0 1991 S. Karger AG. Basel 0042-9007/9V0612-0130 $2.75/0
Vox Sang 1991;61:130-136
Miltenberger Class IX of the MNS Blood Group System Flernrning Skov a, Carole Green ', Geoff Daniels ', Ghizala Khalid b, Patricia Tippett' "Centralsygehuset, Nykobing, Denmark; bMedical Research Council Blood Group Unit, London, UK
Abstract. Mi.IX is a new phenotype in the Miltenberger series of the MNS blood group system with a frequency of 0.43% in Denmark. Mi.IX red cells are Mur+ but do not express any of the other established Miltenberger determinants. They react with a new antibody, anti-DANE, which defines a determinant present on Mi.IX cells but not on cells of other Miltenberger phenotypes. Four Mi.IX propositi have been found. Their families show that Mi'X is inherited with a MS complex (lod score 3.69 at 8 = 0.00) which produces a trypsin-resistant M antigen. DANE has been allotted the ISBT number 002032 (MNS32). Serological and immunochemical studies with human and monoclonal antibodies to various determinants on glycophorin A (GPA) suggest that Mi.IX is associated with an aberrant GPA molecule that lacks the trypsin cleavage site at amino-acid residue 39, retains the chymotrypsin cleavage site at residue 34 and has an apparent M, of about 1,000 less than normal GPA. It is proposed that this Mi.IX molecule has an amino acid and possibly also a glycosylation change in the region of amino-acid residues 35-39.
Introduction MNS is a highly complex blood group system comprising numerous red cell surface antigens [l]. Some of these, including M, N and Ena, are located on the most abundant red cell sialic acid-rich glycoprotein, glycophorin A (GPA); others, including 'N', S, s and U, are on a related glycoprotein, glycophorin B (GPB) [2]. The N-terminal26 aminoacid residues of GPB and of GPA carrying N antigen are identical [2]. The genes producing GPA and GPB have been characterized and are separate [3-51 but closely linked [l]. Some MNS system antigens are associated with hybrid glycophorins consisting of the N-terminus of a GPA molecule and the C-terminus of a GPB molecule, or vice versa [1,2]. At least 22 antigens of low incidence are included in the MNS system [6]. Seven of these antigens define 8 related phenotypes, called the Miltenberger series after the original propositus. Cleghorn [7] proposed a classification for the Miltenberger series after serological investigations of families in which the red cell antigens Vw, Mia, Mur, Hi1 and Hut were
present. Cleghorn's initial description of 4 phenotypes with 5 antigens was expanded to 8 phenotypes following recognition of further specificities associated with the series [8-111. Giles [ll] reappraised and refined the classification of phenotypes in this series. Anti-Hut antibodies described by Cleghorn were considered by Giles to be anti-Hut+Mur. Table 1 shows the expanded Miltenberger series. Anti-Mia probably does not exist as a separate specificity and has been omitted from table 1. The Miltenberger phenotypes are associated with abnormalities of GPA or GPB, or with hybrid molecues comprising part of each of these glycoproteins [2]. Trypsin treatment of intact red cells cleaves GPA between aminoacid residues 39 and 40. Partial trypsin cleavage may also occur between residues 30 and 31 and between residues 31 and 32 [2]. Cells of Mi.1 and Mi.11 phenotypes, associated with the trypsin-sensitive antigens Vw and Hut respectively, have different amino-acid substitutions at residue 28 of GPA resulting in a glycosylation change [12]. Mi.VII1, associated with the trypsin-resistant antigens Hop and Nob, results from a single amino-acid substitution at residue 49;
Miltenberger Class IX
131
Table 1. Serological definition of the Miltenberger series
Class
Vw
Hut'
Hut'
I
I1 I11 IV
+ -
+ + +
+ -
V
-
-
-
-
VI VII VIII IX
+
-
-
Mur
Hi1
Hop
Nob
DANE
-
-
-
-
-
+ + +
-
-
-
-
-
-
-
-
-
-
-
+
+ + +
-
-
+ +
-
+
-
-
-
+
+ -
-
+
' '
Originally called Hut (71. As defined by Giles [ll]. Cells of classes I-IV were positive with sera called anti-Mia which contain a mixture of antibodies.
Mi.VI1, associated with Nob antigen, results from GPA amino-acid substitutions at residues 49 and 52 [13]. Mi.111, Mi.IV and Mi.VI phenotypes are all associated with the trypsin-resistant antigen Mur and with an abnormal GPB molecule of increased apparent molecular weight [14,15]. This abnormal GPB molecule may contain a GPA insert [14,16]. Mi.111 cells are also Hil+, Mi.IV cells Hop+ and Mi.VI cells both Hil+ and Hop+. Mi.V has been fully characterised and is associated with a hybrid glycophorin comprising the first 58 amino-acid residues of GPA at its N-terminal region and residues 27-72 of GPB at its Cterminal domain [15,17]. Mi.V cells are Hil+ but carry no other Miltenberger antigens (table 1). This report adds another class to this already complex and cumbersome series by virtue of the red cells of the propositi reacting with anti-Mur. Cells of the 'new' category, Mi.IX, react only with anti-Mur of the known Miltenberger antibodies, but also react with a previously unidentified antibody, anti-DANE.
Materials and Methods Materials Red cells with rare phenotypes were gifts from colleagues or were obtained via the UK or SCARF exchange schemes. Washed red cells Were prepared from 'fresh' blood or were retrieved from storage in glycerol at -30°C. Antisera were collected over many years, some were gifts, others were obtained from previous investigations involving lowfrequency antigens. Monoclonal antibodies LICR R1.3, LICR RIO, LICR R18, BRIC 116, BRIC 119, BRIC 127, BRIC 163 [18-201 and some anti-M and anti-N were supplied by Dr. D. Anstee, S.W. Regional Transfusion Centre, Bristol, UK; E4 [21] (MAb 150 at second international workshop on monoclonal antibodies against human red
blood cells and related antigens, 1990) from Dr. M. Telen, Duke University Medical Center, Durham, N.C., USA; F84.3E8.E2 (MAb 148 at second monoclonal workshop, 1990) from Dr. D. L. Stone, Immucor Inc., Norcross, GA, USA; anti-M and anti-N from Dr. R. Fraser, Glasgow and West of Scotland Transfusion Centre, Carluke, Lanarkshire, UK. Serological Methods Standard tube techniques with untreated or enzyme-treated red cells were used throughout. Agglutination tests were read microscopically after incubation for 1h or after albumin addition followed by centrifugation at 1,OOO rpm for 1min. Antiglobulin tests were read in tubes after centrifugation: polyspecific anti-human globulin (Blood Products Laboratory (Diagnostics), Elstree, UK) was used for human antibodies and anti-mouse IgG (provided by Dr. R. Knowles) for murine monoclonal antibodies. An enhancement medium (LO-ION, Gamma Biologicals, Houston, Tex., USA) was used with some antisera. Eluates for serological techniques were prepared by an acid elution method (Elu-Kit 11, Gamma Biologicals). Enzyme Treatments Red cells were pretreated with papain, trypsin, chymotrypsin, pronase or sialidase as described previously [22]. Sequential sialidasel chymotrypsin modifications were carried out by sialidase treating the red cells first, washing the cells, and then treating with chymotrypsin. Immunochemical Methods Immunoblotting was performed as described previously [22]. Briefly, red cell ghosts were solubilised in the presence of sodium dodecyl sulphate (SDS) and 2-mercaptoethanol, and the components separated on 10% SDS polyacrylamide gels. The proteins were then electroblotted onto polyvinylidene difluoride filters and unbound sites blocked with bovine milk. The filter was then incubated in monoclonal-antibody supernatant or human antibody (prepared by adsorption onto and elution from antigen-positive red cells) and, after washing, any bound antibody was detected by using horseradish peroxidase-conjugated anti-human globulin or anti-mouse IgG, with 4-chloro-1-naphthol as substrate. Linkage Studies The relationship of the DANE antigen to MNS was investigated by analysing the family data by LIPED [23] using the LYNSKYS v 4.11 program of Attwood and Bryant [24]. MNS frequencies for a Danish population were used in the calculation.
Results and Discussion Anti-DANE and the Mi. IX Phenotype A patient, K.G.R., who had never been transfused, was admitted for cirrhosis Of the liver and acute haematemesis. During 3 months of therapy prior to his death, 40 units of blood were cross-matched with his serum: red cells of onlv (Dane.) reacted when they were tested 6 after his first transfusion. The antibody in the serum of K*G*R.agglutinated suspensions of Dane.% red cells at room temperature, but had a titre of only 2,
132
Investigation of Dane.% red cells (A CDe/CDe) with antibodies to low-frequency antigens led to the finding that his cells express Mur, an antigen of the Miltenberger series of the MNS blood group system. Antisera containing antibodies to the following MNS-associated low-frequency antigens were negative with Dane.% cells: He, Vw, Mg, V‘, Mta, Sta, Ria, Cla, Nya, Hut, Hil, M , Far, sD, Mit, Dantu, Hop, Nob, Or. Dane.’s cells reacted with the original antiMur and with two other examples of this specificity. They also reacted with several sera containing mixtures of Miltenberger antibodies: in each instance this reaction could be attributed to anti-Mur in the serum. Dane.% cells were confirmed as Mur+ by adsorption and elution of anti-Mur from Dane.% cells and from Mi.111 cells. In each case the eluate reacted with Dane.’s cells and with Mi.111, Mi.IV and Mi.VI cells, but not with Mi.11 cells. The pattern of reactions with Miltenberger antisera had not been seen previously and Dane.’s cells, therefore, have a new Miltenberger phenotype, Mi.IX (table 1). Dane.’s cells are M+N+S+s-U+. The M antigen is very unusual since it can still be detected on trypsin-treated cells. The M antigen of normal M+N+ cells and of Mi.111, Mi.IV and Mi.VI cells, which also carry the Mur antigen, is readily destroyed by trypsin. The Mur antigen on Dane.’s cells, like that on other Mur+ cells, was resistant to trypsin treatment. Surprisingly, the new antigen on Dane.’s cells, to be called DANE, was destroyed by trypsin treatment of the cells. Like normal M+N+ cells, the M antigen of M+N+ Mi.IX cells is not destroyed by chymotrypsin treatment. Following the finding that Dane.% cells were Mur+ , serum of K.G.R. was tested with cells of other Miltenberger phenotypes: none reacted. K.G.R. serum, therefore, contains a new Miltenberger antibody, anti-DANE, which detects the Mi.IX phenotype (table 1). Other Mi. IX Propositi A second Mi.IX propositus was disclosed because his SC:1,2 cells were used during the investigation of the K.G.R. antibody (two further examples of SC:1,2 cells were negative). Adsorption and elution tests showed that the reactivity with K.G.R. serum was due to the same antigen as that present on the red cells of Dane. Like those of Dane., red cells of this second propositus were Mur+ and carried a trypsin-resistant M antigen. Red cells from 467 random blood donors from the Nykabing Falstar region in Denmark were treated with trypsin and tested with a monoclonal anti-M (clone BS57, Biotest, Frankfurt, FRG). Red cells from two donors were agglutinated, showing that their red cells possessed a trypsin-
Skov/Green/Daniels/KhalidRippett
resistant M antigen. Both were DANE+ Mur+ and have the Mi.1X phenotype. Mi.IX, therefore, is not uncommon in Denmark, with a frequency of 0.43%. Family Studies Families from all 4 propositi were investigated; figure 1 shows the results of testing with anti-M, -N, -S, -s, -DANE and anti-Mur. In each family, red cells of all DANE+ individuals reacted with anti-Mur and had a trypsin-resistant M antigen. The families showed that DANE was an autosomal dominant character. Inspection of the pedigrees showed clearly that D A N E was inherited with MS.No recombination between D A N E and MNS was observed. The sum of lod scores for D A N E and MNS at a recombination fraction of zero is 3.69: therefore, DANE is produced by a gene at the MNS locus or one that is very closely linked to it. DANE is, therefore, a new low-frequency MNS antigen and has been allotted the ISBT number 002032 (MNS32). Further Serological Characteristics of Mi. IX Red Cells Dane.%cells did not appear to have any obvious reduction in sialic acid levels. They were not agglutinated by Glycine soja lectin or by incomplete (IgG) anti-D. Trypsin treatment of intact red cells cleaves GPA at amino-acid residue 39 and partially at residue 31 [2] (table 2). Dahr et al. [25] showed that GPA has a chymotrypsin cleavage site at position 34, but that this site is blocked on some GPA molecules by the sialic acid of an 0-linked oligosaccharide on threonine at position 33 (table 2). Complete cleavage with chymotrypsin can be obtained after desialylation [25]. Mi.IX cells were tested with a collection of monoclonal antibodies to epitopes on the extracellular domain of GPA (table 3). R10, BRIC 116, BRIC 119, BRIC 127 and E4 all detect trypsin-sensitive determinants on control cells. None of these antibodies reacted with red cells treated with sialidase followed by chymotrypsin; BRIC 119 is sialic acid dependent, the others are not. R.L. is a human alloantibody (anti-EnaTS) made by an individual deficient in GPA [26]. R.L. serum and BRIC 127 did not react with either trypsin-treated or chymotrypsin-treated control cells and presumably only react with those GPA molecules which are not glycosylated at amino-acid residue 33 [25,19]. All these antibodies to trypsin-sensitive epitopes reacted with Mi.IX cells; with the exception of E4, none reacted with trypsin-treated Mi.IX cells. Neither BRIC 127 nor R.L. antibody reacted with chymotrypsin-treated Mi.IX cells (table 3). R18 detects a trypsin-resistant epitope on glycophorin A and R1.3 detects a sialic acid-dependent epitope common
Miltenberger Class IX
133
to the N-terminal26 amino-acid residues of both GPA and GPB. Both antibodies reacted with Mi.IX cells. The association Of a tqpsin-resistant antigen with the Mi.IX phenotype suggests that the trypsin cleavage site at amino-acid residue 39 is altered in some way. Yet despite the apparent loss of the trypsin cleavage site in Mi.1X GPA, most anti-En'TS antibodies failed to react with trypsintreated ~ i . cells, 1 ~suggesting that ~ i . GPA 1 ~lacks the epitopes detected by these antibodies. however, does react with trypsin-treated Mi.IX cells and therefore the E4 E49
Table 2. Segment of GPA molecule from amino-acid residues 30-40
*
*
Lys-Arg~Asp~Th~~Tyr~A~a~Ala-Thr-Pro-Arg~~~~~~~s~ 30 I I 33 I 37 I 41
I
1
I
I
T
dC
(T) (TI
-
* = 0-glycosylated in the majority of molecules. dC = Partial chymotrypsin cleavage site of native GPA, complete cleavage after desialylation. T = trypsi,, cleavage site. (T) = partial trypsin &avage site,
Table 3. Reactions of monoclonal antibodies and a human alloantibody to determinants on GPA with M+N+ Mi.IX and M+N+ control cells Enzyme treatment
Cells
None
Mi.IX M+N+
Trypsin
Mi.IX M+N+ Mi.IX M+N+
Chymotrypsin Sialidase
Mi.IX M+N+
Sialidasel chymotrypsin
Mi.IX M+N+
anti-M
R.L.
R10
BRIC116
BRIC119
BFUC127
E4
R18
R1.3
+ + +
+ +
+ +
+ +
+ +
+ +
+
0
0 0
0 0
0 0
0 0
0 0
+ + +
+ +
0 0
+ + +
+ +
+ +
0 0
+ + + + + +
+
0 0
+ +
0 0
0 0
Fig 1. Pedigrees of four families with Mi.IX members. NT = not tested; + = dead; M* = trypsin-resistant M antigen. 4, *=DANE and Mur positive. 0,0 =DANE and Mur negative or NT. Arrow indicates propositus.
+
+
0
+ + + +
0 0
+ + +
+
+ + + + +
0 0
134
Fig. 2. Immunoblots stained with monoclonal antibodies detecting different epitopes on GPA. a M + N + control. b trypsin-treated Mi.IX cells. c Mi.IX cells. T1 and T2 = additional bands seen with trypsintreated Mi.IX cells. X =additional band seen with untreated Mi.IX cells.
epitope does appear to be present on the Mi.IX GPA molecule. E4 does not react with chymotrypsin-treated desialylated cells (table 3) suggesting that the chymotrypsin cleavage site at amino-acid residue 34 is intact.
Immunochemistry Immunoblotting of M+N+ MI.IX cells with monoclonal anti-M (fig. 2) revealed bands representing GPA, its dimer (GPA2) and heterodimer with GPB (GPAB). However, the GPA band demonstrated a slight increase in mobility, compared with GPA from M+N+ control cells, equivalent to a decrease in apparent MI of about 1,260 (from 8 determinations with 8 monoclonal anti-M). Family studies showed that Mi" is associated with MS and so the GPA molecule detected by anti-M on immunoblots of M+N+ Mi.IX cells must be the GPA molecule produced by the MSMi.'Xgene.Immunoblots with anti-N did not demonstrate any difference between the apparent MI of GPA of Mi.IX cells and M+N+ control cells [results not shown]. R18, a monoclonal antibody to a trypsin-resistant determinant on the extracellular domain of GPA (En'KT), and BRIC 163, a monoclonal antibody to an epitope on the cytoplasmic C-terminal domain of GPA, gave a similar staining pattern with Mi.IX cells to those seen with anti-M,
Skov/Green/Daniels/KhalidiTippett
demonstrating the reduction in M, of GPA (fig. 2, results for BRIC 163 not shown). R1.3, a monoclonal antibody to an epitope common to the N-terminal portions of GPA and GPB also demonstrated the decreased apparent MI of the GPA of Mi.IX cells, as well as revealing an apparently normal GPB (fig. 2). Mi.IX, therefore, appears to be associated with a GPA molecule of slightly decreased apparent MI. A weakly staining band of apparent MI 63,500 (3 determinations) was seen on some immunoblots of Mi.IX cells with antibodies to GPA (X in Fig. 2). The significance of this band is not clear but it may represent a degradation product. When membranes prepared from trypsin-treated cells were used, anti-M, which stained no bands on an immunoblot of trypsin-treated M+N+ control cells, revealed GPA, GPA2 and GPAB on immunoblots of Mi.IX cells (fig. 2). R1.3, which only binds to GPB and GPBz on immunoblots of trypsin-treated control cells [not shown], also stained GPA, GPA2and GPAB on immunoblots of trypsintreated Mi.IX cells. Anti-M (and R1.3) stained an additional two bands, of apparent M, 61,000 and 51,200 (from 7 determinations), with membranes from trypsin-treated Mi.IX cells (fig. 2). These two extra bands were seen on immunoblots of red cells of Mi.IX members of all four families and were absent from those of non-Mi.IX members. The results with trypsin-treated cells confirm that Mi.IX is associated with a GPA molecule in which the N-terminal portion is not cleaved by trypsin. The significance of the extra two bands (TIand T2 in fig. 2) is not clear. One of them could represent a dimer of Mi.IX GPA with trypsin-cleaved normal GPA, or with fragments of some Mi.IX GPA molecules cleaved by trypsin at an alternative site to the main trypsin cleavage site at amino-acid residue 39. R10, a monoclonal antibody to a trypsin-sensitive epitope on GPA between amino-acid residues 26 and 39, also revealed GPA, GPAz and GPAB on Mi.IX cells. However, with R10 the decrease in apparent MI of GPA was not apparent and the GPA of Mi.IX cells showed reduced staining intensity compared with that of control M+N+ cells (fig. 2). R10 revealed no bands on immunoblots of trypsintreated Mi.IX or control cells. These results provide further evidence that R10 does not react with Mi.IX GPA. GPB of Mi.IX cells appeared normal as determined by immunoblots with anti-N, R1.3, human anti-S or monoclonal anti-GPB (F84.3E8.E2). The serological and immunochemical results suggest that Mi.IX is associated with an aberrant GPA molecule which has apparently normal N - and C-terminal regions and normal En'KT, a determinant situated around amino-
Miltenberger Class IX
acid residue 49 [25]. It lacks the trypsin cleavage site at residue 39, probably lacks En'TS (R.L., R10, BRIC 119, BRIC 127, but not for E4) which is situated between residues 26 and 39, and has an apparent M, of about 1,000 less than normal GPA. Based on data obtained with the monoclonal antibody E4, the chymotrypsin cleavage site at residue 34 appears to be intact. A glycosylated threonine at position 33 appears to be present as it is this glycosylation which prevents chymotrypsin cleavage of some GPA molecules [25] (table 2). The partial trypsin cleavage site between residues 30 and 31 of GPA may remain intact in Mi.IX cells as only some of the GPA molecules would be expected to be cleaved at this site and therefore trypsin treatment would not destroy M antigen expression in these cells. The biochemical and serological data are consistant with an alteration to GPA associated with Mi.IX located around the region of amino-acid residues 35-40. If the 1,000 molecular-weight reduction of Mi.IX GPA were due to the loss of one 0-glycan, then this could be from threonine at position 37. Alternatively, the reduction in molecular weight could result from a deletion of a few amino-acid residues from the region of the GPA molecule around the trypsin cleavage site at residue 39. Mur antigen has previously only been found in the phenotypes Mi.111, Mi.IV and Mi.VI. In these phenotypes Mur is associated with the presence of an abnormal GPB molecule, which may contain a GPA insert [14,16]. It is probable that anti-Mur is detecting a determinant on the aberrant GPA molecule in Mi.IX and possible that the same determinant results from the presence of an insert in the unusual GPB molecule of Mi.111, Mi.IV and Mi.VI. The significance of DANE antigen in Mi.IX is more difficult to interpret. DANE is trypsin sensitive, suggesting that it is not carried on the Mi.IX GPA molecule. However, since anti-DANE has a very low titre against Mi.IX cells, it is possible that the cleavage of a small number of DANEactive GPA molecules with trypsin, possibly at less accessible sites than the main trypsin cleavage site at position 39, would reduce the antigen site density sufficiently to prevent agglutination of the cells by anti-DANE. Unfortunately the only anti-DANE serum is in extremely short supply and the producer of the antibody is dead, so until another example is found it is unlikely that the location of the DANE antigen will be clarified. During the course of this investigation it was found that several Mg+ cells were also DANE+. This will be the subject of a future report. The serological and immunochemical findings on cells with the Mi.IX phenotype are in good agreement. The biochemical nature of the Miltenberger phenotypes Mi.1,
135
Mi.11, Mi.V, Mi.VII and Mi.VII1 is now known, whereas the biochemistry of the Mur+ phenotypes Mi.111, Mi.IV, Mi.VI and now Mi.IX remains speculative. It is probable that these will be determined by molecular analysis of the genes producing the aberrant GPA or GPB molecules.
Acknowledgements We are very grateful to Dr. D. Anstee, Dr. M. Telen, Dr. D.L. Stone and Dr. R. Fraser for providing monoclonal anibodies for use in this study.
References 1 Issitt PD: Applied Blood Group Serology, ed 3. Miami, Montgomery Scientific Publications, 1985. 2 Dahr W Immunochemistry of sialoglycoproteins in human red blood cell membranes; in Vengelen-Tyler V, Judd WJ (eds): Recent Advances in Blood Group Biochemistry. Arlington, American Association of Blood Banks, 1986, pp 23-65. 3 Siebert PD, Fukuda M: Isolation and characterization of human glycophorin A cDNA clones by a synthetic oligonucleotide approach: Nucleotide sequence and mRNA structure. Proc Natl Acad Sci USA 1986;83:1665-1669. 4 Siebert PD, Fukuda M: Human glycophorin A and B are encoded by separate, single copy genes coordinately regulated by a tumorpromoting phorbol ester. J Biol Chem 1986;261:12433-12436. 5 Siebert PD, Fukuda M: Molecular cloning of a human glycophorin B cDNA: Nucleotide sequence and genomic relationship to glycophorin A. Proc Natl Acad Sci USA 1987;84:67354739. 6 Lewis M, Anstee DJ, Bird GWG, Brodheim E , Cartron JP, Contreras M, Crookston MC, Dahr W, Daniels GL, Engelfriet CP, Giles CM, Issitt PD, Jorgensen J, Kornstad L, Lubenko A. Marsh WL, McCreary J, Moore BPL, Morel P, Moulds JJ, Nevanlinna H, Nordhagen R, Okubo Y,Rosenfield RE, Rouger P, Rubinstein P, Salmon C, Seidl S, Sistonen P, Tippett P, Walker RH, Woodfield G, Young S: Blood group terminology 1990. From the ISBT Working Party on Terminology for Red Cell Surface Antigens. Vox Sang 1990;58:152-169. 7 Cleghorn TE: A memorandum on the Miltenberger blood groups. Vox Sang 1966;11:219-222. 8 Crossland JD, Pepper MD, Giles CM, Ikin EW A British family possessing two variants of the MNSs blood group system, M' and a new class within the Miltenberger complex. Vox Sang 1970;18:407413. 9 Giles CM, Chandanayingyong D, Webb AJ: Three antibodies of the MNSs system and their association with the Miltenberger complex of antigens. 111. Anek, Raddon and Lane antisera in relation to each other and the Miltenberger complex. Vox Sang 1977;32:277279. 10 Dybkjaer E, Poole J, Giles CM: A new Miltenberger class detected by a second example of Anek type serum. Vox Sang 1981;41:302305. 11 Giles CM: Serological activity of low frequency antigens of the MNSs system and reappraisal of the Miltenberger complex. Vox Sang 1982;42:256-261.
Skov/GreenlDaniels/Khalid/Tippett
136
12 Dahr W, Newman RA, Contreras M, Kordowicz M, Teesdale P, Beyreuther K, Kruger J: Structures of Miltenberger class I and I1 specific major human erythrocyte membrane sialoglycoproteins. Eur J Biochem 1984;138:259-265. 13 Dahr W, Beyreuther K, Dybkjaer E, Moulds J, Vengelen-Tyler V Biochemical characterization of class VII and VIII cells within the Miltenberger system; in Mayr W (ed): Advances in Forensic Haemogenetics 2.Berlin, Springer, 1988,pp 22-25. 14 Dahr W, Longster G, Uhlenbruck G , Schumacher K: Studies on Miltenberger class 111, V, M and M’ red cells. I. Sodium dodecylsulfate polyacrylamide gel electrophoretic investigations. Blut 1978;37:129-138. 15 Anstee DJ, Mawby WJ, Tanner MJA: Abnormal blood-group-Ssactive sialoglycoproteins in the membrane of Miltenberger class 111, IV and V human erythrocytes. Biochem J 1979;183:193-203. 16 King MJ, Poole J, Anstee DJ: An application of immunoblotting in the classification of the Miltenberger series of blood group antigens. Transfusion 1989;29:106-112. 17 Vignal A, Rahuel C, El Maliki B, London J, Le Van Kim C, Blanchard D, Andre C, d’Auriol L, Galibert F, Blajchman MA, Cartron J-P: Molecular analysis of glycophorin A and B gene structure and expression in homozygous Miltenberger class V (Mi.V) human erythrocyes. Eur J Biochem 1989;184:337-344. 18 Anstee DJ, Edwards PAW Monoclonal antibodies to human erythrocytes. Eur J Immunol 1982;12:228-232. 19 Gardner B, Parsons SF, Merry AH, Anstee DJ: Epitopes on sialoglycoprotein a: Evidence for heterogeneity in the molecule. Immunology 1989;68:283-289. 20 Okubo Y, Daniels GL, Parsons SF, Anstee DJ, Yamaguchi H. Tomita T, Seno T A Japanese family with two sisters apparently homozygous for M”.Vox Sang 1988;54:107-111. 21 Telen MJ, Scearce RM, Haynes BF: Human erythrocyte antigens. 111. Characterization of a panel of murine monoclonal antibodies that react with human erythrocyte and erythroid precursor membranes. Vox Sang 1987;52:23&243.
22 Khalid G , Green CA: Immunoblotting of human red cell membranes: Detection of glycophorin B with anti-S and anti-s antibodies. Vox Sang 1990;59:48-54. 23 Ott J: Estimation of the recombination fraction in human pedigrees: Efficient computation of the likelihood for human linkage studies. Ann Hum Genet 1974;26:588-597. 24 Attwood J, Bryant S: A computer program to make linkage analysis with LIPED and LINKAGE easier to perform and less prone to input errors. Ann Hum Genet 1988;52:529. 25 Dahr W, Muller T, Moulds J, Baumeister G , Issitt PD, Wilkinson S, Garratty G: High frequency antigens of human erythrocyte membrane sialoglycoproteins. I. Enareceptors in the glycosylated domain of the MN sialoglycoprotein. Biol Chem Hoppe Seyler 1985;366:41-51. 26 Taliano V, Guevin RM, Hebert D , Daniels GL, Tippett P, Anstee DJ, Mawby WJ, Tanner MJA: The rare phenotype En(a-) in a French-Canadian family. Vox Sang 1980;38:87-93.
Received: December 5,1990 Revised manuscript received: February 15, 1991 Accepted: February 15, 1991 Ms C.A. Green MRC Blood Group Unit Wolfson House 4 Stephenson Way London, NW12HE (UK)
Vox Sang 1991;61:130-136
Miltenberger Class IX of the MNS Blood Group System Flernrning Skov a, Carole Green ', Geoff Daniels ', Ghizala Khalid b, Patricia Tippett' "Centralsygehuset, Nykobing, Denmark; bMedical Research Council Blood Group Unit, London, UK
Abstract. Mi.IX is a new phenotype in the Miltenberger series of the MNS blood group system with a frequency of 0.43% in Denmark. Mi.IX red cells are Mur+ but do not express any of the other established Miltenberger determinants. They react with a new antibody, anti-DANE, which defines a determinant present on Mi.IX cells but not on cells of other Miltenberger phenotypes. Four Mi.IX propositi have been found. Their families show that Mi'X is inherited with a MS complex (lod score 3.69 at 8 = 0.00) which produces a trypsin-resistant M antigen. DANE has been allotted the ISBT number 002032 (MNS32). Serological and immunochemical studies with human and monoclonal antibodies to various determinants on glycophorin A (GPA) suggest that Mi.IX is associated with an aberrant GPA molecule that lacks the trypsin cleavage site at amino-acid residue 39, retains the chymotrypsin cleavage site at residue 34 and has an apparent M, of about 1,000 less than normal GPA. It is proposed that this Mi.IX molecule has an amino acid and possibly also a glycosylation change in the region of amino-acid residues 35-39.
Introduction MNS is a highly complex blood group system comprising numerous red cell surface antigens [l]. Some of these, including M, N and Ena, are located on the most abundant red cell sialic acid-rich glycoprotein, glycophorin A (GPA); others, including 'N', S, s and U, are on a related glycoprotein, glycophorin B (GPB) [2]. The N-terminal26 aminoacid residues of GPB and of GPA carrying N antigen are identical [2]. The genes producing GPA and GPB have been characterized and are separate [3-51 but closely linked [l]. Some MNS system antigens are associated with hybrid glycophorins consisting of the N-terminus of a GPA molecule and the C-terminus of a GPB molecule, or vice versa [1,2]. At least 22 antigens of low incidence are included in the MNS system [6]. Seven of these antigens define 8 related phenotypes, called the Miltenberger series after the original propositus. Cleghorn [7] proposed a classification for the Miltenberger series after serological investigations of families in which the red cell antigens Vw, Mia, Mur, Hi1 and Hut were
present. Cleghorn's initial description of 4 phenotypes with 5 antigens was expanded to 8 phenotypes following recognition of further specificities associated with the series [8-111. Giles [ll] reappraised and refined the classification of phenotypes in this series. Anti-Hut antibodies described by Cleghorn were considered by Giles to be anti-Hut+Mur. Table 1 shows the expanded Miltenberger series. Anti-Mia probably does not exist as a separate specificity and has been omitted from table 1. The Miltenberger phenotypes are associated with abnormalities of GPA or GPB, or with hybrid molecues comprising part of each of these glycoproteins [2]. Trypsin treatment of intact red cells cleaves GPA between aminoacid residues 39 and 40. Partial trypsin cleavage may also occur between residues 30 and 31 and between residues 31 and 32 [2]. Cells of Mi.1 and Mi.11 phenotypes, associated with the trypsin-sensitive antigens Vw and Hut respectively, have different amino-acid substitutions at residue 28 of GPA resulting in a glycosylation change [12]. Mi.VII1, associated with the trypsin-resistant antigens Hop and Nob, results from a single amino-acid substitution at residue 49;
Miltenberger Class IX
131
Table 1. Serological definition of the Miltenberger series
Class
Vw
Hut'
Hut'
I
I1 I11 IV
+ -
+ + +
+ -
V
-
-
-
-
VI VII VIII IX
+
-
-
Mur
Hi1
Hop
Nob
DANE
-
-
-
-
-
+ + +
-
-
-
-
-
-
-
-
-
-
-
+
+ + +
-
-
+ +
-
+
-
-
-
+
+ -
-
+
' '
Originally called Hut (71. As defined by Giles [ll]. Cells of classes I-IV were positive with sera called anti-Mia which contain a mixture of antibodies.
Mi.VI1, associated with Nob antigen, results from GPA amino-acid substitutions at residues 49 and 52 [13]. Mi.111, Mi.IV and Mi.VI phenotypes are all associated with the trypsin-resistant antigen Mur and with an abnormal GPB molecule of increased apparent molecular weight [14,15]. This abnormal GPB molecule may contain a GPA insert [14,16]. Mi.111 cells are also Hil+, Mi.IV cells Hop+ and Mi.VI cells both Hil+ and Hop+. Mi.V has been fully characterised and is associated with a hybrid glycophorin comprising the first 58 amino-acid residues of GPA at its N-terminal region and residues 27-72 of GPB at its Cterminal domain [15,17]. Mi.V cells are Hil+ but carry no other Miltenberger antigens (table 1). This report adds another class to this already complex and cumbersome series by virtue of the red cells of the propositi reacting with anti-Mur. Cells of the 'new' category, Mi.IX, react only with anti-Mur of the known Miltenberger antibodies, but also react with a previously unidentified antibody, anti-DANE.
Materials and Methods Materials Red cells with rare phenotypes were gifts from colleagues or were obtained via the UK or SCARF exchange schemes. Washed red cells Were prepared from 'fresh' blood or were retrieved from storage in glycerol at -30°C. Antisera were collected over many years, some were gifts, others were obtained from previous investigations involving lowfrequency antigens. Monoclonal antibodies LICR R1.3, LICR RIO, LICR R18, BRIC 116, BRIC 119, BRIC 127, BRIC 163 [18-201 and some anti-M and anti-N were supplied by Dr. D. Anstee, S.W. Regional Transfusion Centre, Bristol, UK; E4 [21] (MAb 150 at second international workshop on monoclonal antibodies against human red
blood cells and related antigens, 1990) from Dr. M. Telen, Duke University Medical Center, Durham, N.C., USA; F84.3E8.E2 (MAb 148 at second monoclonal workshop, 1990) from Dr. D. L. Stone, Immucor Inc., Norcross, GA, USA; anti-M and anti-N from Dr. R. Fraser, Glasgow and West of Scotland Transfusion Centre, Carluke, Lanarkshire, UK. Serological Methods Standard tube techniques with untreated or enzyme-treated red cells were used throughout. Agglutination tests were read microscopically after incubation for 1h or after albumin addition followed by centrifugation at 1,OOO rpm for 1min. Antiglobulin tests were read in tubes after centrifugation: polyspecific anti-human globulin (Blood Products Laboratory (Diagnostics), Elstree, UK) was used for human antibodies and anti-mouse IgG (provided by Dr. R. Knowles) for murine monoclonal antibodies. An enhancement medium (LO-ION, Gamma Biologicals, Houston, Tex., USA) was used with some antisera. Eluates for serological techniques were prepared by an acid elution method (Elu-Kit 11, Gamma Biologicals). Enzyme Treatments Red cells were pretreated with papain, trypsin, chymotrypsin, pronase or sialidase as described previously [22]. Sequential sialidasel chymotrypsin modifications were carried out by sialidase treating the red cells first, washing the cells, and then treating with chymotrypsin. Immunochemical Methods Immunoblotting was performed as described previously [22]. Briefly, red cell ghosts were solubilised in the presence of sodium dodecyl sulphate (SDS) and 2-mercaptoethanol, and the components separated on 10% SDS polyacrylamide gels. The proteins were then electroblotted onto polyvinylidene difluoride filters and unbound sites blocked with bovine milk. The filter was then incubated in monoclonal-antibody supernatant or human antibody (prepared by adsorption onto and elution from antigen-positive red cells) and, after washing, any bound antibody was detected by using horseradish peroxidase-conjugated anti-human globulin or anti-mouse IgG, with 4-chloro-1-naphthol as substrate. Linkage Studies The relationship of the DANE antigen to MNS was investigated by analysing the family data by LIPED [23] using the LYNSKYS v 4.11 program of Attwood and Bryant [24]. MNS frequencies for a Danish population were used in the calculation.
Results and Discussion Anti-DANE and the Mi. IX Phenotype A patient, K.G.R., who had never been transfused, was admitted for cirrhosis Of the liver and acute haematemesis. During 3 months of therapy prior to his death, 40 units of blood were cross-matched with his serum: red cells of onlv (Dane.) reacted when they were tested 6 after his first transfusion. The antibody in the serum of K*G*R.agglutinated suspensions of Dane.% red cells at room temperature, but had a titre of only 2,
132
Investigation of Dane.% red cells (A CDe/CDe) with antibodies to low-frequency antigens led to the finding that his cells express Mur, an antigen of the Miltenberger series of the MNS blood group system. Antisera containing antibodies to the following MNS-associated low-frequency antigens were negative with Dane.% cells: He, Vw, Mg, V‘, Mta, Sta, Ria, Cla, Nya, Hut, Hil, M , Far, sD, Mit, Dantu, Hop, Nob, Or. Dane.’s cells reacted with the original antiMur and with two other examples of this specificity. They also reacted with several sera containing mixtures of Miltenberger antibodies: in each instance this reaction could be attributed to anti-Mur in the serum. Dane.% cells were confirmed as Mur+ by adsorption and elution of anti-Mur from Dane.% cells and from Mi.111 cells. In each case the eluate reacted with Dane.’s cells and with Mi.111, Mi.IV and Mi.VI cells, but not with Mi.11 cells. The pattern of reactions with Miltenberger antisera had not been seen previously and Dane.’s cells, therefore, have a new Miltenberger phenotype, Mi.IX (table 1). Dane.’s cells are M+N+S+s-U+. The M antigen is very unusual since it can still be detected on trypsin-treated cells. The M antigen of normal M+N+ cells and of Mi.111, Mi.IV and Mi.VI cells, which also carry the Mur antigen, is readily destroyed by trypsin. The Mur antigen on Dane.’s cells, like that on other Mur+ cells, was resistant to trypsin treatment. Surprisingly, the new antigen on Dane.’s cells, to be called DANE, was destroyed by trypsin treatment of the cells. Like normal M+N+ cells, the M antigen of M+N+ Mi.IX cells is not destroyed by chymotrypsin treatment. Following the finding that Dane.% cells were Mur+ , serum of K.G.R. was tested with cells of other Miltenberger phenotypes: none reacted. K.G.R. serum, therefore, contains a new Miltenberger antibody, anti-DANE, which detects the Mi.IX phenotype (table 1). Other Mi. IX Propositi A second Mi.IX propositus was disclosed because his SC:1,2 cells were used during the investigation of the K.G.R. antibody (two further examples of SC:1,2 cells were negative). Adsorption and elution tests showed that the reactivity with K.G.R. serum was due to the same antigen as that present on the red cells of Dane. Like those of Dane., red cells of this second propositus were Mur+ and carried a trypsin-resistant M antigen. Red cells from 467 random blood donors from the Nykabing Falstar region in Denmark were treated with trypsin and tested with a monoclonal anti-M (clone BS57, Biotest, Frankfurt, FRG). Red cells from two donors were agglutinated, showing that their red cells possessed a trypsin-
Skov/Green/Daniels/KhalidRippett
resistant M antigen. Both were DANE+ Mur+ and have the Mi.1X phenotype. Mi.IX, therefore, is not uncommon in Denmark, with a frequency of 0.43%. Family Studies Families from all 4 propositi were investigated; figure 1 shows the results of testing with anti-M, -N, -S, -s, -DANE and anti-Mur. In each family, red cells of all DANE+ individuals reacted with anti-Mur and had a trypsin-resistant M antigen. The families showed that DANE was an autosomal dominant character. Inspection of the pedigrees showed clearly that D A N E was inherited with MS.No recombination between D A N E and MNS was observed. The sum of lod scores for D A N E and MNS at a recombination fraction of zero is 3.69: therefore, DANE is produced by a gene at the MNS locus or one that is very closely linked to it. DANE is, therefore, a new low-frequency MNS antigen and has been allotted the ISBT number 002032 (MNS32). Further Serological Characteristics of Mi. IX Red Cells Dane.%cells did not appear to have any obvious reduction in sialic acid levels. They were not agglutinated by Glycine soja lectin or by incomplete (IgG) anti-D. Trypsin treatment of intact red cells cleaves GPA at amino-acid residue 39 and partially at residue 31 [2] (table 2). Dahr et al. [25] showed that GPA has a chymotrypsin cleavage site at position 34, but that this site is blocked on some GPA molecules by the sialic acid of an 0-linked oligosaccharide on threonine at position 33 (table 2). Complete cleavage with chymotrypsin can be obtained after desialylation [25]. Mi.IX cells were tested with a collection of monoclonal antibodies to epitopes on the extracellular domain of GPA (table 3). R10, BRIC 116, BRIC 119, BRIC 127 and E4 all detect trypsin-sensitive determinants on control cells. None of these antibodies reacted with red cells treated with sialidase followed by chymotrypsin; BRIC 119 is sialic acid dependent, the others are not. R.L. is a human alloantibody (anti-EnaTS) made by an individual deficient in GPA [26]. R.L. serum and BRIC 127 did not react with either trypsin-treated or chymotrypsin-treated control cells and presumably only react with those GPA molecules which are not glycosylated at amino-acid residue 33 [25,19]. All these antibodies to trypsin-sensitive epitopes reacted with Mi.IX cells; with the exception of E4, none reacted with trypsin-treated Mi.IX cells. Neither BRIC 127 nor R.L. antibody reacted with chymotrypsin-treated Mi.IX cells (table 3). R18 detects a trypsin-resistant epitope on glycophorin A and R1.3 detects a sialic acid-dependent epitope common
Miltenberger Class IX
133
to the N-terminal26 amino-acid residues of both GPA and GPB. Both antibodies reacted with Mi.IX cells. The association Of a tqpsin-resistant antigen with the Mi.IX phenotype suggests that the trypsin cleavage site at amino-acid residue 39 is altered in some way. Yet despite the apparent loss of the trypsin cleavage site in Mi.1X GPA, most anti-En'TS antibodies failed to react with trypsintreated ~ i . cells, 1 ~suggesting that ~ i . GPA 1 ~lacks the epitopes detected by these antibodies. however, does react with trypsin-treated Mi.IX cells and therefore the E4 E49
Table 2. Segment of GPA molecule from amino-acid residues 30-40
*
*
Lys-Arg~Asp~Th~~Tyr~A~a~Ala-Thr-Pro-Arg~~~~~~~s~ 30 I I 33 I 37 I 41
I
1
I
I
T
dC
(T) (TI
-
* = 0-glycosylated in the majority of molecules. dC = Partial chymotrypsin cleavage site of native GPA, complete cleavage after desialylation. T = trypsi,, cleavage site. (T) = partial trypsin &avage site,
Table 3. Reactions of monoclonal antibodies and a human alloantibody to determinants on GPA with M+N+ Mi.IX and M+N+ control cells Enzyme treatment
Cells
None
Mi.IX M+N+
Trypsin
Mi.IX M+N+ Mi.IX M+N+
Chymotrypsin Sialidase
Mi.IX M+N+
Sialidasel chymotrypsin
Mi.IX M+N+
anti-M
R.L.
R10
BRIC116
BRIC119
BFUC127
E4
R18
R1.3
+ + +
+ +
+ +
+ +
+ +
+ +
+
0
0 0
0 0
0 0
0 0
0 0
+ + +
+ +
0 0
+ + +
+ +
+ +
0 0
+ + + + + +
+
0 0
+ +
0 0
0 0
Fig 1. Pedigrees of four families with Mi.IX members. NT = not tested; + = dead; M* = trypsin-resistant M antigen. 4, *=DANE and Mur positive. 0,0 =DANE and Mur negative or NT. Arrow indicates propositus.
+
+
0
+ + + +
0 0
+ + +
+
+ + + + +
0 0
134
Fig. 2. Immunoblots stained with monoclonal antibodies detecting different epitopes on GPA. a M + N + control. b trypsin-treated Mi.IX cells. c Mi.IX cells. T1 and T2 = additional bands seen with trypsintreated Mi.IX cells. X =additional band seen with untreated Mi.IX cells.
epitope does appear to be present on the Mi.IX GPA molecule. E4 does not react with chymotrypsin-treated desialylated cells (table 3) suggesting that the chymotrypsin cleavage site at amino-acid residue 34 is intact.
Immunochemistry Immunoblotting of M+N+ MI.IX cells with monoclonal anti-M (fig. 2) revealed bands representing GPA, its dimer (GPA2) and heterodimer with GPB (GPAB). However, the GPA band demonstrated a slight increase in mobility, compared with GPA from M+N+ control cells, equivalent to a decrease in apparent MI of about 1,260 (from 8 determinations with 8 monoclonal anti-M). Family studies showed that Mi" is associated with MS and so the GPA molecule detected by anti-M on immunoblots of M+N+ Mi.IX cells must be the GPA molecule produced by the MSMi.'Xgene.Immunoblots with anti-N did not demonstrate any difference between the apparent MI of GPA of Mi.IX cells and M+N+ control cells [results not shown]. R18, a monoclonal antibody to a trypsin-resistant determinant on the extracellular domain of GPA (En'KT), and BRIC 163, a monoclonal antibody to an epitope on the cytoplasmic C-terminal domain of GPA, gave a similar staining pattern with Mi.IX cells to those seen with anti-M,
Skov/Green/Daniels/KhalidiTippett
demonstrating the reduction in M, of GPA (fig. 2, results for BRIC 163 not shown). R1.3, a monoclonal antibody to an epitope common to the N-terminal portions of GPA and GPB also demonstrated the decreased apparent MI of the GPA of Mi.IX cells, as well as revealing an apparently normal GPB (fig. 2). Mi.IX, therefore, appears to be associated with a GPA molecule of slightly decreased apparent MI. A weakly staining band of apparent MI 63,500 (3 determinations) was seen on some immunoblots of Mi.IX cells with antibodies to GPA (X in Fig. 2). The significance of this band is not clear but it may represent a degradation product. When membranes prepared from trypsin-treated cells were used, anti-M, which stained no bands on an immunoblot of trypsin-treated M+N+ control cells, revealed GPA, GPA2 and GPAB on immunoblots of Mi.IX cells (fig. 2). R1.3, which only binds to GPB and GPBz on immunoblots of trypsin-treated control cells [not shown], also stained GPA, GPA2and GPAB on immunoblots of trypsintreated Mi.IX cells. Anti-M (and R1.3) stained an additional two bands, of apparent M, 61,000 and 51,200 (from 7 determinations), with membranes from trypsin-treated Mi.IX cells (fig. 2). These two extra bands were seen on immunoblots of red cells of Mi.IX members of all four families and were absent from those of non-Mi.IX members. The results with trypsin-treated cells confirm that Mi.IX is associated with a GPA molecule in which the N-terminal portion is not cleaved by trypsin. The significance of the extra two bands (TIand T2 in fig. 2) is not clear. One of them could represent a dimer of Mi.IX GPA with trypsin-cleaved normal GPA, or with fragments of some Mi.IX GPA molecules cleaved by trypsin at an alternative site to the main trypsin cleavage site at amino-acid residue 39. R10, a monoclonal antibody to a trypsin-sensitive epitope on GPA between amino-acid residues 26 and 39, also revealed GPA, GPAz and GPAB on Mi.IX cells. However, with R10 the decrease in apparent MI of GPA was not apparent and the GPA of Mi.IX cells showed reduced staining intensity compared with that of control M+N+ cells (fig. 2). R10 revealed no bands on immunoblots of trypsintreated Mi.IX or control cells. These results provide further evidence that R10 does not react with Mi.IX GPA. GPB of Mi.IX cells appeared normal as determined by immunoblots with anti-N, R1.3, human anti-S or monoclonal anti-GPB (F84.3E8.E2). The serological and immunochemical results suggest that Mi.IX is associated with an aberrant GPA molecule which has apparently normal N - and C-terminal regions and normal En'KT, a determinant situated around amino-
Miltenberger Class IX
acid residue 49 [25]. It lacks the trypsin cleavage site at residue 39, probably lacks En'TS (R.L., R10, BRIC 119, BRIC 127, but not for E4) which is situated between residues 26 and 39, and has an apparent M, of about 1,000 less than normal GPA. Based on data obtained with the monoclonal antibody E4, the chymotrypsin cleavage site at residue 34 appears to be intact. A glycosylated threonine at position 33 appears to be present as it is this glycosylation which prevents chymotrypsin cleavage of some GPA molecules [25] (table 2). The partial trypsin cleavage site between residues 30 and 31 of GPA may remain intact in Mi.IX cells as only some of the GPA molecules would be expected to be cleaved at this site and therefore trypsin treatment would not destroy M antigen expression in these cells. The biochemical and serological data are consistant with an alteration to GPA associated with Mi.IX located around the region of amino-acid residues 35-40. If the 1,000 molecular-weight reduction of Mi.IX GPA were due to the loss of one 0-glycan, then this could be from threonine at position 37. Alternatively, the reduction in molecular weight could result from a deletion of a few amino-acid residues from the region of the GPA molecule around the trypsin cleavage site at residue 39. Mur antigen has previously only been found in the phenotypes Mi.111, Mi.IV and Mi.VI. In these phenotypes Mur is associated with the presence of an abnormal GPB molecule, which may contain a GPA insert [14,16]. It is probable that anti-Mur is detecting a determinant on the aberrant GPA molecule in Mi.IX and possible that the same determinant results from the presence of an insert in the unusual GPB molecule of Mi.111, Mi.IV and Mi.VI. The significance of DANE antigen in Mi.IX is more difficult to interpret. DANE is trypsin sensitive, suggesting that it is not carried on the Mi.IX GPA molecule. However, since anti-DANE has a very low titre against Mi.IX cells, it is possible that the cleavage of a small number of DANEactive GPA molecules with trypsin, possibly at less accessible sites than the main trypsin cleavage site at position 39, would reduce the antigen site density sufficiently to prevent agglutination of the cells by anti-DANE. Unfortunately the only anti-DANE serum is in extremely short supply and the producer of the antibody is dead, so until another example is found it is unlikely that the location of the DANE antigen will be clarified. During the course of this investigation it was found that several Mg+ cells were also DANE+. This will be the subject of a future report. The serological and immunochemical findings on cells with the Mi.IX phenotype are in good agreement. The biochemical nature of the Miltenberger phenotypes Mi.1,
135
Mi.11, Mi.V, Mi.VII and Mi.VII1 is now known, whereas the biochemistry of the Mur+ phenotypes Mi.111, Mi.IV, Mi.VI and now Mi.IX remains speculative. It is probable that these will be determined by molecular analysis of the genes producing the aberrant GPA or GPB molecules.
Acknowledgements We are very grateful to Dr. D. Anstee, Dr. M. Telen, Dr. D.L. Stone and Dr. R. Fraser for providing monoclonal anibodies for use in this study.
References 1 Issitt PD: Applied Blood Group Serology, ed 3. Miami, Montgomery Scientific Publications, 1985. 2 Dahr W Immunochemistry of sialoglycoproteins in human red blood cell membranes; in Vengelen-Tyler V, Judd WJ (eds): Recent Advances in Blood Group Biochemistry. Arlington, American Association of Blood Banks, 1986, pp 23-65. 3 Siebert PD, Fukuda M: Isolation and characterization of human glycophorin A cDNA clones by a synthetic oligonucleotide approach: Nucleotide sequence and mRNA structure. Proc Natl Acad Sci USA 1986;83:1665-1669. 4 Siebert PD, Fukuda M: Human glycophorin A and B are encoded by separate, single copy genes coordinately regulated by a tumorpromoting phorbol ester. J Biol Chem 1986;261:12433-12436. 5 Siebert PD, Fukuda M: Molecular cloning of a human glycophorin B cDNA: Nucleotide sequence and genomic relationship to glycophorin A. Proc Natl Acad Sci USA 1987;84:67354739. 6 Lewis M, Anstee DJ, Bird GWG, Brodheim E , Cartron JP, Contreras M, Crookston MC, Dahr W, Daniels GL, Engelfriet CP, Giles CM, Issitt PD, Jorgensen J, Kornstad L, Lubenko A. Marsh WL, McCreary J, Moore BPL, Morel P, Moulds JJ, Nevanlinna H, Nordhagen R, Okubo Y,Rosenfield RE, Rouger P, Rubinstein P, Salmon C, Seidl S, Sistonen P, Tippett P, Walker RH, Woodfield G, Young S: Blood group terminology 1990. From the ISBT Working Party on Terminology for Red Cell Surface Antigens. Vox Sang 1990;58:152-169. 7 Cleghorn TE: A memorandum on the Miltenberger blood groups. Vox Sang 1966;11:219-222. 8 Crossland JD, Pepper MD, Giles CM, Ikin EW A British family possessing two variants of the MNSs blood group system, M' and a new class within the Miltenberger complex. Vox Sang 1970;18:407413. 9 Giles CM, Chandanayingyong D, Webb AJ: Three antibodies of the MNSs system and their association with the Miltenberger complex of antigens. 111. Anek, Raddon and Lane antisera in relation to each other and the Miltenberger complex. Vox Sang 1977;32:277279. 10 Dybkjaer E, Poole J, Giles CM: A new Miltenberger class detected by a second example of Anek type serum. Vox Sang 1981;41:302305. 11 Giles CM: Serological activity of low frequency antigens of the MNSs system and reappraisal of the Miltenberger complex. Vox Sang 1982;42:256-261.
Skov/GreenlDaniels/Khalid/Tippett
136
12 Dahr W, Newman RA, Contreras M, Kordowicz M, Teesdale P, Beyreuther K, Kruger J: Structures of Miltenberger class I and I1 specific major human erythrocyte membrane sialoglycoproteins. Eur J Biochem 1984;138:259-265. 13 Dahr W, Beyreuther K, Dybkjaer E, Moulds J, Vengelen-Tyler V Biochemical characterization of class VII and VIII cells within the Miltenberger system; in Mayr W (ed): Advances in Forensic Haemogenetics 2.Berlin, Springer, 1988,pp 22-25. 14 Dahr W, Longster G, Uhlenbruck G , Schumacher K: Studies on Miltenberger class 111, V, M and M’ red cells. I. Sodium dodecylsulfate polyacrylamide gel electrophoretic investigations. Blut 1978;37:129-138. 15 Anstee DJ, Mawby WJ, Tanner MJA: Abnormal blood-group-Ssactive sialoglycoproteins in the membrane of Miltenberger class 111, IV and V human erythrocytes. Biochem J 1979;183:193-203. 16 King MJ, Poole J, Anstee DJ: An application of immunoblotting in the classification of the Miltenberger series of blood group antigens. Transfusion 1989;29:106-112. 17 Vignal A, Rahuel C, El Maliki B, London J, Le Van Kim C, Blanchard D, Andre C, d’Auriol L, Galibert F, Blajchman MA, Cartron J-P: Molecular analysis of glycophorin A and B gene structure and expression in homozygous Miltenberger class V (Mi.V) human erythrocyes. Eur J Biochem 1989;184:337-344. 18 Anstee DJ, Edwards PAW Monoclonal antibodies to human erythrocytes. Eur J Immunol 1982;12:228-232. 19 Gardner B, Parsons SF, Merry AH, Anstee DJ: Epitopes on sialoglycoprotein a: Evidence for heterogeneity in the molecule. Immunology 1989;68:283-289. 20 Okubo Y, Daniels GL, Parsons SF, Anstee DJ, Yamaguchi H. Tomita T, Seno T A Japanese family with two sisters apparently homozygous for M”.Vox Sang 1988;54:107-111. 21 Telen MJ, Scearce RM, Haynes BF: Human erythrocyte antigens. 111. Characterization of a panel of murine monoclonal antibodies that react with human erythrocyte and erythroid precursor membranes. Vox Sang 1987;52:23&243.
22 Khalid G , Green CA: Immunoblotting of human red cell membranes: Detection of glycophorin B with anti-S and anti-s antibodies. Vox Sang 1990;59:48-54. 23 Ott J: Estimation of the recombination fraction in human pedigrees: Efficient computation of the likelihood for human linkage studies. Ann Hum Genet 1974;26:588-597. 24 Attwood J, Bryant S: A computer program to make linkage analysis with LIPED and LINKAGE easier to perform and less prone to input errors. Ann Hum Genet 1988;52:529. 25 Dahr W, Muller T, Moulds J, Baumeister G , Issitt PD, Wilkinson S, Garratty G: High frequency antigens of human erythrocyte membrane sialoglycoproteins. I. Enareceptors in the glycosylated domain of the MN sialoglycoprotein. Biol Chem Hoppe Seyler 1985;366:41-51. 26 Taliano V, Guevin RM, Hebert D , Daniels GL, Tippett P, Anstee DJ, Mawby WJ, Tanner MJA: The rare phenotype En(a-) in a French-Canadian family. Vox Sang 1980;38:87-93.
Received: December 5,1990 Revised manuscript received: February 15, 1991 Accepted: February 15, 1991 Ms C.A. Green MRC Blood Group Unit Wolfson House 4 Stephenson Way London, NW12HE (UK)
