Journal of Proteomics 127 (2015) 103–113

Contents lists available at ScienceDirect

Journal of Proteomics

j ourna l homepage: www.e lsev ie r .com/ locate / jp rot

Proteomic analysis of Plasmodium falciparum induced alterations inhumans from different endemic regions of India to decipher malariapathogenesis and identify surrogate markers of severity☆

Sandipan Ray a,1, Vipin Kumar a,1, Amruta Bhave a, Vaidhvi Singh a, Nithya J. Gogtay b, Urmila M. Thatte b,Arunansu Talukdar c, Sanjay K. Kochar d, Swati Patankar a, Sanjeeva Srivastava a,⁎a Department of Biosciences and Bioengineering, Indian Institute of Technology Bombay, Powai, Mumbai 400076, Indiab Department of Clinical Pharmacology, Seth GS Medical College & KEM Hospital, Parel, Mumbai 400012, Indiac Department of Medicine, Medical College and Hospital Kolkata, 88, College Street, Kolkata 700073, Indiad Department of Medicine, Malaria Research Center, S.P. Medical College, Bikaner 334003, India

☆ This article is part of a Special Issue entitled: Proteom⁎ Corresponding author.

E-mail address: sanjeeva@iitb.ac.in (S. Srivastava).1 Both authors contributed equally to the preparation o

http://dx.doi.org/10.1016/j.jprot.2015.04.0321874-3919/© 2015 Elsevier B.V. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 26 February 2015Received in revised form 21 April 2015Accepted 29 April 2015Available online 14 May 2015

Keywords:IndiaMalaria pathogenesisPlasmodium falciparumProteomicsSevere falciparummalariaSurrogate markers

India significantly contributes to the global malaria burden and has the largest population in the world at risk ofmalaria. This study aims to analyze alterations in the human serum proteome as a consequence of non-severeand severe infections by themalaria parasite Plasmodium falciparum to identify markers related to disease sever-ity and to obtainmechanistic insights about disease pathogenesis andhost immune responses. In discovery phaseof the study, a comprehensive quantitative proteomic analysis was performed using gel-based (2D-DIGE) andgel-free (iTRAQ) techniques on two independent mass spectrometry platforms (ESI-Q-TOF and Q-Exactivemass spectrometry), and selected targets were validated by ELISA. Proteins showing altered serum abundancein falciparummalaria patients revealed themodulation of different physiological pathways including chemokineand cytokine signaling, IL-12 signaling and production inmacrophages, complement cascades, blood coagulation,and protein ubiquitination pathways. Somemuscle related and cytoskeletal proteins such as titin and galectin-3-binding protein were found to be up-regulated in severe malaria patients. Hemoglobin levels and platelet countswere also found to be drastically lower in severemalaria patients. Identified proteins including serum amyloid A,C-reactive protein, apolipoprotein E and haptoglobin,which exhibited sequential alterations in their serumabun-dance in different severity levels of malaria, could serve as potential predictive markers for disease severity. Tothe best of our information, we report here the first comprehensive analysis describing the serum proteomicalterations observed in severe P. falciparum infected patients from different malaria endemic regions of India.This article is part of a Special Issue entitled: Proteomics in India.

© 2015 Elsevier B.V. All rights reserved.

1. Introduction

Plasmodium falciparum infection represents the major cause ofmalaria-associated mortality worldwide [1]. This lethal species of ma-laria parasite is responsible for approximately 247 million cases andaround one million deaths each year, particularly in the sub-SaharanAfrica [2]. India notably contributes to the global malaria burden andhas the largest population in the world at risk of malaria [3]. Moreover,due to the extremely variablemalaria epidemiology in India, it is consid-ered as an important country for malaria research [4]. Importantly, in

ics in India.

f this manuscript.

recent years there is an increased incidence of P. falciparum comparedto Plasmodium vivax malaria in different endemic regions of India [4].Severe falciparum malaria often leads to fatal and complicated clinicalmanifestations including hepatic dysfunction, renal dysfunction, severeanemia, hypoglycaemia, acute respiratory distress syndrome (ARDS),cerebral manifestation, and multiple organ involvement [5].

Proteomic techniques pose tremendous potential to provide awealth of new information to accelerate malaria research [6,7]. In-depth analysis of the differential abundances of serum/plasma proteinsduring the febrile stage of the infectionmay help in identification of sur-rogate markers of infection and disease severity and can provide valu-able information regarding disease pathogenesis and host immuneresponses [8,9]. A fewprevious studies have investigated the alterationsin plasma proteome profiles in cerebral falciparum malaria in childrenfrom different endemic and holoendemic regions of Africa [10–12].However, there is a dearth of similar proteomic analysis of severe

http://crossmark.crossref.org/dialog/?doi=10.1016/j.jprot.2015.04.032&domain=pdfhttp://dx.doi.org/10.1016/j.jprot.2015.04.032mailto:sanjeeva@iitb.ac.inhttp://dx.doi.org/10.1016/j.jprot.2015.04.032http://www.sciencedirect.com/science/journal/18743919

104 S. Ray et al. / Journal of Proteomics 127 (2015) 103–113

falciparummalaria in Indian populations. In an earlier studywe have re-ported the modulations in human serum proteome and various physio-logical pathways in uncomplicated non-severe falciparummalaria in anadult population from India [13]. In this study, serum samples fromadult severe and non-severe falciparum malaria (SFM and NSFM) pa-tients along with healthy community controls from three different en-demic regions of India were investigated using 2D-DIGE and iTRAQ-based quantitative proteomics in combination with ESI-Q-TOF and Q-Exactive mass spectrometry. Different hematological and liver functionparameters were measured in malaria patients and controls for a com-parative statistical analysis. Bioinformatic analysis involving the identi-fied proteins showing altered abundance in the serum samples offalciparum malaria patients revealed the modulation of different vitalphysiological pathways. This study revealed potential biomarkers formonitoring disease severity of P. falciparum infection and enhancedour understanding regarding pathogenesis of falciparum malaria.

2. Materials and methods

2.1. Subject recruitment, blood collection, and serum separation

This proteomics study was conducted involving non-severe andsevere falciparum malaria patients and healthy controls from threedifferent endemic regions of India; Mumbai, Kolkata, and Bikaner. Thismulti-centric analysis was performed with the approval of the institu-tional ethics committees of Seth GS Medical College and King EdwardMemorial Hospital—Mumbai, Medical College Hospital Kolkata —Kolkata, and Malaria Research Center, S.P. Medical College — Bikaner.Written informed consent was obtained from each participant (malariapatients and controls) prior to sample collection. A total of 104 adult pa-tients with falciparummalaria (67 with non-severe and 37 with severeinfection) confirmed throughmicroscopic examination of a thin periph-eral blood smear followed by rapid diagnostic test (RDT) and 146 age-matched healthy controls were enrolled for this proteomic study. Casedefinition for severe malaria was adopted from standard WHO guide-lines [5]. Patients co-infected with any other infectious diseases orother plasmodial infections such as vivax malaria or mixed infections(infected with both P. falciparum and P. vivax) were excluded fromthis study. Sample collection, serum separation, and storage were per-formed following the same protocol as reported earlier [13]. Serumsamples were stored as multiple aliquots and uniformity was main-tained in terms of number of freeze/thaw cycles while selection of thetest (NSFM/SFM) and control (HC) samples for the comparative prote-omic analysis to minimize the pre-analytical variables.

2.2. Sample processing, 2DE and 2D-DIGE

Sample processing, 2DE and 2D-DIGE were performed as describedpreviously [13]. In brief, the top two high abundance proteins (albuminand IgG) were removed, and 400 μg amount of depleted serumproteinswas focused on linear pH 4–7 IPG strips (18 cm) and then separated on12.5% polyacrylamide gels. 2DE gels were stained using Sypro RubyStain Gel (Life Technologies, Molecular Probes, CA, USA) following themanufacturer's protocol. In 2D-DIGE analysis, serum samples (HC,SFM and NSFM) were labeled with Cy3, Cy5, and the mixture of equalamount of all samples (internal standard) was labeled with Cy2. Dyeswitching strategywas applied during labeling the test and control sam-ples to avoid the dye-specific biasness/effects. DIGE gels were scannedusing a Typhoon 9550 Variable Mode Imager (GE Healthcare) andimage acquisition and data analysis was performed using DeCyder 2Dsoftware; version 7.0 (GE Healthcare) as described previously [13].

2.3. In-gel digestion and MALDI TOF/TOF analysis for protein identification

The identity of differentially abundant serum proteins (p b 0.05) inNSFM and SFM was established using an AB Sciex 4800 MALDI-TOF/

TOF mass spectrometer. In-gel digestion and mass spectrometric analy-sis was performed as described previously [13].

2.4. In-solution digestion, iTRAQ labeling and OFFGEL fractionation

Extracted protein samples suspended in rehydration buffer weresubjected to buffer exchange process to transfer the proteins in TEABbuffer using Amicon Ultra 0.5 mL centrifugal 3 kDa filters (Millipore,Watford, UK). After the buffer exchange process, each protein samplewas quantified using QuickStart Bradford reagent (BioRad, USA).75 μg of the pooled protein from each group (HC, NSFM, and SFM)was reduced, alkylated, and trypsinized following themanufacturer's instructions. Trypsin (Trypsin Gold, mass spectrome-try grade, Promega, Madison, WI, USA) was applied in 1:20 trypsin:protein ratio for performing in-solution digestion. Trypsin digestedsamples were then labeled with 4-plex iTRAQ reagents (AB SciexUK Limited, UK) solubilized in 70% ethanol at room temperatureand were incubated for 1:30 h at RT. After incubation, reactionswere quenched with milliQ water, and incubated at RT for 30 min.Sample labeling strategy for differential proteomic analysis was:control-114, NSFM-115, and SFM-116. All the labeled samples werepooled, vacuum dried and subjected to OFFGEL fractionation usinga 3100 OFFGEL fractionator (Agilent Technologies, Santa Clara, CA)with high resolution (pH 3–10, 24 cm) IPG strips. After OFFGEL fraction-ation iTRAQ labeled fractionated samples were Zip-tipped for the pep-tide enrichment.

2.5. LC-MS/MS analysis for protein identification and quantitation

iTRAQ-based quantitative proteomic analysis was carried out in twostages; initial analysis was performed using pooled samples (each poolcontaining 20 samples) and the results obtained from the pooled co-horts were further validated on 3 individual subjects randomly selectedfrom each group (i.e. HC, NSFM and SFM). Fold-change values for differ-ential abundance of the serum proteins across the experimental groupswere derived from the three most consistent data-sets (n = 3). Twodifferent types of mass spectrometer, Agilent 6550 Q-TOF and ThermoScientific Q-Exactive were used to analyze iTRAQ-labeled samples.Technical details for the mass spectrometric analysis, data acquisitionand analysis have been reported previously [14,15]. In brief, data wasacquired from an Agilent 6550 iFunnel Q-TOF LC MS/MS instrument(Agilent Technologies, USA) in a positive ion mode, monitored byMass Hunter acquisition software. All searches were performed usingthe SpectrumMill Protein Identification software (Agilent Technologies,USA) against the Uniprot database. Data was extracted between MH+600 and 4000. Alkylation of cysteine (carbamidomethylation (C)) andiTRAQ (N-term, K)were defined as thefixedmodifications, while oxida-tion of methionine was specified as a variable modification. 20 ppmprecursor mass tolerance and 50 ppm fragment mass error tolerancewas specified. Peptides identified with confidence interval (CI) valuesabove 95% were used for protein identification and quantification. TheiTRAQ report peak areas (RPAs) corresponding to quantification ionsm/z 114–117 were extracted from the raw spectra and corrected forisotopic carryover using GPS Explorer.

iTRAQ labeled samples were also analyzed using Q-Exactive massspectrometer (Thermo Fisher Scientific, Waltham, MA, USA). Thevacuum dried iTRAQ labeled samples were resuspended in 0.1%formic acid in water and were injected into nano flow HPLC pumpcoupled online with the Q-Exactive Orbitrap mass spectrometer witha nanoelectrospray ion source. All the .raw files were processed usingproteome discoverer 1.4 (Thermo Fisher Scientific) software package;MASCOT 2.2.4 and SEQUEST were used for database searching againstthe Uniprot Homo sapiens FASTA. Database searching parameters in-cluded precursor ionmass tolerance of 5 ppm and fragmentmass toler-ance of 0.02 Da. N-terminal modifications were assigned as iTRAQ 4plex reaction, with dynamic modifications at oxidation (M) and

Fig. 1. Differentially abundant serum proteins in severe and non-severe falciparummalaria (SFM andNSFM) identified using 2DE and 2D-DIGE. (A) Representative 2D gels of serum fromhealthy controls and SFMandNSFMpatients. 400 μg of total serumproteinswere focused on linear pH4–7 IPG strips (18 cm) and then separated on 12.5% polyacrylamidegels,whichwerestained with Sypro Ruby Stain. The 3D images of some selected statistically significant (p b 0.05; paired t-test) differentially abundant proteins identified in 2DE are presented.(B) Schematic representation of experimental strategy for CyDye labeling and2D-DIGE. (C) Representative 2D-DIGE image to compare serumproteome of HC and SFMandNSFMpatients.Graphical and 3D fluorescence intensity representations of few selected statistically significant (p b 0.05; paired t-test and one-wayANOVA)differentially abundant proteins in SFM/NSFMpatients identified in biological variation analysis (BVA) using DeCyder 2D software.

105S. Ray et al. / Journal of Proteomics 127 (2015) 103–113

deamination (N,Q) in addition to the static modification at methylthio(C). A decoy database search was executed for calculation of the falsediscovery rate (FDR) and 1% cut-off was used to report identifications.Only the peptides uniquely identifying individual proteinswere utilizedfor calculation of fold-change values.

2.6. Protein networks and bioinformatic analysis

Functional annotation and clustering of the differentially abundantproteins (p b 0.05) identified in malaria patients (SFM and NSFM)in our quantitative proteomic analysis were performed using

Image of Fig. 1

Fig. 2.Differentially abundant serum proteins identified in severe and non-severe falciparummalaria using iTRAQ-based quantitative proteomics. (A) Schematic representation of exper-imental strategy for iTRAQ-based quantitative proteomic analysis. (B) Representative MS/MS spectrum of a few selected differentially abundant serum proteins identified in falciparummalaria patients. Inset presenting the iTRAQ reporter ion intensities for representative peptides in healthy community control, severe and non-severe falciparum malaria. (C) Venn dia-gram showing the unique and overlapping proteins identified by Q-Exactive and ESI-Q-TOF instruments. (D) Venn diagram presenting the unique and common differentially abundantproteins in severe and non-severe falciparummalaria (differentially abundant proteins (fold-change N 1.5) identified in the iTRAQ analysis using pooled samples are represented).

106 S. Ray et al. / Journal of Proteomics 127 (2015) 103–113

Image of Fig. 2

Table 1Clinicopathological details of the severe and non-severe falciparum malaria patients and healthy controls enrolled for this study.

Hemoglobin (g/dL) Platelets/μL (thousands) Creatinine (mg/dL) Total bilirubin (mg %) SGOT (IU/L) SGPT (IU/L) ALP (IU/L)

HC (n = 146) 12.9 ± 1.35 297.0 ± 94.3 0.96 ± 0.58 0.80 ± 0.37 29.5 ± 11.5 31.7 ± 12.3 96.0 ± 30.4NSFM (n = 67) 10.7 ± 2.23 140.2 ± 99.9 1.03 ± 0.43 1.24 ± 0.58 38.7 ± 24.1 37.8 ± 20.1 109.1 ± 53.6SFM (n = 37) 9.4 ± 2.96 97.4 ± 81.3 2.8 ± 3.19 5.9 ± 6.84 119.2 ± 211.6 144.0 ± 300.7 291.7 ± 311.6

107S. Ray et al. / Journal of Proteomics 127 (2015) 103–113

ingenuity pathway analysis (IPA) version 9.0 (Ingenuity® Systems,www.ingenuity.com), PANTHER system, version 7 (http://www.pantherdb.org) [16] and DAVID database version 6.7 (http://david.abcc.ncifcrf.gov/home.jsp) [17]. UniProt accession number files for thedifferentially abundant proteins were uploaded and mapped underthe Homo sapiens dataset to extract and summarize the pathways andfunctional associations of the individual or group of genes and proteins.

2.7. ELISA and receiver operating characteristic (ROC) curve analysis

Validation of eight selected targets; haptoglobin, serum amyloidA, apolipoprotein A-I, hemopexin, apolipoprotein E, retinol binding pro-tein 4, ceruloplasmin, and plasminogen was performed using ELISA.Concentrations of the candidate proteins in serum of control (n =103) and falciparum malaria patients (including NSFM (n = 42) andSFM(n=39))weremeasured using AssayMaxELISA kits fromAssayPro(USA) following the manufacturer's instructions. Receiver operatingcharacteristic (ROC) curve analysis was performed to evaluate the effi-ciency of the differentially abundant proteins for prediction of severeand non-severe malaria patients using GraphPad Prism software pack-age (version 6) following the same protocol as described previously [13].

2.8. SPR-based quantification of active protein concentration of serumamyloid A in serum samples

Quantification of active serum protein concentration for one of thedifferentially expressed candidates; serum amyloid A (SAA), wasperformed using surface plasmon resonance (SPR) by calibration-freeconcentration analysis (CFCA) method using a Biacore T200 system(GE Healthcare). Anti-SAA was immobilized covalently on the surfaceof CM5 sensor chip using amine coupling chemistry to prepare the reac-tion surface appropriate for CFCA. The antibody was passed through theactive flow cell for 720 s at 30 μg/mL in 10mM sodium acetate pH 5.0 ata flow rate of 10 μL/min, providing an immobilization level of 5011 RU.The remaining active esters were blocked with 1 M ethanolamine-HClpH 8.5. The CFCA for SAA protein was performed on serum samples ofsevere and non-severe falciparum malaria patients and healthy con-trols. A total of 3 pools (each pool containing 10 samples) were testedfor healthy controls and patients suffering from each of non-severeand severe falciparum malaria. The results obtained from the pooledsamples were further validated by performing CFCA on 7 individualserum samples randomly selected from each group. Each serum samplewas serially diluted to 100 and 200 folds (in running buffer) forperforming the concentration analysis, and injected in duplicates atflow rates of 5 μL/min and 100 μL/min for 120 s over the active and ref-erenceflow cells at 25 °C. Data obtained from the referenceflow cellwasautomatically subtracted from experimental measurements to yield thespecific signal. Protein concentration was then determined from thebinding data using the CFCA evaluation feature of the software.

3. Results

3.1. Alterations in clinicopathological parameters in falciparum malariapatients

A total of 250 subjects (HC: n = 146, NSFM: n = 67, SFM: n = 37)were analyzed in this multi-centric study. Platelet counts and Hb levelswere found to be significantly lower (p b 0.05) inmalaria patients (both

NSFM and SFM) as compared to the healthy controls (Table 1 andFig. S1). Sequential decreases in those two hematological parameterswere observed alongwith the increase in disease severity. Liver functionparameters including total bilirubin, serum glutamic oxaloacetic trans-aminase (SGOT), serum glutamic pyruvic transaminase (SGPT), and al-kaline phosphatase (ALP) were found to be higher in the severe malariapatients in comparison to the non-severe malaria patients and healthycontrol cohorts (Table S1 and Fig. S1A). Comparative analysis of the clin-icopathological parameters among the study cohorts from the three dif-ferent endemic regions of India indicates that malaria patients (bothNSFM and SFM) from Bikaner have a lower level of hemoglobin com-pared to the patients fromMumbai or Kolkata. Interestingly, lever func-tion parameters (SGPT, SGOT, ALP and total bilirubin) were found to behigher among the severe malaria patients from Bikaner. However, nospecific trend was observed for platelet count and serum creatininelevel in the malaria patient cohorts from different endemic regions(Fig. S1B).

3.2. Differential abundance of serum proteins in falciparum malariaidentified in gel-based proteomics analysis

Comparative proteome analysis of healthy control and falciparummalaria patient cohorts by classical 2DE and DIGE experiments indicat-ed altered serum abundance of multiple proteins, as a consequenceof P. falciparum infection. In 2DE analysis, 22 statistically significant(p b 0.05) differentially abundant (17 down-regulated and 5 up-regulated) serum proteinswere identified in SFM population comparedto theHC. A comparative analysis betweenNSFMand SFM study cohortsindicated differential abundance of 19 protein spots. Representative2DE images of serum proteome profile of falciparum malaria patientsand healthy individuals and 3D views of some selected differentiallyexpressed proteins are shown in Fig. 1A. In HC vs. SFM comparison by2D-DIGE, a total of 1183 protein spots were detected among which 55spots exhibited significant alterations (p ≤ 0.05, FC N 1.5) in theirserum abundances in SFM patients, while 37 protein spots were foundto be exhibiting differential serum abundance in severemalaria patientscompared to those who are suffering from a non-severe infection(NSFM) (Table S2). Fig. 1C depicts 3D views and graphical representa-tion of a few selected differentially abundant protein spots in NSFMand SFM patients. 2DE/2D-DIGE identified differentially abundantserum protein spots that could be excised from the gels were subse-quently subjected to MS and MS/MS analysis, which successfullyestablished the identity of 9 (HC vs. SFM analysis) and 6 (NSFM vs.SFM analysis) protein spots (Table S3). Proteins identified in regular2DE experiment were also visualized in 2D-DIGE; in addition, new can-didates were also identified by 2D-DIGE due to its higher sensitivity andreproducibility. Three proteins (alpha-1-antitrypsin, Ig alpha-1 chain Cregion and serumamyloid A)were found to bewith higher serumabun-dance in malaria patients; whereas several proteins including apolipo-protein A-I, sorcin (22 kDa protein), serum albumin, transthyretin,clusterin, haptoglobin, retinol binding protein exhibited reducedserum abundance after the plasmodial infection.

3.3. Identification of differentially abundant serum proteins in falciparummalaria by employing iTRAQ-based quantitative proteomics analysis

iTRAQ labeled samples were analyzed using Q-TOF and Q-Exactivemass spectrometers in order to obtain a better coverage of serum

http://www.ingenuity.comhttp://www.pantherdb.orghttp://www.pantherdb.orghttp://david.abcc.ncifcrf.gov/home.jsphttp://david.abcc.ncifcrf.gov/home.jsp

108 S. Ray et al. / Journal of Proteomics 127 (2015) 103–113

proteome. The signature of proteinswas searched against the Swissprotdatabase, and 367 proteins were identified based on the relative inten-sity with 95% confidence interval and 1% FDR from Q-TOF (Table S4).Among the 367 identified proteins, 169 proteins in NSFM (55 up-regulated and 114 down-regulated) and 179 proteins in SFM (60 up-regulated and 119 down-regulated) were found to be differentiallyabundant in diseased serum (fold-change ≥ 1.5) (Table S4). InQ-Exactive analysis, a total of 369 proteins were identified, out ofwhich 162 were with ≥2 peptides (Table S5). Significant increase inthe serum levels of 9 proteins was observed in NSFM and 10 proteinsin SFM patients, whereas serum abundance of 65 and 119 proteinswere found to be reduced in NSFM and SFM, respectively, comparedto the healthy controls (Table S5). S-curve distributions of the differ-entially expressed proteins in NSFM and SFM are represented inFig. S2. MS/MS spectra for the few selected proteins with inset show-ing the iTRAQ reporter ion intensities for representative peptides incontrols and malaria patients (NSFM and SFM) are presented inFig. 2B. iTRAQ data obtained from the Q-TOF and Q-Exactive were com-pared for identifying the common proteins. Comparative analysis of thefindings obtained from these two different mass spectrometersindicated significant overlap of 62 proteins, among which 53 werefound to be with ≥2 peptides (Fig. 2C, Tables S4 and S5). Most of theproteins exhibited similar trend of differential abundance in both themass spectrometric analysis. Some proteins including apolipoproteinA-I, apolipoprotein E, ceruloplasmin, haptoglobin, plasminogen, serumamyloid A exhibited sequential alterations in their serum abundancewith respect to the increased severity (Tables 2 and S4 and S5).Among the differentially abundant proteins (fold-change ≥1.5; FDR1%) identified in ESI-Q-TOF, 109 candidates were found to be commonbetween NSFM and SFM, while in Q-Exactive analysis, 139 proteinswere found to be common between NSFM and SFM. iTRAQ ratios ofentire protein list, protein score and unique peptide information areprovided under the supplementary information (Tables S4 and S5).

3.4. Modulation of physiological pathways in severe and non-severefalciparum malaria

Identified differentially abundant proteins were subjected to path-way analysis using IPA, PANTHER, and DAVID software packages fordetermining the connection of the identified proteins with variousphysiological pathways, biological functions, and their involvement inmalaria pathogenesis. Even though most of the pathways were foundto be common in severe and non-severemalaria, the number associatedmolecules and their alterations in serum abundance were higher in se-vere infection. The highest scoring networks identified in IPA analysisincluded 25 out of 53 focus molecules in NSFM, while in case of SFM23 out of 44 focus molecules were mapped (Table S6). IPA analysisindicated association of 16 significant canonical pathways with thedifferentially abundant proteins including acute phase response sig-naling, coagulation system, clathrin mediated endocytosis signaling,production of nitric oxide oxygen species in macrophage etc. (Fig. 4and Table S6).

PANTHER analysis revealed association of the identified proteinswith blood coagulation, cytoskeletal regulation by Rho GTPase, andplasminogen activating cascade pathways (Table S6). In DAVID analysis,the KEGG category revealed modulation of coagulation pathway, com-plement pathway, and prothrombin activation pathway, the Reactomecategory indicated association of hemostasis, signaling in immunesystem, metabolism of lipids and lipoproteins; while complementpathway, intrinsic and extrinsic prothrombin activation pathway wereidentified in the Biocarta category (Table S6; Fig. S6).

3.5. Validation of selected differentially abundant serum proteins

Eight selected differentially abundant serum proteins; serum amy-loid A, haptoglobin, hemopexin, apolipoprotein E, apolipoprotein A-1,

retinol binding protein 4, ceruloplasmin, and plasminogen were vali-dated by ELISA. Selection of the proteins for validation was performedon the basis of their fold-changes, possible association of the proteinswith malaria pathogenesis and severity, and availability of the requiredELISA kits. Compared to the healthy controls, serum amyloid A,hemopexin, plasminogen, apolipoprotein E, and ceruloplasmin exhibit-ed higher serum abundance (p b 0.05 in a Mann–Whitney U test) infalciparum malaria patients (Fig. 4A; Table S7). However, alteration inserum level of hemopexin in NSFM cohort was found to be statisticallyinsignificant (p N 0.05). In accordance with the quantitative proteomicfindings, lower serum levels of haptoglobin, apolipoprotein A-1 and ret-inol binding protein 4 were observed in SFM and NSFM patients inELISA assays (Fig. 4A; Table S7).

No statistically significant difference was observed in the serumabundances of these eight proteins among the malaria patients fromthe three different endemic regions of India (Fig. S5). However, slightlyhigher (p N 0.05) serum levels of SAA, CP, ApoE, and HPXwere observedin the malaria patients from Bikaner compared to the patients fromMumbai or Kolkata. Decrease in the serum abundances of HP and ApoA1 was found to be more apparent in the severe malaria patients fromMumbai and Bikaner than those from Kolkata.

Further, quantification of one protein, SAA was performed bycalibration-free concentration analysis (CFCA) using surface plasmonresonance. Serumabundance of SAAwas found be increased sequential-ly along with the severity of the infection (Fig. 4C). ROC curve analysisindicated that SAA and HP are good predictors for falciparum malaria,both at the non-severe and severe stages of the infection. Hemopexin,apolipoprotein E, and RBP 4 exhibited good potential for discriminationbetween the SFM and NSFM patient cohorts (Figs. 4B, S6).

4. Discussion

Specific parasite and host factors, which attribute the complexity ofthe disease in falciparummalaria have not been clearly deciphered [18].Hitherto, there is no clear mechanism by which we can comprehend oranticipate the severity of disease at an early stage of the infection [19].The present study aims to contribute towards unraveling the complex-ity of severe falciparum malaria pathogenesis. Moreover, the intent ofthis study was to identify potential predictive and disease monitoringmarkers for severity in falciparummalaria. Comparative proteomic pro-filing of serum from healthy community controls and non-severe andsevere falciparum malaria patients from different malaria endemic re-gions from India provided a plethora of information to complementthe earlier similar proteomics studies reported on the African popula-tions [10–12,20], and provided some novel mechanistic insights intothe pathogenicmechanism for disease severity in P. falciparum infection(Fig. 3). Importantly, differential serum abundance of some of our identi-fied serum proteins including titin, carbonic anhydrase 2, galectin-3-binding proteinwerenot reported earlier in the context of falciparumma-laria (Table 2).

A series of studies over the past decade has provided valuable infor-mation regarding the proteomic alterations in humanhost in falciparumand vivax malaria [10–12,20,21]. In previously reported studies, differ-ential abundance of multiple serum proteins including haptoglobin,serum amyloid A, serum albumin, apolipoprotein A-1, apolipoproteinE, alpha-2-macroglobulin, and vitronectin has been described in adultssuffering from non-severe falciparummalaria [13,20] (Table 2). In a re-cent study, Bachmann et al. have reported that oxidative markers indi-cate severe malaria anemia complications, while markers for plateletadhesion, muscular damage and endothelial activation are relatedwith cerebral malaria in children [11]. Gillespie et al. showed that mea-surement of serumamyloid A alongwith C-reactive protein (CRP) couldserve as a valuablemarker to access the severity of acute falciparumma-laria [22]. Ceruloplasmin and hemopexin act as antioxidants and trapthe free radicals generated during hemolysis [23]. In our study, serumlevels of both of the proteins were found to be elevated in severe

Table 2Partial list of the differentially abundant serum proteins identified in non-severe and severe falciparum malaria using iTRAQ-based quantitative proteomics analysisa.

Protein name Uniprot accession Unique Peptidesb MW [kDa] Fold-change of HC vs. NSFMc Fold-change of HC vs. SFMc References

1 C-reactive protein P02741 5 27.22 2.38 ± 0.98 5.26 ± 2.79 [11,13]2 Lipopolysaccharide-binding protein P18428 7 56.72 2.81 ± 0.5 3.06 ± 0.93 [42]3 von Willebrand factor P04275 4 338.17 2.62 ± 0.23 2.61 ± 0.32 [17]4 Alpha-1-acid glycoprotein 1 P02763 10 25.75 2.3 ± 0.45 2.17 ± 0.25 [43]5 Thyroxine-binding globulin P05543 4 50.70 2.08 ± 0.79 1.99 ± 0.47 NR6 Protein Z-dependent protease inhibitor Q9UK55 2 55.43 1.89 ± 0.14 1.94 ± 0.52 NR7 Leucine-rich alpha-2-glycoproteind P02750 8 40.27 3.31 ± 1.1 1.92 ± 0.27 [13]8 Inter-alpha-trypsin inhibitor heavy chain H3 Q06033 8 108.63 2.23 ± 0.78 1.83 ± 0.28 NR9 Serum amyloid A-4 protein P35542 1 15.86 1.39 ± 0.15 1.79 ± 0.46 [13]10 Apolipoprotein Ed, e P02649 9 38.14 1.46 ± 0.13 1.68 ± 0.51 [13,20]11 Plasminogene P00747 5 100.36 1.44 ± 0.12 1.67 ± 0.45 [13,44]12 Plasma protease C1 inhibitor P05155 14 59.56 1.47 ± 0.05 1.61 ± 0.07 [45]13 Serum amyloid P-component P02743 6 27.23 1.38 ± 0.18 1.45 ± 0.12 [46]14 Titin Q8WZ42 4 51.06 1.24 ± 0.27 1.26 ± 0.54 NR15 Galectin-3-binding protein Q08380 4 68.98 0.81 ± 0.06 1.24 ± 0.49 NR16 Ceruloplasmine P00450 31 132.57 1.33 ± 0.03 1.16 ± 0.04 [13,21]17 Alpha-2-macroglobulin P01023 55 177.54 0.74 ± 0.07 0.7 ± 0.04 [13,20]18 Gelsolin P06396 12 92.58 0.84 ± 0.08 0.66 ± 0.04 2019 Afamin P43652 12 78.35 0.6 ± 0.09 0.64 ± 0.12 NR20 Carbonic anhydrase 1 P00915 3 31.52 0.63 ± 0.05 0.64 ± 0.03 [47]21 Plasma retinol-binding protein precursord, e P02753 11 23.02 0.6 ± 0.05 0.56 ± 0.07 [13]22 Hemoglobin subunit alpha P69905 5 16.90 0.56 ± 0.09 0.57 ± 0.1 NR23 Transthyretind P02766 8 17.15 0.77 ± 0.08 0.55 ± 0.04 [48]24 Carbonic anhydrase 2 P00918 3 32.76 0.67 ± 0.09 0.45 ± 0.1 NR25 Apolipoprotein A-II P02652 6 12.73 0.66 ± 0.05 0.43 ± 0.13 [20]26 Serum paraoxonase/arylesterase 1d P27169 6 42.78 1.23 ± 0.01 0.38 ± 0.08 [13]27 Haptoglobind, e P00738 16 50.93 0.32 ± 0.04 0.42 ± 0.07 [11,13]28 Apolipoprotein A-Id, e P02647 25 33.94 0.4 ± 0.04 0.33 ± 0.08 [20]

NR: not reported earlier in the context of falciparummalaria.a This is a partial list for a few selected candidates (proteins identified in four biological replicates) identified in iTRAQ-based quantitative proteomics analysis using Q Exactive Orbitrap

and ESI-Q-TOF instruments; complete lists are provided under supplementary information (Tables S4 and S5).b Median value for the identified unique peptides in different biological replicates is represented.c Data is represented as mean ± SEM (where n = 3) [at least three most consistent values among the four replicates are used to calculate the fold-change values].d Differential abundance for these candidates is also identified in 2D-DIGE (details are provided in Table S3 and S3).e These candidates are validated by ELISA (details are provided in Table S7).

109S. Ray et al. / Journal of Proteomics 127 (2015) 103–113

P. falciparum infection as compared to the non-severe patients indicat-ing oxidative stress increases in severe cases possibly due to massivehemolysis. Moreover, increased serum abundances of serum amyloidassociated proteins, plasminogen, alpha 1 antitrypsin, CRP can be con-sidered as a consequence of inflammation-mediated acute phasereactions due to severe falciparum malaria [24,25].

Even though serum level of CRP is increased in many infectiveetiologies, elevated serum abundance of CRP has been demonstratedas a potential marker, which can be used to differentiate betweenmalaria and other infections such as dengue fever [26]. CRP binds tophosphocholine epitopes and activates the classical complement path-way, and has been studied as a prognostic marker in malaria from East-ern India [27]. C3b plays an important role in opsonisation of pathogensand the other fragments (C5a, C4a, and C3a) acting as anaphylatoxinshelp in activation of neutrophils and macrophages. Earlier studieshave demonstrated altered serum levels of C1, C2, C3 [28], and C4[29,30] and terminal complement complex [31] in malaria patients.In our proteomic analysis we have identified elevated serum levelsof various complement components, immunoglobulin components(including rheumatoid factor components), and factors involved incoagulation cascade in falciparum malaria patients. Malaria causes anacute systemic human disease leading to the release of inflammatorycytokines [32,33], and our findings indicate that falciparummalaria pa-tients exhibit higher level of inflammationwhen the transition happensfrom non-severe to severe infection.

Dysregulation in coagulation system is a feature of severe malaria,and consumptive coagulopathy, thrombocytopenia and impaired syn-thesis of the clotting factors are pathological reasons behind it. Severemalaria is associated with systemic inflammation, increased capillarypermeability, hypoxia, acidosis, endothelial activation and microvascu-lar coagulopathy. Malaria infection influences blood coagulation by

various pathobiological mechanisms and it can be a critical componentof infection [34]. Mohanty et al. proposed that the reduced level of pro-tein S and antithrombin III clotting factor consumption is due to themi-crovascular thrombosis rather than reduced synthesis in the liver [35].In our study, both the proteins exhibited reduced serum levels infalciparum malaria patients. Clemens et al. reported that the levels offactor XII and pre-kallikrein activities are significantly reduced in severefalciparum malaria [36]. In our study, similar trend was observed infalciparum malaria patients.

Interestingly, we have indentified quite a fewproteins such as serumamyloid A, C-reactive protein, lipopolysaccharide-binding protein,serum paraoxonase, haptoglobin, carbonic anhydrase 2, which exhibit-ed alterations in their serum abundance in severe falciparum as com-pared to non-severe falciparum malaria, which could be promisingindicators of disease severity. Muscle related proteins such as titin wasfound to be differentially abundant in severe malaria patients. Interest-ingly, alteration in serum abundance of galectin-3-binding protein wasfound to be non-significant in non-severe malaria, but was increased insera of severe falciparummalaria. Galectin-3-binding protein promotescell adhesion via integrin mediated pathway [37]. Consequently, its al-tered serum level may play some important role in protection of the in-fected RBCs against host responses.

Analysis of clinicopathological parameters indicated that severefalciparum malaria patients have lower level of mean hemoglobin(p b 0.05), WBC (p N 0.05) and platelet count (p b 0.05) in compar-ison to the patients suffering from non-severe infection (Table 1).To this end, our serum proteomic analysis indicated higherserum levels of hemoglobin chains in the severe falciparum malariacohorts pointing towards a greater level of haemolysis that iscorroborative with the finding of McKenzie et al. in P. vivax infec-tion [38].

ncbi-p:P02741ncbi-p:P18428ncbi-p:P04275ncbi-p:P02763ncbi-p:P05543ncbi-p:Q9UK55ncbi-p:P02750ncbi-p:Q06033ncbi-p:P35542ncbi-p:P02649ncbi-p:P00747ncbi-p:P05155ncbi-p:P02743ncbi-p:Q8WZ42ncbi-p:Q08380ncbi-p:P00450ncbi-p:P01023ncbi-p:P06396ncbi-p:P43652ncbi-p:P00915ncbi-p:P02753ncbi-p:P69905ncbi-p:P02766ncbi-p:P00918ncbi-p:P02652ncbi-p:P27169ncbi-p:P00738ncbi-p:P02647

Fig. 3. Synopsis of variousmodulated physiological pathways and panels of differentially abundant proteins in severe andnon-severe falciparummalaria. Green and red symbols representproteins that were down or up-regulated in falciparum malaria, respectively.

110 S. Ray et al. / Journal of Proteomics 127 (2015) 103–113

It is important to emphasize that several serum/plasma proteins,particularly those which are associated with acute phase responsesignaling cascades, frequently exhibit differential abundant in variousdiseased conditions [8]. Usually such candidates alone cannot be imple-mented as reliable diagnostic markers for any specific disease or infec-tion in a general population. However, the level and trend (up ordown) of alteration in serum abundances of the marker proteins is

also imperative while evaluating their clinical significance. For instance,serum level of haptoglobin (Hp) was found to be significantly lower inmalaria patients, whereas being a positive acute phase protein Hpexhibits elevated serum abundance in different cancers, infectious,and cardiovascular diseases [39]. Of note, establishment of a panel ofmarkers rather than a single candidate is much more effective fordetection of any specific infection and its discrimination from the

Image of Fig. 3

Fig. 4. Immunoassay-based and SPR validation, and ROC curve analysis of some selected differentially abundant proteins. (A) Measurement of serum levels of differentially abundantproteins (identified in the discovery phase of the study) in healthy controls and severe and non-severe falciparum malaria patients by ELISA assay. HC: Healthy control (n = 103),NSFM: non-severe falciparum malaria (n = 42), SFM: severe falciparum malaria (n = 39). (B) Receiver operating characteristic (ROC) curves indicating the accuracy of differentserum proteins for prediction of severe and non-severe falciparum malaria. (C) Calibration-free concentration analysis (CFCA) of serum amyloid A in healthy controls and severe andnon-severe falciparum malaria patients (pooled and individual samples) using surface plasmon resonance. ** Indicates p b 0.001, * indicates 0.001 b p b 0.05 and NS indicates p N 0.05based on a Mann–Whitney U test.

111S. Ray et al. / Journal of Proteomics 127 (2015) 103–113

other clinically related illnesses. To this end, in our earlier reports, wehave demonstrated that a panel of serum/plasma proteins consistingof six candidates (serum amyloid A, hemopexin, apolipoprotein E,

haptoglobin, retinol-binding protein 4 and apolipoprotein A-I) can dis-tinguish malaria from the other infectious diseases with overlappingclinical manifestations (dengue fever and leptospirosis) [13,40,41].

Image of Fig. 4

112 S. Ray et al. / Journal of Proteomics 127 (2015) 103–113

Consequently, it is rational to conjecture that a combination of serum/plasmamarkers, which display unique disease trend, alongwith clinico-pathological parameters can offer improved prediction accuracy.

Even though there are quite a few earlier reports of proteomic anal-ysis of plasma proteome alterations in African children with severe orcerebral falciparum malaria, there is a lack of similar high-resolutionproteomic datasets from Indian population. To this end, it should beemphasized that due to the extremely variable malaria epidemiologyin India and seemingly variable patterns of treatment response amongpatient populations, findings obtained from the populations fromother parts of the globe cannot be extrapolated directly on Indianpopulations for specifying the case definitions of severe and non-severe malaria. To the best of our information, this is the first compre-hensive study describing the serum proteomic alterations indentifiedin severe P. falciparum infected patients from different malaria endemicregions of India. This study identified potential biomarkers for monitor-ing disease severity of P. falciparum infection and enhanced our under-standing of pathogenesis and host responses in this fatal parasiticdisease. Longitudinal analysis involving a follow-up analysis of the pa-tients after therapeutic interventions may provide additional insightsregarding the correlation of the identified markers with disease pro-gression and their efficacy as disease monitoring or prognostic markers,which could be an informative future continuance of the presentinvestigation.

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.jprot.2015.04.032.

Competing interests

The authors have declared that no competing interests exist.

Acknowledgments

The active support from Prajakta Gandhe from the Department ofClinical Pharmacology, Seth GS Medical College & KEM Hospital,Mumbai, Sumit Verma from theMedicine Department, Medical CollegeHospital Kolkata, and Dharmendra Rojh from the Department ofMedicine, Malaria Research Center, S.P. Medical College, Bikaner in clin-ical sample collection process is gratefully acknowledged.Wewould liketo thank Mayuri N. Gandhi and Manali Jadhav from the Centre forResearch in Nanotechnology & Science (CRNTS), Indian Institute ofTechnology Bombay, Mumbai for their help in performing the Q-TOFLC/MS-MS experiments, and Krishnatej Nishtala and Bini Ramachandranfrom Thermo Fisher Scientific India Pvt Ltd, Mumbai for their support inperforming Q-Exactivemass spectrometric analysis. We are also gratefulto Sandip K. Patel, Apoorva Venkatesh, and Veenita G. Shah from theDepartment of Biosciences and Bioengineering, Indian Institute ofTechnology Bombay for their assistance in performing the MS and SPRexperiments. This research was supported by the Board of Research inNuclear Sciences (BRNS) DAE young scientist award (2009/20/37/4/BRNS and 2013/37B/24/BRNS) and a seed fund (12IRSGHC002)from the Healthcare Initiative of Industrial Research and ConsultancyCentre -IIT Bombay (IRCC-IITB) to SP and SS. SR was supported by theIIT Bombay fellowships and VK was supported by a CSIR fellowship.

References

[1] R.W. Snow, C.A. Guerra, A.M. Noor, H.Y. Myint, S.I. Hay, The global distribution ofclinical episodes of Plasmodium falciparum malaria, Nature 434 (2005) 214–217.

[2] WHO, WHO World Malaria Report, WHO, Geneva, 2008. (Available at http://www.who.int/malaria/wmr2008/malaria2008.pdf. Accessed, 2012).

[3] N.K. Shah, G.P. Dhillon, A.P. Dash, U. Arora, S.R. Meshnick, N. Valecha, Antimalarialdrug resistance of Plasmodium falciparum in India: changes over time and space,Lancet Infect. Dis. 11 (2011) 57–64.

[4] V. Singh, N. Mishra, G. Awasthi, A.P. Dash, A. Das, Why is it important to study ma-laria epidemiology in India? Trends Parasitol. 25 (2009) 452–457.

[5] Severe falciparum malaria. World Health Organization, Communicable DiseasesCluster, Trans. R. Soc. Trop. Med. Hyg. 94 (2000) S1–S90.

[6] E.A. Winzeler, Malaria research in the post-genomic era, Nature 455 (2008)751–756.

[7] J.M. Bautista, P. Marin-Garcia, A. Diez, I.G. Azcarate, A. Puyet, Malaria proteomics: in-sights into the parasite-host interactions in the pathogenic space, J. Proteome 97(2014) 107–125.

[8] S. Ray, S.K. Patel, V. Kumar, J. Damahe, S. Srivastava, Differential expression ofserum/plasma proteins in various infectious diseases: specific or nonspecific signa-tures, Proteomics Clin. Appl. 53–72 (2014).

[9] S. Ray, P.J. Reddy, R. Jain, K. Gollapalli, A. Moiyadi, S. Srivastava, Proteomic technol-ogies for the identification of disease biomarkers in serum: advances and challengesahead, Proteomics 11 (2011) 2139–2161.

[10] E.N. Gitau, G.O. Kokwaro, H. Karanja, C.R. Newton, S.A.Ward, Plasma and cerebrospi-nal proteomes from children with cerebral malaria differ from those of childrenwith other encephalopathies, J. Infect. Dis. 208 (2013) 1494–1503.

[11] J. Bachmann, F. Burte, S. Pramana, I. Conte, B.J. Brown, A.E. Orimadegun, et al., Affin-ity proteomics reveals elevated muscle proteins in plasma of children with cerebralmalaria, PLoS Pathog. 10 (2014) e1004038.

[12] F. Burte, B.J. Brown, A.E. Orimadegun, W.A. Ajetunmobi, F. Battaglia, B.K. Ely, et al.,Severe childhood malaria syndromes defined by plasma proteome profiles, PLoSOne 7 (2012) e49778.

[13] S. Ray, D. Renu, R. Srivastava, K. Gollapalli, S. Taur, T. Jhaveri, et al., Proteomic inves-tigation of falciparum and vivax malaria for identification of surrogate proteinmarkers, PLoS One 7 (2012) e41751.

[14] S. Sharma, S. Ray, A. Moiyadi, E. Sridhar, S. Srivastava, Quantitative proteomic anal-ysis of meningiomas for the identification of surrogate protein markers, Sci. Rep. 4(2014) 7140.

[15] S. Sharma, S. Ray, S. Mukherjee, A. Moiyadi, E. Sridhar, S. Srivastava, Multiprongedquantitative proteomic analyses indicate modulation of various signal transductionpathways in human meningiomas, Proteomics 15 (2015) 394–407.

[16] H. Mi, A. Muruganujan, J.T. Casagrande, P.D. Thomas, Large-scale gene function anal-ysis with the PANTHER classification system, Nat. Protoc. 8 (2013) 1551–1566.

[17] dWHuang, B.T. Sherman, R.A. Lempicki, Systematic and integrative analysis of largegene lists using DAVID bioinformatics resources, Nat. Protoc. 4 (2009) 44–57.

[18] L.H. Miller, D.I. Baruch, K. Marsh, O.K. Doumbo, The pathogenic basis of malaria,Nature 415 (2002) 673–679.

[19] K.E. Lyke, R. Burges, Y. Cissoko, L. Sangare, M. Dao, I. Diarra, et al., Serum levels of theproinflammatory cytokines interleukin-1 beta (IL-1beta), IL-6, IL-8, IL-10, tumor ne-crosis factor alpha, and IL-12(p70) in Malian children with severe Plasmodiumfalciparum malaria and matched uncomplicated malaria or healthy controls, Infect.Immun. 72 (2004) 5630–5637.

[20] F.A. Kassa, M.T. Shio, M.J. Bellemare, B. Faye, M. Ndao, M. Olivier, New inflammation-related biomarkers during malaria infection, PLoS One 6 (2011) e26495.

[21] S. Ray, K.S. Kamath, R. Srivastava, D. Raghu, K. Gollapalli, R. Jain, et al., Serum prote-ome analysis of vivax malaria: an insight into the disease pathogenesis and hostimmune response, J. Proteome 75 (2012) 3063–3080.

[22] S.H. Gillespie, C. Dow, J.G. Raynes, R.H. Behrens, P.L. Chiodini, K.P. McAdam, Mea-surement of acute phase proteins for assessing severity of Plasmodium falciparummalaria, J. Clin. Pathol. 44 (1991) 228–231.

[23] N.I. Krinsky, Mechanism of action of biological antioxidants, Proc. Soc. Exp. Biol.Med. 200 (1992) 248–254.

[24] B.B. Andrade, A. Reis-Filho, S.M. Souza-Neto, J. Clarencio, L.M. Camargo, A. Barral,et al., Severe Plasmodium vivax malaria exhibits marked inflammatory imbalance,Malar. J. 9 (2010) 13.

[25] A.W. Taylor-Robinson, Increased production of acute-phase proteins corresponds tothe peak parasitaemia of primary malaria infection, Parasitol. Int. 48 (2000)297–301.

[26] L. Epelboin, C. Boulle, S. Ouar-Epelboin, M. Hanf, P. Dussart, F. Djossou, et al., Dis-criminating malaria from dengue fever in endemic areas: clinical and biologicalcriteria, prognostic score and utility of the C-reactive protein: a retrospectivematched-pair study in French Guiana, PLoS Negl. Trop. Dis. 7 (2013) e2420.

[27] R. Paul, P.K. Sinha, R. Bhattacharya, A.K. Banerjee, P. Raychaudhuri, J. Mondal, Studyof C reactive protein as a prognostic marker in malaria from Eastern India, Adv.Biomed. Res. 1 (2012) 41.

[28] B.M. Greenwood, M.J. Brueton, Complement activation in childrenwith acute malar-ia, Clin. Exp. Immunol. 18 (1974) 267–272.

[29] S. Biryukov, J.A. Stoute, Complement activation in malaria: friend or foe? TrendsMol. Med. 20 (2014) 293–301.

[30] F.A. Neva, W.A. Howard, R.H. Glew, W.A. Krotoski, A.A. Gam, W.E. Collins, et al., Re-lationship of serum complement levels to events of the malarial paroxysm, J. Clin.Invest. 54 (1974) 451–460.

[31] M. Roestenberg, M. McCall, T.E. Mollnes, D.M. van, T. Sprong, I. Klasen, et al., Comple-ment activation in experimental human malaria infection, Trans. R. Soc. Trop. Med.Hyg. 101 (2007) 643–649.

[32] I.A. Clark, A.C. Budd, L.M. Alleva, W.B. Cowden, Human malarial disease: a conse-quence of inflammatory cytokine release, Malar. J. 5 (2006) 85.

[33] I.A. Clark, L.M. Alleva, A.C. Budd, W.B. Cowden, Understanding the role of inflamma-tory cytokines in malaria and related diseases, Travel Med. Infect. Dis. 6 (2008)67–81.

[34] I.M. Francischetti, K.B. Seydel, R.Q. Monteiro, Blood coagulation, inflammation, andmalaria, Microcirculation 15 (2008) 81–107.

[35] D.Mohanty, K. Ghosh, S.K. Nandwani, S. Shetty, C. Phillips, S. Rizvi, et al., Fibrinolysis,inhibitors of blood coagulation, and monocyte derived coagulant activity in acutemalaria, Am. J. Hematol. 54 (1997) 23–29.

[36] R. Clemens, C. Pramoolsinsap, R. Lorenz, S. Pukrittayakamee, H.L. Bock, N.J. White,Activation of the coagulation cascade in severe falciparum malaria through the in-trinsic pathway, Br. J. Haematol. 87 (1994) 100–105.

http://dx.doi.org/10.1016/j.jprot.2015.04.032http://dx.doi.org/10.1016/j.jprot.2015.04.032http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0005http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0005http://www.who.int/malaria/wmr2008/malaria2008.pdfhttp://www.who.int/malaria/wmr2008/malaria2008.pdfhttp://refhub.elsevier.com/S1874-3919(15)00232-8/rf0010http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0010http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0010http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0015http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0015http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0230http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0230http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0020http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0020http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0025http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0025http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0025http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0030http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0030http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0030http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0035http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0035http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0035http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0040http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0040http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0040http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0045http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0045http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0045http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0050http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0050http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0050http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0055http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0055http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0055http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0060http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0060http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0060http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0065http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0065http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0065http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0070http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0070http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0235http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0235http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0075http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0075http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0080http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0080http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0080http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0080http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0080http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0085http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0085http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0090http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0090http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0090http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0095http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0095http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0095http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0100http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0100http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0105http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0105http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0105http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0110http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0110http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0110http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0115http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0115http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0115http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0115http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0120http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0120http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0120http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0125http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0125http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0130http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0130http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0135http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0135http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0135http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0240http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0240http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0240http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0140http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0140http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0145http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0145http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0145http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0150http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0150http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0155http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0155http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0155http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0160http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0160http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0160

113S. Ray et al. / Journal of Proteomics 127 (2015) 103–113

[37] R.C. Hughes, Galectins asmodulators of cell adhesion, Biochimie 83 (2001) 667–676.[38] F.E. McKenzie, G.M. Jeffery, W.E. Collins, Plasmodium vivax blood-stage dynamics, J.

Parasitol. 88 (2002) 521–535.[39] I.K. Quaye, Haptoglobin, inflammation and disease, Trans. R. Soc. Trop. Med. Hyg.

102 (2008) 735–742.[40] S. Ray, R. Srivastava, K. Tripathi, V. Vaibhav, S. Patankar, S. Srivastava, Serum prote-

ome changes in dengue virus-infected patients from a dengue-endemic area ofIndia: towards new molecular targets? OMICS 16 (2012) 527–536.

[41] R. Srivastava, S. Ray, V. Vaibhav, K. Gollapalli, T. Jhaveri, S. Taur, et al., Serum profilingof leptospirosis patients to investigate proteomic alterations, J. Proteome 76 (2012)56–68.

[42] O.A. Skorokhod, E. Schwarzer, M. Ceretto, P. Arese, Malarial pigment haemozoin,IFN-gamma, TNF-alpha, IL-1beta and LPS do not stimulate expression of induciblenitric oxide synthase and production of nitric oxide in immuno-purified humanmonocytes, Malar. J. 6 (2007) 73.

[43] W. Graninger, F. Thalhammer, U. Hollenstein, G.M. Zotter, P.G. Kremsner, Serum pro-tein concentrations in Plasmodium falciparummalaria, Acta Trop. 52 (1992) 121–128.

[44] E. Baum, K. Badu, D.M.Molina, X. Liang, P.L. Felgner, G. Yan, Protein microarray anal-ysis of antibody responses to Plasmodium falciparum in western Kenyan highlandsites with differing transmission levels, PLoS One 8 (2013) e82246.

[45] G. Arinola, C. Ezeh, C1 inhibitor, C3 activator, IgG, IgA, and IgM titers in Nigeriansickle cell disease patients with Plasmodium falciparum, Iran. J. Immunol. 4 (2007)44–49.

[46] P. Balmer, F. McMonagle, J. Alexander, P.R. Stephen, Experimental erythrocytic ma-laria infection induces elevated serum amyloid P production in mice, Immunol. Lett.72 (2000) 147–152.

[47] S. Reungprapavut, S.R. Krungkrai, J. Krungkrai, Plasmodium falciparum carbonicanhydrase is a possible target for malaria chemotherapy, J. Enzyme Inhib. Med.Chem. 19 (2004) 249–256.

[48] F.J. Rosales, J.D. Topping, J.E. Smith, A.H. Shankar, A.C. Ross, Relation of serum retinolto acute phase proteins andmalarial morbidity in Papua NewGuinea children, Am. J.Clin. Nutr. 71 (2000) 1582–1588.

http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0165http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0170http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0170http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0175http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0175http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0180http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0180http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0180http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0185http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0185http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0185http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0190http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0190http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0190http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0190http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0195http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0195http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0200http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0200http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0200http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0205http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0205http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0205http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0210http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0210http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0210http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0215http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0215http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0215http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0220http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0220http://refhub.elsevier.com/S1874-3919(15)00232-8/rf0220

Proteomic analysis of Plasmodium falciparum induced alterations in humans from different endemic regions of India to deciph...1. Introduction2. Materials and methods2.1. Subject recruitment, blood collection, and serum separation2.2. Sample processing, 2DE and 2D-DIGE2.3. In-gel digestion and MALDI TOF/TOF analysis for protein identification2.4. In-solution digestion, iTRAQ labeling and OFFGEL fractionation2.5. LC-MS/MS analysis for protein identification and quantitation2.6. Protein networks and bioinformatic analysis2.7. ELISA and receiver operating characteristic (ROC) curve analysis2.8. SPR-based quantification of active protein concentration of serum amyloid A in serum samples

3. Results3.1. Alterations in clinicopathological parameters in falciparum malaria patients3.2. Differential abundance of serum proteins in falciparum malaria identified in gel-based proteomics analysis3.3. Identification of differentially abundant serum proteins in falciparum malaria by employing iTRAQ-based quantitative p...3.4. Modulation of physiological pathways in severe and non-severe falciparum malaria3.5. Validation of selected differentially abundant serum proteins

4. DiscussionCompeting interestsAcknowledgmentsReferences