Genetic Screening for the Vel- Phenotype Circumvents Difficult Serological Screening Due to Variable Vel Expression Levels

UvA-DARE (Digital Academic Repository)

Genetic basis of rare blood group variants

Wigman, L.

Publication date 2013

Link to publication

Citation for published version (APA): Wigman, L. (2013). Genetic basis of rare blood group variants.

General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.

UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

Download date:28 Sep 2021 Chapter 6

Genetic screening for the Vel- phenotype circumvents difficult serological screening due to variable Vel expression levels

Lonneke Haer-Wigman1 Shabnam Solati1 Aïcha Ait Soussan1 Erik Beckers2 Pim van der Harst3 Marga van Hulst-Sundermeijer4 Peter Ligthart1 Dick van Rhenen5 Hein Schepers3 Tamara Stegmann1 Masja de Haas1 C. Ellen van der Schoot1

1 Sanquin Research, Amsterdam and Landsteiner Laboratory, Academic Medical Centre, University of Amsterdam, Amsterdam, the Netherlands 2 Maastricht University Medical Centre+, Maastricht, The Netherlands 3 University Medical Center Groningen, University of Groningen, Groningen, The Netherlands 4 Sanquin Diagnostic Services, Amsterdam, The Netherlands 5 Sanquin Blood Bank, Rotterdam, Netherlands

Manuscript under preparation Chapter 6

Abstract Background: Serological determination of the Vel- phenotype is challenging due to variable Vel expression levels. In this study we investigated the genetic basis for variable Vel expression levels and developed a high-throughput genotyping assay to detect Vel- donors.

Methods: In 107 Vel+w and 548 random Caucasian donors genetic variation in the SMIM1 gene was studied and correlated to the level of Vel expression. A total of 3,366 Caucasian, 621 Black and 333 Chinese donors were screened with a high-throughput genotyping assay targeting the SMIM1*64_80del allele.

Results: The Vel+w phenotype is caused by the presence of one SMIM1 allele carrying the major allele of the rs1175550 SNP in combination with in most cases a SMIM1*64_80del allele or in few cases the SMIM1*152T>A or SMIM1*152T>G allele. In ~6% of Vel+w donors genetic factors in SMIM1 could not explain the weak expression. We excluded the possibility that lack of expression of another blood group system was correlated with variable Vel expression levels. Furthermore, using a high-throughput Vel genotyping assay we detected two Caucasian Vel- donors.

Conclusions: Variable Vel expression levels are influenced by multiple genetic factors in SMIM1 and also by genetic or environmental factors. Due to this variable Vel expression serological typing of the Vel- phenotype is difficult and a genotyping assay targeting the c.64_80del deletion in SMIM1 should be considered to screen donors for the Vel- phenotype.

124 of rs1175550 on the non-deletion haplotype. of rs1175550onthenon-deletion weakened Vel thec.64_80del expression andthemajorallele deletionheterozygous carried effect ofrs1175550onSMIM1/Vel thatallindividualswith regulatory by thefact issupported reaction canoccurdueto intravascular hemolysisof Vel+ red bloodcells. with anti-Vel istransfusedwith Vel+ red bloodcellsasevere immediate hemolytictransfusion genetic variation thatleadsto the genetic variation Vel- phenotype. Recently, we othergroups andtwo identifiedthegeneticbasisof Vel antigenandthe phenotype. inthedonorpopulationwillcircumventphenotype thedifficultserological typingofthe Vel- human anti-Vel sera show heterogeneous levels of reactivity. due to large in individual variation Vel expression levels and because the scarcely available are notroutinelyfor typed Vel. Serological screening for the Vel- iscumbersome, phenotype of Variation in Vel expression levels may also berelated inintron to thers1175550variation 2 transcript levelsis associated compared withhigherSMIM1transcript to themajor Aallele. a bloodtransfusionhave atransfusionhazard situation. inanemergency Patients withantibodiesdirected to the Vel bloodgroup antigenandwhoare inneedof Introduction The geneticbasisofweak aimsofthecurrent studyisto theunderlying characterize Vel [Vel+ rare. Vel cantherefore onlybetransfusedwith Vel- red bloodcells, the Vel- is, phenotype however, SMIM1 (c.64_80del) thatabolishes thecomplete expression oftheSMIM1protein. the underlies Vel antigen and the Vel- is causedby a17-basepair deletion in phenotype molecular typing assay tomolecular typing identifydonorswiththe Vel- phenotype. levels ofthe Vel antigen.Furthermore, we developed andvalidated ahigh-throughput expression ofSMIM1ontheexpression andto theeffect characterize ofthegeneticvariation p.Met51Arg). acid substitution, were detected (c.152T>A encoding p.Met51Lys and c.152T>G encoding mutations, atthesamenucleotidepositionofSMIM1,butresulting inadifferent amino Furthermore, individualswithweak intwo Vel expression missense singleheterozygous two expression ofthe Vel antigeniscorrelated presence withtheheterozygous ofthisdeletion. variation is a regulatory region isaregulatory forvariation levels. SMIM1transcript Chilcotin Indians inCanada. Chilcotin Indians and the Vel- occasionallyintheAsian hasbeendescribed population andthe phenotype SMIM1. 1,6 The overall ofthe frequency Vel- intheCaucasian populationis~0.025% phenotype 11,14 Anexpression quantitative traitlocus[eQTL] studyshowed thatthers1175550 11

6-9 The availability of The availability Vel- donorsiseven lower, becausedonors 11 11-13 The SMIM1protein encodedby SMIM1 11,14 The minorGalleleofrs1175550 10 Genetic screening Genetic for the Vel- Molecular basisofweak Molecular Vel expression 3-5 Patients withanti- 1,2 When apatient 11-13 14 Weak This 11,12 125 w ]

Chapter 6 Chapter 6

Material and methods Samples and DNA isolation All cases were included after written informed consent was provided. In 2005, 141 donors with weak Vel expression and five Vel- donors were identified upon serological screening of 10,500 donors with a human polyclonal anti-Vel serum. Nineteen of the 141 Vel+w donors have also been reported in previous work.11 For 107 of the 141 donors material collected in 2005 was available. A total of 3,366 Caucasian, 621 Black and 333 Black Asian random donor samples were included. Samples of Dutch Caucasian blood donor samples were collected by Sanquin Blood Supply, The Netherlands. Asian and Black donor samples were collected by The South African Blood Transfusion Service in Johannesburg South-Africa. Additional Black donor samples were collected by The Ethiopian Red Cross Society Transfusion Service in Addis Ababa, Ethiopia and The Red Cross Blood Bank Curaçao in Willemstad, Curaçao. DNA was extracted from white blood cells using a DNA extraction kit (QIAamp DNA Blood Mini Kit).

Serology Screening for the Vel- phenotype on 10,500 donors with blood group O was performed using undiluted human serum (NL15) in the Cellbind card technique. All donors with a negative agglutination reaction were tested in a second round using standard tube agglutination technique with NL15 and one to three additional human anti-Vel sera (NL25, NLVzO and/or NLVzA). A donor was determined to have weak Vel expression levels when at least one of the Vel sera gave a positive reaction. Vel antigen expression was examined via agglutination using Liss/Coombs cards (Bio-Rad, Veenendaal, The Netherlands) with anti-Vel serum of NL25 on En(a-), S-s-U-, Rh- (regulator type), In(Lu), Lu(a-b-), K0, Kx-, Fy(a-b-), Jk(a-b-), Gy(a-), Co(a-b-), JMH-, p and Ge:-2,-3 red blood cells and on red blood cells of a patient with paroxysmal nocturnal hemoglobinuria. Agglutination strength was assessed after serum and red blood cells were incubated for 15 minutes at 37ºC with a subsequent spin for 10 minutes.

Flow cytometry Vel expression on red blood cells was measured using flow cytometry with immunopurified human polyclonal anti-Vel (NL25 and/or NL04-01). The purified human polyclonal anti-Vel was obtained after incubation of human serum containing anti-Vel with group O Vel+ red blood cells. After washing, the antibodies were eluted from the red blood cells using the Gamma ELU-KIT II (ImmucorGamma). The purified anti-Vel was incubated for 30 minutes at room temperature in a 3% (vol/vol) red blood cell suspension. Subsequently, red blood cells were washed and Alexa-488 labelled goat anti-human-IgG (Molecular Probes) was added in a 1:500 dilution and incubated for 30 minutes at room temperature. Red blood cells were washed and analyzed by flow cytometry (LSRFortessa cell analyzer, Becton Dickinson). Statistical

126 (M1613, Pelicluster), CD47(427603, B.V.,CD44 (G4426Dickinson 559942,Becton Breda, (LA18.18),CD36 RhAG theNetherlands), (BS56),GPD(BRIC4),CD35(E11), Duffy(64-4A8),Band3(BRIC71andBRIC90),ICAM-4 (Bric18), [GP]A (AME1, Pelicluster), (Lu-4F2), (BRIC69)B-CAM Kell (703.7),Rh GPB(OSK29-4), RhCE red cellmembraneproteins; specific monoclonalantibodies[Moabs] murine forGlycophorin Flow to above, measure asdescribed expression cytometry, wasperformed levels ofother were repeated andthesecondmeasurement wasusedfor analysis. Samples withameanfluorescence oflessthan400(n=36)or intensity over 3500(n=17) with the average per experiment intensity mean fluorescence of all experiments. intensity theaverage by correcting of thesampleswasnormalized perexperiment meanfluorescence batch wise;themeanfluorescence red intensity bloodcellsofthe548donorswasperformed t-test. usinganunpaired two-tailed analysis wasperformed The Vel expression levels ofthe Genetic typing of the minor G and major A allele of the typing rs1175550 SNP of Genetic TaqMan assay antibody.secondary Alexa-488 labelledgoatanti-human-IgGorPEanti-mouseIgGwasusedasa (AB5) wereagainst RhD and human MoAbs used. When antibodies were not directly labeled, B.V.), B.V.), Dickinson CD59(p282,Becton CD71(M3162,Pelicluster) (67A4) andE-Cadherin allele confirmation sequencingwasperformed. SMIM1*64_80del alleleconfirmation TaqMan assay. for the minorGalleleofrs1775550orpositive allcaseshomozygous for In the pool was positive all samples of the pool were tested separately using the of apoolwasinthesamerange astheDCtvalueofpositive control sample. When a (average Ct valueofthec.64_80del TaqMan assay –average CtoftheALBTaqManassay) StepOnePlus software. Apoolwas “positive” for theSMIM1*64_80delallelewhenDCtvalue for themutation(negative control) usingthe were along. taken Dataanalysiswasperformed negative for the mutation (positive control) and a sample containing 150 ng DNA negative themutationinabackground of150ngDNA containing 1.6ngDNAfrom adonorcarrying fluorescence signal wasmeasured. Pools were tested induplo.In eachrunapositive control the of15secondsat95°Cand1minute at60 °C attheendofeachcycle 95°C, 50cycles Biosystems) andreverse and0.3μMforward primers. PCRconditionswere: 10minutes at volume (Applied of25mL,containing100-200ngDNA,12.5 mL SYBR®Green Mix PCRMaster Biosystems) according to manufacturers’ inatotal reactions were protocol. performed short In a finalconcentration of150 ng/mL.AllassaysonaStepOnePlus were (Applied performed These ofDNA of 45Caucasian poolscontainedamixture or90Asian orBlackdonorswith and theALB TaqMan andtheSMIM1*64_80del TaqMan onpoolsofDNA. were alsoperformed to measure total DNA concentration. All TaqMan assays on single samples were performed for theSMIM1*64_80del allelewas developed. A TaqMan assay for theALBgenewasused A allele. For high-throughput of the genotyping Vel- a phenotype TaqMan assay specific usingtwo performed TaqMan assays containing, respectively, specific primers for theGor R&D Systems Europe Ltd.), Dickinson CD55(1A10,Becton Molecular basisofweak Molecular Vel expression SMIM1*64_80del SMIM1 was 127

Chapter 6 Chapter 6

DNA sequencing Primers were developed to sequence all exons and introns of SMIM1. The polymerase chain reaction [PCR] was performed on a Veriti thermocycler (Applied Biosystems,) in a total volume of 20 mL, containing 50-150 ng DNA, 10 mL of 2x GeneAmp Fast PCR Master Mix (Applied Biosystems), 0.5 mM forward and reverse primer. PCR conditions were: 10 seconds at 95°C, 35 cycles of 10 seconds at 95°C and 15 seconds at 64°C, followed by 1 minute at 72°C. PCR products were purified using illustra ExoProStar (GE Healthcare) according to manufacturer’s protocol. The sequence reaction was performed on a Veriti thermocycler in a total volume of 20 mL, containing 1 mL of purified PCR product, 1 mL 2.5x BigDye Terminator v1.1 Cycle (Applied Biosystems), 3.5 ml 5x BigDye Terminator Buffer (Applied Biosystems) and 0.25m M forward or reverse primer. Sequence conditions were: 25 cycles of 15 seconds at 95°C, 10 seconds at 50°C and 4 minutes at 60°C. Sequence products were analyzed on a 3130 Genetic Analyzer (Applied Biosystems).

SMIM1 transduction The mLwpRRLsSMIM1itNGFR vector was used as a wild-type SMIM1 construct. Vectors containing the c.152T>A and c.152T>G were made from the wild-type SMIM1 construct using site directed mutagenesis. HEK293T cells (7.5 × 105) were plated in 6-cm2 plates and lentivirally transduced with the three different constructs and a mock construct. Forty-eight hours after transduction, cells were harvested and screened by flow cytometry using purified human polyclonal anti-Vel (NL25 and NL04-01) (see flow cytometry) and an antibody to CD271 (Milteny Biotech B.V.) was taken along as a transduction control.

Results The major allele of rs1175550 is correlated with reduced Vel expression levels To determine the effect of the rs1175550 SNP on Vel expression levels, 548 random Caucasian blood donors were genotyped for the rs1175550 SNP and SMIM1*64_80del allele. The frequency of the major A allele of rs1175550 (78.6%) and of the minor G allele (21.4%) of this cohort was the same as the European population frequency (see Web Resources). Twenty of the 548 donors were heterozygous for the SMIM1*64_80del allele. Vel expression levels in the 548 donors were measured via flow cytometry using purified anti-Vel (NL25). Donors homozygous for the minor G allele had significantly higher Vel expression levels compared to donors heterozygous for rs1175550 and Vel expression of these donors was significantly higher compared to donors homozygous of the major A allele of rs1175550 (Figure 1). Furthermore, donors who were heterozygous for the SMIM1*64_80del allele had significantly lower Vel expression levels compared to donors with the same rs1175550 haplotype, but without the c.64_80del deletion (Figure 1). Interestingly, donors who were homozygous for the major A allele of rs1175550 had roughly the same Vel expression levels as donors who had one

128 SMIM1*64_80del alleleon Vel expression levels, aswe already postulated. thatthepresence oftheminorGallelers1175550neutralizesIndicating theeffect ofthe theminorGalleleofrs1175550. haplotype SMIM1*64_80del alleleandonthenon-deletion 80del allele and the minor G allele of rs1175550 on the non-deletion haplotype (p=0.67). haplotype 80del alleleandtheminorG ofrs1175550onthenon-deletion for the A allele of homozygous rs1175550 and compared to SMIM1*64- donors with a heterozygous rs1175550). No difference in Vel expression level was determinedin donors without the c.64_80del and for forin donorsheterozygous majorAalleleof rs1175550andp=<0.0001 indonorshomozygous levels compared to for donorsheterozygous c.64_80del (p=0.0467 withthesamers1175550haplotype withoutthec.64_80delallele ofrs1175550(p=<0.0001).Donors hadsignificantly higher Vel expression these donorshadsignificantly higher Vel expression levels compared for theA to donorshomozygous for theGallelecompareddonors homozygous to for donorsheterozygous rs1175550(p=0.0004), 80del allele)were alongasnegative controls. taken Vel expression levels were significantly higherin unpaired two-sided T-test. Four donorsdetermined Vel for negative theSMIM1*64- (homozygous with an SMIM*152T>Aallele. heterozygous Statisticalanalysiswasperformed donorswhocarry two fluorescence intensity, error barsindicate the2.5%and97.5%percentile. two blackdotsdepict The levels werecytometry. Box usingflow determined plotshows 25%and75%percentile and average in SMIM1. via inthedonorswasdetermined The SMIM1genotype TaqMan assays and Vel expression Figure 1.Vel expression for levels andc.64_80del in552donorsgenotyped thers1175550variation

n = 35 162 c.64_80del wt/wt wt/wt wt/wt wt/ rs1175550 Mean Fluorsence Intensity (arbitrary units) 10000 1000 100 GA AA AAA AA AG AA AG GG *** 331 3 *** * p=0.67 genotype *** *** wt/ 17 4 *** Molecular basisofweak Molecular Vel expression / 11 129

Chapter 6 Chapter 6

The heterozygous expression of SMIM1*64_80del allele and the major A allele of rs1175550 on the non-deletion haplotype causes the Vel+w phenotype In 2005, 10,500 Dutch donors were serologically typed for the Vel antigen. In five (0.047%) donors the Vel- phenotype was detected and in 141 (1.3%) donors weak Vel expression was detected. In 107 of the 141 Vel+w donors leukocytes were available for DNA isolation. To investigate whether a genetic variation in SMIM1 underlies the Vel+w phenotype, all exons and part of intron 2 (hg19, chr. 1: g.3691316-3691832) in SMIM1 were sequenced. In 101 of the 107 donors the Vel+w phenotype could be assigned to the heterozygous presence of the SMIM1*64_80del allele and the presence of the major A allele of rs117550 on the non-deletion haplotype. The remaining six Vel+w donors were all homozygous for the major A allele of rs1175550, but were negative for the c.64_80del deletion. None of the Vel+w donors was positive for the minor G allele of rs1175550. The serological typing results from the series in 2005 showed that the weak agglutination reactions in five cases without the deletion were of comparable strength to the average levels of agglutination observed with the donors heterozygous for the SMIM1*64_80del allele (Table 1). The original data of UCN13 could not be retrieved anymore. Furthermore, we verified the weak Vel expression levels of these six donors by flow cytometry analysis with immunopurified NL25. Red blood cells of the donation in 2005 were available of the six donors without the c.64_80del deletion in SMIM1 and of 91 Vel+w donors heterozygous for the SMIM1*64_80del allele, which were taken along as controls. As an additional control, red blood cells of ten random donors homozygous positive for the major allele of rs1175550 were taken along. The Vel expression of donors without the c.64_80del deletion did not significantly differ from that of the 91 Vel+w donors heterozygous for the SMIM1*64_80del allele, while expression levels were significantly lower compared to ten control samples homozygous for the major allele of rs1175550 allele (Supplementary Figure S1).

Table 1. Vel expression levels of six donors with weak Vel expression levels without the c.64-80del in SMIM1 PEG technique PEG technique PEG technique PEG technique NL15 anti-Vel in NL15 anti-Vel in NL15 anti-Vel in NL25 anti-Vel NLVzA anti-Vel in anti-Vel NLVzA NLVzO anti-Vel in anti-Vel NLVzO

UCN technique Cellbind 13 0 42 0 0 0.5 0 1 57 0 0 2 0 0 30 0 0 2 0.5 1 40 0 0.5 1 2 2 26 0 0.5 2 1 2 Vel+w donors heterozygous for 0 0.3 1.8 1.1 0.9 SMIM1*64_80del Agglutination in Vel+ donors 1-2 2-3

130 these donorshasalready previously beendescribed c.152T>A missense mutation (encoding p.Met51Lys) in these ninedonorswere sequencedfor allexons donorsthe two andintrons ofSMIM1.In investigate inSMIM1are whetherothergeneticfactors responsible for weak Vel expression, to levelpresence the heterozygous could not be linked of the Genetic analysisGenetic ofnine Vel+ (Table 2).Except(Table inone Vel+ ofintronicand thedistribution SNPs wasnotdifferent compared to the European population (Figure theremaining ninedonorsnomutationswere 1).In present inthecodingsequence donorsislowerin thesetwo compared to for donorsheterozygous theSMIM1*64_80del of rs1175550. for for theminorallele in expression the majororhomozygous donorshomozygous between on expression levels oftheothermembrane proteins, becausewe difference didnotdetect Vel+ and Vel- red bloodcells(Figure 2).Moreover, thelevel of Vel expression hadalsonoeffect red cellmembraneproteins CD36, CD47,CD55,CD71and E-cadherin wasdetected between levels of the GPA, GPB, RhD, Kell,Duffy, B-CAM, RhCE, Band GPD, 3,ICAM-4, RhAG, CD35, CD44, membrane proteins are influenced by thelevel of Vel expression. Nodifference inexpression proteins. addition,we investigated In whethertheexpression levels ofotherred blood cell whichlackexpression hemoglobinuria, ofallGPI-linked a patientwithparoxysmal nocturnal or JMH. CD55,RhAG Aquaporin 1,Xk, Vel expression onthered wasalsonormal bloodcellsof GPB, RhD, Kell,Duffy, B-CAM, RhCE, UreaB, GPC,GPD, transporter Do,Acetylcholinesterase, that themembraneexpression ofthe Vel antigenisnot dependentontheexpression ofGPA, of redwas detected onallthese types blood cells(data notshown) andwe therefore conclude Colton, Kx, Cromer,Dombrock, RHAG or the JMH blood group systems. Normal Vel expression complete expression Lu,blood cellsthatlacked oftheMNS,Rh, Kell,Duffy,Kidd,Gerbich, Yt, expression ofotherred bloodcellmembraneproteins, Vel expression onred wasdetermined withotherproteins.interact To investigate whether Vel expression is dependent on the The ofthe protein thatcould Vel part residues attheextracellular cysteine antigenhastwo membrane proteins Vel expression levels are not correlated withexpression levels ofother red bloodcell elementasrs1175550. is present inthesameregulatory Dutch donorsorin279European inEnsembl(See donorsdescribed Web Resources). This SNP intronic SNPrs191041962. The minoralleleofrs191041962wasnotdetected in91random with the same haplotype. Furthermore, in six of the 107 Vel+ hadreproducibly haplotype) lowernon-deletion Vel expression levels compared to donors heterozygous forheterozygous the for theminoralleleofrs1175550andonedonor allele ofrs1175550,onedonorhomozygous of548randomdonors, for three the major donors(one thecohort donorhomozygous In SMIM1*64_80del w donor, positive whowasheterozygous for theminoralleleof w donorswithunexplained weak expression allele and carrying the minor allele of rs1175550 on the allele and carrying 11 . Interestingly, the Vel expression levels SMIM1 was detected (Table 2), one of w donors the weak expression Molecular basisofweak Molecular Vel expression SMIM1*64_80del allele. To allele 131

Chapter 6

Chapter 6 c.*58G>A

A GA GA GG GG GG GG GG GG GG

3’ UTR 3’ c.152 T>A c.152

A TT TT TT TT TT TT TT T A T A

Exon 4 Exon rs71634364

GA GA GA 8% GG GG GG GG GG GG 0.06 rs2282455 Intron 3 Intron

C TT TC TC TC TC TC

CC CC CC 0.25 42% c.64_80del

wt/ D wt/wt wt/wt wt/wt wt/wt wt/wt wt/wt wt/wt wt/wt

Exon 3 Exon rs1175549

CC AA AA AA AA AA AA AA AA 0.54 21% rs9424296

CC CC CC CC CC CC CC CC CA 7% 0.30 10% rs1175550

AA AA AA AA AA AA AA AA GG 0.54 19% 21% rs6673829

AA GA GA GA GG GG GG GG GG 0.78 29% 32% rs191041962

C T CC CC CC CC CC CC CC CC 0% 0% 0.00 rs70940313

ins ins ins wt/ ins/ 0.97 21%

Intron 2 Intron wt/wt wt/wt wt/wt wt/wt wt/wt wt/wt wt/wt rs2794732

AA AA GA GA GG GG GG GG GG 0.87 45% rs143872648

D /

wt/ D wt/ D wt/wt wt/wt wt/wt wt/wt wt/wt wt/wt donors rs1184341 w

T TT

CT CT CC CC CC CC CC CC 0.96 22% rs2840327

AA GA GG GG GG GG GG GG GG 0.72 10% rs34961462

T TT

CC CC CC CC CC CC CC CC 0.79 12% rs2275819 A AA GA GG GG GG GG GG GG GG 0.21 10% Intron 1 Intron test) 2 Minor allele nucleotide 26-566 UCN 42 Dutch frequencyDutch of minor allele† population European frequency of minor allele 26-561 26-175 13 30 40 26 57 p value (Chi genotype determined in nine Vel+ 2. SMIM1 genotype determined in nine Table † Determined donors on 91 random Dutch

132 has previously been described heterozygous in a donor with very weak inadonorwithvery has previouslyheterozygous beendescribed Vel expression levels. p.Met51Lys SMIM1allele. to awildtype next The c.152T>G mutationencodingp.Met51Arg In this study we identified two thisstudywe identified In Vel+ The c.152T>A andc.152T>G mutations inSMIM1causeweak Vel expression red cellmembraneproteinsCD36, CD47,CD55,CD71andE-cadherin wasdetected. expression ofthe GPA, GPB, RhD, Kell,Duffy, B-CAM, RhCE, GPD, Band3,ICAM-4, RhAG, CD35,CD44, allele of rs1175550 (p = <0.001) and between Vel+ donors and Vel- donors (p = 0.035). No differential Vel+ for theminorGalleleofrs1175550and donorshomozygous Vel+ for themajor donorshomozygous with anunpaired two-sided T-test. The expression ofthe Vel antigenwassignificantly different between donors. Statistical analysis of the expression levels between Vel+ and Vel- red blood cells was performed - dotted black line).Expression levels were oneighttill24 determined Vel+ donorsandfour tilleight Vel- - grey line)andon Vel- for red theSMIM1*64_80del allele(open squares; bloodcellshomozygous mean - blackline),on Vel+ for red themajorAalleleofrs1175550(grey bloodcellshomozygous rounds; mean proteins on Vel+ for red theminorGalleleofrs1175550(blackdiamonds;mean bloodcellshomozygous Scatter plots showcytometry. the mean fluorescence of the different intensity red blood cell membrane Figure 2.Expression levels ofnineteen red bloodcellmembraneproteins were viaflow determined specific mutations were expressed inHEK T293cells. The c152T>G(p.Met51Arg) variant containing the assay. constructs andtwo amockconstruct construct, ASMIM1wild-type in the transmembrane region (Figure 3a). The effect of the c.152T>A and c.152T>G mutations regionThese missensemutations are located oftheprotein inaconserved nearorjustwithin Mean Fluoresence Intensity (arbitrary units) 100000 SMIM1 ontheexpression level of Vel studiedusingaheterologous expression wasfurther 10000 1000

Vel (25 100 10

0 GpA 646 ) (AME-1 GpB (OSK29-4

)

RhD (AB5))

RhC E (7 03-7) Rh B-CAM (Bri c6 9) (LU-4F2)

Kell (B

R Duffy ( IC 1 8)

w 6 Band3 (B4-4A8) donorswiththec.152T>A missensemutationencoding

R Band3 (B IC 7

Antigen (Antibody) 1 ) R IC IC AM-4 (BS 956 0 )

GpD (Br )

ic4) CD35 (

CD44 ( E1 1 ) G RhA 4 4 G 26) ( LA18.18) CD36 (

M1 Molecular basisofweak Molecular Vel expression CD47 ( 613 ) 4 27603) CD55 (

1 A1 CD59 ( 0 ) p CD71 ( 2 8 2) Vel- Vel+(rs1175550 AA) GG) Vel+ (rs1175550 M3 162 E Ca ) d he rin 133 11

Chapter 6 Chapter 6 caused almost the complete absence of the Vel antigen, while the c.152T>A (p.Met51Lys) variant caused a reduction of ~70% of the Vel antigen (Figure 3b). The PolyPhen software predicts that the p.Met51Arg substitution has a damaging score of 0.799 (the closer to 1, the more predicted damaging effect of the mutation) and the p.Met51Lys substitution has a damaging score of 0.5.15 a)

b) Negative non-transduced control Positive wild type SMIM1 control SMIM1*152G allele SMIM1*152A allele Count

Fluorescence Intensity (Arbitrary Units) Figure 3. Effect of missense mutations in SMIM1. a) Conservation of the SMIM1 gene via cross-species sequence alignment of protein homologues in ten different species. Conserved amino acids are colored black, amino acids with the same properties are colored grey, non-conserved amino acids are colored white. Bars below indicate the degree of conservation. The amino acid location bar indicates the putative position of the amino acid in the

Genetic screening for Vel- blood donors Because the Vel- phenotype is caused by a single genetic defect, genetic screening for donors with the Vel- phenotype can be easily performed. A TaqMan assay, sensitive enough to detect one positive sample in a pool of 90 samples (data not shown), was developed for the SMIM1*64_80del allele. A total of 3,366 Caucasian (including the cohort of 548 donors mentioned in the first paragraph of the results), 621 African black and 333 Asian donors were screened with this TaqMan assay in pools containing 45 or 90 donors. In total 94 Caucasian,

134 gene (due to presence heterozygous oftheSMIM64_80delallele)have roughly halfthe Vel gene dosageeffect for theSMIM1gene. RBCofdonorswhohave onlyonefunctional Vel forheterozygous rs1175550 hadintermediate expression addition,we levels. seeaclear In correlated with increased Vel expression levels compared to themajor A allele, anddonors responsible for variable Vel expression levels onred bloodcells;theminorGalleleisstrongly thisstudy, levels.and higherSMIM1transcript In we show thatthers1175550SNPisalso Furthermore, Fermann region, has lower withnuclear binding affinity proteins compared to the major A allele. previously region,contains aregulatory thatmay influence the and Vel expressionwe inerythroblasts positiveand Chinese donors heterozygous for the their race, theSNPs surrounding thec.64_80del were sequencedinCaucasian, black African hence allpersonspositive forthesameancestralallele, regardless thec.64_80del, of carry To decidewhetherthe17nucleotidedeletioninSMIM1occurred inasinglegeneticevent, phenotype. but larger population studiesare ofthe frequency neededto thecorrect determine Vel- donors is Vel-. ofthe The frequency Vel- population in Black and Asian populations is lower, (95% CI0.01%–1.19%).Usingthisallelefrequency, we calculate that~1in5000Caucasian (95% CIblackpopulation 0.15% –0.98%)and0.60% in theAfrican Chinese population SMIM1*64_80del inatube agglutination test withNL25serum.allele wasconfirmed ofthe The allelefrequency allele. The expected forVel- theSMIM1*64_80del donorshomozygous ofthetwo phenotype in allcases, except Caucasian for two samplesthatwere theSMIM1*64_80del homozygous presence ofthesecasesdetected theheterozygous Sequencing oftheSMIM1*64_80del seven blackandfour African Asian donorswere positive for theSMIM1*64_80del genetic variations in SMIM1 genetic variations difficult.phenotype The Velantigeniscoded bytheSMIM1 expressionThe highlyvariable levels ofthe Vel serological ofthe antigen makes typing Vel used to screen donorsfor the Vel- phenotype. Vel expression levels. Furthermore, thedeveloped highthroughputassay canbe genotyping Thehaplotype. deletion presence of the this study we showIn that weak Vel expression is caused most byoften the heterozygous Discussion to thec.64_80dellinked deletioninSMIM1. rs9424296 andtheminorallelesofrs2282455rs71634364(data notshown), thatare themajorallelesofrs1181893,rs6673829,rs1175550and orhomozygous), (heterozygous 11 showed thattheminorGalleleofrs1175550SNP, located inthis regulatory SMIM1*64_80del allele and the major A allele of the rs1175550 SNP on the non- allele was1.46%(95%CI1.17%-1.74%)intheCaucasian population,0.56% et al. mutations also cause weakened SMIM1*152T>A andSMIM1*152T>Gmutationsalsocauseweakened 14 are responsible for expression. this variable 2 of Intron determined a determined strong the minor G allele associationbetween SMIM1*64_80del gene andwe investigated which Molecular basisofweak Molecular Vel expression allele. All donors carried carried SMIM1 allele. allele 135

Chapter 6 Chapter 6 expression compared to RBC of donors with two functional SMIM1 alleles with the same rs1175550 haplotype. The lowest Vel expression levels were found in donors carrying one SMIM1 allele with major A allele of rs1175550 and one SMIM1*64_80del allele. Interestingly, the minor G allele of rs1175550 on the non-deletion haplotype overcame low expression caused by the presence of the SMIM64_80del allele. In a cohort of >10,000 donors screened for the Vel phenotype we identified the Vel+w phenotype in 1.3% of the donors. The frequency of donors with a SMIM1*64_80del allele and the major A allele of rs1175550 on the non-deletion haplotype (which gives the lowest Vel expression levels) is also ~1.3% (taken in account an allele frequency of 1.6% for the SMIM64_80del allele and 79% for the major A allele of rs1175550). Indeed, 94% of the Vel+w donors was heterozygous for the c.64_80del deletion and homozygous for the major A allele of rs1175550. Hence, in most individuals the Vel+w phenotype is caused by one SMIM1 allele carrying the major A allele of rs1175550 in combination with a SMIM1*64_80del allele. The remaining six Vel+w donors were homozygous for the major A allele of rs1175550, but were negative for the c.64_80del deletion. Furthermore, in 548 random Caucasian donors another three Vel+w donors were detected of whom the low Vel expression could not be accounted to presence of the c.64_80del deletion. Two of these nine Vel+w donors were heterozygous positive for the SMIM1*152T>A allele, that is already described to weaken Vel expression levels. In the other seven donors no mutations were present in the coding sequence and no common intronic variation was detected. In one donor the minor allele of rs191041962 was heterozygous present, while the minor allele of rs191041962 was not detected in 367 Caucasian control samples. The rs191041962 is located in the same regulatory element as rs1175550 and it is possible that the minor allele of rs191041962 has an additional enhancing effect on transcription binding, and thereby further reduces Vel expression levels. Nevertheless, in at least six Vel+w donors no mutations and variations in SMIM1 could explain the weak Vel expression levels. We therefore conclude that other genetic factors outside the introns and exons of SMIM1 or environmental factors also influence Vel expression levels. For that reason, we investigated whether other blood group antigens have an effect on Vel expression levels. No differential Vel expression levels were discovered on red blood cells that lacked the complete expression of the MNS, Rh, Lu, Kell, Duffy, Kidd, Gerbich, Yt, Dombrock, Colton, Kx, Cromer, RHAG or the JMH blood group systems. We therefore conclude that Vel expression is not dependent on the expression of these other blood group antigens. The other way around, no difference was detected in expression levels of GPA, GPB, RhD, RhCE, B-CAM, Kell, Duffy, Band 3, ICAM-4, GPD, CD35, CD44, RhAG, CD36, CD47, CD55, CD71 and E-cadherin red cell membrane proteins between Vel+ and Vel- red blood cells. Two Vel+w donors were identified to be heterozygous positive for the SMIM*152T>A allele encoding p.Met51Lys and we previously11 described one donor heterozygous positive for the SMIM*152T>G allele encoding p.Met51Arg. In a heterologous expression assay the c.152T>G

136 Storry Storry blackandAsianthe African seemsto belower (~1in30,000donorshasthe Vel- phenotype). c.64_80del in deletion,aswasdetermined Vel- donors. deletion were positive for allSNPs inintron 2 andintron 3ofSMIM1thatareto the linked blackand Chinesedonorsthedeletionwasidenticaland withthe as intheAfrican as a nucleotide single deletion genetic occurredevent, because most likely upon reduction. [kD] protein31 kiloDalton complex,whereas thisbandmigrates to asinglebandof18 kD Vel protein. oftheSMIM1/ motif ataminoacidposition67till71indicates apossiblehomodimerization allele, whichcanbeexplainedwhenSMIM1formstype ahomodimer. theGXXXG Indeed, that themutationhasadominantnegative effect on Vel expression from the remaining wild- one functional Vel presence allele (heterozygous of the with a p.Met51Lys mutation had far lower expression levels compared than donors with only amino acid, respectively, atthisposition.Unexpectedly, thered donors blood cellsofthetwo while thec.152T>A andc.152T>G mutationsencodethepositively charged lysineorarginine SMIM1/Velwild type protein anon-polarmethionineaminoacidispresent atposition51, that thep.Met51Lys substitution is lessdamaging thanthep.Met51Arg the substitution.In reduced the Vel expression levels with ~70%. Moreover, the PolyPhen also predicts software mutation almost completely abolished Vel expression levels, whilethec.152T>A mutation serology; ~1in4,000 serology; Vel- individualsintheCaucasian population. in theCaucasian population, whichisroughly by determined thesameasfrequency SMIM1*64_80del allelewe thattheoccurrence determine ofthe Vel- is~1in5,000 phenotype Caucasianand detected two Vel- donors. Furthermore, ofthe usingtheallelefrequency and validated ahighthroughputassay to genotyping screen donorsfor the Vel- phenotype assay canmore donorswiththe easilydetect Vel- thanserology. phenotype We developed Because onlytheSMIM1*64_80delalleleiscorrelated withthe Vel- agenotyping phenotype reduction of Vel expression levels. the c.152T>A mutation is present on the wild-type any inEnsembl, oftheotherpopulationsdescribed SMIM1 (c.*58G>A), which is not detected in the 279 Caucasian individuals or individuals of of SMIM1/Vel. Althoughwe cannotformally mutationinthe3’UTRof exclude thattheextra oftheSMIM1protein,dimerization red henceinhibitcorrect cellmembraneincorporation (0.56%, 95%CI0.15%-0.98%)ofthisstudy. Furthermore, aswasalready proposed (0.09%,95%CI0.02%-0.17%)compared Americans African to blackpopulation theAfrican population. The American [NHLBI] Exome Project, Sequencing thattheSMIM1*64_80delalleleispresent intheAfrican et al. 12 12 already Lung, demonstrated, usingdataoftheNationalHeart, andBloodInstitute Moreover, undernon-reducing circumstances the Vel by a antigeniscarried 13 The missensemutationscausing thep.Met51 alteration mightinhibit SMIM1*64_80del allelehad, however, alower inthe frequency SMIM1 allele and responsible for additional observed inbothindividualspositive for observed SMIM64_80del allele). This suggests Molecular basisofweak Molecular Vel expression 6,11 within The frequency in the Caucasian 12,13 , the17 137

Chapter 6 Chapter 6

In conclusion, weakened expression of the Vel antigen is most often caused by the heterozygous SMIM1*64_80del allele in combination with the minor A allele of rs117555 on the non-deletion haplotype. This weakened Vel expression makes serological screening for the Vel- phenotype difficult. An easy to perform high throughput genotyping assay can correctly identify Vel- donors and using this assay the availability of Vel- donors and therefore Vel- red blood cells can be increased, making safe transfusion possible for individuals who have anti-Vel.

Acknowledgements The authors like to thank R. Bijman for her assistance with flow cytometry experiments, J. Hooydonk (The South African Blood Transfusion Service (Johannesburg, South Africa) for collection of samples from Black and Asian donors, G. Tesfaye (The Ethiopian Red Cross Society Transfusion Service Addis Ababa, Ethiopia) and A.J. Duits (The Red Cross Blood bank Curaçao, Willemstad, Curaçao) for the collection of samples from black donors.

Web Resources Population frequency of the rs1175550 and rs191041962 SNPs, respectively. http://www.ensembl.org/Homo_sapiens/Variation/Population?db=core;r=1:3691028- 3692028;v=rs1175550;vdb=variation;vf=928760, http://www.ensembl.org/Homo_sapiens/ Variation/Population?db=core;r=1:3690982-3691982;v=rs191041962;vdb=variation; vf=51904433. Accessed 160813.

138 15. 14. 13. 12. 11. 10. 9. 8. 7. 6. 5. 4. 3. 2. 1. References

Mol.Genet. 2001;10(6):591-7. Mol.Genet. Sunyaev S,Ramensky V, I,Lathe Koch W, P. AS,Bork Kondrashov Prediction humanalleles. of deleterious Hum. intermediate genes, withamajorrole for theHLA.PLoS.Genet. 2011;7(8):e1002197. al. Trans-eQTLs reveal converge associated withacomplexphenotype that independent geneticvariants on Fehrmann RS, Jansen RC, Veldink JH, Westra HJ, Arends D, Bonder MJ, Fu J, P, Deelen Groen A, et HJ, Smolonska Disruption ofSMIM1causesthe Vel- EMBOMol.Med. bloodtype. 2013;5(5):751-61. Ballif BA,Helias V, Peyrard T, C,SaisonLucien Menanteau N,Bourgouin JP, S,Le M,Cartron Gall L. Arnaud for anullalleleofSMIM1definesthe Vel-negative blood group 2013;45(5):537-41. Nat.Genet. phenotype. MK, M,Christophersen JR,Joud B,Thuresson OlssonML.Homozygosity BN,Nilsson B,Storry AkerstromB, Sojka 2013;45(5):542-5. Nat.Genet. E,Farrow S,etal. the SMIM1underlies Bielczyk-Maczynska Vel bloodgroup andinfluences red bloodcelltraits. Cvejic A,Haer-Wigman L,Stephens J, M,SmethurstPA, Kostadima P, Frontini E,Bertone M,vandenAkker of Vel system antibodies. Vox Sang. 1968;15(2):125-32. Adebahr ME,AllenFH,Jr., JK, Issitt PD, R,Reihart Kuhns Oyen WJ. Anti-Vel antibodyshowing 2,anew heterogeneity Cleghorn TE. rareThe bloodgroup occurence 1961. inBritain ofcertain factors the Anaham(Chilcotin) Am.J.Phys.Anthropol. Indians. 1970;32(3):329-37. Alfred BM,Stout TD, J, Lee Petrakis NL.Bloodgroups, M,Birkbeck of phosphoglucomutase, andcerumentypes Chandanayingyong D, Sasaki GreenwaltTT, TJ. Bloodgroups of the Thais. Transfusion 1967;7(4):269-76. 2012. ME,Lomas-FrancisReid C,OlssonML. The BloodGroup AntigenFacts 3ed. SanDiego:Academic Book. Press; 1998;75(1):70-1. ofaninvivo detection haemolyticanti-vel by thegeltest. L,ClasenC.Unsatisfactory J, Bartz VoxNeppert Sang. 1961;1:111-5. Levine P, White Ja,Stroup negative M.Seven membersinthree generationsofafamily.Ve-a (Vel) Transfusion apropos ofacaseanti-VEL immunization].Bibl.Haematol. 1965;23:309-11. [Studyofthe Battaglini Pf, Rm. J, population C,SalmonNicoli intheMarseilles Ranque Bridonneau VEL factor EB. bloodfactor: [New Sussman LN,Miller Vel.]. Rev.Hematol. 1952;7(3):368-71. Daniels G.HumanBloodGroupsoud. 2ed. Oxford: 2002. Science; Blackwell Molecular basisofweak Molecular Vel expression 139

Chapter 6 Chapter 6

140 Chapter 6 information Supporting Molecular basisofweak Molecular Vel expression 141

Chapter 6 Chapter 6

10000 * ***

p = 0.45

1000 Mode Fluoresence Intensity (arbitrary units) (arbitrary ModeIntensity Fluoresence

100 rs1175550 AA AA AA c.64_80del wt/wt wt/ wt/wt SMIM1 genotype Supplementary Figure 1. Vel expression levels in 97 Vel+w donors and ten Vel+ donors. The SMIM1 genotype was determined via sequencing of SMIM1 and Vel expression levels were determined using flow cytometry. Box plot shows 25% and 75% percentile and average fluorescence intensity, error bars indicate the 2.5% and 97.5% percentile. Statistical analysis was performed with an unpaired two-sided T-test. The six donors in which agglutination determined weakened Vel expression, while in these cases the SMIM1*64-80del allele was absent (white boxplot), had significantly lower Vel expression levels compared to control samples with same genotype (dark grey box plot) (p = 0.0114). Moreover, no significant difference was detected between these six donors and 91 donors heterozygous positive for the SMIM1*64_80del allele (light grey boxplot) (p = 0.4503).

142