Sausan Al Kawas , Zubaidah H.A. Rahim , David B. Ferguson * in the diagnosis of disease Potential uses of human salivary protein and peptideanalysis Review

(1)

Review

Potential uses of human salivary protein and peptide analysis in the diagnosis of disease

Sausan Al Kawas

^a,b

, Zubaidah H.A. Rahim

^b

, David B. Ferguson

^b,

*

aDepartmentofOral&CraniofacialHealthSciences,CollegeofDentistry,UniversityofSharjah,Sharjah,UnitedArabEmirates

bDepartmentofOralBiology,FacultyofDentistry,KualaLumpur,Malaysia

Salivaisanimportantbodyfluidcontainingacomplexmixtureofproteins,peptidesand othersubstances.Thesearenotonlyimportantinmaintainingthehealthoftheoralcavity butalsomayyieldinformationaboutoralandsystemicdisease.Comprehensiveanalysis andidentificationoftheproteomiccontentofhumansalivamaycontributetotheunder- standingoforalpathophysiologyandprovideafoundationfortherecognitionofpotential biomarkersofhumandisease.Thecollectionofsalivasamplesisnon-invasive,safe,and inexpensive.Itseemslikelythattestingmethodscanbedevelopedwhichcanbeusedin generalmedicalordentalpractice.However,itisimportanttorealizethatthecollectionof salivamustbecarefullycontrolled.Inthispaperwereviewtheprogressintheanalysisof thehumansalivaryproteomeandsummarisethediagnosticpossibilitiesthathavebeen explored.Theprecautionsincollectingsaliva,andsomeofthefactorswhichwouldhaveto beconsideredifadiagnostictestweretobegenerallyadoptedarediscussed.

#2011ElsevierLtd.Allrightsreserved.

*Correspondingauthor.

E-mailaddress:dbferguson@supanet.com(D.B.Ferguson).

a v a i l a bl e a tw w w. sc i e nce d i re c t. c om

journalhomepage:http://www.elsevier.com/locate/aob

0003–9969/$–seefrontmatter#2011ElsevierLtd.Allrightsreserved.

doi:10.1016/j.archoralbio.2011.06.013

(2)

1. Introduction

Human saliva is a fluid with many biological functions essentialforthemaintenanceoforalhealth.Scientistshave beenmoreinterested inthepastinstudyingthe biological functionsofsalivainthemouththanintryingtoassessits possibleroleasanindicatorofsystemicororaldisease.The recentuseofsalivainthediagnosisofhumanimmunodefi- ciencyvirus(HIV),carcinomainanumberoftissues,cardiac disease and autoimmune diseases has demonstrated that salivacanbeausefulaidtoclinicaldiagnosis,andhasdrawn attention to the possibilities of detecting other diseases throughsalivaryanalysis.¹

Although there is now a very considerable body of literatureonthediagnosticpossibilitiesofsaliva,thenumber ofpapers whichprovide actualscientificevidence ismuch smaller. We have concentrated in this review on papers reportingactualdata.

Wholesalivaismainlyamixtureofthesecretionsfromthe three pairs of major salivary glands, each secreting a characteristictypeofsaliva.Therearefurthercontributions fromthemanysmallminorsalivaryglandssituatedbeneath theoralmucosa.Inadditionitcontainsconstituentsfromthe gingivalcrevicularfluid,frommanymicrobialcontaminants inthe mouth,and fromthe desquamated cells ofthe oral epithelium. Even the relativecontributions ofthe different glandstowholesalivaarevariable,dependinguponthetypes and degree of stimulation and even the time of day. The variablenatureofwholesalivasecretionsmeansthatdifferent approaches may have to be adopted when studying its compositionorthepossibilitiesofusingitforthedetection ofdiseasebiomarkers.Manyresearchworkershaveconcen- tratedonstudyingwholesaliva^2–4becauseitcanbeobtained bysimplyspittingintoatesttubeorallowingittodribblefrom the mouth. Others have concentrated on the ductal saliva obtainedfromdifferentsalivaryglands.^5,6Thedevelopmentof whole saliva proteome analysis, as well as that of saliva obtainedfromthedifferentglands,mayyieldimportantclues to the health of the oral cavity and even the wider pathogenesisofsystemicdisease.⁶

The rich variety of molecules present in the salivary secretions renders saliva an attractive possible source of diseasebiomarkers.Overthelastfewyearssalivaryresearch workers have been developing salivary diagnostic tools to monitor both oral and systemic disease. Saliva has many apparent advantages over serum as a medium forclinical diagnosis.Wholesalivaiseasytocollectandcanbestoredand transportedatlowcost–althoughtheneedtomaintainitina frozenstateimposesahiddencost.Unlikebloodplasma,it doesnotclotandisthereforeeasiertohandle.Thecollection techniquesforwholesalivaarenon-invasive,thusreducing patientdiscomfortandanxietywhenrepeatedsamplesare requiredoveraperiodoftime.Onthenegativeside,whole salivausuallyrequirescentrifugationorfiltrationtoremove precipitatedmucinsandcellularcontaminants.Suchcentri- fugationmayalsoremoveotherproteins.Somelife-threaten- ing diseases such as neoplasia and some cardiovascular diseases are difficult to diagnose without invasive and complicatedclinicaltests:currentresearchistryingtodevelop

easier diagnostic tests using saliva. Moreover, saliva in comparison with blood may demonstrate more sensitive andmorespecificmarkersfororaldiseasessuchassquamous cellcarcinomaofthemouth.⁷

Currently,salivaresearchworkersarebeginningtoregard salivaasavaluablefluidwhichcanitselfprovideinformation aboutdiseaseandnotmerelyasanadjuncttothestandard laboratorytestsinvolvingbloodorurine.Themainbarriersto themorewidespreadadoptionofdiagnosticmethodsusing salivaarefromthetechnologyinvolvedandthecostandspeed of the diagnosticmethods. Modern techniquesand rapidly emerging scientificresearch areeliminatingmanyofthese problems. Recentreportshave identifiedseveralpotentially usefulbiomarkersinsaliva.Thedevelopmentofspecifictests for particular biomarkers and the use of small hand-held devices forthesetestscouldreducetheanalyticalcostand speedupthediagnosticprocess tobecomeoneusedatthe pointofcare.⁸Itshouldberealizedthatitmaybedifficultto changethemindsetofthetypicalphysician,whoautomati- callythinksfirstoftakingabloodsampleandthenofsending it off to a standard clinical chemistry laboratory with its batteryofanalyticalmachinesspecificallydesignedforblood analysis.

Salivary proteins and peptides have been studied with traditionalbiochemicaltechniques,includingliquidchroma- tography, gel electrophoresis, capillary electrophoresis, nuclear magnetic resonance, mass spectrography, immunoassays (radio-immune assays, immunoradiometric assays,enzymeimmuno-assays,enzyme-linkedimmunosor- bent assays) and lectin probe analysis.^2,3 Most of these analyses have been aimed to investigate specific salivary proteingroups,but someefforts havebeenmadetoobtain complete analyses of salivary proteins and peptides with proteomictechniques.Comprehensiveanalysisandidentifi- cationofthesalivaryproteomemaybenecessarytounder- standfully oralpathophysiology,ontheonehand,andthe possibilityofusingsalivaryproteinsandpeptidesasbiomark- ersofsystemicdisease,ontheother.Thepresentproteomic technologies for the comprehensive identification of the proteinsandpeptidesinhumanwholesalivaandtheprogress being made in identifying potentialsalivary biomarkers of humanoralandsystemicdiseasewillnowbediscussed.

2. Analysis of the proteome of saliva

Repeatedattemptsoverthelast40yearstocharacteriseand cataloguetheproteinsandpeptidesofsalivahaverevealedan abundance of proteins with a wide range of functional properties.Thesehaveincludedimmuneresponses(immunoglobulins)andantimicrobialactivity(lysozyme,lactoferrin, sialoperoxidase,histatins,defensins).Lubricationandphysi- calprotectionoftheoraltissuesareduetothemucins.The proline-richproteins,whosefunctionsincludetheprecipita- tionofpotentiallyharmfultanninsfromthediet,arepresent inmanyisoforms,and,likeothersalivaryproteins,aresubject to the proteolytic activity of both salivary and bacterial proteasesinthemouth.⁹

Inrecentyearstechnologicaldevelopmentshaveresulted innewapproachestomapoutthetotalproteinandpeptide

(3)

compositionof saliva– the salivaryproteome. In areview paper Huqet al.¹⁰statethat over1380 proteinshave been detectedinsalivabydifferentresearchworkersusingdifferent techniques,althoughonlyaround100ofthesearepresentin relativelyhighabundance.Itbecomespossiblewithsuchan arrayofproteinstoenvisagesalivaryproteomeanalysisasa means of determining changes which result from localor systemicdisease.

Manydifferenttechniqueshavebeenused,eithersingly, or, more often, in combination. One technique involves separation of the salivary proteins by two-dimensional electrophoresis and detecting them with a suitable stain.

Proteins and their isoforms are usually separated by this technique.Eachproteinspotcanthenbeexcisedfromthegel anddigestedwithtrypticenzymes.Theproteinsseparatedby thepolyacryalmidegelelectrophoresiscanbemoreaccurately characterizedandidentifiedusingmassspectrometry.Mass spectrometry has become one of the core technologies in proteomicsbecauseofitssensitivityinmassmeasurementof peptidesandproteinswithahighdegreeofaccuracy.Further characterizationoftheproteinsandpeptidescanbecarried outusingionizationmethodssuchaselectrosprayionization (ESI)andmatrix-assistedlaserdesorptionionization(MALDI).

Asmanyofthehighlyabundant salivaryproteins, suchas amylase,cystatinandimmunoglobulin,appearasfragments onthetypicaltwo-dimensionalgel,itisnecessarytodeplete them in order to improve identification of the lower- abundanceproteins.⁷

Anothertechniquecombinesliquidchromatographywith mass spectrometry. All these techniques together have extendedtheproteomeofwholesalivatocompriseofmore than1050proteins.^2,11,12Thisfigureislowerthanthefigureof 1380 citedbyHuqetal.¹⁰; however1380 wasa cumulative figure fromdifferentstudies andis likelytohave included substances which appear in secretions taken from single glands but not in whole saliva where they are subject to proteolysis(bybacterialaswellassalivaryenzymeactivity).

Shotgunproteomeanalysisbasedonadvancedmolecular spectrometric techniques, such a quadruple time-of-flight, matrix-assisted laser desorption/ionisation, linear ion trap, andlineariontrap–Orbitrap–providessignificantlyenhanced resolutionforidentificationincomparisontotwo-dimension-

algelelectrophoresisfollowedbysimplemassspectrometry.¹³ Combining surface chromatography with matrix-assisted laserdesorption/ionisationtimeofflightmassspectrometry enables rapid and high-throughput detection of important proteins and peptides.¹⁴ Finally, additional methodologies suchashighperformanceliquidchromatographycombined with mass spectrometry have been shown to be useful complementary methods inthe evaluation of the smallest salivaryproteinsandpeptides.^12,15

Themostabundantproteinsinsalivaalreadywellknown andnowrecognisedbythesetechniquesaretheproline-rich peptides,alpha-amylase,cystatins,histatins,mucins,secre- toryIgAandcarbonicanhydrase.⁶

3. Proteome of normal human saliva (Table 1)

TheUnitedStatesNationalInstituteforDentalandCraniofa- cialresearch(NIDCR)hasfundedthesettingupofadatabase (www.hspp.ucla.edu)tocollecttogetherasmuchinformation aspossibleontheproteinsandpeptidesofsaliva.Beforethis, despite the extensive workthat has been carried out, few researchershavepublishedcompletelistsofthewholesaliva proteins and peptides which they have found. A notable exceptionwasthepaperbyXieetal.¹⁶Theyreportedfinding 918molecularspeciesinwholesalivafromonehealthyfemale subject,andwereabletomaketentativeidentificationof437 proteinswithhighconfidence(falsepositiveratebelow1%).

Several othergroups have participatedinestablishinga proteomemapfromwholeorductalsaliva.^1,2,6Aconsortium ofthreeresearchgroupswasformedtoproduceacatalogueof proteinsinhumansalivacollectedfromparotidandsubman- dibularglands.⁶Thesupplementarydataforthispublication showsthattheauthorsidentified914proteinsinparotidsaliva and917proteinsinsubmandibular/sublingualsalivafroma totalof23subjects(12femalesand11males).Theresultsfrom thiscollaborative workshowedthat174ofthe657proteins foundinplasmaand236ofthe467proteinsfoundintearsare alsoexpressedinsaliva.However,LeeandWong¹statethat

‘‘formostoftheseproteins,theirexpressioninsalivaisquite distinct from that in serum or tears, and have already demonstrated clinical diagnostic valuesfor diseases mani-

Table1–Proteomeanalysisofsaliva.

Author Ref Subjects Saliva Stimulation Numberofpeptides

Huetal.(2005) ³ 1 Whole Unstim 309Identified

Huetal.(2004) ⁵ 9 Submandibular,

sublingual

Stim Notgiven

Dennyetal.(2006/2008) ⁶ 23 Parotid,submandibular, sublingual

Stim P914,SM/SL917

Huangetal.(2004) ² 4gingivitis 4control

Whole Unstim 200

Ghafourietal.(2003) ¹¹ 5 Whole Unstim 100

Xieetal.(2005) ¹⁶ 1 Whole Unstim 918(437identified)

Huetal.(2006) ⁶¹ 10 Whole Unstim 282Identified,297unknown

570genesknown417unknown

Siqueiraetal.(2008) ¹⁷ 10 Labialgland Unstim 56Identified

Abbreviations:stim,stimulated;unstim,unstimulated;P,parotid;SM,submandibular;SL,sublingual.

(4)

festedintheoralcavity.’’Theproteinsidentifiedthusfarcover awiderangeofmolecularmassesandhavefunctionsranging fromcellandtissuestructuralrolestocatalyticorenzymatic activities.Themajorproteinsreportedbyallthreegroupsare extracellularand secreted proteins suchas alpha-amylase, proline-richproteins,cystatin,histatinsandmucins.Howev- er, the total analysis revealed that many proteins were amongstthoseinvolvedinmetabolicprocesses,complement andclottingcascades,celladhesionandcommunication,cell cycleprogressionandregulationoftheactincytoskeleton.A comprehensive catalogue ofsalivaryproteins and peptides withsomeindicationoftheirquantitiesisessentialfortheoral biologiststudyingsalivarysecretionsaswellasforthosewho wouldwishtousesalivaasadiagnosticaid.TheworkofNIHin fundingthispresentcatalogueisamassivestepforwardin thisdirection.Otherworkersneedtoaddtheirresultstothis catalogue.

Inanotherstudyontheproteome oflabialglandsaliva, reportedfromOppenheim’slaboratory,¹⁷about100proteins wereidentifiedinthissecretionandgroupedaccordingtotheir propertiesasbeing about 50%secreted proteins,12% cyto- plasmic, 9% nuclear,7% cell membrane and about 20%

unidentified.Theirfunctionswerecategorisedashostdefence (32%), signal pathways (11%), ion transport (11%), nuclear functions (4%), cell metabolism (13%) and nearly 30%

unidentified.

4. Problems of standardisation

Although superficially it mayappear that the collection of salivaisasimplenon-invasiveprocedure,thereareproblems which maynotbe obvious.Thecollection ofunstimulated wholesaliva,bysimplyallowingsalivatoaccumulateinthe mouth and then allowing it to dribble out, is subject to considerablevariationasthepatienttries,tovaryingextents, speed up the collection by tongue and cheek movements.

Whetherthisincreasesthenumberofproteinsinthesaliva,or whether it simply alters the concentrations has not been studied.However,researchersneedtogiveclearinstructions towhich volunteers must adhere beforeand during saliva collections.Ifsalivaistobeusedfordiagnosticpurposesmore widely,similarcontrolmustbeexertedovercollectionsfrom patients.Theprocessmight,infact,bemoreeasilystandar- dised using wax stimulated saliva, or giving the patient a sterilisedpieceofsoftplasticsurgicalcannulatochew.Wax pelletsareavailablefromanumberofcompanies.GCprovides sterilisedwaxpelletsaspartoftheGCSalivaTestkit.Whether stimulated or unstimulated, wholesaliva contains desquamated epithelial and bacterial cells, as well as a varying contributionfromgingivalcrevicularfluid.Thesecellularand fluidcomponentswillbemoreinstimulatedsaliva.Datafrom standard textbooks¹⁸ suggests thatcentrifugation of whole salivamayyieldapelletofasmuchas10%ofthevolume.The epithelial cell content has been reported as up to 600cell fragments/ml, and leucocytes up to 600 cell fragments/ml whilstthebacterialcontentmaybeashighas10⁸CFU/ml.The dataareforstimulatedwholesaliva:wearenotawareofany similar data for unstimulated whole saliva. Even after centrifugationmuchofthebacterialcontentremainsinthe

supernatant.¹⁹ If one could filter whole saliva with filters capableofretainingbacteriathismighthelp.Unfortunately such filters rapidly clog with mucins and other protein material.Amorecomplex filtrationtechniquedescribedby DeJongandvanderHoeven,²⁰inwhichthegel-likestructure ofsalivaisdegradedbyreduction,issaidtoyieldabacteria- free saliva. All this assumes that a whole saliva free of bacterialandepithelialcellcontaminationisrequired.Itmay be,withthesensitivityofcurrentmethodsofproteindetection thatsuchcontaminationcanbeallowedforintheanalysis,or eventhatitsproteinmaterialmaybeofvalueinsystemicor oraldiagnosis.Nonetheless,investigatorsshouldbeawareof theseproblems.Itisrecommendedthatsalivasamplesshould alwaysbecollectedatthesametimeofdaytoreducecircadian variation²¹andthedurationofstimulationorsecretionmay alsoaffectsalivacomposition.²²

Theextent of contamination withgingivalfluid compo- nents(mainlyblood-derived)isequallyunknown.Asserum albuminispresentinbloodatconcentrations1000-foldhigher than those in saliva, some indication of contamination of salivawithbloodproteinsmaybeassessedfromtheserum albumin content. The presence of markers from gingival crevicular fluid may be of value in assessing periodontal disease(seelater).

The protein and peptide content of individual gland secretionshastobecarefullyassessed.Identificationofmajor secretedsalivaryproteinsisnottoodifficultandtheyprobably accountforaround100ofthepeptidesthusfaridentified.The remainder may arise from interstitial fluid, from minor secretedmaterials andtheirtransportmolecules,orsimply fromthewallsoftheductsalongwhichthesecretionpassesas itleavestheacinus.¹⁶

5. Concept of biomarkers

Thetermbiomarkerhascomeintouseoverthelasttenyears todenoteasignificantmolecularspeciesorcombination of speciesinatestfluidsuchasbloodorsalivawhichisunique to,and thereforediagnosticof,aparticularphysiologicalor pathologicalstate.Atitsmostprecise,thetermshouldreferto asinglemolecularspecieswhichispresentinsamplesfroma subjectwithaparticulardiseaseorstatus,andisnotpresentin other subjects. Slightly less precisely, abiomarker maybe detectableatconcentrationssignificantlydifferentfromthose incontrolsubjects.Thetermisnowbeingused,muchless precisely,todescribecombinationsofmolecularspeciesinthe sampleshowingsignificantvariationinconcentrationfrom thoseregardedasnormal.Thus,ifcErb2isfoundinthesaliva ofpatientswithmammarycarcinomabutatverymuchlower concentrationsinthesalivaofalargenumberofpersonswith healthymammaryglands,thisclearlyindicatesitasagood biomarker. Italsopassesthetestofbiological significance:

cErb2isproducedbymammarycarcinomas.^23–25

Thesituationisdifferentwithcombinationsofmolecular specieswhichmayberegardedasbiomarkers.Therangeof variationwhich can beregardedasnormalmust beascer- tainedforanumberofsubstancesinalargepopulation.The extentofvariationfromthisnormalgroupwhichisregarded assignificanthastobedefined.Thereisanadditionalproblem

(5)

withmanysalivaryproteinsappearinginpolymorphicforms.⁹ The polymorphisms of the proline-rich proteins provide a goodexample. Degradationofproteinsinwholesalivaalso contributestovarietyintheproteinsandpeptidesobserved.

Someexplanationofthecauseofthevariationshouldbe sought,ifonlytoavoidthepossibilityofchanceassociation,or alternativeexplanationsoftheanalyticalresult.Theultimate test of a biomarker, or combination of biomarkers, is the percentageofcasescorrectlydiagnosedandthepercentageof casesincorrectly diagnosedaspossessing thedisease.Very fewofthebiomarkerssofarproposedhavebeentestedinthis way.

6. Salivary proteomes in relation to oral diseases (Table 2)

Theuseofapatient’ssalivaryproteomeshouldtheoretically beofuseinassessingthestatusofdiseaseintheoralcavity.

Unlessthediseaseinvolvesthesalivaryglandsthemselves, wholesalivawouldappeartobethemostappropriatefluidto use in assessingoral disease. Thetwo most common oral diseasesaredentalcariesandperiodontaldisease.Boththese diseasesare diagnosedroutinely bythe dental practitioner andthevalueofanysalivaryanalysisliesinprognosisrather than diagnosis. The situation is different, however, with squamouscellcarcinomaoftheoralcavity,whereanumberof non-malignantorpre-malignantlesionsmaybepresent,anda biomarkertoidentifythepre-malignantwouldbevaluable.

6.1. Dentalcariessusceptibility(Table2)

There have been many studies attempting to relate caries prevalenceandthesalivarygeneticphenotype,butcontradic- toryresultshavebeenreported.²⁶Salivarytestshaverecently been developed to evaluate caries risk by measuring the amountsofselectedoligosaccharideswhoseconcentrations have shown a correlation with caries experience inyoung adults.^27,28 Significant associations have been reported be- tween dental caries experience, age of subjects, and the concentrationsinsubmandibularorsublingualglandsalivaof

lactoferrin, albumin, lysozyme, mucin, and cystatins, in additiontotheconcentrationsofpotassium,calcium,sodium andchloride.²⁹Mungiaetal.²⁹alsocommentedthatchanges inthesalivaoutputduringageingcorrelatedwithgreaterrisk of dental caries and may beindicators ofcaries risk. Two recentstudieshavereportedthatchildrenwithcaries have lowermedianvaluesofalpha-defensins^30,31suggestingthat theseproteinsmightbeusedtoassesscariesrisk.Inallstudies such as these, the question arises as to whether the component measuredisrelated tothesusceptibility tothe disease,orisaresultofthedisease.Asimilarcommentapplies to the data of Vitorino et al.^32,33 showing higher levels of statherinandcystatinSincaries-freechildren.Rudneyetal.³⁴ have used a proteomic approach to test whether salivary proteins can actas biomarkers forcaries risk assessment.

TheirdatasuggestthatstatherinandcystatinSarethebest predictors of occlusal caries in saliva although they also considered supragingival levels of total plaque DNA, and variations inthe totalnumbers ofsomeoralbacteria. Itis possiblethatthesefindingssuggestarelationshipbetweenthe molecular functions of statherin and cystatin S and the antimicrobialpropertiesofsaliva.³⁴

6.2. Evaluationofperiodontaldisease(Table2)

There are a number of recent reports on salivary protein variationsinperiodontaldisease.^35–47Biomarkersstudiedin relation to the diagnosis, or more usually, prognosis, of periodontaldiseasehaveincludedhostenzymesandimmu- noglobulins,phenotypicmarkerssuchasepithelialkeratins, hostcells,salivaryions,hormones,andbacteria.^38,39Enzyme immunosorbent assays have been used to test for three potential biomarkers of aspects of periodontal disease:

interleukin-1 beta as a marker of inflammation, matrix metalloproteinase-8 as a marker of collagen breakdown, andosteoprotegerinasamarkerforboneturnover.^40,41Levels of interleukin-1 beta and matrix metalloproteinase-8 were both raisedinwholesaliva frompatientswithperiodontal disease.Aswholesalivacontainsacontributionfromgingival crevicular fluid,and these substances,like matrix metallo- proteases-2 and -9, are known to be present in gingival

Table2–Proteinsassociatedwithdentalcariesorperiodontaldisease.

Author Ref Subjects Disease Saliva Associations

Mungiaetal.(2008) ²⁹ 811 Caries SM/SLstimand

unstim

Lactoferrin,albumin,lysozyme, mucin,cystatin

Rudneyetal.(2009) ³⁴ 18lowcaries28highcaries Caries Unstimwhole Statherin,cystatin5 Aureretal.(2005) ³⁷ 18aggressive,10chronic,

14controls,9edentulous

Periodontal Wholeunstim Decreasedcomplement,alpha-2 macroglobulin,TNF

Milleretal.(2006) ⁴⁰ 57 Periodontal Whole MMP8,IL-1beta

Raietal.(2008) ⁴¹ 18gingivitis,15control, 20periodontitis

Periodontal MMP2,8,9

Bassimetal.(2008) ⁴⁴ 9severe,11moderate Periodontaldiabetic Wholeunstim Procalcitoninraised Wuetal.(2009) ⁴⁵ 5aggressiveperio5control Periodontitis Wholeunstim 6Proteinsincreased,

5decreased

Goncalesetal.(2010) ⁴⁶ 5m5F Periodontitis Unstimwhole Bloodproteinsincluding immunoglobulins,lesscystatin Abbreviations: stim, stimulated; unstim, unstimulated; SM, submandibular; SL, sublingual; TNF, tumour necrosis factor; MMP, matrix metalloproteinase;IL,interleukin.

(6)

crevicular fluid,their detectionis notsurprising. Similarly, Lamsteretal.⁴²foundthat salivaryconcentrationsofbeta- glucuronidase,anenzymeknowntobeamarkeringingival crevicular fluid, were related to the extent of periodontal damage,Therearealreadydiagnostickitsonthemarketfor theanalysisofgingivalcrevicularfluidwhichcanbeusedby thegeneraldentalpractitioner.⁴³Anotherstudyhassuggested thatsalivaryprocalcitonin,amediatorofsystemicinflamma- tion,couldreflectthedegreeofperiodontitisand hypergly- caemiaindiabetestype2patients.⁴⁴Inageneralreviewofthe proteomic profile of subjects with generalised aggressive periodontitis compared with that of healthy volunteers,⁴⁵ elevensalivaryproteins(serumalbumin,immunoglobulin(Ig) d2chainC region, Iga2chainC region, vitaminD-binding protein,salivarya-amylaseandzinc-a2glycoprotein,lacto- transferrin,elongation factor 2,14-3-3 sigma, shortpalate, lung and nasal epithelium carcinoma-associated protein 2 precursorandcarbonicanhydrase6)werefoundtodifferin concentration in the two groups. Gonc¸alis et al.⁴⁶ also comparedsalivaryproteomesfrompatientswithperiodontitis with those from healthy subjects by two-dimensional gel electrophoresisandliquidchromatographyfollowedbymass spectrometry.Theperiodontitispatientshadincreasedrela- tiveconcentrationsofbloodproteinsandimmunoglobulins, andalsolowerlevelsofcystatins.Alpha-amylaseconcentra- tions were increased in the saliva of the patients with periodontitis.ThereareanumberofpapersfromGiannobile’s groupondevelopmentofmicrochipassaysystems,e.g.,Herr etal.⁸Inarecentstudywholesalivasampleswerecollected from18healthysubjects(lessthan20%ofsitesbleedingon probing),32subjectswithgingivitis (morethan20%ofsites bleedingonprobing),28subjectswithmildchronicperiodon- titis(lessthan30%ofsiteswithclinicalattachmentlossless than3mm)and21subjectswithseverechronicperiodontitis (more than 30% of sites with clinical attachment loss exceeding 3mm).⁴⁷ Biomarkers in saliva for three distinct phases of periodontal breakdown were identified. In peri- odontalinflammationInterleukins6and7wereidentified.As matrixdegradationoccurred,matrixmetalloproteinases8and 9appeared andalveolar boneturnoveror degradationwas associated with osteoprotegerin and ICTP presence in the saliva.

6.3. Diagnosisoforalsquamouscellcarcinoma(Table3)

Severalgroups^7,13,48–50havereportedthatsomeproteinsare increasedinamountinwholesalivafrompatientswithoral squamouscellcarcinoma.Theproteinsincludedinterleukin- 8, and thioredoxin.^50,51 Thioredoxin was also found to be

significantly increased in the saliva of heavy cigarette smokers,suggestingthatitmaybeapredisposingfactorfor oral cancer in these subjects.⁵² Elevated levels of salivary solubleCD44wereshowninthemajorityoforalsquamouscell carcinomaandcoulddistinguishcancerfrombenigntumours withhigh specificity.⁴⁸Otherstudies havesuggestedthata panelofbiomarkersfororalsquamouscellcarcinomawould besensitiveandspecific.^49,50Threetumourmarkers,cytoker- atin 19 fragment, tissue polypeptide antigen, and cancer antigen 125, were foundtobe significantlyelevated inthe saliva oforalsquamouscell carcinomapatients.Combined measurements of these markers in the saliva of oral squamouscellcarcinomapatientsshowedsimilardiagnostic valuetothoseobtainedwhenmeasuringthemintheseraof thosepatients.⁴⁹Agroupof5proteinshasbeenidentifiedas possible candidates: M2BP, MRP14, CD59, catalase and profilin.¹³ This combination gave a high sensitivity (90%) andspecificity(83%)indetectingthisneoplasia.Moreworkis neededtovalidatethispanelofbiomarkers.Thediscoveryand validation ofpanelof biomarkers insaliva could leadtoa simplediagnostictestfororalcancerdetection.Wong⁵¹has suggested that application of appropriate biomarkers in a microfluidic device could be a powerful tool in the future diagnosisoforalcarcinomas.

6.4. Salivaryvariationinotheroralinflammatory diseases(Table4)

Salivaryproteinshavealsobeenproposedasbiomarkersfor other oral diseases. Patients with oral lichen planus have increasedexpressionofurinaryprokallikrein,anddecreased expressionofthepalate,lungandnasalepitheliumcarcinoma associated protein (PLUNC).⁵³ The salivary proteome of Sjo¨grenSyndromepatientsshowsincreasesintheinflamma- toryproteins,beta-2microglobulin,lactoferrin,immunoglobulin (Ig) kappa-light chain, polymeric immunoglobulin receptor, lysozyme C and cystatinC,and decreases in the acinarproteins,proline-richproteins,amylaseandcarbonic anhydrase VI, in comparison with patients without the disease.⁵⁴

7. Salivary proteomes in relation to systemic diseases (Table 4)

Thereisconsiderableinterestinthepossibilityofusingsaliva samples inthe diagnosis ofsystemic disease.Thesalivary proteomehasbeeninvestigatedindiabetesmellitus,⁵⁵cystic fibrosis⁵⁶anddiffusesystemicsclerosis.⁵⁷Analyseshavebeen

Table3–Proteinsassociatedwithoralsquamouscellcarcinoma.

Author Ref Subjects Saliva Proteins

Huetal.(2008) ¹³ 64+64matched Wholeunstim M2BP(tumourantigen),MRP14,

CD59,profilm1,catalase

Franzmannetal.(2007) ⁴⁸ 102+69control Oralrinse CD44increased

Lietal.(2004) ⁵⁰ 32+32control Wholeunstim OAZ-1,SAT,IL-1bIL-8

Xieetal.(2008) ⁶² 4 Cellsinwholeunstimsaliva Proteinanalysis

(7)

carriedouttodetectdifferencesinthesalivaryproteomein thesediseasesbutitisnotyetclearwhetherthesedifferences willhaveadiagnosticpotential.Salivahasbeenusedinthe diagnosisofauto-immunedeficiencysyndrome(AIDS)⁵⁸and otherviralinfectionssuchashepatitisBandC.⁵⁹Thehepatitis virusistransmissibleviasalivabutthehumanimmunodefi- ciencyvirusisnottransmittedinthisway.

At this time, the most useful diagnostic application of salivarypeptideanalysishasbeeninthedetectionofcErb2in the saliva of patients with mammary carcinoma.^23–25 This peptideisproducedbyneoplasticcellsofexocrineglandsand its presence in blood has been regarded as diagnostic of mammarycarcinoma.Streckfusandhisco-workers^23–25found significantlyhigherlevelsofcErb2insalivafromwomenwith malignantbreastlesionsthaninhealthywomenandwomen withbenignbreastlesions.Thetestisusefulinthecontextof suspected carcinoma of the breast, although theoretically otherexocrineglandtumoursmightalsoraisesalivarycErb2 levelsinsaliva.⁶⁰Inthisparticularinstance,thesalivatestis lookingforaproductspecificforthedisease.Itispossiblethat othercarcinomaproductsmightbefoundinsaliva.Itwouldbe very useful to have a saliva test to distinguish prostatic hyperplasiafromcarcinomaofthe prostate:this possibility shouldbeinvestigated.

8. Practical issues

Itis instructive toconsider the mostprecise and selective salivarydiagnosticteststhathavebeendevelopedpreviously– thosefor steroidhormones.³⁸Thesearesimple tests using wholesaliva, unstimulated orstimulated, andthey can be performed easily in standard clinical laboratories. Indeed, they have been adapted for patient use – for example, in determiningperiods forconception. Yetthe testshave not beenwidelyadoptedbythemedicalprofession,whopreferto usethetraditionalbloodsample,andsenditawayforaresult in 24h or so from a clinical laboratory with an array of machinesdesignedtomeasurebloodchemistry.Ifadiagnos- tictest using salivais togain wide acceptance it must be simple touseand preferablyadaptedforuse insurgeryor clinic.Thedevelopment ofmicrofluidicdevices maybe an essentialstepinthisprocess.

9. Conclusions

1. Itisessentialthatacomprehensivedatabasebeestablished fortheproteomeofwholesalivacollectedandprocessed understandardconditions.TheinitiativeoftheNIDCRin fundingthedatabasesforsecretionsofindividualglands hasbeenexactlywhatwasrequired,butsimilarinforma- tionisneededforwholesaliva,particularlyrestingwhole saliva,fromalargenumberofhealthysubjects.

2. Thebiologicalrationaleforthechoiceofbiomarkersshould besound–notthepossiblychanceassociationofmolecules withdisease–andthebiomarkersmustbevalidated.Are theirpredictionsverifiedbylatereventsoranalyses?

3. Ifabiomarkertestistobeusefulitmustshowadvantages overothertests,besufficientlyeasyinuse,and,preferably, able to give a rapid and accurate result in the clinical consultation.Itshouldalsobepatientfriendly.

In summary, saliva is an upcoming area for basic and clinical research withsubstantial potential forgrowth and development.Moreresearchisrequiredtovalidatethevarious discovered potentialbiomarkers forearly diseasedetection that will lead to more effective treatment. Oral fluid diagnostics can be aided by new technology which may become a powerful tool for oral and systemic diseases diagnosisinthefuture.

Funding

None.

Ethical approval statement

Notrequired.

Conflict of interest

D.B.FergusonisapreviousEditorinChiefofArchivesofOral Biology.

Table4–Proteinsorpeptidesassociatedwithotherconditions.

Authors Ref Condition Subjects Saliva Proteins

Yangetal.(2006) ⁵³ Orallichenplanus 6+6control Wholeunstim Atleast14proteinsincreased over2

Ryuetal.(2006) ⁵⁴ Sjogrens 41+20control Stimparotid Beta-2microglobulin,lactoferrin, Iglightchain,polymericIgreceptor, lysozymeC,cystatinC

Giustietal.(2007) ⁵⁷ Sjogrens 15+15controls Unstimwhole 19Proteinsincreased(seetext)

Landrumetal.(2005) ⁵⁸ HIV CompareEIAand

oraQuicktests

Unstimwhole oraQuickvalidated

Streckfusetal.(1999) ²³ Mammarycarcinoma 135Controls Stimulatedwhole c-erbB-2andCA15-3lowin controlpatients

Streckfusetal.(2000) ²⁴ Mammarycarcinoma 57Controls,41benign breastlesions,30breast carcinoma

Stimulatedwhole saliva

c-erbB-2andCA15-3concentrations diagnosticofmammary

carcinoma

(8)

references

1. LeeY-H,WongDT.Saliva:anemergingbiofluidforearly detectionofdiseases.AmJDent2009;22:241–8.

2. HuangCM.Comparativeproteomicanalysisofhuman wholesaliva.ArchOralBiol2004;49:951–62.

3. HuS,XieY,RamachandranP,OgorzalekLooRR,LiY,LooJA, etal.Large-scaleidentificationofproteinsinhuman salivaryproteomebyliquidchromatography/mass spectrometryandtwo-dimensionalgelelectrophoresis- massspectrometry.Proteomics2005;5:1714–28.

4. GuoT,RudnickPA,WangW,LeeCS,DevoeDL,BalgleyBM.

Characterizationofthehumansalivaryproteomeby capillaryisoelectricfocusing/nanoreversed-phaseliquid chromatographycoupledwithESI-tandemMS.JProteome Res2006;5:1469–78.

5. HuS,DennyP,XieY,LooJA,WolinskyLE,LiL,etal.

Differentiallyexpressedproteinmarkersinhuman submandibularandsublingualsecretions.IntJOncol 2004;25:1423–30.

6. DennyP,HagenFK,HardtM,LiaoL,YanW,ArellannoM, etal.Theproteomesofhumanparotidandsubmandibular/

sublingualglandsalivacollectedastheductalsecretions.J Proteome2008;7:1994–2006.

7. HuS,YuT,XieY,YangY,LiY,ZhouX,etal.Discoveryof oralfluidbiomarkersforhumanoralcancerbymass spectrometry.CancerGenomicsProteomics2007;4:55–64.

8. HerrAE,HatchAV,ThreckmartinDJ,tranHM,BrennanJS, GiannobileWV,etal.Microfluidicassaysasrapidsaliva- basedclinicaldiagnostics.ProcNatlAcadSciUSA 2007;104:5268–73.

9. OppenheimFG,SalihE,SiqueiraWL,ZhangW,Helmerhost EJ.Salivaryproteomeanditsgeneticspolymorphisms.Ann NYAcadSci2007;1098:22–50.

10.HuqNL,CrossKJ,UngM,MyroforidisH,VeithPD,ChenD, etal.Areviewofthesalivaryproteomeandpeptidomeand saliva-derivedpeptidetherapeutics.IntJPeptResTher 2007;13:547–64.

11.GhafouriB,TagessonC,LindahlM.Mappingofproteinsin humansalivausingtwo-dimensionalgelelectrophoresis andpeptidemassfingerprinting.Proteomics2003;3:1003–15.

12.BaldiniC,GiustiCL,BazzichiL,LucacchiniA,BombardieriS.

Proteomicanalysisofthesaliva.Aclueforunderstanding primaryfromsecondarySjogrensyndrome?AutoimmunRev 2008;7:185–91.

13.HuS,ArellanoM,BoontheungP,WangJ,ZhouH,JiangJ, etal.Salivaryproteomicsfororalcancerbiomarker discovery.ClinCancerRes2008;14:6246–52.

14.SchipperRG,SillettiE,VingerhoedsMH.Salivaasresearch material:biochemical,physicochemicalandpractical aspects.ArchOralBiol2007;52:1114–35.

15.HardtM,ThomasLR,DixonSE,NewportG,Agabian.

PrakobpholNA,etal.Towarddefiningthehumanparotid glandsalivaryproteomeandpeptidome:identificationand characterisationusing2DSDS-PAGE,ultrafiltration,HPLC, andmassspectrometry.Biochemistry2005;44:2885–99.

16.XieH,RhoduNL,GriffinRJ,CarlisJV,GriffinTJ.Acatalogue ofhumansalivaproteinsidentifiedbyfreeflow

electrophoresis-basedpeptideseparationandtandemmass spectrometry.MolCellProteomics2005;4:1826–30.

17.SiqueiraWL,SalihTE,WanEJ,HelmerhorstEJ,Oppenheim G.Theproteomeofhumanminorglandsecretion.JDentRes 2008;87:445–50.

18.FergusonDB.Oralbioscience.UnitedKingdom:AuthorsOn LineLtd.;2006.

19.FergusonDB.Dehydrogenaseenzymesinhumansaliva.

ArchOralBiol1968;13:582–8.

20.DeJong.vanderHoeven.Thegrowthoforalbacteriaon saliva.JDentRes1987;66:498–505.

21.FergusonDB,BotchwayCA.Circadianvariationsintheflow rateandcompositionofwholesalivastimulatedby mastication.ArchOralBiol1979;24:877–81.

22.DawesC.Theeffectsofflowrateanddurationofstimulation ontheconcentrationsofproteinandmainelectrolytesin humanparotidsaliva.ArchOralBiol1969;14:277–94.

23.StreckfusC,BiglerL,DellingerT,PfeiferM,RoseA,Thigpen JT.CA15-3andc-erbB-2presenceinthesalivaofwomen.

ClinOralInvest1999;3:138–43.

24.StreckfusC,BiglerL,DellingerT,DaiX,KingmanA,Thigpen JT.Thepresenceofsolublec-erbB-2insalivaandserum amongwomenwithbreastcarcinoma:apreliminarystudy.

ClinCancerRes2000;6:2363–70.

25.StreckfusC,BiglerL.Theuseofsoluble,salivaryc-erbB-2for thedetectionandpost-operativefollow-upofbreastcancer inwomen:theresultsofafive-yeartranslationalresearch study.AdvDentRes2005;18:17–24.

26.DoddsMW,JohnsonDA,YehCK.Healthbenefitsofsaliva:a review.JDent2005;33:223–33.

27.DennyPC,DennyPA,TakashimaJ,SiY,NavazeshM, GalliganJM.Anovelsalivatestforcariesriskassessment.J CanDentAssoc2006;34:287–94.

28.DennyPC,DennyPA,TakashimaJ,GalliganJ,NavazeshM.A novelcariesrisktest.AnnNYAcadSci2007;1098:204–15.

29.MungiaR,CanoSM,JohnsonDA,DangH,BrownJP.

Interactionofageandspecificsalivacomponentoutputon caries.AgingClinExpRes2008;20:503–8.

30.TaoR,JurevicRJ,CoultonKK,TsutsuiMT,RobertsMC, KimballJR,etal.Salivaryantimicrobialpeptideexpression anddentalcariesexperienceinchildren.AntimicrobAgents Chemother2005;49:3883–8.

31.DaleBA,TaoR,KimballR,JurevicRJ.Oralantimicrobial peptidesandbiologicalcontrolofcaries.BMCOralHealth 2006;6(Suppl.1):S13.

32.VitorinoR,LoboMJC,DuarteJR,Ferrer-CorreiaAJ,

DominiguesPM,AmadoFML.Theroleofsalivarypeptidein dentalcaries.BiomedChromatogr2005;19:214–22.

33.VitorinoR,deMoraisGuedesS,FerreiraR,LoboMJC,Duarte J,Ferre-CorreiaAJ,etal.Two-dimensionalelectrophoresis studyofinvitropellicleformationanddentalcaries susceptibility.EurJOralSci2006;114:147–53.

34.RudneyJD,StaikovRK,JohnsonJD.Potentialbiomarkersof humansalivaryfunction:amodifiedproteomicapproach.

ArchOralBiol2009;54:91–100.

35.KornmanKS,CraneA,WangHY,diGiovlneS,NewmanMG, PirkFW,etal.Theinterleukin-1genotypeasaseverity factorinadultperiodontaldisease.JClinPeriodontol 1997;24:72–7.

36.SocranskySS,HaffajeeAD.Dentalbiofilms:difficult therapeutictargets.Periodontology20002002;28:12–53.

37.AurerA,Jorgic-SrdjakK,PlancakD,Stavljenik-RukavinaA, Aurer-KoseljJ.Proinflammatoryfactorsinsalivaaspossible markersforperiodontaldisease.CollAntropol2005;29:435–9.

38.FergusonDB.Currentdiagnosticusesofsaliva.JDentRes 1987;66:420–4.

39.KaufmanE,LamsterIB.Analysisofsalivaforperiodontal diagnosis–areview.JClinPeriodontol2000;27:453–65.

40.MillerCS,KingCPJr.LangubMC,KryscioRJ,ThomasMV.

Salivarybiomarkersofexistingperiodontaldisease:across sectionalstudy.JAmDentAssoc2006;137:322–9.

41.RaiB,KharbS,JainR,AnandSC.Biomarkersofperiodontitis inoralfluids.JOralSci2008;50:53–6.

42.LamsterIB,KaufmanE,GrbicJT,WinstonLJ,SingerRE.Beta- glucuronidseactivityinsaliva:relationshiptoclinical periodontalparameters.JPeriodontol2003;74:353–9.

43.Ma¨ntyla¨ P,StenmanM,KinaneDF,TikanojaS,LuotoSH, SaloT,etal.Gingivalcrevicularfluidcollagenase-2(MMP-8)

(9)

teststickforchairsidemonitoringofperiodontitis.J PeriodontolRes2003;38:436–9.

44.BassimCW,RedmanRS,DeNucciDJ,BeckerKL,NylenES.

Salivaryprocalcitoninandperiodontitisindiabetes.JDent Res2008;87:630–4.

45.WuY,ShuR,LuoL-J,GeL-H,XieY-F.Initialcomparisonof proteomicprofilesofwholeunstimulatedsalivaobtained fromgeneralizedaggressiveperiodontitispatientsand healthycontrolsubjects.JPeriodontalRes2009;44:636–44.

46.DaRo´sGonc¸alvesL,SoaresMR,NogueiraFCS,GarciaC, CamisascaDR,DomontG,etal.Comparativeproteomic analysisofwholesalivafromchronicperiodontitispatients.

JProteomics2010;73:1334–41.

47.RamseierCA,KinneyJS,HerrAE,BraunT,SugaiJV,D’Silva NJ,etal.Identificationofpathogenandhostresponse markerscorrelatedwithperiodontaldisease.JPeriodontol 2009;80:436–46.

48.FranzmannEJ,ReateguiEP,PedrosoF,PernasFG, KarakullukcuBM,CarrawayKL,etal.SolubleCD44isa potentialmarkerfortheearlydetectionofheadandneck cancer.CancerEpidemiolBiomarkersPrev2007;16:1348–55.

49.NaglerR,BaharG,ShpitzerT,FeinmesserR.Concomitant analysisofsalivarytumormarkers–anewdiagnostictool fororalcancer.ClinCancerRes2006;12:3979–84.

50.LiY,StJohnMA,ZhouX,KimY,SinhaU,JordanRCK,etal.

Salivarytranscriptomediagnosticsfororalcancerdetection.

ClinCancerRes2004;10:8442–50.

51.WongDT.Salivarydiagnosticpoweredby

nanotechnologies,proteomicsandgenomics.JAmDent Assoc2006;137:313–21.

52.JessieK,PangWW,HajiZ,RahimA,HashimOH.Proteomic analysisofwholehumansalivadetectsenhanced

expressionofinterleukin-1receptorantagonist,thioredoxin andlipocalin-1incigarettesmokerscomparedtonon- smokers.IntJMolSci2010;11:4488–505.

53.YangLL,LiuW,ChengB,LiMT.Comparativeanalysisof wholesalivaproteomesforthescreeningofbiomarkersfor orallichenplanus.InflammRes2006;55:405–7.

54.RyuOH,AtkinsonJC,HoehnGT,IlleiGG,HartTC.

IdentificationofparotidsalivarybiomarkersinSjogren’s syndromebysurface-enhancedlaserdesorption/ionization time-of-flightmassspectrometryandtwo-dimensional differencegelelectrophoresis.Rheumatology2006;45:1077–86.

55.RaoPV,ReddyAP,LuX,DasariS,KrishnaprasadA,BiggsE, etal.Proteomicidentificationofsalivarybiomarkersof type-2diabetes.JProteomeRes2009;8:239–45.

56.MinarowskiL,MinarowskiA,KarwowskaA,SandsD.

ActivityoflysosomalcarboxypeptidaseAinmixedsalivaof cysticfibrosispatients.JCysticFibrosis2008;7:S19.

57.GiustiL,BaldiniC,BazzichiL,CiregiaF,TonazziniI,Mascia G,etal.Proteomeanalysisofwholesaliva:anewtoolfor rheumaticdiseases–theexampleofSjogrensyndrome.

Proteomics2007;7:1634–43.

58.LandrumMI,WilsonCH,PerriLP,HannibalSL,O’ConnellRJ.

Usefulnessofarapidhumanimmunodeficiencyvirus-1 antibodytestforthemanagementofoccupationalexposure tobloodandbodyfluid.InfectControlHospitalEpidemiol 2005;26:768–74.

59.MalamudD,NagashunmugamT,DavisC,KennedyS, AbramsWR,KreamR,etal.InhibitionofHIVinfectivityby humansaliva.OralDis1997;Suppl.3:58–63.

60.FergusonDB,Salivaasadiagnosticfluid. ActaStomatolCroat 2009;43:77–82.

61.HuS,LiY,WangJ,XieY,TjonK,WolinskyL,etal.Human salivaproteomeandtranscriptome.JDentRes2006;85:1129–33.

62.XieH,OsongoG,PopkoJ,deJongEP,CaoJ,CarlisJV,etal.

Proteomicsanalysisofcellsinwholesalivafromoralcancer patientsviavalue-addedthree-dimensionalpeptide fractionationandtandemmassspectroscopy.MolCell Proteomics2008;7:486–98.