foster City, CA). Carbamidomethylation of cysteine, partial oxidation of methionine residues and one missed trypsin cleavage were allowed during the search. Sequences with a confidence level of 99% were chosen. Pathway Analysis The proteins having significantly different expression levels were studied for their biological interaction network related to AD and apoptosis. All the information of genes corresponding to these proteins and their related functions were searched from NCBI and BioGRID [9]. The program Cytoscape Ver. 2.7.0 [10] and the plugin MiMI [11] were used for determining the interactions for all proteins. Cytoscape plugin retrieves molecular interactions from Michigan Molecular Interactions (MiMI) database and displays the interaction network with Cytoscape. MiMI gathers and merges data from well-known protein interaction databases including BIND, DIP, HPRD, RefSeq, SwissProt, IPI, KEGG and CCSB-HI1, etc. [12-18]. The plugin also integrates with other NCBI tools for literature information, document summarization and pathway matching. RESULTS CSF proteins from 8 cases of AD were compared with 8 cognitively normal subjects by 2D-DIGE. Demographic details of patients are listed in (Table 1). Experimental design of 2D-DIGE is presented in Supplementary (Table 1). A representative DIGE gel is shown in Fig. (1). 70 spots were found with altered expression between the two groups. The number of spots for MS/MS analysis was selected on the basis of following criteria: (1) the fold change of the spot volume between AD and cognitively normal was 1.5 with a p value
