Identification of Plasma Proteomic Biomarkers in Patients with Mild Cognitive Impairment

Abstract

Plasma proteomic profiling may provide novel biomarkers for the identification of mild cognitive impairment (MCI). The early diagnosis of MCI still remains a challenging task due to its diverse origin. Currently, molecular approaches have been used to identify MCI diversified origin as its onset is governed by a variety of molecular changes. Therefore, we aimed to find out molecular alteration in plasma using proteomics in patients with MCI for early detection of prodromal Alzheimer’s disease (AD). To achieve this, we performed two-dimensional (2-D) gel electrophoresis coupled with MALDI-TOF/MS, which is used to analyze the differentially expressed proteins. In our study, we found three significantly altered proteins. Out of three differentially expressed proteins, one was downregulated and two were upregulated in MCI individuals as compared to control. Further, In silico analysis showed that identified proteins are involved in pathways such as complement and coagulation cascades, platelet activation and AD. STRING interaction network analysis revealed that the majority of proteins including apolipoprotein E (APO-E) have a common association with Transthyretin (TTR) and fibrinogen chain beta (FGB) protein. This suggests that APO-E, TTR and FGB are the key proteins with which other proteins interact to exert other biological functions. Conclusively, these proteins showing differential expression in the plasma might be used as a potent signature in blood for the diagnosis of MCI individuals.

Background

Mild cognitive impairment (MCI) is a transitional stage between healthy ageing and very mild Alzheimer’s disease (AD). The main symptom of MCI is memory impairment which is more than that expected for one’s age. The conversion rate of MCI to AD dementia is 10–15% per year [1, 2]. The preliminary diagnosis of prodromal AD includes a neurological examination, mental status test and brain imaging [1]. The researchers are trying to identify the specific biomarkers, which may provide a conclusive diagnosis for early AD identification. The early detection of AD is difficult due to its long asymptomatic prodromal stages and variability in clinical features, genetic polymorphism and molecular etiologies such as Presenilin-1, Presenilin-2 and Amyloid precursor protein mutation, apolipoprotein (APO-E) and sporadic AD [3]. Blood is a non-invasive, less expensive and time-saving source of biomarker entity [4]. Blood plasma is an easily accessible body fluid, which can be available with minimal discomfort to the patient, and enables sampling of large cohorts and serial sampling. Plasma proteome is found very promising for the identification of biomarkers for a range of diseases, as it is in contact and molecular exchange with every organ and tissue including the brain and hence reflects many physiological and pathophysiological changes [5, 6]. In the present study, we have attempted to identify the plasma proteomic alterations, which could be used for the identification of MCI or early AD.

Materials and Methods

The study included elderly individuals who reported from December 2015 to September 2017 in the outpatient department (OPD) and inpatient department (IPD) of Sir Sundar Lal Hospital, Institute of Medical Sciences, Banaras Hindu University Varanasi. These individuals complained of forgetfulness for the last 2–6 months. Trained doctors and mental health clinicians conducted the examinations. Following the screening of 2000 elderly individuals for cognitive function, 200 were found to have MCI. As only ten amnestic-MCI individuals participated in our follow-up study, we conducted a detailed investigation on them. Age and sex-matched ten healthy control (HC) were taken as control.

MCI Inclusion Criteria

• Memory complaint

• Normal general cognitive function.

• Intact Activities of Daily Living.

• Not demented.

• Have normal

• Able to give informed consent, or assent

• MRI findings must be normal or unremarkable for the age of the patient.

Exclusion Criteria

• Other neuropsychiatric diagnoses

• Major medical illness including potential secondary causes of cognitive decline.

• Disease, combination of disease, or presentation that, in the clinician's judgment, could introduce intolerable variance into the PET brain scan image

• Current substance or alcohol dependence or history of same, and no alcohol or substance abuse within the last eight weeks.

Neuropsychological Assessment. MCI was identified according to the modified Petersen's criteria [7]. Two thousand elderly individuals were screened using HMSE and MOCA examination scale [8, 9]. The score of HMSE/MOCA ranged from 0 to 30, with higher scores indicating better cognition. The cognitive score was categorized into four groups based on the standard classification system: no cognitive impairment (scores 25–30); MCI (scores 18–24); moderate cognitive impairment (scores 10–17) and severe cognitive impairment (scores 0–9). Healthy control individuals scored 25–30 in HMSE and MOCA. Ten individuals with a complaint of forgetfulness were screened by psychological test. They further participated in the follow-up study.

Plasma Processing

Blood collection was done after the neuropsychological screening of MCI individuals and healthy control. Venous blood (2.5 mL) was drawn using a syringe from both MCI individuals and healthy control in ethylene diamine tetra acetic acid (EDTA) anticoagulant tubes. Plasma was separated from the blood by centrifugation at 2500 ×g for 15 min. It was collected in a new vial. Protease inhibitor was added to the plasma and frozen at − 80 °C until utilized for the proteomic study.

Preparation of Plasma Sample for the Proteomic Study

Stored plasma samples were brought to room temperature and protein concentration was determined by Bradford’s method using bovine serum albumin (BSA) as a standard [10]. Blood plasma protein (500 µg) was precipitated in four volumes of acetone at − 20 °C for 2 h. After precipitation, it was centrifuged at 7500 ×g for 10 min. Then the supernatant was removed and air-dried at 37 °C.

Two-dimensional Gel Electrophoresis

Two-dimensional gel electrophoresis was performed as described previously with small amendments [11]. The bood plasma protein pellet (above mentioned) was resuspended in 250 µL rehydration buffer (8 M urea, 2% w/v CHAPS, 15 mM DTT, 0.5% v/v IPG buffer pH 3–10 and 0.001% bromophenol blue). For the first dimension, the resuspended protein was used to rehydrate IPG strips passively at least for 12–16 h at room temperature. The isoelectric focusing was performed with rehydrated IPG strips to resolve the proteins based on their isoelectric point. The focusing of rehydrated proteins was performed at 40,000 Volt-Hour at 20 °C under mineral oil in IEF Cell (Bio-Rad, USA). After focusing, the strips were incubated for 15 min in 2.5 mL of equilibration buffer I (6 M urea, 30% w/v glycerol, 2% w/v SDS, and 1% w/v DTT in 50 mM Tris–HCl buffer, pH 8.8) followed by equilibration buffer II (6 M urea, 30% w/v glycerol, 2% w/v SDS, and 4% w/v iodoacetamide in 375 mM Tris–HCl buffer, pH 8.8). After the equilibration step, the strips were incubated for 10 min in SDS running buffer and finally transferred onto Tris–glycine 10% SDS-PAGE, where proteins were separated in the second dimension according to their molecular mass.

Protein Visualization and Image Analysis

Protein spots were detected by staining with Coomassie Brilliant Blue (5% aluminium sulfate, 10% ethanol, 0.02% CBB G-250, and 2% orthophosphoric acid). Gel images were captured by gel documentation system (Alpha Innotech, USA). Three replicate images were used for automatic spot detection using PD-Quest 2D gel analysis software (Hercules, CA, USA). The spot intensities of three biological replicates from control and treated groups were subjected to independent t-test analysis using SPSS software. Protein spots showing expression alteration between the control and experimental group (ratio1.5) were marked and excised for MALDI-TOF/MS analysis.

In-gel Protein Digestion

The in-gel tryptic digestion of the protein was carried out as described previously with slight modifications [12]. Briefly, excised pieces from 2-D gels were destained at room temperature with destaining solution 500 mL [50% acetonitrile (ACN)/50 mM NH₄HCO₃] thrice for 15 min each with gentle vortexing. The gel pieces were then reduced by reducing solution (10 mM DTT in 100 mM NH₄HCO₃), alkylated using alkylation solution (2% iodoacetamide in 50 mM NH₄HCO₃), dried, and then digested with a minimal amount of trypsin (20 mg/mL) solution at 37 °C overnight. Peptides were extracted with 50% acetonitrile and 0.1% trifluoroacetic acid (TFA), dried, and resuspended in 5 mL resuspension solution (50% ACN and 0.1% TFA) before MALDI analysis.

Matrix-Assisted Laser Desorption/Ionization-Time of Flight Mass Spectrometry (MALDI-TOF/MS)

The samples were spotted (1 mL) on MALDI target plate [MTP 384 ground steel (Bruker Daltonics, Germany)] followed by 0.5 mL of alpha cyano-4-hydroxycinnamic acid matrix [10 mg/mL in 50% ACN, 0.1% TFA (Sigma-Aldrich, USA)]. The spectra of the peptides were attained in an AUTOFLEX speed MALDI TOF/TOF instrument (Bruker Daltonics, Germany) taking Nd:YAG smart Laser beam of 335 nm wavelength. External calibration was done with a standard peptide supplied by Bruker, with masses ranging from 1046 to 3147 Da. The spectra were acquired using Flex Control version 3.3 software in reflectron ion mode with an average of 2000 laser shots at mass detection range between 700 and 4000 m/z. Here three upmost plentiful peaks were subjected to further fragmentation using the LIFT method. The data were examined by Flex analysis software version 3.3 (Bruker Daltonics, Germany) and explored in MASCOT webserver (Matrix Science; http://www.matrixscience.com) using Biotools version 3.2 software (Bruker Daltonics, Germany).

In Silico Analysis

The online bioinformatics tool Panther (www.pantherdb.org) was used to classify the function, protein class and pathway of identified proteins altered during MCI [13]. Further, bio-computational interaction network analysis was performed using the STRING tool (www.string-db.org) [14]. This provided additional important information about the complex interactions among these proteins in MCI individuals.

Statistical analysis

Student's t-test was used to determine the statistical significance by SPSS 16.0 (Statistical Products and Service Solutions, IBM Corporation, Armonk, NY, USA). The p-value < 0.05 was considered statistically significant.

Result

Baseline characteristics

Our data showed that MCI individuals scored 24.65 for HMSE and 22.16 for the MOCA test, whereas control individuals scored 29 for both HMSE and MOCA tests. Most of the individuals from the MCI (60%) and control (70%) group belonged to the rural area. Majority of the MCI (60%) individuals were vegetarian as compared to the control (40%). Further, a small number of MCI individuals were having alcohol (20%) and tobacco addiction (40%) in comparison to control (15% and 30%, respectively) (Table 1).

Table 1.

Demographic and psychological screening details

Variables	Group 1 (Control) (Mean ± SD)	Group 2 (Case) (Mean ± SD)	p value
Age (years)	59.86 ± 4.398	59.92 ± 6.51	0.959
HMSE	29.00 ± 0.971	24.65 ± 1.60	< 0.05
MOCA	29.00 ± 0.248	22.16 ± 2.30	< 0.05
Demographic location	30% Urban	40% Urban	–
	70% Rural	60% Rural	–
Diet	60% Vegetarian	60% Vegetarian	–
	40% Non-vegetarian	40% Non-vegetarian
Addiction
Alcohol	15%	20%	–
Tabacco	30%	40%	–

*p value < 0.05 for comparison between MCI individuals and healthy control

Identification of Altered Proteins

MCI plasma proteomic changes were analyzed using MALDI-TOF/MS. Total plasma protein was resolved by 2-D gel electrophoresis (Fig. 1) and protein spots were detected automatically and matched by PD-Quest software. The protein quantity of each spot was determined by measuring spot volume and intensity (Fig. 2a). The ratio of protein quantity was calculated for an individual spot between control and MCI individual. These spots were further identified by MALDI-TOF/MS analysis. These identified protein spots are listed as calculated fold change (Fig. 2b) (Fig. 3a–c). The downregulated protein included APO-E and upregulated proteins included fibrinogen chain beta (FGB) and Transthyretin (TTR) (Table 2).

Fig. 1.

 Two dimensional gel of the plasma of MCI and healthy control individuals (Equal amount of total protein from the plasma of healthy control (a) and MCI (b) were separated by 2-DE. Differentially expressed protein spots were marked and excised for identification by MALDI-TOF/MS)

Fig. 2.

a Three spots altered in MCI patients compared to healthy individuals indicated by the arrow in gel, b Bar plot showing fold change in MCI

Fig. 3.

a–c MALDI-TOF/MS analysis of MCI individual’s plasma proteins

Table 2.

List of probable proteins in the plasma of MCI individuals based on MASCOT Score

Spot No	Accession id	Probable protein	Mass	Score	Expect	Matches	Expression Upregulation/downregulation
SPOT1	2::TTHY_HUMAN	Transthyretin OS = Homo sapiens OX = 9606 GN = TTR PE = 1 SV = 1	15,991	96	8.5e−05	13	$\uparrow$
SPOT 2	1::AAB59518.1	Apolipoprotein E [Homo sapiens]	36,302	60	0.39	12	$\downarrow$
SPOT 3	2::FIBB_HUMAN	Fibrinogen beta chain OS = Homo sapiens OX = 9606 GN = FGB PE = 1 SV = 2	56,577	92	0.00021	13	$\uparrow$

Bio-computational Classification and Pathway Analysis of Identified Proteins

Next, we classified all the proteins according to their biological processes, molecular functions, cellular components and protein classes using the Panther database. All three proteins are involved in molecular function. Further, the Panther pathway analysis predicted that the above-mentioned identified proteins are associated with different pathways. For instance, FGB protein is involved in several pathways such as platelet activation cascade and blood coagulation pathway. This suggests that FGB protein might be playing an important role in cognitive impairment by platelet activation observed in cognitive impairment Fig. 4(a, b).

Fig. 4.

a, b Molecular function and associated pathways identification by Gene’s ontology tool (Panther)

Bio-computational Interaction Network Analysis of Identified Proteins

On the contrary to the proteomic study, we have used Vennya 2.1 tool to identify common genes between AD (CUI:C0002395) and MCI (CUI:C1270972) from DisGeNET database [15]. We found that all three identified proteins were coded by both AD and MCI related genes. To further explore the interaction network of the identified proteins, we used the online STITCH 5.0 tool. STITCH 5.0 showed that APO-E has association with TTR and FGB protein. This suggests that APO-E is a key protein with which other proteins interacts to exert the function. Moreover, STITCH 5.0 showed pathways such as complement cascade, coagulation cascades, platelet activation, and AD pathway which leads to the occurrence of MCI and AD. This indicates that APO-E might be involved in the onset of MCI (Fig. 5).

Fig. 5.

Functional network of plasma proteins using STITCH 5.0 database

Discussion

The MCI subjects reported in our studies have shown a significant decline in cognitive function as evaluated by HMSE and MOCA neuropsychological tests. Moreover, there was no difference between MCI and control group with respect to their demographic location, dietary habit and type of addiction. Further, we have observed increased plasma TTR in MCI individuals. TTR is a thyroid hormone carrier and plasma retinol transporter [16]. It is encoded by a single gene copy on chromosome number 18 and expressed in the human liver, kidney, pancreas, retinal epithelium, lepto-meningeal epithelium, choroid plexus and potentially in neurons, which are not independently related to AD. Earlier study reported an increased level of TTR in AD patients [17]. The significance of increased TTR in MCI and decreased level in AD is not clear. However, one potential hypothesis is that the increased level in MCI represents a microglial or neuronal reaction to early amyloid deposition, which then fails in the later stage of AD. Another alternative hypothesis is that AD patients may have a genetic disposition or an acquired low CSF-TTR level [17].

Also, we found reduced plasma APO-E levels in MCI individuals. APO-E was involved in the Aβ clearance and lipid transport. A significantly decreased level of plasma APO-E in an AD group was also reported earlier [2]. A reduced level of APO-E in plasma and CNS was associated with the development of AD [18]. It is also involved in the cholinergic pathway as lipid homeostasis maintainer, which is crucial for the synthesis of acetylcholine [19–21]. Cholinergic dysfunction is a well-documented feature of AD, whereas many treatment strategies have revolved around augmenting levels of this particular neurotransmitter [22]. The link between cholinesterase dysfunction and amyloid deposition is lipid maintenance, which illustrates the potential importance of APO-E in this pathway [23–25].

In addition, we found increased plasma FGB level in MCI individuals. Previously it was reported that dysregulation of plasma proteins FGA, FGB and FGG exist in MCI individuals [2]. Fibrin is the primary protein component of a blood clot [26]. Earlier, it was reported that fibrin inactive precursor (fibrinogen) circulates in the blood as a large complex molecule of 340 kDa. In a healthy condition of individuals, fibrinogen is excluded from the brain by the blood–brain barrier, but under AD progression state it has been found to accumulate in the extravascular space over time [27, 28]. Elevated fibrinogen levels are associated with the increased risk for AD and dementia [29]. An in vitro study reported that the interaction between Aβ and fibrinogen modifies fibrinogen’s structure, which might lead to abnormal fibrin clot formation and vascular abnormalities in AD [30].

Conclusion

In brief, the data from the present study suggest that FGB, TTR and APO-E might contribute to the onset of AD. Plasma TTR, APOE and FGB alteration might be associated with the MCI and AD. Further, the In silico study evidenced their role in the MCI and AD pathogenesis.

Funding

Indian Council of Medical Research-New Delhi, Institute of Medical Sciences and UGC-CRET, BHU, Varanasi.

Declarations

Conflict of interest

All authors declare that they have no conflict of interest.

Ethical Approval

Ethical approval was taken from the institutional ethical committee (Institute of Medical Sciences, BHU, Varanasi).

Consent to Participate

Consent was taken from both healthy control and MCI individuals.