Du is the weak expression of D antigen. The cells which are not immediately agglutinted by Anti-D sera cannot be easily classified as D negative because some of these agglutinate after addition of antiglobulin sera. This weak reactivity is termed as Du. The genetically transmissible Du is more common in blacks and is transmitted in medelian dominant pattern of inheritance. The gene in this case appears to be Ro(cDe) and this is reffered as low grade Du. The more commonly occuring Du represents CDe gene which is due to the position effect and is commonly seen in whites.  This is reffered as high grade Du.

The Du positive cells are likely to elicit an immune response in D negative individuals and the Du cells could be destroyed if the recepient is already immunised.  Therefore, Du positive donor is treated as D positive and recepient is treated as negative. Hemolytic disease of the new born has also been reported in a D negative mother with D antibodies due to earlier Du positive baby.

* Informative blood typing procedure videos…

1.

2. 

3.

The D antigen, whether from a D positive or Du positive individual, is considered to be mosaic of many parts or many antigenic determinants, called EPITOPES.  Under certain circumstances, perhaps by mutation or unequal crossing over, one or more of these epitopes is missing or changed, or a substitution has occured, so that the shape of total D antigen is altered. Presumably this change involves a very small portion of the D antigen.  Since most of the antigen remains intact and most Rh antisera are heterogenous, the red cells react only in the test for Du. If the person is lacking common D positive cells either by pregnancy or tramsfusion, that epitope of the total antigen may be recognized as foreign by the reception and an antibody may be produced against that epitope cells because they have the total antigen.  It is not known how many epitopes there are to the mosaic.  Based on the work of Tippett in England and work done in the Phillip Levine laboratories, no fewer than eight varients of D and Du have been identified.  Some varients may be the result of more than one missing epitope.

Du Testing

Not all red cells can be classified as Rh positive or negative by direct agglutination tests.  The cells of a few persons react weakly with anti-D or requires a longer reaction time than most Rh positive cells.  An even smaller number of persons have red cells that are not agglutinated by antiglobulin serum.  These cells are called Du.  Cells of the Du phenotype may fall anywhere within this spectrum of reactivity with anti-D.

Because Du is a form of D, red cells of the Du phenotype can stimulate the production of anti-D in Rh negative receipients and, more importantly, react with anti-D in vivo.  It is for these reasons that donor blood must be shown to be negative not only in the test for D but also in the test for Du.  In general,testing the red cells of recipients for Du is considered unnecessary.  The recipient’s welfare is not compromised if he or she is of the Du phenotype but is typed as D negative and recieves Rh negative red cells.  In such circumstances Rh negative donor blood may be used unnecessarily.

It is important that the Du status of the D negative pregnant woman be established early in pregnancy.  If the mother is found to be Rh positive, Du varient, she is not a canditate for Rh immuneglobulin prophylaxis- either antepartum or postpartum- whereas the Rh negative (D and Du negative) mother is a canditate.  The reason for performing the Du test early in pregnancy is to avoid mis-interpreting the cause of a positive fetal cell screening test at the time of delivery.

In addition to prenatal patients, newborn babies are also tested for Du if they type as D negative.  Again, this relates to the need for Rh immune globulin: the D negative, Du negative baby cannot immunize its mother; for this reason she does not need Rhimmune globulin protection.   However, the mother should receive Rh immune globulin if the baby is of the Du phenotype.

Du red cells fall into a wide spectrum of reactivity when tested with anti-D reagents.  How each cell is detected depends on the type of anti-D that is used and the kind of test that is performed. To test for Du, red cells are incubated at 37C with an IgG anti-D and an antiglobulin test is performed.  If serum suspended cells are used, some blood samples at the upper end of the Du spectrum will be agglutinated weakly by most anti-D reagents prior to the antiglobulin test, either at room temperature or at 37C.  When the same red cells are suspended in saline, direct agglutination may not be observed, or it may be seen with one reagent and not another.  Regardless of whether the red cells are agglutinated directly by anti-D or they absorb abti-D and it is detected in the antiglobulin phase of the test, they are Rh positive, provided both controls for D typing and the Du test are negative

Unusually Weak Agglutination

 Each of the common antigens of the Rh system has been found in a varient form.  The varient form of D (Du) has already been described.  There are also varients of C, c, E and e as well as individual whose red cells show depression of all the Rh antigens.  In some cases varient antigen may react with one antiserum of a given specificity and not with another of the same apparent specificity.  This occurs because the two antisera are not infact identical, although the difference between them is insignificant when testing the vast majority of blood samples.

Cells sensitized with anti-D in vivo, such as cells of babies with hemolytic disease due to anti-D, may not type correctly by any direct agglutination method because all or almost all D antigen sites may be blocked.  When all the D antigen sites on the cells take up anti-D in vivo, no sites are available for the attachment of the anti-D in the reagent and the cells appear to be Rh negative.  false negative tests due to such heavy sensitisation are extreamely rare.  If an eluate prepared from sensitized red cells is shown to contain anti-D, the red cells are then established as Rh positive; were they really Rh negative, they would not have absorbed anti-D in vivo and eluates could not contain anti-D.

(LINK: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC528079/pdf/pnas00720-0031.pdf)

VOL.41,1955 GENETICS: CEPPELLINI ET AL. 283

ance data than was previously possible in those cases where the fraction of ions associatedissmall. Workonthisproblemandonthecaseofunsymmetrical electrolytesisinprogress.

* Contribution No. 1275. 1L.Onsager,Physik.Z.,27,388,1926. 2 L.Onsager,ibid.,28,277,1927. 3L.OnsagerandR.M.Fuoss,J.Phys.Chem.,36,2689,1932. Inordertosavespace,mate-

rialinpp.2689-2704and2735-2744ofthisreferencewillbeusedwithoutrederivation. A re- view of this work is available in H. S. Harned and B. B. Owen, Physical Chemistry of Electrolytic Solutions,(NewYork: ReinholdPublishingCorp.,1950),pp.75-85.

4P.DebyeandE.Hhckel,Physik.Z.,24,185,1923. 5P.DebyeandE.Hfickel,Ibid.,24,305,1923. 6We plantopresentthedetailsofthiscomputationlater.

H. Falkenhagen, M. Leist, and G. Kelbg, Ann. Physik, 6th ser., 11, 51, 1952. 8M.EigenandE.Wicke,Naturwissenschaften,38,453,1951. 9 R. M. Fuoss, J. Phys. Chem.. 58, 682, 1954.

10T.Shedlovsky,J.Am. Chem.Soc.,54,1411,1932.

AN INTERACTION BETWEEN ALLELES AT THE Rh LOCUS IN MAN WHICH WEAKENS THE REACTIVITY OF THE Rho FACTOR (Du)

BY R. CEPPELLINI, L. C. DUNN, AND M. TURRI

INSTITUTE FOR THE STUDY OF HUMAN VARIATION, COLUMBIA UNIVERSITY, NEW YORK; ISTITUTO DI GENETICA AND ISTITUTO SIEROTERAPICO MILANESE S. BELFANTI, UNlVERSITA DI MILANO

Communicated March 25, 1955

Occasionally bloods are found which, when tested with different anti-Rho (anti- D) sera, give some but not althe serological reactions expected from the presence oftheRho(D)factor,ortheintensityofsuchreactionsisconsistentlyweakerthan theaverage. OwingtotheclinicalimportanceoftheRhofactor,onwhichthe condition of Rh-positivity or Rh-negativity depends, such “intermediate vari- ants” (Wiener1), usually symbolized as DU (Stratton2) or Who (Wiener3), represent adificultproblemforblood-groupinglaboratories. Forareviewofthesubject werefertoRaceandSanger.4

From a genetical point of view, a number of family investigations, actually not very large (Stratton and Renton;5 Dunsford;6 7 Race, Sanger, and Lawler8), have led to the conclusion that the DU variants are to be regarded as products of mutationoftheDalleleandarethusinheritedliketheotherRhbloodfactors. Ow- ingtothefactthatoftentheDu variantsinmembersofthesamefamilyshowidenti- calserologicalpeculiarities,whiletheymay differbroadlybetween differentfamilies, aseriesofDualeleshasbeensupposed. Thisviewreceivedadditionalindirect support from the knowledge that generally blood groups behave as a direct, not mediate,productofthedetermininggeneandarelitleinfluencedbyotherenviron- mentalorgeneticalagents.

While the interpretation given by the British authors is certainly true in some andmay betrueinthemajorityofDu cases,othergeneticalinterpretationscannot bedisregarded.284 GENETICS: CEPPELLINI ET AL. PROC. N. A. S.

In 1952 one of us,9 discussing the data on some of the individuals now presented infamilyA (Fig.1andTable2),statedthat”thematernalchromosomeCDeseems toproduceintheoffspringanormalD antigenwhenpairedwiththepaternal chromosome cde, but a Du antigen when paired with the paternal chromosome Cde.” ThustheDuphenotypesinthatfamilyappearedtobenotthedirectprod- uctofamutantDu(orST)genebuttheresultofinteractionoftheD chromosome withthepartnerCdechromosome.

This observation was in agreement with the excess of CCDue phenotypes noted in their material by Race et al.,8 who, following their hypothesis of an uncompli- catedinheritanceofDu,wereforcedtoattributetosuchphenotypestherareCDUe! Cde genotype (the other possible alternative CDUe/CDue should be exceedingly rare) because, on a serological level and in the absence of the hypothetical anti- serum,DUmustberegardedtobedominanttodbutrecessivetoD inthesame wayasA2isdominantto0butrecessivetoAl. Theauthorsconcluded”thatthe relatively large number of CD’e/Cde bloods demands an explanation but we have notanyreallysatisfactoryonetoofer.”8

Eldon’0alsoobservedanexcessofCCDUe individualsandsuggestedtheexistence ofamodifyinggeneU. AlthoughEldon’sviewsarenoteasilydiscussedonthe basisofthescantygeneticalandserologicaldatawhichhepresents,thehypothesis ofamodifierlinkedwiththeRh locusorindependentofitcannotbediscardedon thebasisofpresentevidence.

An unusual opportunity for investigating this problem was presented when we found in a Roman Jewish community” an unusually high frequency of the Cde chromosome (r’gene), over 4 per cent as compared with less than 1 per cent in mostItaliansamples. Owingtothefactthatweweredealingmainlywithcom- plete families, the genotypes were often ascertained unambiguously; out of 642 individuals tested, the Cde chromosome was present at least in the 43 following combinations: 24Cde/cde;16Cde/CDe;2Cde/cDE;1Cde/cDe.

Moreover, according to the first Rh typing, where three different anti-D sera wereroutinelyusedinparalel,besides545Rh-positiveand83Rh-negativepersons, unequivocally clasified as such, in fourteen cases a weakened or discordant D- positivity toward the three sera was observed, which was taken as sufficient evi- denceforclassifyingthesebloodsasDU. Inalcasesthesereactions,confirmedby quantitative titrations with several anti-D sera, were found in families in which the Cde chromosome was present, and thus Ceppellini’s suggestion, previously quoted,wasprovedtobeofmoregeneralvalue. Elevensuchfamilies(eightfrom the Roman Jewish community and three from other Italian districts) have now beenstudied. Sixofthesefamilieswerechosenbecauseaproposituswasidentified as Du in the first typing; in the other five cases the families were chosen through a Cedepropositus.

ThepresenceofanyweaklyreactingformoftheD factorwasexcludedforal bloods finally classified as Rh-negative (cde or Ccde) by means of the indirect antiglobulin test after sensitization with at least three (sometimes as many as nine)potentincompleteanti-Dsera.

ThelevelatwhichaDu classificationisjustifiedisratheran arbitraryone;itis thus of fundamental importance in presenting such an investigation to state exactly theserologicalcriteriaandtechniquesused. Forlackofspace,suchdetailswill

VOL.41,1955 GENETICS: CEPPELLINI ET AL. 285

begivenonlyinthefulaccountofourstudy,tobepublishedelsewhere. Inthis preliminary note we shall present only a few examples of some of the families studied.

In Table 1 an example of titration-is reported in detail for one family against oneanti-Dserum: accordingtotheintensityoftheagglutination,differentpoint

TABLE 1 TITRATION AGAINST ONE ANTI-D SERUM (V) OF THE Rh-POSITIVE MEMBERS OF FAMILY A*

DILUTION OF ANTI-D SERUM (V) FAMILYMEMBERS 1 2 4 8 16 32 64 128 SCORE

1-2 + ++v++c++c++v++ 4 W 43

II-6

- – +++- – — 8 II-8 ++++V++C++V++V+ W – 39

I4

-W+W—-5 II-10 ++ ++’ ++1 ++C ++’ + _1- 44

AGGLUTINATION SYMBOLS AND SCORING

++e = Completeagglutinationintoafewsolidclumps ++v = Clumpseasilyrecognizedwiththenakedeye ++ = Largeclumpsseenunderlow-powermag.;rareunagglu-

= 9 points = 7points

= 5points = 3 points = 2points

= 1point

+ I

w

tinatedcels = Clumpsof8-15elements,someunagglutinatedcels = Rareclumpsof4-6elements,manyunagglutinatedcels = Occasionalclumpsof2-4elements,majorityofcelsun-

agglutinated

* See Figure 1 and Table 2. Note the prozone, most marked with the two weaklyreactingbloods (I-6and I-4).

scores are attributed for each dilution (from 10 to 1), and the sum of points makes uptheindividualscoreforagivenserum. Usuallyatleastfiveanti-Dsaline agglutinatingseraforeachsetofexperimentswereused;theserawereselectedonly inasmuch as they did not contain any antibody other than anti-D; strict precau- tionsforavoidingbiasinreadingweretaken.

InTable2thescoresforthemembersoffamilyA (Fig.1)arepresented: the total scores (sum of the scores for the six different antisera) show that the reac-

TABLE 2

FAMILYA(FIG.1): INDIVIDUALSCORESOFTHERh-POSITIVEMEMBERSAGAINST6DIFFERENT ANTI-D SALINE SERA (ALL Rh-NEGATIVE INDIVIDUALS RECHECKED WITH INDIRECT ANTIGLOBULIN TEST)

______________ ANTI-DSERA PHENOTYPE* GENOTYPEt (i) (i) (ii) (iv) (v) (vi) TOTAL

I-1 Ccde Cde/cde … … 0

I-2 CcDe CDe/cde 115 108 75 62 43 37 440 II-5 Ccde Cde/cde … … 0 II-6t CCDue Cde/CDe 64 62 23 44 8 18 219 II-7 cde cde/cde … … 0 11-8 CcDe cde/CDe 108 110 80 74 39 39 450

I-3 cde cde/cde … … 0

I-4 CCDue Cde/CDe 60 71 19 48 5 12 215 11-9 Ccde Cde/cde … 0 II-10 CcDe CDe/cde 121 110 69 59 44 41 444

* Phenotypes symbolized by the letters corresponding to the blood factors actually found by using the 5 Rh anti-sera:anti-C,anti-c,anti-D,anti-E,anti-e;dsymbolizesnegativitytowardanti-D. Du reportedwhenonthe firsttypinganabnormallyweakreactionofD wasobserved.

t Genotypes unambiguously determined from the family segregations; in the offspring the chromosome inherited from the father is printed in italic; the chromosome from the mother, in italic boldface.

Propositus.

tivityoftheD antigensproducedbythesameCDechromosomes(2,6,8and4,10) is markedly changed according to whether they are paired with a cde-or a Cde chromosome.

286 GENETICS: CEPPELLINI ET AL. PROC.N.A.S.

Itisworthwhiletoemphasizethatwe aimedmainlyatinvestigatingwhetherthe presence of a Cde chromosome in the genotype quantitatively depressed the reac- tivityoftheD antigen,afactwhichmaynotberelatedtotheappearanceofatrue DUphenotypeasoriginallydefinedbyRaceetal.8 Actually,whileafewofour DiCdeindividuals(seeTables1and2)showamarkedlydifferentreactivitytoward different anti-D sera (a typical feature of DU phenotypes, which seems to be of a qualitativeorder),themajorityofthemmerelyscorelowerthannormalD bloods withalanti-Dsera. Thissuggeststhatthephenomenonheredescribedisinthe mainaquantitativeone.

Family A I3

cdelcde Cde/CCD de/jjj e/cdc

II 9(SA 5 6 7 8 Cde/cde CDe/cde Cde/cde Cde/CDe cde/cde cde/CDO

Family B Family C

12 12 CWe/cde Cdecde cde/cde Cde/cDE

3; 4\44 5( 3 4 5 6 CDe/cde CNe/Cde CNe/Cde cde/Cde cde/cDE cde/cDE cde/Cde

Family D Family E 1211@

34

4(f

Cde/CDe

3§ CDe/Ce cde/Cde cde/CDe

/CDOCDe

WelcCde/ICde

FIG.1.-* = D-positiveblood,ofnormalstrengthwhentitratedagainst severalanti-Dsera. X = D-positiveblood,significantlyweakerthanaverage ontitration(likeaDuvariant). 0 = D-negativeblood(I.C.T.-negative, withseveralstronganti-D).- = Propositus.ForfamilyAseeTable2. ForfamiliesB,C,D,andEseeTable3. Ingenotypesofoffspringpaternal chromosomeiswrittenfirst(fromCepelinietal.”5).

A proper answer to such a quantitative problem can only be obtained by sub- jectingthedatatoananalysisofvariance,aspresentedinTable3. TheRh- positivemembers offourfamiliesw4 titratedon two differentdays(experiments IandII)againstthesamefiveanti-Dsera. Thetwelveindividualsaredivided intotwo clases,accordingtowhether,intheirgenotype,unambiguouslydefined byfamilysegregations,aCdechromosomeispresentornot. Variabilityofreac- tionsbetweenindividualsofthesame genotypewithrespecttoCde-notCde,tested onthesameday,withthesameserum,wasusedastheerror: atwith8degreesof freedomisthuscalculatedforeachserum (Table3).

The statistical analysis proves beyond any doubt that in this experiment the fiveDiCdegenotypesshowalowerD reactivitythanthesevenD/non-Cdegeno-

VOL.41,1955 GENETICS: CEPPELLINI ET AL.

287

types;thefivedifferentantiserashownosignificantlydifferentdiscriminatingabil- ities(F4.40= 1.93).

Foursimilarsetsofexperimentshavebeenanalyzed,withthesameresults,that is, altogether, a total of 19 certain D/Cde, 20 D/not-Cde, and 4 D/? genotypes (detaileddatatobepublishedelsewhere).

We concludethatthepresenceofCdepairedtoanormalD chromosome (inour investigation 15 Cde/CDe, 3 Cde/cDE, and 1 Cde/cDe genotypes were studied) pro- duces a significant weakening of the D reactivity, which in extreme cases may be ofthesameorderastheserologicalbehaviorshownbytypicalDUbloods. There- fore,D’phenotypesarenotalwaysduetothesamegeneticalmechanism.

Such interpretation applies also to certain of the families reported by Race et al.8 (families Lowe, Green, Curtis; probably also the two individuals Gillham and Pope-the last one a typical example of low-grade DU) and by Eldon.10

ClassD/Cde:

1stdayD-3 62 8 aylE-2 50

B-4 39 2dday B-5 37 (C-2 39

I

II III

43 18 41 30

26 28 34 32 25 29 33.81 27.40

71 56 56 42 64 52 42

IV

43 26

VI

TOTAL SCORE

222 201

189 206 188 201.00

ClawsD/not-Cde: D-2 86

325 284 315 265 273 274 271

1stday E-1 68 E-580

2d

a

55 50 52 51.43 68.00 286.71

m 47.20

56 66 56

TABLE 3 STATISTICALANALYSISOFDATA

INDIVIDUAL SCORES AGAINST 5 DIFFERENT ANTI-D SERA

B-1 51 B-3 66 C-4 57

C-5 54 m 66.00 51.57 49.71

45 42 41

50 45 53 52

71 62 72 72

t(8d-f.) 5.23* 6.54* 7.58* 5.39* 4.16t 11.21*

Statisticalanalysisofthescoresrepresentingthestrengthofagglutinationof12Rh-positiveindividualsbelonging to4familiesagainst5differentanti-Dsalinesera(familiesB,C,D,E [Fig.1]). Thesamepanelofanti-Dsera wasusedinthetwoexperimentscarriedonontwodifferentdays. Theindividualshavebeengroupedintotwo classes according to the presence or absence of Cde in their genotype (ascertained unambiguously through family segregations). D-2comesfromtwoCcDeparentsandhasacdesib. E-3,whoscores52,35,32,34,44(total197), hasnotbeenincludedintheanalysisbecauseoftheuncertaingenotype.

* Significant at .001 P level. fSignificantat.01P level.

We donotdiscussheretheobviouspracticalimplicationsofourfindings,forin- stance,withregardtotheuseofRhbloodgroupsinpaternitycases. Wewish only to point out that, for the blood groups as for other characters, the phenotype cannotberegardedinanycaseasadirectreflectionofthegenotype.

From a formal point of view the phenomenon described could be explained in a varietyofways. Wementiononlytheinterpretationsforwhichatleastabasisfor discussionexists:

a) If the weakening effect of Cde were confined to the CCDUe phenotype, it couldberegardedasduetoadosageeffectbetweencompetingRhaleles: two CalelesversusonlyoneDproduceDu. Thefactthatthesamephenomenonhas been observed in Cde/cDE and Cde/cDe as well as in CDe/Cde genotypes does not justifysuchahypothesis.

40 36 39 36.80 55.8

49 65 51 67 52 .67 51

56 45

288 GENETICS: CEPPELLINI ET AL. PROC.N.A.S.

b) The weakening effect could be produced not by the Cde chromosome as a whole but by a mutant form of d (say, dW) capable of interacting with a partner aleleD. Uptonow,thiswouldhavebeenrecognizedonlyintheCdwecombi- nation. Itshouldbeinvestigatedwhetheracdwechromosomeexists,thatisto say, whether some CcDue or cDue phenotypes are to be regarded as CDe/cdwe or cDe/cdwe. Surprisinglyenough,anumberofpublishedDufamiliesdonotcon- tradict,atleastfromaformalpointofview,thispossibility.

Inconclusion,theweaklyreactingD phenotypewhichoccurswhenaCdechro- mosomeispairedtoanormalD chromosomeis,inouropinion,oneinstanceofthe manyinteractionsbetweenalelestakingplaceattheRhlocus. Similarinterac- tionsbetween CDe and cDE (forexample) have been assumed by Cameron etal.’2 toproduceadepressionoftheanti-Ereactivity. Thiskindofinteractionappears more reasonable ifwe regard Rh as a complex locus, inherited as a unit but built up from more than one site of mutation (Ceppellini;13 Dunn’4).

Summary.-In eleven Italian families, eight from the “ghetto community” of Rome,aweakenedreactionoftheD antigenhasbeenshowntobeassociatedwith thesimultaneouspresenceofthechromosomeCde(r’)inthegenotype. Thisis assumedtobeduetointeractionbetweenD andthecombinedeffectsofCdein juxtapositionortootherpeculiaritiesofparticularCdechromosomes.

1A.S.Wiener,Science,100,595,1944. 2 F.Stratton,Nature,155,542,1946. 3A.S.Wiener,AnRh-HrSyllabus(NewYork: Grune&Stratton,1954). 4R.R.RaceandR.Sanger,BloodGroupsinMan (“BlackwellScientificPublications” [2ded.;

Oxford,1954]). 5 P. H. Renton and F. Stratton, Ann. Eugen. (London), 15, 189, 1950. 6 T.Dunsford,Ann.Eugen.(London),14,142,1948. 7I.Dunsford,ibid.,17,283,1953. 8 R. R. Race, R. Sanger, S. D. Lawler, Ann. Eugen. (London), 14, 171, 1948. 9 R. Ceppellini, “Le Varianti Rh,” in R. Ceppellini, S. Nasso, and F. Tezilacich, La Malattia

emolitica del neonato (Milan: Ed. I.S.M.S.B., 1952), p. 140. 10K.Eldon,Rev.d’hfmatol.,5,294,1950. “L.C.Dunn,Am.J.Phys.Anthropol.1955(inpress). 12 R. R. Race, R. Sanger, P. Levine, R. T. McGee, J. J. van Loghem, M. van der Hart, and

C. Cameron, Nature, 174,460, 1954. 13 R. Ceppellini, “La Natura dei geni Rh,” in R. Ceppellini, S. Nasso, and F. Tezilacich, La

Malattiaemoliticadelneonato(Milan: Ed.I.S.M.S.B.,1952),p.169. 14L.C.Dunn, Caryologia,supi.vol.,p.15,1954. 15R.Ceppellini,L.C.Dunn,M.Turri. Reportto7thInt.Cong.BloodTransfusion,Paris,

Sept. 11, 1954.