Characterization of salivary proteins of schizophrenic and bipolar disorder patients by top-down proteomics

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/authorsrights

http://www.elsevier.com/authorsrights

Author's personal copy

Characterization of salivary proteins ofschizophrenic and bipolar disorder patients bytop-down proteomics

Federica Iavaronea, Marianna Melisb, Giovanna Plataniab, Tiziana Cabrasc,Barbara Manconic, Raffaele Petruzzellid, Massimo Cordaroe, Alberto Siracusanob,Gavino Faaf, Irene Messanac, Marco Zanasib, Massimo Castagnolaa,⁎aIstituto di Biochimica e Biochimica Clinica, Facoltà di Medicina, Università Cattolica e Istituto di Chimica del Riconoscimento Molecolare CNR,Roma, ItalybDip. di Medicina dei Sistemi, Unità Operativa di Psichiatria, Univ. di Tor Vergata, Roma, ItalycDip. di Scienze della Vita e dell'Ambiente, Sezione di Biochimica e Biologia Molecolare Università di Cagliari, ItalydDip. di Scienze Sperimentali e Cliniche, Univ. G. D'Annunzio, Chieti, ItalyeIstituto di Clinica Odontostomatologica, Facoltà di Medicina, Università Cattolica, Roma, ItalyfDip. di Scienze Chirurgiche, Sez. Anatomia Patologica, Università di Cagliari, Italy

A R T I C L E I N F O A B S T R A C T

Article history:Received 23 December 2013Accepted 9 March 2014Available online 30 March 2014

The analysis of whole saliva of 32 subjects with diagnosis of schizophrenia (SZ), 17 withdiagnosis of bipolar disorder (BD), and 31healthy subjects divided in non-smokers (HN; n = 19)and smokers (HS; n = 12) using an HPLC–ESI-MS top-down platform is reported in this study.Both SZ and BD revealed more than 10 fold mean increase of α-defensins 1–4, S100A12,cystatin A and S-derivatives of cystatin B levels with respect to the HN and HS control groups.Nodifferences of protein levelswere observed between SZandBDgroups andbetweenHNandHS groups. Moreover, the correlation coefficients among the different proteins weresignificantly better in BD group than in SZ group.

Biological significanceThis study on whole saliva confirms a schizophrenia-associated dysregulation of immunepathway of peripheral white blood cells and suggests that the dysregulation of BD groupcould involve the activation of more specific cell type than that of SZ group.

© 2014 Elsevier B.V. All rights reserved.

Keywords:Whole salivaAlpha-defensinsS100 A12CystatinsSchizophreniaBipolar disorder

J O U R N A L O F P R O T E O M I C S 1 0 3 ( 2 0 1 4 ) 1 5 – 2 2

Abbreviations: SZ, schizophrenia; BD, bipolar disorder; HN, healthy non-smokers; HS, healthy smokers; DUP, duration of untreatedpsychosis; XIC, extracted ion current.⁎ Corresponding author at: Istituto di Biochimica e Biochimica Clinica, Facoltà di Medicina, Università Cattolica, Largo F. Vito, 00168 Roma,

Italy. Tel./fax: +39 06 3053598.E-mail address: [email protected] (M. Castagnola).

http://dx.doi.org/10.1016/j.jprot.2014.03.0201874-3919/© 2014 Elsevier B.V. All rights reserved.

Ava i l ab l e on l i ne a t www.sc i enced i r ec t . com

ScienceDirect

www.e l sev i e r . com/ loca te / j p ro t


1. Introduction

Schizophrenia is a severe psychotic illness affecting about 0.3–0.7% (prevalence lifetime) of the general population and one ofthemajor causes of disability in young and adult people, aswellas a serious social burden worldwide. Because of a greatoverlapping of the symptomatology with other psychiatricillness, it is not characterized by specific pathological features.The disorder is diagnosed on the basis of a clinical interviewand using criteria of the diagnosis manuals, such as Diagnosticand Statistical Manual of Mental Disorders IV TR; 2000 (DSM)published by the American Psychiatric Association and Inter-national Statistical Classification of Diseases and RelatedHealth Problems (ICD) produced by theWorld Health Organiza-tion (WHO).

Because of the variable symptomatology, clinicians com-monly get to diagnose schizophrenia after many months andsometimes more. The duration of untreated psychosis ispositively relied to worse disease progression and outcomeand lasts one year on average.

Bipolar mood disorder is another severe psychiatric illness,affecting the 1% of the population; the lifetime risk of sufferingfromabipolar spectrumdisorder is 2% in the general population(DSM IV). Bipolar mood disorder is associated with severedisability in lifetime and increased risk of committing suicide.Despite more specific pathological features of this disorder,the duration of untreated illness is, as in schizophrenia, animportant predictive element for the disease progression andoutcomes.

These preliminary remarks follows the importance of charac-terizing analyte variations in bodily fluids of subjects sufferingfrom these diseases that could be used in the future for thedevelopment of predictive diagnostic tests and could help tounderstand the molecular events underlying these disorders [1].

Recently, several studies showed the presence of significantalterations in the immune system of bipolar and schizophrenicpatients. Hope et al. found that the elevated plasma levels ofinflammatory markers, particularly IL-1 receptor antagonist(IL-1Ra) and soluble tumor necrosis factor receptor (sTNF-R1),were associated with general disease severity and psychoticfeatures in schizophrenia and bipolar disorder [2]. This isconsistent with the results of Dickerson et al. [3] that showeda relationship between the elevated level of C-reactive proteinand schizophrenia and the rate of C-reactive protein and theseverity of affective symptoms in patients with bipolar disorderin manic phase [4]. These studies support a central role ofimmune activation in the core pathological mechanisms ofsevere mental disorders.

Also some proteomic studies have been carried out asrecently reviewed by de-Souza and colleagues [5] with theintent to identify specific markers of the disease.

A proteomic study carried out with SELDI-TOF-MS and ELISAonanti-CD3 stimulated andunstimulated peripheral bloodT celllysates evidenced that α-defensins contributed significantly tothe separation of schizophrenic patients and control groups [6].ELISA analysis of plasma samples in affected and unaffectedmonozygotic twins confirmed significantly elevated α-defensinlevels when compared to healthy unaffected twins. Stimulatedby these findings we have carried out a top-down proteomicanalysison the acidic soluble solutionofwhole saliva of different

individualswith psychiatric illness (schizophrenia, bipolarmooddisorder) comparing them against a group of healthy controlsfurther divided in smokers and non-smokers. Indeed, saliva is abodily fluid attractive for the non-invasive specimen collection[7] and it can be sometimes considered a good substitute ofplasma and blood, with a particular concern for the detection ofbiomarkers also in the pediatric age [8]. Top-down proteomicanalysis of this bodily fluid has allowed us the characterizationin the last years ofmore than250naturally peptides andproteins[9]. Top-down proteomic platform can discriminate betweendifferent isoforms and it is able to distinguish their levels by theextracted ion current (XIC) procedure [10]. Attracted by theadvantages of top-down platforms other groups are applyingthem to the study of salivary proteome [11–14]. This studyconfirmed increased levels of α-defensins 1–4 and evidencedincreased levels of S100A12, cystatin A and S-derivatives ofcystatin B [15] in whole saliva of individuals with psychiatricillness, confirming some of the results obtained by bloodanalysis and, once again, demonstrating that whole saliva canbe a good alternate for the detection of some relevant plasmaclinical analytes.

2. Methods and materials

2.1. Sample collection

Resting whole saliva (from 0.2 to 1.0 mL) was collected with asoft plastic aspirator at the basis of the tongue from 10.00 to12.00 a.m. Samples were collected at least 30 min after anyfood or beverage had been consumed and teeth had beencleaned. After collection salivary samples were immediatelymixed with an equal volume of 0.2% 2,2,2 trifluoroacetic acid(v/v; TFA) in an ice bath. After stirring, the acidic solution wascentrifuged at 9000 g for 5 min to remove the precipitate andthe acidic clear solution was either immediately analyzed byHPLC–ESI-MS (100 μL, corresponding to 50 μL of saliva) orstored at −80 °C until analysis.

2.2. Participants and ethics statements

The study protocol and written consent forms were approvedby the Medical Ethics Committee of the Faculty of Medicine ofthe Catholic University of Rome. Informed consent forms werefilled out and all the rules have been complied according to theinstructions of the Declaration of Helsinki. Exclusion criterionwas the use of drugs of abuse. Whole saliva was collectedaccording to the protocol described in the previous section in 32subjects with a diagnosis of schizophrenia (SZ) and 17 subjectswith a diagnosis of bipolar disorder (BD) classified according tothe guidelines of Diagnostic and Statistical Manual of MentalDisorders IV TR, (2000) American Psychiatric Association, and31 healthy subjects (HT), further divided into non-smoker (HN;n = 19) and smoker (HS; n = 12) groups.

2.3. Reagents and apparatus

Chemicals and reagents, all of LC–MS grade, were purchasedfrom J.T.Baker (Deventer the Netherlands), Merck (Darmstadt,

16 J O U R N A L O F P R O T E O M I C S 1 0 3 ( 2 0 1 4 ) 1 5 – 2 2


Germany) andSigmaAldrich (St. Louis,MI, USA).HPLC–ESI-IT-MSapparatuswas a Surveyor HPLC system (ThermoFisher, San Jose,CA, USA) connected by a T splitter to a PDA diode-array detectorand to an LCQ Deca XP Plus mass spectrometer. The massspectrometer was equipped with an ESI source. The chromato-graphic column was a Zorbax SB300 C8 (Agilent) column, with5 μm particle diameter (column dimensions 150 × 2.1 mm).

2.4. RP-HPLC–ESI-MS analysis

The following solutions were utilized for the low resolutionchromatographic separation: (eluent A) 0.056% aqueous TFA and(eluent B) 0.050% TFA in acetonitrile–water 80/20 (v/v). Thegradient appliedwas linear from0 to 55% in 40 min, at a flow rateof 0.30 mL/min. The T splitter addressed a flow-rate of about

Fig. 1 – Example of a XIC procedure for the relative quantitation of α-defensins 1–4 in one typical sample of schizophrenicwhole saliva. The panels from A(XIC) to D(XIC) show the extraction from the HPLC–ESI-MS profile of the m/z values characteristicof α-defensin 1, α-defensin 2, α-defensin 3 and α-defensin 4, respectively. The panels A, B, C(ESI) and D(ESI) show thecorresponding ESI spectra and the panels A, B, C(Decon) and D(Decon) show their deconvolution. The A, B, C(ESI) spectrum ofα-defensins 1–3 (and its deconvolution) is unique for the coelution of the three peptides. RT: retention time; MA: measuredarea; AV: average; NL: normalization level.

17J O U R N A L O F P R O T E O M I C S 1 0 3 ( 2 0 1 4 ) 1 5 – 2 2


0.20 mL/min towards the diode array detector and 0.10 mL/mintowards the ESI source. During the first 5 min of separation theeluate was not addressed to the mass spectrometer to avoidinstrument damage due to the high salt concentration. Thediode array detector was set at wavelengths of 214 and 276 nm.Mass spectrawere collected every 3 ms in the positive ionmode.MS spray voltagewas 4.50 kV and the capillary temperature was250 °C.

2.5. Data analysis

Deconvolution of averaged ESI mass spectra was automaticallyperformed by Xcalibur or by MagTran 1.0 software [16].

Experimental mass values were compared with average theo-retical values available at the Swiss-Prot data bank (http://us.expasy.org/tools), with the following accession numbers:α-defensins 1/2: P59665; α-defensins 2/3: P59666; α-defensin 4:P12838; S100A12: P80511; cystatin A P01040; cystatin B: P04080.Relative abundance of the salivary proteins and peptides wasdetermined by considering the XIC peak area and relating it to1.0 mL of saliva. This value is linearly proportional to thepeptide concentration and it can be used to monitor relativeabundances, under constant analytical conditions [17].

In determining the XIC peak area the right choice of m/zvalues for the detection of the protein of interest is relevantto avoid m/z of ESI potentially overlapping spectra belonging

Fig. 2 – Values of the [XIC peak area]/mL of saliva (arbitrary unit) of α-defensin 1, α-defensin 2, α-defensin 3, α-defensin 4,S100A12, cystatin A, cystatin B S-glutathionyl and S-cysteinyl derivativesmeasured in the different groups (schizophrenia (SZ),bipolar disorder (BD), healthy subjects (HT) divided into non-smokers (HN) and smokers (HS)).



to other proteins that elute very close in crowded areas ofchromatographic elution,. The m/z values chosen were: α-defensin 1, 1722.0 (2+), 1148.4 (3+), 861.5 (4+) m/z (Exp. averagemass: 3442 ± 1, Theor. 3442.0 Da, Elution time: 23.2–23.8 min);α-defensin 2, 1646.5 (2+), 1124.7 (3+), 843.7 (4+)m/z (Exp. averagemass: 3371 ± 1, Theor. 3371.0 Da, Elution time: 23.2–23.8 min);α-defensin 3, 1744.0 (2+), 1163.0 (3+), 872.5 (4+)m/z (Exp. averagemass: 3486 ± 1, Theor. 3486.1 Da, Elution time: 23.2–23.8 min);α-defensin 4, 1855.7 (2+), 1237.5 (3+), 928.5 (4+)m/z (Exp. averagemass: 3709 ± 1, Theor. 3709.4 Da, Elution time: 26.9–27.5 min);S100A12, 1306.5 (8+), 1161.4 (9+), 1045.4 (10+), 950.4 (11+)m/z (Exp.average mass: 10,444 ± 2, Theor. 10,443.8 Da, Elution time: 39.8–40.2 min); cystatin A, 1835.4 (6+), 1573.4 (7+), 1376.8 (8+), 1224.0(9+), 1101.7 (10+), 1001.6 (11+)m/z (Exp. averagemass: 11,006 ± 2,Theor. 11,006.5 Da, Elution time: 31.4–32.2 min); cystatin BS-glutathionylated, 1915.5 (6+), 1642.0 (7+), 1436.9 (8+), 1277.3(9+), 1149.7 (10+), 1045.3 (11+)m/z (Exp. averagemass: 11,487 ± 2,Theor. 11,486.7 Da, Elution time: 32.5–33.1 min); and cystatin BS-cysteinylated, 1884.5 (6+), 1615.4 (7+), 1413.6 (8+), 1256.6 (9+),1131.1 (10+), 1028.3 (11+) m/z (Exp. average mass: 11,301 ± 2,Theor. 11,300.7 Da, Elution time: 32.6–33.2 min). The window forall these values was in a range of ±0.5 m/z. The percentage errorof the measurements was less than 10%. Student T-test wasperformed to establish the statistical significant differencesamong the different groups and Pearson's R linear correlationcoefficient was computed to establish the relationship existingbetween the different proteins and peptides within any group.

3. Results

Proteins and peptides specific of saliva (acidic and basisproline-rich proteins, histatins, statherin, PB peptide andS type cystatins), as well as α-defensins 1–4, cystatin A,S-glutathionylated and S-cysteinylated cystatin B, the mainderivatives of cystatin B [15] and proteins belonging to theclass of S100, such as S100A7, S100A8, S100A9 and S100A12and all their known derivatives [8,19] were detected asnaturally occurring peptides in the HPLC–ESI-MS profile ofwhole saliva acidic soluble fraction of healthy subjects,schizophrenic and bipolar patients and quantified by XICprocedures as previously described [8,18]. The levels of theseproteins did not show any statistically significant differenceamong the groups under study except for α-defensins 1–4,S100A12, cystatin A and the S-derivatives of cystatin B. Asimilar concentration of the major salivary proteins in thehealthy subjects and in the patients was allowed assumingthat the flow rate was not different in the groups under study.

The average masses and the m/z values utilized to extractthe ion current (XIC) peaks of the peptides and proteinsdisplaying a significant difference among the groups arereported in the section Data analysis of Material and methods.Some examples of the XIC procedure are shown in Fig. 1. Thevalues of the [XIC peak area]/mL saliva of the peptides andproteins measured in the different groups are reported in Fig. 2and Table 1. Student T-test analysis results are reported inTable 2. Table 3 shows the Pearson's correlation coefficients (R)

Table 1 – Mean values of the [XIC peak area]/mL saliva (±standard deviation) (arbitrary unity × 109) detected on the differentgroups under study.

SZ (N = 32) BD (N = 17) HT (N = 31) HN (N = 19) HS (N = 12)

α-Defensin 1 50.4 ± 61.3 (32/32) 36.4 ± 42.7 (17/17) 4.56 ± 9.27 (18/31) 2.28 ± 3.59 (9/19) 8.18 ± 13.8 (9/12)α-Defensin 2 33.8 ± 36.9 (32/32) 25.2 ± 32.9 (17/17) 2.92 ± 5.61 (16/31) 1.71 ± 2.72 (8/19) 4.82 ± 8.18 (8/12)α-Defensin 3 29.0 ± 27.8 (31/32) 29.3 ± 52.1 (17/17) 1.65 ± 2.89 (13/31) 1.63 ± 2.88 (8/19) 1.69 ± 3.06 (5/12)α-Defensin 4 10.2 ± 9.62 (30/32) 7.75 ± 8.78 (16/17) 0.25 ± 0.65 (6/31) 0.41 ± 0.80 (6/19) −(0/12)S100 A12 17.0 ± 19.6 (22/32) 17.5 ± 19.7 (16/17) 0.14 ± 0.35 (5/31) 0.09 ± 0.28 (2/19) 0.21 ± 0.42 (3/12)Cystatin A 16.6 ± 10.7 (32/32) 24.2 ± 22.6 (17/17) 4.50 ± 4.48 (31/31) 3.03 ± 3.33 (19/19) 6.45 ± 5.37 (12/12)Cystatin B glutathionyl 3.93 ± 3.43 (28/32) 5.86 ± 5.71 (17/17) 1.72 ± 2.12 (30/31) 1.59 ± 2.08 (18/19) 1.77 ± 2.24 (12/12)Cystatin B cysteinyl 1.70 ± 1.59 (28/32) 2.90 ± 3.12 (17/17) 0.73 ± 0.89 (30/31) 0.71 ± 1.01 (18/19) 0.68 ± 0.69 (12/12)

Table 2 – Cross Student T-test analysis among the differ-ent groups.

BD HN HS HT

α-Defensin 1SZ n.s <0.0002 <0.0008 <0.0003BD <0.005 <0.02 <0.008HN n.s.

α-Defensin 2SZ n.s <0.00003 <0.0002 <0.00006BD <0.01 <0.03 <0.02HN n.s.

α-Defensin 3SZ n.s. <0.00001 <0.00001 <0.00001BD <0.06 <0.06 <0.06HN n.s.

α-Defensin 4SZ n.s. <0.000005 <0.000005 <0.000005BD <0.004 <0.003 <0.003HN <0.04

S100 A12SZ n.s. <0.00003 <0.00004 <0.00004BD <0.003 <0.003 <0.003HN n.s.

Cystatin ASZ n.s. <0.0000002 <0.0002 <0.0000006BD <0.002 <0.006 <0.003HN <0.08

Cystatin B glutathionylatedSZ n.s. <0.007 <0.03 <0.003BD <0.002 <0.002 <0.001HN n.s.

Cystatin B cysteinylatedSZ n.s. <0.02 <0.006 <0.005BD <0.02 <0.02 <0.02HN n.s.



computed among the different proteins in the same group,where the control group (HT) comprehends both smokers (HS)and non-smokers (HN). The R computation was performedeliminating from the statistical analysis the couples whereat least one of the values resulted under the sensitivity ofthe method. Data indicated a very significant increase ofconcentration for all the peptides and proteins listed inTables 1 and 2 in SZ and BD patients when compared againstthe control groups (HN and HS), with the exception ofα-defensin 3 which showed in the BD group a concentrationslightly lower, even though not significant, with respect to thecontrol groups. No significant differences for all the proteinswere observed between the HN and HS groups, except forα-defensin 4, which was not detected in the HS group. Nostatistical significant difference was observed between SZ andBD groups. Of particular interest were the Pearson's correlationcoefficients reported in Table 3. In all the three groups thecorrelation between α-defensin 1 and α-defensin 2 was alwaysfound more significant than between α-defensin 2 andα-defensin 3. Because α-defensin 2 can be generated by eitherα-defensin 1 or α-defensin 3 after the loss of the N-terminalamino acid residue, the higher correlation between α-defensins1 and 2 strongly suggests that the latter derives mainly fromα-defensin 1. The significant correlations determined betweencystatin A and S-derivatives of cystatin B in all the threegroups confirmed, as previously reported [15], a relationshipbetween the expression of the two cystatins. The very highcorrelation observed among all the proteins under study in theBD group with respect to the SZ and HT groups was alsoremarkable.

4. Discussion

All the proteins and peptideswhich significantly increased in theSZ and BD groups with respect to control are involved in theinnate immunity. Defensins are small cationic peptides withanti-bacterial, antiviral and immunomodulatoryproperties. Theyare divided in three subfamilies, named α, β and θ-defensinsclassified according to the connectivity of three intramoleculardisulfide bonds generated by six conserved cysteine residues [20].The four α-defensins 1–4 detected in this study are also calledHNP-1 to HPN-4 (HNP: human neutrophil peptide) because theywere firstly isolated from neutrophils [21], but they have beensubsequently recognized as the secretion products ofmonocytes,B cells, αβ and γδ T cells, and natural killer cells [21,22].α-Defensins 5 and 6 (also called HD-5 and HD-6) are the productsof Paneth cells and are commonly secreted in the intestine [23].α-Defensins 1 and 3 are the products of DEFA1 and DEFA3 genesanddiffer only at theN-terminal residue (Ala inHNP-1 andAsp inHPN-3). α-Defensin 2which is not a product of a specific gene canoriginate both from HNP-1 and HPN-3 by enzymatic loss of theN-terminal amino acid residue. DEFA1 and DEFA3 genes showextensive copy number polymorphism, and the number of genecopy per diploid genome varies between 4 and 14 with a meanvalue of 10. On the contrary, the DEFA4 gene, responsible for theexpression of α-defensin 4, is not duplicated and only two copiesexist in a diploid genome. The relative amount of α-defensin 4 istherefore noticeably lower and its structure is somewhatdifferent from that of α-defensins 1–3 [20]. The amount ofα-defensin 4 in whole saliva of HT was frequently under the

Table 3 – Pearson's corr. coeff. (R) computed among the different proteins in the same group.

α-Def 2 α-Def 3 α-Def 4 S100A12 Cyst A Cyst B Glut Cyst B Cyst

SZ group (N = 32)α-Def 1 0.981*** (32) 0.679*** (30) 0.563*** (29) 0.794*** (22) 0.247 (31) 0.395* (28) 0.279 (28)α-Def 2 0.696*** (30) 0.545** (29) 0.776*** (22) 0.233 (31) 0.412* (28) 0.269 (28)α-Def 3 0.524* (28) 0.784*** (22) 0.059 (29) 0.481* (26) 0.218 (26)α-Def 4 0.574** (21) 0.295 (28) 0.294 (26) 0.453* (25)S100A12 −0.038 (22) 0.297 (21) 0.198 (21)Cyst A 0.408* (28) 0.391* (28)Cyst B Glut 0.759*** (28)

BD group (N = 17)α-Def 1 0.979*** (17) 0.799*** (17) 0.918*** (16) 0.865*** (16) 0.741*** (17) 0.705** (17) 0.586** (17)α-Def 2 0.859*** (17) 0.933*** (16) 0.860*** (16) 0.665** (17) 0.710*** (17) 0.590** (17)α-Def 3 0.894*** (16) 0.708*** (16) 0.519* (17) 0.599** (17) 0.601** (17)α-Def 4 0.831*** (15) 0.697** (17) 0.691** (17) 0.602** (16)S100A12 0.769*** (16) 0.831*** (16) 0.819*** (16)Cyst A 0.698*** (17) 0.770*** (17)Cyst B Glut 0.859*** (17)

HT group (N = 31)α-Def 1 0.993*** (16) 0.793*** (13) 0.964*** (6) – (3) 0.200 (18) 0.392 (18) −0.016 (18)α-Def 2 0.789*** (13) 0.989*** (5) – (3) 0.170 (16) 0.374 (16) −0.070 (16)α-Def 3 0.535 (5) – (3) 0.700** (13) 0.901*** (13) 0.143 (13)α-Def 4 – (2) 0.297 (7) 0.369 (6) 0.060 (5)S100A12 −0.097 (5) −0.588 (5) 0.130 (5)Cyst A 0.642*** (30) 0.543*** (30)Cyst B Glut 0.276 (30)

p < 0.05 = *; p < 0.01 = **; p < 0.001 = ***.



detection limit of the method (26 out of 31). On the contrary inwhole saliva of SZ and BD α-defensin 4 was almost alwaysdetectable (45 out of 49). In SZ and DB the relative amount ofα-defensin 4 was on average 5 times lower than that ofα-defensins 1–3. It is reported that in human neutrophils therelative amounts of α-defensins 1–3 and α-defensin 4 occur in aratio sensibly higher than the modal ratio [24]. In this study, thelevels ofα-defensins 1–3 versusα-defensin 4 in SZandBDwere ina ratio of ca. 5:1, in agreement with the mean modal ratio of thehuman genes, suggesting that their increased concentrationcould not only be linked to an increased neutrophil activity, butcould rather be derived also from the contribution of other cellsdevoted to the adaptive immunity.

S100A12, also called calgranulin C, is a member of the S100family of EF-hand calcium-binding proteins. Human S100A12 ispredominantly expressed and secreted by neutrophil too and,therefore, has been assigned to the S100 protein subfamily ofcalgranulins or myeloid-related proteins. Intracellular S100A12exists as an anti-parallel homodimer and upon calcium-dependent activation interacts with target proteins to regulatecellular functions. Extracellular S100A12 shows cytokine-likecharacteristics and recruits inflammatory cells to the sites oftissue damage [25]. Human S100A12 ismarkedly overexpressedin inflammatory compartments where neutrophils and mono-cytes belong to the most abundant cell types [26], and elevatedserum levels of S100A12 are found in patients suffering fromvarious inflammatory, neurodegenerative, metabolic, and neo-plastic disorders. Recent clinical evidence suggests a highpotential of S100A12 as a sensitive and specific diagnosticmarker of localized inflammatory processes [27].

Cystatins A and B (also called stefins) are members of thetype 1 cystatin (inhibitors of cysteine proteinase family). Type 1cystatins are α/β-type globular proteins differing from type 2cystatins for size (about 100 amino acid residues in type 1 versus120 residues in type 2), absence of disulfide bonds (two bonds intype 2 cystatins) and phosphorylation [28]. Cystatin A wasinitially detected as secretion product of polymorphonucleargranulocytes, subsequently was detected in all leukocytefamilies, further than in many other organ and tissues [29].Cystatin B is also expressed by different leukocytes, such asneutrophils, monocyte macrophages and lymphoid cells. Manystudies are evidencing the role of cystatins A and B in theimmune system. Recently, we have verified that cystatin B isusually present inwholehumansalivamainly as S-glutathionyland S-cysteinyl derivatives [15].

The significant increase of salivary concentrations of allthe above reported proteins and peptides in SZ and BDsubjects confirms that these patients display a strong activationof the innate immune system. Moreover, the very significantand higher correlations found between the levels of theseproteins in BD subjects with respect to the SZ group stronglysuggest that the activation of the immune system in BDsubjects could be linked to more specific cell types than in SZsubjects. The results of this study are stimulus for furtherstudies, in order to establish the different ratios of expression ofα-defensins, S100A12 and stefins in the different leukocytefamilies.

Conjectures about the involvement of the immune systemin schizophrenia and bipolar disorder are controversial. It couldbe a consequence of the pathology, could contribute to the

pathology or could be one of the events at the basis of thedisease.

Several studies showed the presence of significant alter-ations in the immune system in schizophrenic and bipolarpatients [2–4]. Although the concentration in saliva does notreflect directly from the concentration in blood or serum, theresults of our study are in agreement with those of Craddockand colleagues' [6] increased α-defensin levels in plasma ofschizophrenic patients. The same authors from the study ofplasma of affected and non-affected discordant monozygotictwins hypothesized that α-defensins are more associated tothe risk of developing rather than an outcome of the disease.

Increased levels of α-defensins, S100A12 and cystatinsA and B in different bodily fluids and tissues have beenreported in other pathologies, such as neoplasies and infec-tious diseases [20,27,29]. Thus, the use of these salivarypeptides and proteins as potential predictive biomarkers inSZ and BD could be characterized by low selectivity andspecificity.

The mean levels of α-defensins 1–3 in whole saliva ofschizophrenic patients are remarkable and about one order ofmagnitude higher with respect to controls, while the increaseof α-defensin 4 was not computable, because this defensinwas frequently undetectable in the healthy subjects. Furtherstudies will be necessary to verify whether so remarkableincrease is comparable or higher than that observed in otherlocal or systemic affections.

5. Conclusion

Conclusively, this study demonstrates again that the HPLC–ESI-MS top-downplatforms providedetailed identifications andrelative quantitation of peptides with very similar structures,and evidences that human saliva could be a precious bodilyfluid for the characterization of putative biomarkers of systemicand multifactorial diseases, provided well standardized proce-dures. In adifference to theblood and cerebrospinal fluid,wholesaliva collection is non-invasive; it can be carried out withoutthe direct intervention of health-care specialists and it does notprovoke any pain. Results obtained are preliminary and addresstowards further studies to investigate the role of confoundingfactors, as the use of drugs of abuse, common in BPD in all thedisease stages (maniacal, depressive and euthymic). In order toconsider α-defensins 1–4, S100A12 and cystatins A and B aspotential markers of psychiatric pathology it is necessary in thefuture to evaluate their sensitivity and specificity, especiallywith respect to neurodegenerative diseases as Parkinson andAlzheimer diseases.

Conflict of interest

The authors have declared no conflicts of interest.

Acknowledgment

The authors acknowledge the financial support of CagliariUniversity, the Catholic University of Rome, the MIUR, the



Italian National Research Council (CNR), the Regione Sardegnaand the Nando Peretti Foundation, according to their programsof scientific diffusion.

R E F E R E N C E S

[1] Holmes E, Tsang TM, Huang JT, Leweke FM, Koethe D, GerthCW, et al. Metabolic profiling of CSF: evidence that earlyintervention may impact on disease progression andoutcome in schizophrenia. PLoS Med 2006;3:1420–8.

[2] Hope S, Ueland T, Steen NE, Dieset I, Lorentzen S, Berg AO,et al. Interleukin 1 receptor antagonist and soluble tumornecrosis factor receptor 1 are associated with general severityand psychotic symptoms in schizophrenia and bipolardisorder. Schizophr Res 2013;145:36–42.

[3] Dickerson F, Stallings C, Origoni A, Vaughan C, Khushalani S,Yang S, et al. C-reactive protein is elevated in schizophrenia.Schizophr Res 2013;143:198–202.

[4] Dickerson F, Stallings C, Origoni A, Vaughan C, Khushalani S,Yolken R. Elevated C-reactive protein and cognitive deficits inindividuals with bipolar disorder. J Affect Disord2013;150:456–9.

[5] Martins-de-Souza D, Guest PC, Rahmoune H, Bahn S.Proteomic approaches to unravel the complexity ofschizophrenia. Expert Rev Proteomics 2012;9:97–108.

[6] Craddock RM, Huang JT, Jackson E, Harris N, Torrey EF,Herberth M, et al. Increased alpha-defensins as a bloodmarker for schizophrenia susceptibility. Mol Cell Proteomics2008;7:1204–13.

[7] Tabak LA. Point-of-care diagnostics enter themouth. Ann N YAcad Sci 2007;1098:7–14.

[8] Castagnola M, Inzitari R, Fanali C, Iavarone F, Vitali A,Desiderio C, et al. The surprising composition of the salivaryproteome of preterm human newborn. Mol Cell Proteomics2011;10 [M110.003467].

[9] Castagnola M, Cabras T, Iavarone F, Vincenzoni F, Vitali A,Pisano E, et al. Top-down platform for deciphering the humansalivary proteome. J Matern Fetal Neonatal Med 2012;25:27–43.

[10] Messana I, Cabras T, Iavarone F, Vincenzoni F, Urbani A,Castagnola M. Unraveling the different proteomic platforms.J Sep Sci 2013;36:128–39.

[11] Whitelegge JP, Zabrouskov V, Halgand F, Souda P, Bassilian S,Yan W, et al. Protein-sequence polymorphisms andpost-translational modifications in proteins from human salivausing top-down Fourier-transform ion cyclotron resonancemass spectrometry. Int J Mass Spectrom 2007;268:190–7.

[12] Halgand F, Zabrouskov V, Bassilian S, Souda P, Wong DT, LooJA, et al. Micro-heterogeneity of human saliva Peptide P–Ccharacterized by high-resolution top-downFourier-transform mass spectrometry. J Am Soc MassSpectrom 2010;21:868–77.

[13] Halgand F, Zabrouskov V, Bassilian S, Souda P, Loo JA, FaullKF, et al. Defining intact protein primary structures from

saliva: a step toward the human proteome project. AnalChem 2012;84:4383–95.

[14] Peng Y, Chen X, Sato T, Rankin SA, Tsuji RF, Ge Y. Purificationand high-resolution top-down mass spectrometriccharacterization of human salivary α-amylase. Anal Chem2012;84:3339–46.

[15] Cabras T, Manconi B, Iavarone F, Fanali C, Nemolato S, FioritaA, et al. RP-HPLC–ESI-MS evidenced that salivary cystatin B isdetectable in adult human whole saliva mostly as S-modifiedderivatives: S-glutathionyl, S-cysteinyl and S-S 2-mer.J Proteomics 2012;75:908–13.

[16] Zhang Z, Marshall AG. A universal algorithm for fast andautomated charge state deconvolution of electrospraymass-to-charge ratio spectra. J Am Soc Mass Spectrom1998;9:225–33.

[17] Levin Y, Schwarz E, Wang L, Leweke FM, Bahn S. Label-freeLC–MS/MS quantitative proteomics for large-scale biomarkerdiscovery in complex samples. J Sep Sci 2007;30:2198–203.

[18] Pisano E, Cabras T, Montaldo C, Piras V, Inzitari R, Olmi C,et al. Peptides of human gingival crevicular fluid determinedby HPLC–ESI-MS. Eur J Oral Sci 2005;113:462–8.

[19] Castagnola M, Cabras T, Iavarone F, Fanali C, Messana I.Detection of Ca2+-binding S100 proteins in human saliva byHPLC–ESI-MS. Methods Mol Biol 2013;963:357–71.

[20] Lehrer RI, Lu W. α-Defensins in human innate immunity.Immunol Rev 2012;245:84–112.

[21] Agerberth B, Charo J, Werr J, Olsson B, Idali F, Lindbom L, et al.The human antimicrobial and chemotactic peptides LL-37and apha-defensins are expressed by specific lymphocyteand monocyte populations. Blood 2000;96:3086–93.

[22] Chalifour A, Jeannin P, Gauchat JF, Blaecke A, Malissard M,N'Guyen T, et al. Direct bacterial protein PAMP recognition byhuman NK cells involves TLRs and triggers alpha-defensinproduction. Blood 2004;104:1778–83.

[23] Grigat J, Soruri A, Forssmann U, Riggert J, Zwirner J.Chemoattraction of macrophages, T lymphocytes, and mastcells is evolutionarily conserved within the humanalpha-defensin family. J Immunol 2007;179:3958–65.

[24] Harwig SS, Park AS, Lehrer RI. Characterization of defensinprecursors in mature human neutrophils. Blood1992;79:1532–7.

[25] Foell D, Wittkowski H, Vogl T, Roth J. S100 proteins expressedin phagocytes: a novel group of damage-associatedmolecularpattern molecules. J Leukoc Biol 2007;81:28–37.

[26] Tang Y, Xu H, Du X, Lit L, Walker W, Lu A, et al. Geneexpression in blood changes rapidly in neutrophils andmonocytes after ischemic stroke in humans: a microarraystudy. J Cereb Blood Flow Metab 2006;26:1089–102.

[27] Pietzsch J, Hoppmann S. Human S100A12: a novel key playerin inflammation? Amino Acids 2009;36:381–9.

[28] Turk V, Stoka V, Turk D. Cystatins: biochemical andstructural properties, and medical relevance. Front Biosci2008;13:5406–20.

[29] Magister S, Kos J. Cystatins in immune system. J Cancer2013;4:45–56.


Characterization of salivary proteins of schizophrenic and bipolar disorder patients by top-down proteomics

Documents