Association of rs1319868, rs1567811 and rs8041224 of gene with infection among sickle cell anemia Tunisian patients

(1)

Original research article/ Praca oryginalna

Association of rs1319868, rs1567811 and rs8041224 of IGF1R gene with infection among sickle cell

anemia Tunisian patients

Mouna Ben Sassi, Leila Chaouch *, Miniar Kalai, Imen Moumni, Houyem Ouragini, Imen Darragi, Dorra Chaouachi, Imen Boudrigua, Raouf Hafsia, Salem Abbes

UniversitédeTunisElManar,InstitutPasteurdeTunis,Laboratoired’HématologieMoléculaireetCellulaire,Tunis, Tunisia

Introduction

Despite sharing the same b globin mutation, the range of severity inthe phenotype of SCAis striking,with patients

disabled by many complications. SCA patients are also differently predisposed to several pathological manifesta- tions.Over thelast 20yearsscientiﬁc researchhastriedto elucidate the role of several factors responsible for this clinical variability[1]. In factsickle cell phenotypeappears article info

Articlehistory:

Received:21.09.2015 Accepted:04.10.2016 Availableonline:14.10.2016

Keywords:

IGF1R

SCA

Infection

abstract

Background and aim:Sickle cellanemia (SCA) is characterized byvariable patternsof clinicalexpression. Polymorphismslinkedtodifferentgeneshavebeen associatedwith specificcomplicationsofthedisease.Herein,wefocusedonthestudyoftheassociation of4polymorphismsofInsulinlikeGrowthFactor1receptor(IGF1R)genewithinfections, whicharethemajorcauseofdeathinSCA.Materialandmethods:Thisstudyinvolved 116sicklecellpatientsamongwhom58SShavethesameconfirmedinfectiouspheno- type. Allele-SpecificPCR was performed for the studyof rs1319868, whereas the PCR/

sequencingmethodwascarriedoutforrs1567811,rs2872060andrs8041224.Results: The resultsshowedthatrs1319868and rs1567811wereassociated witha decreased riskof infectionamongSS patients(p=0.038,RR=0.54;p=0.044,RR=0.56,respectively). Inte- restingly,thecombinationofdifferentgenotypesshowedtheassociationofthegenotype GTofrs1319868andthegenotypeCCofrs8041224withfurtherdecreasedinfectionrisk in SCA (p=0.028, RR=0.04). Conclusion:These signiﬁcant associations of IGF1R SNPs withinfectionsuggestthatthisgenecouldplayanimportantroleintheimmunefunc- tioninSCA.

*Correspondingauthorat:InstitutPasteurdeTunis,laboratoired’hématologiemoléculaireetcellulaire,13,placepasteurBP74Tunis leBelvédére,Tunis1002,Tunisia.Tel.:+21671783022.

E-mailaddress:Leila.chaouch@gmail.com(L.Chaouch).

ContentslistsavailableatScienceDirect

Acta Haematologica Polonica

journal homepage:www.elsevier.com/locate/achaem

http://dx.doi.org/10.1016/j.achaem.2016.10.004

zo.o.Allrightsreserved.

(2)

to be modulated by polymorphisms of genes involved in several biological mechanisms whose understanding will elucidate the pathophysiology of the disease [2]. This modulationhasbeen firstlyattributed tothestructureof b globincluster.Asamatteroffactthevariationinhemoglo- bin F (HbF) levels is linked with b globin gene cluster haplotypesandotherchromosomalsites,significantlyinflu- encing the severity of the disease [3]. Additionally, coex- istenceof a-thalassemia with SCAproduces hematological and clinical consequences that are beneficial in some complications but damaging in others [4]. Nevertheless theseglobinmodifiergenesoftheSCAseveritydonotseem tobeenoughtoexplainallthephenotypicheterogeneityof thisdisease.

In the last decade, many studies suggest that SCA phenotype ismultigenic,includingothermutations located in non-globin genes that appear to be involved in the clinicalvariabilityof the disease [5]. Infact, someof these genes may interact to modulate several complications simultaneously. Among these the BMP6 affecting both stroke,legulcerandinfections,TGFBR3involvedinvascular necrosis,legulcers,priapismandinfections[6].Othergenes maybespeciﬁc toasinglecomplicationincludethecaseof UGT1A1incholelithiasis[7–11]andIGF1Rininfection[12].

Indeed,polymorphismsinanumberofgenesinvolvedin theimmune responsehavebeen suggestedas contributing factorstoincreasedsusceptibilitytoinfectioninSCA.These genetic events are an important part of the infectious phenotype and explain the inter-individual differences in spiteof allthepatientshavingthesame sickleglobingene inthehomozygous form.Infection iscommoninSCA,and severebacterialinfectionsarethemajorcausesofmorbidity andmortalityinSCApatients[13].ParticularHLAIIsubtypes have been shown to be predisposing or protecting factors for infectiouscomplications [14]while somesingle nucleo- tide polymorphisms in the IGF1R and TGFb/BMP pathway genes have been associated with an increased risk of infection [15]. IGF1R ispervasively expressed byalmost all cells, regulates major biological processes (angiogenesis,

apoptosis,celldifferentiation,etc.)throughinteractionswith several signaling pathways [16]. Variation in IGF1R gene might modulate the risk of infection by contributing to abnormal signaling in the TGFb pathway and affect the responseofBandTlymphocytestobacteria[15].

Inthispaper,weaimedtoexploretheassociationoffour known polymorphisms of IGF1R gene namely: rs1319868, rs1567811, rs2872060andrs8041224withaninfection-prone phenotypeamongTunisianSCApatients.

Material and methods

Material

This study was performed from November 2009 untilMay 2010,inPasteurInstituteofTunis,Tunis,Tunisia.Thestudy involved 116unrelated SCA subjects,all Tunisians, among whom 58 were patients with reported infection cases.

Control subjects had no history of infection. All of them have homozygosity for bT/A globin mutation which called SSandtheyhavethesameBeninhaplotypefor microsatel- lite and restriction haplotype. Extensive patients’ clinical information was gathered in order to establish a case ﬁle database. Table I summarizes the main characteristics of thestudiedpopulation.

Clinicaleventsanalyzed

Data and clinicalevents weretaken from patient’s history via search of theclinical registry. Onthe otherhand all of the patients chosen have developed the same type of infectious events including: pulmonary, meningitis, osteo- myelitisandurinaryinfection.

Laboratorymethods

Venousbloodsamplesof2.5mlvolumewerecollectedfrom 116 sickle cell subjects enrolled in this study and was

TableI–Hematological,demographicandclinicaldataofstudiedpopulation

CaseSCApatientswithinfection ControlSCApatientswithoutinfection p

Number 58SS 58SS

Age(mean) 272.9 313.6 0.425

Sexratio(M/F) 19/39 22/36 0.423

Hb(g/dl) 7.30.9 7.91.3 0.521

RBC(1012/L) 2.891.02 3.290.9 0.270

MCV(ﬂ) 77.21.3 79.70.9 0.560

MCH(pg) 35.71.02 34.92.1 0.100

RDW(%) 5.291.02 4.830.5 0.579

HbA 0 0 1

HbS(%) 86.40.4 860.3 1

HbF(%) 10.60.3 110.1 1

HbA2 30.1 30.2 1

Thedemographicandhematologicvaluesareindicatedasmeanstandarddeviation.

SS–homozygousofbglobingenemutation;Hb–hemoglobin;RBC–redbloodcell;MCV–meancorpuscularvolume;MCH–meancorpuscular hemoglobin;RDW–redblooddistributionwidth.

Statisticsforthecomparisonofdemographicandhematologicalvariablesbetweenthetwogroupswereperformedusingthettestandchi- squaretestasappropriate(SPSS16.0).

(3)

collected in K2-EDTA anticoagulant containers. SCA was diagnosed on the basis of cation-exchange high perfor- mance liquid chromatography (HPLC) (D10, Biorad) and further conﬁrmation by means of DNA studies. The com- pletebloodcountsincludingcountsofredbloodcells(RBC), whitebloodcells(WBC),andthemeasurementofhemoglo- bin(Hb),meancorpuscularvolume(MCV),meancorpuscular hemoglobin (MCH), mean corpuscular hemoglobin concen- tration(MCHC),and redcell distributionwidth (RDW)were performedusinganautomatedcellcounter(ABXpentra60c +).GenomicDNAwasextractedfromperipheralbloodusing the standard phenol-chloroform procedure. bs-globin gene was performed by restriction fragment length polymorphism(RFLP)aspreviouslydescribedbyRomanaMetal.2000 [17]. We determined total and fetal hemoglobin (Hb F) concentrationsbyHPLC(D10BioRad).

Molecularmethods

116 patients chosen for the molecular methods were selected on the basis of homozygosity for b globin gene.

Genomic DNA was extracted from peripheral blood bythe standardphenol-chloroformprocedure.rs1567811,rs2872060 andrs8041224genotypingwasperformedusingPCR/sequen- cing. Whereas rs1319868 genotyping was performed using AS-PCR.

rs1567811,rs2872060andrs8041224genotyping

The analysis of rs1567811, rs2872060 and rs8041224 was performed by means of simplex PCR using primers shown in TableII. Each PCR reaction contained 100ng of template genomic DNA. After an initial denaturing at 948C for 10min, PCR was carried out for 35 cycles each consistingofdenaturingat 948Cfor1minfollowed by an annealing step at different temperatures for each primer pair during 1min and extension at 728C for 1min. The ﬁnal extension was performed at 728C for 10min. PCR products were puriﬁed using Exonuclease I and Shrimp Alkaline Phosphatase andsubsequentlydoublysequenced by ABI PRISM Big Dye Terminator Ready Reaction using ABI310DNASequencer(PEAppliedBiosystems,Fosterlity, USA).

rs1319868genotyping

An allele-speciﬁc PCR was performed for rs1319868 using 3primers, onecommon(13C) andtwo allelespeciﬁc(13W:

wildtype;13M:mutant).Theprimersequencesarelistedin the table below(Tab.II). Theampliﬁcationconsisted of an initial denaturing at 948Cfor 10min followed by35 cycles of denaturing at 948C for 30s, annealing at 608C for 30s andextensionat728Cfor45s.Theﬁnalextensionstepwas 728Cfor10min.PCRfragmentswererunouton2%agarose gelfor1hour,andtheexpected221pbbandwasvisualized underUV.

Dataanalysis

The116SSpatientsweredividedintotwogroupsaccording to the presence or absence of the infection. Firstly, we compared demographic, hematological and clinical data between the two groups of patients (case and control).

Secondly, we test for trait association with the candidate SNPs, genotypeand allele frequenciesbetween twogroups accordingtothepresenceortheabsenceoftheinfection.

Statisticalanalysis

Thesampleofpatientswasdividedintotwogroupsaccord- ing to the presence or absence of each complication. The demographicand hematologicdatawerenormallydistribu- ted,sotheauthorcalculatedmeansandstandarddeviations using SPSS (18.0). Then we compared demographic and hematological andclinical databetweenthe twogroups of patientsapplyingthet-test.AllSNPsweretestedfor devia- tion from the Hardy-Weinberg equilibrium using the soft- ware package Arlequin (version 3.01). Chi Square test or ﬁsher test was used to determine genetic differences between patients using Beyond compare 2 (version 1.02).

Stratification of different combination of genotypes found according to the presence or absence of infection was evaluatedbylogisticregressionmodelusingSPSS(18.0)and statisticalsignificancewasdefinedasp<0.05.

For the haplotype analysis, linkage disequilibrium between SNPs and the presentation of marker haplotypes wasinvestigatedwithHaploview(version4.2).

TableII–conditionsofPCR

SNPID SNPlocation Primerset Primers Ampliconsizepb Annealtemperature

rs1319868 57825⁰UTR 13C 5⁰CACCGCAAATGACAAGTGAAAA3⁰ 221pb 608C

13W 5⁰TGGGCCAAGAGTGATAGGCAC3⁰ 13M 5⁰TGGGCCAAGAGTGATAGGCAA3⁰

rs1567811 +35496intron2 15F 5⁰GCTTTGAGAGACCGAGACAGGA3⁰ 362pb 598C

15R 5⁰AATGACCGGCAAACTGTGAAAT3⁰

rs8041224 +51328intron2 80F 5⁰GGGAACCTGAGGATTTGGAAGT3⁰ 377pb 588C

80R 5⁰GAACCAAAACTTCCCTCCCTGT3⁰

rs2872060 +4203exon21 28F 5⁰CGCTCATGTAAGAGAGGGCTGT3⁰ 400pb 598C

28R 5⁰CCTGAAGCCAGATTCCAACAAG3⁰

SNPpositionwassetusingthereferencesequencefromNCBIdatabase.C–commonprimer;W–wildtypeprimer;M–mutantprimer;F– forwardprimer;R–reverseprimer.

(4)

Results

Demographic,hematologicalandbiochemicalanalysis

Thehemoglobinproﬁlesofthesampleofpatientsshowthe presenceof 1phenotypewhichisthe homozygousmutant SSpatients.

The distribution of each continuous variable was performed using the t-test. Theresultsshow that there isno signiﬁcantdifferencebetweenthetwogroupsofSCApatient according to the presence or the absence of infection (p>0.05).

Polymorphismsanalysis

For each polymorphism the samples were found to be in Hardy-Weinbergequilibrium(p>0.05).

The analysis of the rs1319868 showed the presence of three genotypes namely: GG, GT and TT in both patient groups. The results showed that the mutant allele T was signiﬁcantly associated withthe infection in SCAand that this allele seemstodecrease the susceptibilitytoinfection inSCA(Tab.III).

The analysis of the rs1567811 showed the presence of three genotypes namely: CC, CG and GG in both patient groups. The results showed that the mutant allele G was signiﬁcantly associated withthe infection in SCAand that

this allele seems todecrease thesusceptibility toinfection inSCA(Tab.III).

The analysis of the rs8041224 showed the presence of three genotypes namely: TT, TC and CC in both patient groups. The findings showed no significant association between patients and controls according to genotypic and allelicprofile(Tab.III).

Theanalysisof rs2872060 showedthe presence ofthree genotypes namely: GG, GTand TT inboth patient groups.

The findings showed no significant association between patients and controls according to genotypic and allelic profile(Tab.III).

Significantfinding wasfoundingenotypiccombinations of different SNPs. Interestingly, stratification of genotypes foundusinglogisticregressionaccordingtothepresenceor absence of infection revealed that the combination of rs1319868inheterozygousstateGTandrs8041224inmutant homozygous state CC was the most significant (25.86%

controland3.44%case,datanotshown)andhenceappears to decrease the susceptibility to infection in SCA with a globalp value of 0.028yielded RR of 0.04(95% CI, 0.001– 1.048) (Tab. IV). Next, haplotype probability was performed through deductionof differentpossibilities of allelecombi- nation.Thestatisticalmethodsassumethatallsubjectsare unrelated and that haplotypes are ambiguous (due to unknownlinkagephaseofthegeneticmarkers).Thegenetic markers are supposedtobecodominant(one-to-onecorre- spondence betweentheirgenotypes andtheir phenotypes),

TableIII–DistributionofpolymorphismsgenotypesandallelefrequencyaccordingtothepresenceornotofInfectionin SCApatients

Polymorphisms ControlN=58 CaseN=58 pvalue RRCI95%

rs1319868

GG 20 32 1* –

GT 28 20 0.074 –

TT 10 6 0.160 –

G 68 84 1* –

T 48 32 0.038 0.54(0.299–0.969)

rs1567811

CC 6 10 1* –

CG 26 34 0.898 –

GG 26 14 0.114 –

C 38 54 1* –

G 78 62 0.044 0.56(0.317–0.985)

rs8041224

TT 5 10 1* 1*

TC 23 28 0.608 –

CC 30 20 0.128 –

T 33 48 1* –

C 83 68 0.054 –

rs2872060

GG 16 11 1* 1*

GT 29 28 0.628 –

TT 13 29 0.244 –

G 61 50 1* 1*

T 55 66 0.189 –

1*–referencegroup;P–indexofsigniﬁcance;RR–relativerisk;CI–intervalofconﬁdence.

Case:SCApatientswithinfection.

Control:SCApatientswithoutinfection.

(5)

andsowereferedtothemeasurementsofgeneticmarkers asgenotypes.ThefindingsshowedthatthehaplotypeGGTC was major incase (0.387) and control (0.295) and that the haplotype TGGT was minor in all subjects (0.112 in cases and 0.13 in controls). However no significant differences weredetectedbetweenthetwogroupsofSCApatients(data not shown). For linkage disequilibrium, expressed as r², evaluation between pairwise SNPs was determinate. Inter- estingly, significantfindings werefoundwith2IGF1R markers namely: rs1567811 and rs8041224 in strong linkage disequilibriumwiththemostsignificantassociation(r²=29).

Discussion

Previousstudies have suggested thatgenetic heterogeneity inﬂuencethe susceptibilityto infection in SCA[14, 15, 18– 20]. Polymorphisms in a number of gene involved in immune response have been suggested as contributing modify thissusceptibility[1]. HLAIIhas beenshown tobe protecting for infections complications [14] while polymorphisms of the manose-bending lectin [21], Fc receptor [19] and gene of the TGFb/BMP pathway have been associated withanincreased riskof infection. Asfor theIGF1R gene, Adewoye et al. have showed that the mutant allele T of rs1319868 increase the riskof infection withp=0.059 andRR=1.92(95%CI,1.20–3.03).Bycons,ourresultsshowed that the mutant allele T seems to have a protective role against infection p=0.027; RR=0.54 (95% CI, 0.317–0.985).

This result can be conﬁrmed by the fact that this poly- morphismisaffectingabindingsitefortranscriptionfactors at5⁰UTRofIGF1Rgene.

ThesamestudyhasreportedthatthehomozygousGGof rs1567811 is not associated with infection (p=0.0586).

Whereas our ﬁndings showed that the mutant allele G is associatedwithaprotective factoragainstinfectionamong SCA patients withp=0.032; RR=0.56 (95% CI, 0.317–0.985).

Furthermore,Adewoyeet al.haveshowed that themutant homozygousgenotypeCC ofrs8041224, appearstoincrease theriskofinfectioninSCAwithp=0.0133;RR=1.96(95%CI, 1.15–3.33) and that the homozygous mutant TT of polymorphism rs2872060 presented a risk factor for infection with p=0.0321; RR=3.03, (95% CI, 1.10–8.33). Our ﬁndings showed that rs2872060 and rs8041224 were notassociated withinfection inSCA. This resultcan beexplained bythe localizationof thers2872060inSTSwhichbeeliminatedby

splicingactivityandthereforeithasnoeffectonexpression or activity of the IGF1 receptor and the localization of rs8041224inintronicregion.

ThelimitofthestudyreportedbyAdewoyeetal.isthat theyenrolledintheirstudydifferentethnicgroups;however oursubjectsareall Tunisians.Moreovertheyhaveenrolled in theirstudyonlypatientswithbacteremia, whereas,this study included pulmonary, meningitis, osteomyelitis and urinaryinfection.

In addition the additive effectof the mutations studied by different genotypiccombinations was tested. A correla- tion betweenthe combination GT/CC (rs1319868/rs8041224) and infectionin SCA suggesta protective role againstthis complication with p=0.028; RR=0.04 (95% CI, 0.001–1.048).

Tothe bestofourknowledgethis resultisreportedforthe ﬁrsttimeamongSCApatients.

Conclusion

The novelty of this report is that it is the ﬁrst time that a similar study was made on the SCA Tunisian patients.

Theresultsshowaclearassociationofrs1319868,rs1567811 and rs8041224 of IGF1R gene with Infection among SCA Tunisianpatients.Interestinglytheﬁndingsshowtheasso- ciation of the mutantallele G of rs1567811with protective effect for infection, association notdescribed previouslyin other population.Theresultswillrequireadditional valida- tionsandfunctionalstudiestounderstandthemechanisms bywhichIGF1Rmightplayaroleinthe modulationof this complication.

Authors’ contributions/ Wkład autorów

LC – study design, statistical analysis, data interpretation, manuscript preparation. MK – data collection. MBS – data collectionandinterpretation,statisticalanalysis,manuscript preparation, literature search,funds collection. IM – statistical analysis. HO – data interpretation. SA – manuscript preparation.FM,RH,IB,DC–fundscollection.

Conﬂict of interest/ Konﬂikt interesu

Nonedeclared.

Financial support/ Finansowanie

ThisstudyissupportedbyLaboratoryLR11IPT07.

Ethics/ Etyka

The work described in this article hasbeen carried out in accordance withThe Code of Ethics of the World Medical Association (Declarationof Helsinki)for experiments invol- ving humans; EU Directive 2010/63/EU for animal experiments;UniformRequirementsformanuscriptssubmittedto TableIV–Repartitionofcombinatedgenotypesaccording

tothepresenceofinfection

Combinatedgenotypes Infection

p OR(CI95%)

(rs1319868/rs1567811) 0.230 –

(rs1319868/rs2872060) 0.120 –

(rs1319868/rs8041224GT/CC) 0.028 0.04 (0.001–1.048)

(rs1567811/rs2872060) 0.560 –

(rs1567811/rs8041224) 0.740 –

P–indexofsigniﬁcance;OR–oddsratio,CI–intervalofconﬁdence.

(6)

Biomedicaljournals.Thisstudywasapproval bythe ethics committeeofPasteurInstituteofTunis.

references/ pi smiennictwo

[1] DrissA,AsareKO,HibbertJM,GeeBE,AdamkiewiczTV, StilesJK.Sicklecelldiseaseinthepostgenomicera:a monogenicdiseasewithapolygenicphenotype.Lapresse 2009;2:23–48.

[2] FertrinKY,CostaF.Genomicpolymorphismsinsicklecell disease:implicationsforclinicaldiversityandtreatment.

ExpertRevHematol2010;3:443–458.

[3] BenMustaphaM,MoumniI,ZoraiA,DouziK,GhanemA, AbbesS.Microsatelliteandsinglenucleotide

polymorphismsintheb-globinlocuscontrolregion- hypersensitivesite2:speciﬁcityofTunisianbS chromosomes.Hemoglobin2012;36:533–544.

[4] NagelRL,SteinbergMH.Roleofepistatic(modiﬁer)genesin themodulationofthephenotypicdiversityofsickle-cell anemia.PediatrPatholMolMed2001;20:123–136.

[5] SteinbergMH.Predictingclinicalseverityinsicklecell anaemia.BrJHaematol2005;129:465–481.

[6] ChaouchL,KalaiM,BenJbaraM,BenChaabeneA,DarragiI, ChaouachiD,etal.Associationbetweenrs267196and rs267201ofBMP6geneandosteonecrosisamongSickleCell Aneamiapatients.BiomedPapMedFacUnivPalacky OlomoucCzechRepub2015;159:145–149.

[7] ChaouchL,TalbiE,MoumniI,BenChaabeneA,KalaiM, ChaouachiD,etal.EarlycomplicationinSickleCellAnemia childrenduetoA(TA)nTAApolymorphismatthepromoter ofUGT1A1gene.DisMarkers2013;35:1–6.

[8] CarpenterSL,LieffS,HowardTA,HowardTA,EgglestonB, WareRE,etal.UGT1A1promoterpolymorphismsandthe developmentofhyperbilirubenemiaandgallbladder diseaseinchildrenwithsicklecellanemia.AmJHematol 2008;83:800–803.

[9] ChaarV,KeclardL,DiaraJP,LeturduC,ElionJ, KrishnanoorthyR,etal.AssociationofUGT1A1 polymorphismwithprevalenceandageatonsetof cholelithiasisinsicklecellanemia.Haematologica 2005;2:188–199.

[10] HaverﬁeldEV,McKenzieCA,ForresterT,BouzekriN, HardingR,SerjeantG,etal.UGT1A1variationand gallstoneformationinsicklecelldisease.Blood 2005;105:968–972.

[11] FertrinKY,MeloMB,AssisAM,SaadST,CostaFF.UDP- glucuronosyltransferase1genepromoterpolymorphismis associatedwithincreasedserumbilirubinlevelsand cholecystectomyinpatientswithsicklecellanemia.Clin Genet2003;64:160–162.

[12] SebastianiP,SolovieffN,HartleySW,MiltonJN,RivaA, DworkisDA,etal.Geneticmodiﬁersoftheseverityofsickle cellanemiaidentiﬁedthroughgenome-wideassociation study.AmJHematol2010;85:29–35.

[13] BoothC,InusaB,ObaroSA.Infectioninsicklecelldisease:a review.IntJInfectDis2010;14:2–12.

[14] TamouzaR,NeonatoMG,BussonM,MarzaisF,GirotR, LabieD,etal.Infectiouscomplicationsinsicklecelldisease areinﬂuencedbyHLAclassIIalleles.HumImmunol 2002;63:194–199.

[15] AdewoyeAH,NolanVG,MaQ,BaldwinC,WyszynskiDF, FarrellJJ,etal.AssociationofpolymorphismsofIGF1Rand genesinthetransforminggrowthfactor-beta/bone morphogeneticproteinpathwaywithbacteremiainsickle cellanemia.ClinInfectDis2006;43:593–598.

[16] BaudlerS,BaumgartlJ,HampelB,BuchT,WaismanA, SnapperCM,etal.Insulin-likegrowthfactor-1controlstype 2Tcell-independentBcellresponse.JImmunol

2005;174:5516–5525.

[17] RomanaM,KéclardL,FrogerA,LavocatE,Saint-MartinC, BerchelC,etal.Diversegeneticmechanismsoperateto generateatypicalbShaplotypesinthepopulationof Guadeloupe.Hemoglobin2000;24:77–87.

[18] NorrisCF,SurreyS,BuninGR,SchwartzE,BuchananGR, McKenzieSE.RelationshipbetweenFcreceptorIIA polymorphismandinfectioninchildrenwithsicklecell disease.JPediatr1996;128:813–819.

[19] Dossou-YovoOP,ZaccariaI,BenkerrouM,HauchecorneM, AlbertiC,RahimyMC,etal.EffectsofRANTESandMBL2 genepolymorphismsinsicklecelldiseaseclinical outocomes:Associationoftheg.In1.1T>CRANTESvariant withprotectionagainstinfections.AmJHematol

2013;84:378–380.

[20] KalaiM,ChaouchL,MansourIB,HafsiaR,GhanemA,Abbes S.FrequencyofthreepolymorpismsofCCL5genenamely rs2107538,rs2280788andrs2280789andtheirimplication onthephenotypicexpressionofsicklecellanemiain Tunisia.PolJPathol2013;2:84–89.

[21] NeonatoMG,LuCY,Guilloud-BatailleM,LapouméroulieC, Nabeel-JassimH,DabitD,etal.Geneticpolymorphismof themannose-bindingproteingeneinchildrenwithsickle celldisease:identiﬁcationofthreenewvariantallelesand relationshiptoinfections.EurJHumGenet1999;7:679–686.