The occurrence of the feline A blood group antigen on lymphocytes

Sera from 300 cats were tested for the presence of anti-lymphocytic antibodies. One hundred and nineteen sera showed some activity with the majority (79) reacting only with lymphocytes from blood group A cats. Absorption of two such sera with A, AB and B erythrocytes and absorption of AB system reagents with lymphocytes from A and B blood group cats demonstrated that the A antigen is expressed on both erythrocytes and lymphocytes. Blood group and lymphocyte typing tests of foetuses indicated that the A antigen is present on these tissues as early as 46 days gestation. The erythrocytic B antigen could not be demonstrated on lymphocytes although a single antiserum, which reacted against lymphocytes from group B cats, was found. Several sera containing anti-lymphocytic antibodies which were not related to the AB type were also detected.     

The AB blood group system of cats

Holmes (1950) and Eyquem. Podliachouk & Milot (1962) classified feline erythrocytes into two types according to their reactions with naturally occurring antibodies in cats' plasmas. Eyquem et al. (1962) designated the two antigens, A and B. and this nomenclature has been retained in the present study. The blood group system. AB. was investigated in more detail, both genetically and serologically. Frequencies of 73.3 % A and 26.3 % B were found in a survey of 1895 Brisbane cats and in addition, a new phenotype. AB. was discovered with a low incidence of 0.4 %.The results of the serological testing and limited family information suggested that the AB phenotype is inherited and not due to blood chimaerism. Preliminary genetic studies indicated that the A gene is dominant to the B in the usual situation and hypotheses to explain the occurrence of the AB phenotype are discussed.
The incidence of naturally occurring antibodies was investigated in cats, with 1895 of blood type B having anti-A and only 35 % of type A having anti-B. No subgroups of the A and B antigens were detected and no blood group substances were found in the salivas of 37 cats. There was no evidence of any serological relationship of the feline A and B antigens with the human ABO antigens.     

Biochemical characterization of the feline AB blood group system

The biochemical nature of the feline AB blood group system was characterized by analysing red blood cells from homozygous (genotype A/A) and heterozygous (A/B) type A, type B (B/B), and type AB cats. High performance thin layer chromatography (HPTLC) of red cell gly-colipids revealed that specific neuraminic acids (NA) on gangliosides, containing ceramide dihexoside (CDH) as a backbone, correlated with the feline AB blood group antigens. Although disialogangliosides predominated, mono- and trisialogangliosides were also isolated. B cats expressed solely N-acetyl-NA (NeuNAc) on these gangliosides. In addition to expressing N-glycolyl-NA (NeuNGc) containing gangliosides, A red cells have gangliosides with only NeuNAc or mixtures of both NA. HPTLC profiles of disialogangliosides from homozygous and heterozygous A cats differed slightly in the quantity of disialogangliosides. Equal amounts of NeuNAc and NeuNGc containing disialogangliosides, as well as two intermediary forms, were recovered from AB erythrocytes. Analysing disialogangliosides from red cells belonging to 17 genetically related cats, we consistently obtained the expected disialoganglio-side profile, based on blood typing and pedigree information. SDS-PAGE of red cell membrane proteins and blotting with Triticum vulgaris, a lectin recognizing NeuNAc, revealed glycoproteins of approximately 51, 53, and 80 kD in B and AB cats but only a faint band of approximately 53 kD in A cats. By haemagglutination, Triticum vulgaris could also distinguish different blood types by specifically binding to B and AB cells. Flow cytometry showed that more anti-B bound to B cells than anti-A bound to A cells. Although AB cells had a broad range of fluorescence when compared to the profiles seen with A and B cells, the mean fluorescence with AB cells was half of that seen with A or B. These results further characterize the antigens determining the feline AB blood group system illustrating differences between A, B and AB cats, and between homozygous and heterozygous A cats.     

免疫磁珠技术分离唾液A/B血型抗原并用于吸附血清A/B抗体

目的:应用免疫磁珠分离技术获得具有良好抗原性的A/B血型抗原,并探究其作为ABO血型抗体吸附剂去除A/B抗体的可行性。方法:将含有血型物质的唾液进行预处理,再与包被了抗体的磁珠混合,分离出纯度较高的A/B抗原,运用酶联免疫及凝集抑制试验验证所得抗原的抗原性及是否存在交叉反应。用未纯化A/B抗原和纯化A/B抗原包被磁珠,对含有抗A/B IgM、IgG的血清进行抗体吸附,用纯化A/B抗原对100份来自O型血孕妇的临床血清样本进行抗体吸附,分别评价其吸附效果。结果:纯化抗原与对应抗体反应后,其吸光度显著高于对照组(A抗原与A抗体0.85±0.12 vs.0.27±0.03,P0.01;B抗原与B抗体0.86±0.09 vs.0.24±0.06,P0.01),与其它类型抗体反应后的吸光度值与对照组比较差异无统计学意义(P0.05)。进行红细胞凝集抑制试验时,纯化抗原可显著抑制相应抗体与红细胞的凝集反应,对其它类型抗体与红细胞的凝集没有抑制作用。血清抗体吸附实验表明纯化抗原的吸附效率比未纯化抗原的高(97.00%vs.88.00%,P0.001)。临床样本抗体吸附实验显示,纯化A抗原对抗A IgM/IgG的吸附效率分别为96.88%、98.44%;纯化B抗原对抗B IgM/IgG的吸附效率分别为96.88%、98.44%。结论:磁珠纯化抗原能特异性地与对应抗体结合,有效吸附血清中的血型抗体,有望作为合成A/B抗原的替代品。     

Changes in the expression of blood-group carbohydrates during oral mucosal development in human fetuses

Abstract. The distribution of blood group carbohydrate chains with antigen A, B, H type 2 chain (A and B precursor), and N-acetyllactosamine (H type 2 precursor) specificity was studied in human oral epithelium from different anatomical regions. These represented various epithelial differentiation patterns such as non-keratinized, parakeratinized, and orthokeratinized stratified squamous epithelium. The material included buccal and palatal epithelium from 20 persons with blood group A or O, gingival, and alveolar epithelium from 10 persons with blood group A or B, and buccal metaplastically keratinized epithelium from nine blood group A, two blood group B, and nine blood group O individuals. The blood group carbohydrate chains were examined in tissue sections by immunofluorescence microscopy. The A and B blood group antigens were detected by human blood group sera, and antigen H type 2 chains and N-acetyllactosamine by murine monoclonal antibodies. Each antigen showed a similar staining pattern in buccal and alveolar epithelium (non-keratinized) which differed considerably from that seen in palatal and gingival epithelium (ortho- and parakeratinized). The expression of blood group antigens A or B and the precursor antigen H type 2 chains in metaplastically keratinized buccal epithelium was found to differ significantly from that seen in normal non-keratinized buccal epithelium. The regional variations demonstrated in cell surface carbohydrates are suggested to reflect differences in tissue differentiation.     

Regional variations of cell surface carbohydrates in human oral stratified epithelium

The distribution of blood group carbohydrate chains with antigen A, B, H type 2 chain (A and B precursor), and N-acetyllactosamine (H type 2 precursor) specificity was studied in human oral epithelium from different anatomical regions. These represented various epithelial differentiation patterns such as non-keratinized, parakeratinized, and orthokeratinized stratified squamous epithelium. The material included buccal and palatal epithelium from 20 persons with blood group A or O, gingival, and alveolar epithelium from 10 persons with blood group A or B, and buccal metaplastically keratinized epithelium from nine blood group A, two blood group B, and nine blood group O individuals. The blood group carbohydrate chains were examined in tissue sections by immunofluorescence microscopy. The A and B blood group antigens were detected by human blood group sera, and antigen H type 2 chains and N-acetyllactosamine by murine monoclonal antibodies. Each antigen showed a similar staining pattern in buccal and alveolar epithelium (non-keratinized) which differed considerably from that seen in palatal and gingival epithelium (ortho- and parakeratinized). The expression of blood group antigens A or B and the precursor antigen H type 2 chains in metaplastically keratinized buccal epithelium was found to differ significantly from that seen in normal non-keratinized buccal epithelium. The regional variations demonstrated in cell surface carbohydrates are suggested to reflect differences in tissue differentiation.     

Dolichos biflorus agglutinin (DBA) reacts selectively with mast cells in human connective tissues

Dolichos biflorus agglutinin (DBA) binds to N-acetyl-D-galactosamine (GalNAc) residues in glycoconjugates and agglutinates erythrocytes carrying blood group antigen A. In cryostat sections of various tissues from blood group-specified humans, fluorochrome-coupled DBA bound preferentially to fusiform connective tissue cells and to certain epithelial cells. The connective tissue cells were identified as mast cells by their typical metachromasia in consecutive staining with toluidine blue. Double labeling with DBA and conjugated avidin revealed two distinct populations of mast cells. In several tissues the DBA-reactive cells likewise displayed uniform avidin reactivity. In intestinal mucosa, however, morphologically distinct DBA-binding mast cells were found, which were labeled with the avidin conjugates only in specially fixed paraffin sections. DBA did not bind to vascular endothelial cells, which could be identified by double staining with antibodies to factor VIII-related antigen. Labeling with Helix pomatia agglutinin (HPA), another blood group A-reactive lectin, resulted in distinct blood group-dependent fluorescence of the endothelia. Sophora japonica agglutinin (SJA), a blood group B-reactive lectin, labeled vascular endothelial cells in tissues from blood group A, AB, and B donors. HPA and SJA reacted with small mast cells in the gastrointestinal mucosa but failed to label large mast cells in any of the tissues. These results indicate that the blood group reactivity of lectins, as determined by erythroagglutination, is not necessarily consistent with their reactivity with blood group determinants in tissue sections. Moreover, DBA conjugates appear to be a reliable probe for detection of mast cells in various human connective tissues.     

A血型单克隆抗体的抗原结合部位结构多样性研究

应用杂交瘤技术,以Ａ型红细胞,Ａ１血型物质ＭＳＭ（Ａ１）和Ａ－ＲＢＣ＋ＭＳＭ（Ａ１）为免疫原,制备了一组抗人Ａ血型单克隆抗体：Ａ１２１８,Ｂ５７,ＤＥ９２３－Ｇ８,Ｄ２８６－Ｅ１２经Ｔａｋａｔｓｙ微量血细胞凝集试验证明：这组单抗仅能凝集Ａ１,Ａ２及ＡＢ型红细胞,不能凝集Ｂ,Ｏ型红细胞．采用ＥＬＩＳＡ定量抑制试验法,精确测定了它们抗原结合部位的结构,互补于Ａ活性寡糖。Ａ１２１８互补于具有双岩藻糖结构的Ａ活性五糖（Ａ－Ｐｅｎｔａ）;Ｂ５７,ＤＥ９２３－Ｇ８互补于具有单岩藻糖结构的Ａ活性六糖（Ａ－Ｈｅｘａ）;而Ｄ２８６－Ｅ１２则互补于具有单岩藻糖的Ａ活性四糖（Ａ－Ｔｅｔｒａ）．结果表明：血凝特异性相同的抗Ａ单抗,其抗原结合部位的结构可呈现多样性。即Ａ活性寡糖的糖基组成数目和含有岩藻糖数目均可不相同,各种抑制剂对不同单抗的抑制作用强弱也不相同。     

Distribution of blood-group-related carbohydrate antigens on oral endothelial cells

Summary Recent studies indicate that carbohydrate structures are involved in various endothelial functions such as inflammatory processes, adhesion of metastatic cells to endothelium and endothelial differentiation. In this paper we report the endothelial expression of various blood-group-related carbohydrate structures in normal oral mucosa using 15 different monoclonal antibodies reacting with type 1, 2 or 3 carbohydrate chains. Twenty biopsies, including normal oral mucosa, from secretor individuals comprised nine blood group O, nine A, one B and one AB. Endothelial staining was compared with epithelial staining in the same biopsies. Five blood-group-related carbohydrate antigens were detected on endothelial cells. The H type 2 antigen, which is the precursor of A, B and AB antigens, has previously been believed to be a universal marker for endothelial cells. All blood group O individuals (n = 9) showed strong H antigen staining in the endothelium of most vessels. However, of blood group A, B and AB individuals (n = 11), four showed heterogenous H antigen staining. In addition, we found that six out of ten blood group A or AB individuals, who expressed A or A and B antigens on spinous squamous cells and glandular epithelium, showed either heterogenous or no staining for these structures on their endothelial cells. It is concluded that there are differences between the biosynthesis of blood-group-related carbohydrate antigens in oral endothelium and epithelium.     

Distribution of Feline Blood Types Detected in the Copenhagen Area of Denmark

The purpose of the present study was to make the first survey of the distribution of feline AB blood types in the Copenhagen area of Denmark. A total of 244 cats (139 purebred cats and 105 Domestic Shorthair cats) were tested. 93% of all tested cats had blood type A. Neither an AB nor an O type cat was detected and thus, the frequency of blood type B among all tested cats was 7%. Most type B cats were purebred cats (Birman, British Shorthair and Persian cats). No association between sex and blood type could be demonstrated among British Shorthair and Persian cats. Thus, the present study indicates that cats in Denmark predominantly have blood type A, and that blood type B cats are rare, except for certain breeds such as Birman and British Shorthair cats.     

Aberrant glycosylation based on the neo-expression of poly- N-acetyllactosamine structures in squamous cell carcinomas of the head and neck

An immuno- and lectin-histochemical study was performed to investigate the aberrant expression of blood group-related antigens and poly-N- acetyllactosamine structures in squamous cell carcinomas of the maxillary sinus, the larynx, the apipharynx, the hypopharynx, the oral cavity, the parotid gland and the tonsil from 52 patients using monoclonal antibodies against A, B and H antigens, and six lectins, UEA-I, PNA, VVA-B4, PWM, LEA and DSA. In addition, GSA- II staining following endo-·-galactosidase digestion procedure was also applied. A, B and H antigens were expressed in most normal epithelial cells of head and neck organs, and depended on the patient blood type. However, in squamous cell carcinoma, A antigen was not detected in eight out of 25 individuals of blood groups A and AB, although B antigen was consistently expressed in carcinoma cells from all the B and AB individuals. On the other hand, H antigen was expressed in carcinoma cells not only from all blood group O individuals, but from 32 out of 35 individuals of blood groups A, B and AB. T and Tn antigens, which are recognized by PNA and VVA-B4, were strongly expressed in carcinoma cells from 40 and 42 out of 52 individuals respectively. Reactivity with GSA-II staining following endo-·-galactosidase digestion, which recognizes linear poly-N-acetyllactosamine structures, was found in a few malignant cells from 21 individuals. Staining with anti-A, -B and -H monoclonal antibodies and UEA-I lectin was diminished after endo-·-galactosidase digestion in some cases. Lectins specific for poly-N-acetyllactosamine, such as PWM, LEA and DSA, exhibited reactivity in some malignant cells from 30, 22 and 32 out of 52 individuals respectively. These results suggested that the expression of the blood group-related antigens is suppressed and immature carbohydrate chains, that is H, T and Tn antigens, are accumulated in squamous cell carcinomas of the head and neck. The results further suggested that poly-N-acetyllactosamine structures are simultaneously synthesized along with the deletion of A antigen and the accumulation of precursors This revised version was published online in November 2006 with corrections to the Cover Date.     

Blood group A and B glycosyltransferases synthesize A and B determinants on different acceptor polyglycosyl peptidesin vitro

The glycosylation of polyglycosyl chains from human erythrocytes by human plasma blood group A and B glycosyltransferases was studied in order to clarify why human blood group AB erythrocyte polyglycosyl peptides carry only either A or B determinants [Eur J Biochem (1981) 113:259–65].The blood group A transferase was able to add radioactiveN-acetylgalactosamine from labeled UDP-N-acetylgalactosamine to B-type erythrocytes which had been treated with -galactosidase in order to cleave the B determinant sugar from the erythrocytes. This suggests that the enzymes specified by theA andB genes utilize the same acceptor molecules on erythrocyte membranes. Polyglycosyl peptides isolated from blood group B erythrocytes acted as acceptors for blood group A glycosyltransferase and the generation of hybrid structures containing both A and B determinants was demon-strated. When blood group O polyglycosyl peptides were used as acceptors in the simultaneous presence of both blood group A and B glycosyltransferases, however, the A and B determinant sugars were found in different polyglycosyl peptides. It is suggested that the enzyme-acceptor complex does not dissociate until the final number of determinants has been added.     

Alloantigens expressed on activated human T cells different from HLA-A,B, C,and DR antigens

This study investigates alloantisera containing antibodies directed against antigens which are expressed on alloactivated human T lymphocytes but are absent on resting T and B cells. Among 39 defined anti-HLA-DR sera from multiparous women we found six sera giving positive reactions (more than 25 percent cytotoxicity) on in vitro alloactivated T cells, though negative reactions with resting B or T cells from the donors of either the responding or stimulating cell populations used for alloactivation. Two such sera were submitted to absorption and elution studies. Absorption of these sera with activated T cells did not remove the anti-HLA-DR activity. Furthermore, the antibodies eluted from activated T cells did not react with B cells but were positive only on activated T cells. In addition, we absorbed the sera with B cells and observed that they remained positive on activated T cells. The positive reactions do not seem to be due to either the passive acquisition of antigens from the stimulating population or to low levels of HLA-specific antibodies. As one of the sera we studied intensively gave clear positive and negative reactions on a panel of activated T lymphocytes, we believe it may recognize an antigen of an allogeneic system expressed on alloactivated human T cells.     

Differences Between Brucella Antigens Involved in Indirect Hemagglutination Tests with Normal and Tanned Red Blood Cells

Brucella antigens capable of sensitizing normal and tanned sheep red blood cells for indirect hemagglutination were compared with antigens involved in agglutination, gel diffusion, and immunoelectrophoresis. Hyperimmune rabbit sera, before and after absorption with various antigenic preparations from smooth and rough B. abortus, were used in the tests. Normal erythrocytes could be sensitized with an NaOH-treated ether-water extract (EW-T) of smooth Brucella. Tanned erythrocytes could be sensitized with a water-soluble extract from ultrasonically disrupted smooth or rough Brucella. The EW-T produced a single precipitation band and the water-soluble antigens produce 6 to 23 bands in immunoelectrophoresis with unabsorbed sera. After absorption of antisera with water-soluble extracts from smooth or rough Brucella cells or from smooth or rough cell walls, the hemagglutinins for sensitized tanned erythrocytes and the precipitins for water-soluble antigens were removed. Absorption with living smooth or rough Brucella cells or with EW-T did not remove these antibodies. The precipitins and hemagglutinins for the antigen EW-T, and agglutinins for smooth cells, were absorbed by smooth antigens but not by rough antigens. It appears that the antigens which sensitize tanned erythrocytes and diffuse through agar gels are present in both smooth and rough forms and may be situated in the cytoplasm or in the internal part of the cell wall, whereas the agglutinogen and the antigen which attaches to normal erythrocytes are surface antigens found only on the smooth Brucella cell.     

Serological and genetic identification of a bovine B lymphocyte alloantigen system

A two-colour fluorescence micro cytotoxicity test was used to screen antisera for antibodies specific for bovine B lymphocytes. A total of 114 cattle alloantisera were screened against peripheral blood lymphocytes from 100 unrelated individuals. Anti-B lymphocyte activity was detected in 47 antisera. Cytotoxic antibodies to antigens other than B lymphocyte specific antigens were removed by absorbing the antisera with buffy coat cells or platelets isolated from whole blood. Selected antisera were used to type paternal half-sib families. The presence of a polymorphic, MHS-linked antigen system on B lymphocytes was demonstrated. The tissue distribution and MHS linkage of these antigens suggests this system is analogous to the class II or Ia antigens of other species.     

Serological and genetic identification of a bovine B lymphocyte alloantigen system

A two-colour fluorescence microcytotoxicity test was used to screen antisera for antibodies specific for bovine B lymphocytes. A total of 114 cattle alloantisera were screened against peripheral blood lymphocytes from 100 unrelated individuals. Anti-B lymphocyte activity was detected in 47 antisera. Cytotoxic antibodies to antigens other than B lymphocyte specific antigens were removed by absorbing the antisera with buffy coat cells or platelets isolated from whole blood. Selected antisera were used to type paternal half-sib families. The presence of a polymorphic, MHS-linked antigen system on B lymphocytes was demonstrated. The tissue distribution and MHS linkage of these antigens suggests this system is analogous to the class II or Ia antigens of other species.     

Identification of human peripheral T cell antigens.

A new type of differentiation antigens on human T cells was demonstrated by using a heterologous anti-human T cell serum (ATS). This type of antigen, referred to as human peripheral T cell antigen (HPTA), was found on peripheral T cells and medullary thymocytes, but not on cortical thymocytes and B cells. The percentage of ATS-reactive lymphocytes in human peripheral lymphoid organs was correlated with that of cells rosetting with sheep erythrocytes, but contrasted with the number of B cells defined by the presence of a complement (C) receptor or by rabbit anti-human B cell serum (ABS). ATS also reacted with T cells purified by nylon fiber column filtration but ABS did not. Chronic lymphocytic leukemia cells rosetted with either sheep erythrocytes or erythrocyte-antibody-complement complexes were lysed by ATS and ABS, respectively. Mitogenic responses of blood lymphocytes to phytohemagglutinin (PHA) and concanavalin-A (Con A) were abrogated by treating them with ATS and C, whereas ABS suppressed only their response to Con A. Although numerous thymus cells rosetted with SRBC, only 14% were reactive with ATS. Quantitative absorption studies demonstrated that HPTA content of the thymus cells was much lower than that of lymph node cells. Anatomical localization of ATS-reactive lymphocytes in human lymphoid organs studied by immunofluorescence indicated that they were present in the thymus-dependent paracortical areas of lymph node and in the medullary region of thymus. ABS, on the other hand, did not stain thymocytes but reacted selectively with the cells located in the lymphoid follicles of lymph node. These data, together with that from cell suspension studies, confirmed that HPTA were shared between medullary thymocytes and peripheral T cells.     

Expression of the major red cell sialoglycoprotein, glycophorin A, in the human leukemic cell line K562.

We have found that the human leukemic cell line K562 (Lozzio, C.B., and Lozzio, B.B. (1975) Blood 45, 321-334) synthesizes a surface membrane glycoprotein which is identical or closely similar to the major red cell sialoglycoprotein, glycophorin A. The protein can be precipitated by specific anti-glycophorin A antiserum both from surface-labeled and metabolically labeled K562 cells. Cyanogen bromide cleavage of glycophorin A from red cells and the K562 cell protein gives apparently identical fragments, and the glycopeptides and oligosaccharides obtained after Pronase and mild alkaline treatment are closely similar. An antiserum made against intact K562 cells and absorbed with normal human white blood cells precipitated surface-labeled glycophorin A from erythrocytes. The amount of glycophorin A per cell in erythrocytes and K562 cells was very similar when determined by radioimmunoassay. The K562 cells contained blood group MN activity when tested with rabbit anti-M and anti-N sera. When incubated at 37 degrees C with rabbit anti-glycophorin A F(AB)2 fragments and fluorescent sheep anti-rabbit IgG, partial redistribution of glycophorin A (patching and capping) was seen in K562 cells but not in erythrocytes.     

Analysis of the specificity of five murine anti-blood group A monoclonal antibodies, including one that identifies type 3 and type 4 A determinants

The specificity of five mouse monoclonal anti-A blood group antibodies (Ab), four of which were produced by immunization with cultured human cancer cells and one with a synthetic antigen, has been determined by examining their reactivity with purified A glycolipids, erythrocyte glycolipids, oligosaccharides, ovarian cyst glycoproteins, and salivary glycoproteins. Two of the antibodies (HT29-36 and CB) reacted with all A variant structures tested and have a broad anti-A reactivity. Ab CLH6 did not agglutinate A erythrocytes and reacted preferentially with the type 1A structure. Ab S12 agglutinated all A1 erythrocytes and reacted best with simple, monofucosyl type 2 A structures, such as Aa-2, Ab-2, and A tetrasaccharide. Ab M2 has a novel, but complex, spectrum of reactivity. It reacts with type 3 and type 4 A chains and not with type 1 and type 2 A chains. It appears to recognize both an external A structure (formula; see text) (I) (found) in type 3 and type 4 chains) and also an internal structure (II) found in type 3 chains. Ab M2 agglutinates all A and AB erythrocytes but does not react with salivary glycoproteins.     

Frequency and inheritance of A and B blood types in feline breeds of the United States

Using a simple hemagglutination assay to determine A and B blood types, we surveyed 1,072 domestic short- and longhair (DSH/DLH) cats and 1,100 purebred cats in the United States. Data from 234 matings with 552 offspring were consistent with the hypothesis that feline blood types A and B are due to the action of two different alleles at the same gene locus and that A is completely dominant over B. Neither an AB nor an O type cat was encountered. No type B cats were found in the Siamese and related breeds or in American Shorthair and Norwegian Forest cats. Among the breeds with type B blood, the proportion was lowest in DSH/DLH cats (0.0028) and variably higher in Abyssinian, Birman, British Shorthair, Devon Rex, Himalayan, Persian, Scottish Fold, and Somali, ranging from 0.15 to 0.59. Since all type B cats have strong, naturally occurring anti-A alloantibodies, fatal neonatal isoerythrolyses occur in type A offspring of type B mothers bred to type A males. The gene frequency of the B allele and the proportion of mating at risk of neonatal isoerythrolysis were estimated in a number of breeds. In most breeds, the frequency of the B allele was less than 0.5. Since the kittens at risk for neonatal isoerythrolysis always have the genotype AB, there is constant natural selection against heterozygotes. Heterozygote disadvantage in the cat AB system represents an unusual form of negative selection, similar to that in Rh blood group incompatibility in humans.