Blood group systems

Abstract

The red cell membrane carries a great variety of surface proteins, as well as proteins that cross the lipid layer of the cell membrane itself. It is these surface proteins and glycoproteins that carry the blood group antigens and their specificity is mostly determined by the sequence of oligosaccharides (e.g. ABO) or the sequence of amino acids (e.g. Kell, Duffy, Kidd, MNS). These antigens are assigned to blood group systems or collections based on their relationship to each other as determined by serological or genetic studies. As of June 2019, the Red Cell Immunogenetics and Blood Group Terminology working party of the International Society of Blood Transfusion (ISBT) recorded that 38 blood group system genes have been identified and all known polymorphisms (alleles) sequenced. This section will primarily cover the ten major blood groups systems (ABO, Rh, MNS, P1PK, Kell, Duffy, Kidd, Lewis, Lutheran, and I) and provide some information on some of the other blood group systems. Also included is basic information about Human Leucocyte Antigens (HLA) and platelet antigens (HPA). Some references will be made to basic molecular structures, but detailed molecular structures and recent advances in DNA technology are not within the scope of this publication. Since blood group antigens are genetically determined, the frequency distribution of the antigens often varies in different populations. It is important to know what antigen frequencies apply in your local population. Testing should relate to the antigen frequency in the population, and this may present a challenge when using red cell reagents produced in other parts of the world. Where feasible, two references have been used for the percentages of different groups in the major blood group systems; those appearing in the first edition of this publication, and those from the Blood Group Antigen Facts Book (full reference needed here). Where possible we have identified the geographical location from which the information is derived, and the local ethnic groups are identified as “Black”, “Caucasian” and “Asian”. As of June 2019, 38 different blood group systems are known, ten of which are considered in this publication to be major blood group systems. In addition, there are various blood group antigens that have been allocated to collections (the 200 series), low incidence antigens (the 700 series) and high incidence antigens (the 900 series). The ISBT Working Party on Red Cell Immunogenetics and Blood Group Terminology develops and maintains guidelines for blood antigen and alleles nomenclature and assigns newly recognized antigens and alleles to the appropriate system. Currently known system and alleles are shown on the working party section of the ISBT website. Tables 1 and 2 list blood group systems and their main antigens. Various terminologies have been used to describe the different blood group systems and their antigens and respective antibodies ever since the ABO blood group system was first described in 1900 by Karl Landsteiner. In 1980 an ISBT committee was tasked to devise a genetically based numerical terminology for red cell antigens. This is an ongoing process and new information regarding the antigens and candidate new antigens are reviewed by the committee on a regular basis. The numerical terminology was primarily designed to facilitate computer input. The alternative terminologies are commonly used, both in everyday communication, in laboratories and in publications. In this section the ISBT terminology for the blood group system will be shown in brackets, preceded by “ISBT” for clarity. This terminology consists of one or more letters, a space and three digits e.g. ABO 001. We will use the more ‘user-friendly’ alternative names. Note: The term group or type can be used interchangeably when discussing blood groups or types. Further notes on Rh terminology will be found in the Rh section. The number of antigens within a blood group system, collection, and series varies tremendously from 1 in the I (ISBT I 027) system to 55 in the Rh (ISBT RH 004) and 49 in the MNS (ISBT MNS 002) system. The cluster of differentiation (abbreviated to CD) is a protocol for the identification of cell surface molecules that provide targets for the phenotyping of cells. CD molecules often act as receptors for various other molecules, and some play a role in cell signalling. Some examples are CD8 found on cytotoxic T-cells and NK cells, and CD4 found on T-helper cells. In order to distinguish these cells from one another, they may be referred to as CD8 cells and CD4 cells. Some CD molecules carry blood group antigens, such as CD235a which carries the MN antigens of the MNS blood group system (ISBT MNS 002) and CD235b which carries the Ss antigens of the MNS blood group system. At the time of this publication, more than 370 unique CD clusters and sub-clusters have been identified. Drawn by Elisabet Sjöberg Webster and reproduced with permission. The structures of the different blood group carrier molecules and their antigens have been studied extensively, and a wealth of information has become available, particularly since the development of molecular genetic techniques and the data from the human genome project. However, only a little is known about the function of the blood groups. The red cell is a complex structure, and the red cell membrane contains many surface proteins that are anchored to the membrane, cross the lipid bilayer one or more times or are adsorbed onto the surface of the red cells. Many of the proteins expressed on the surface of the red cells are polymorphic and carry the different blood groups. Figure 1 shows the red cell membrane with representative blood groups. The functions of some of the red cell membrane proteins have been identified, such as the carrier molecule of the Jra antigen of the JR blood group system (ISBT JR 032) that was identified as ABCG2, a breast cancer resistance protein (BCRP) that makes cancer cells more resistant to anti-cancer drug therapy. It has been designated as CD338. Studies on the null phenotypes that occur in most blood group systems have contributed to knowledge of their function. For example, the Rh protein, which assists in the transport of carbon dioxide across the cell membrane, also has a structural role in maintaining the flexibility and flattened shape of the red cell. Absence of Rh antigens (i.e. the Rh null phenotype) is associated with to structural changes to the red cell membrane that can produce haemolytic anaemia. The ABO, H, I, P1PK blood groups are carbohydrate structures on the red cell membrane glycolipids and glycoproteins and less is known about their function. Table 3 provides a list of the functions of the blood groups. The ABO and the Rh blood group systems are the most clinically significant blood group systems. Table 1 shows the major blood group systems. Note that H antigen is in a separate system, H (ISBT H 018), and is not part of the ABO system (ISBT ABO 001) Table 2 provides information on blood group systems other than the ten major systems. Although the ABO and H are two different blood group systems genetically, they will be described together as they are closely related, both at the biochemical and phenotype level. The ABO system is the most important blood group system in transfusion therapy and was the first blood group system to be described. This great contribution to medicine was made by Karl Landsteiner in Vienna, Austria, in 1900 when he observed that ‘the serum of healthy humans not only has an agglutinating effect on animal blood corpuscles, but also on human blood corpuscles from different individuals'. The following year, in 1901, Landsteiner was able to recognise two antigens on the red cells by separating and mixing the cells and sera of several individuals. He called the antigens A and B. Those individuals with the A antigen on their red cells were called Group A; those with the B antigen, Group B. Many individuals lack the A and the B antigens and were termed Group C, which was later termed Group O (for the German “ohne” meaning “without” or “null”). The least common group, called AB, was found by two of Landsteiner's students in 1902. Group AB individuals express both the A and the B antigens on their red cells. Landsteiner found that the serum of an individual always contained antibodies to the antigen which was not expressed on that individual’s red cells. Thus, Group A individuals will have anti-B antibodies in their serum and Group B individuals will have anti-A antibodies in their serum. These facts became known as Landsteiner’s Rule which states, ‘(In the ABO system) the antibody to the antigen lacking on the red cells is always present in the serum or plasma.’ The regular presence of anti-A and/or anti-B antibodies means that it is critical for patient safety and good transfusion practice that ABO groups are performed, recorded and interpreted correctly prior to transfusion. ABO incompatibilities are responsible for the majority of serious and/or fatal transfusion reactions and are usually caused by technical, clerical or administrative errors. In 1930 Karl Landsteiner received the Nobel Prize in Physiology or Medicine for his work on blood types. As mentioned above, the ABO system is unique in that whenever the A or B antigens are not present on the red cells, the corresponding antibody is present in the plasma. Anti-A and anti-B isoagglutinins (also known as isohaemagglutinins) are often referred to as being ‘naturally occurring’. It should be noted that the anti-A,B produced by a group O individual is different from anti-A + anti-B, which is a mixture of anti-A from one source and anti-B from another source. Anti-A,B detected in group O individuals is an antibody that will react with group A and group B cells. More information on ABO typing can be found in Section 10: Donation testing. The ABO genes are located on chromosome number 9 (9q34.1-q34.2). The inheritance in the ABO system is controlled by various alleles, four of which are common: A1, A2, B and O and a series of rare alleles, for example A3, Ax and Am with a total of 286 alleles and 537 variants being reported by March 2019. The O allele (which does not produce an antigenic product) is recessive to the A and B alleles, which are co-dominant. The ABO phenotype is shown by the grouping laboratory with ABO testing of a blood specimen, but the genotype of the individual is not obvious from these results. For example, the phenotype A1 can result from one of several genotypes such as A1A1, A1A2, A1A3, A1Ax, and A1Am or A1O. Although each individual has two ABO genes, serological tests do not reveal the O allele in the A and B phenotypes, nor can an allele producing a weak form of A be recognised if an allele higher in the scale of A antigen production is simultaneously present. The genotype can, however, be determined by DNA analysis of the gene or may be determined by family studies. Table 5 shows ABO blood group phenotypes with possible genotypes (simplified), including some of the rare alleles. The frequency of the ABO blood group genes varies between different populations. Note the variation shown in Table 6 as an example of ABO blood group distribution. The ABO red cell antigens expressed on the red cells are dependent on the presence of both the H (or FUT1 gene) as described below, and the ABO genes. The loci for the ABO and FUT1 genes are not linked (although they are functionally related) and they are therefore allocated to two separate blood group systems. The FUT1, A and B genes do not code directly for red cell antigens, but for enzymes known as transferases. The H-transferase (fucosyltransferase 1, hence FUT1 as the proper name for the gene) adds the sugar L-fucose to a precursor substrate, which is a carbohydrate chain already expressed on the red cell membrane. Once this has been performed, the 3-α-N-acetyl-galactosaminyltransferse (the enzyme produced by the A gene and for simplicity called A-transferase) and 3-α-galactosyltransferase (the enzyme produced by the B gene and for simplicity called B-transferase) can act. The A-transferase adds another sugar residue called N-acetyl-D-galactosamine, which results in the expression of A antigen on the red cells. Similarly, the B-transferase adds the sugar residue D-galactose and the cells then also express the B antigen. These red cells, as a result of the actions of the H-transferase, the A-transferase and the B-transferase, type as group AB. Therefore, the group A antigen is expressed when the H- and A-transferases are the two enzymes present; the group B antigen is expressed when the H- and B-transferases are the enzymes present, and in the case of group O only the H-transferase is present. Figure 2 shows a simplified diagram to indicate the structural differences in the molecules that result in ABH antigen expression. The expression of A, B or AB antigens results in a relative “masking” of the H antigen. Thus, A1, B or A1B cells express only small quantities of H antigen and Group O cells express the most. The A2 allele is less effective than the A1 allele in masking the H determinant. A2 cells therefore express considerably more H antigen and less A antigen than do A1 cells. A1 individuals express approximately 1 000 000 A antigens per red cell whereas A2 individuals express only around 250 000 A antigens per red cell. The O allele in the homozygous state leads to the expression of H specificity alone, resulting in group O individuals having abundant H antigen. The amount of H antigen that is detectable on red cells of different ABO groups, from left to right in decreasing order is as follows: most H antigen: O → Weak A → A2 → A2B → B → A1 → A1B → least H antigen. The A, B and H antigens are detectable long before birth, although are expressed less strongly on the red cells of children than those of adults. The ABH antigen strength usually peaks at between two and four years of age and then remains relatively constant in most individuals. It may not be possible to distinguish serologically between group A1 and A2 groups at birth as the antigens may not yet be fully expressed. Although the ABO and H are two different blood group systems genetically (ABO Blood Group System: Number 001 and H Blood Group System: Number 018), they are closely related at the biochemical and phenotype level. The H-deficient phenotypes are very rare and include a total deficiency in H antigen (the Oh phenotype, often called the “Bombay” phenotype) or a partial deficiency (“para-Bombay” phenotype). The Oh phenotype, in which the cells lack the H antigen, arises when the individual has not inherited the very common FUT1 gene. As there is no FUT1 gene present, the H-transferase enzyme is absent. The precursor substance on the red cell remains unchanged and no molecules of L-fucose are present on the precursor substrate in the red cell membrane. The individual may have inherited the A and/or B genes, which code normally for the appropriate transferases. However, without the single terminal carbohydrate L-fucose at the α-1,2- position of the substrate protein, these transferases are non-reactive. The Oh phenotype, therefore, results when the individual has inherited homozygosity for the rare null allele h. The null h gene does not code for H-transferase. Individuals who have inherited one or two FUT1 produce normal amounts of H-transferase. Oh individuals are extremely rare. Those who were originally shown to carry the trait were individuals born in India, whose ancestors originated in Bombay (now Mumbai), hence the “Bombay blood group”. There have been rare cases of Oh phenotype individual throughout the world, some with no apparent southeast Asian heritage, such as Italians living in Europe. Their red cells are not agglutinated by anti-A, -B, -A,B or -H. Oh individuals usually have powerful anti-H and -A,B antibodies in their serum/plasma. To avoid serious transfusion reaction, recipients can therefore only be transfused with group Oh blood. Table 7 shows the difference between group O and group Oh blood. About 10 years after the description of the ABO groups, the first subgroup of A was described. It was observed that not all group A bloods gave identical results when tested with anti-A from Group B individuals. It was realised, furthermore, that the common A antigen occurred in two forms: A1 and A2. Later studies on transferase enzymes of A1 and A2 individuals showed that fewer antigenic sites are produced in group A2 individuals as the enzyme is less effective in converting the precursor H substance into A antigen. However, with the use of monoclonal anti-A blood grouping reagents, little if any, difference between the reactions of A1 and A2 cells can be detected in the laboratory. In one survey in Southern Africa, about 99·9% of all group A bloods from Caucasians and about 96% of group A bloods from Blacks, were either A1 or A2, with A1 being more frequent than A2 in both populations. A higher incidence of A1 was detected in the Black population. The anti-A found in the serum/plasma of group B individuals consists of two separate antibody specificities, anti-A and anti-A1, the latter being specific for the A1 type. Group A or AB individuals who do not express the A1 antigen may form an irregular, normally “cold-reacting”, anti-A1 antibody in their serum/plasma. The lectin Dolichos biflorus (from which an anti-A1 reagent can be prepared and standardised) or monoclonal anti-A1 reagents, are usually used to type red cells for the A1 antigen. A number of other subgroups of group A and B have been described. The subgroups are caused by genetic variations that result in a variety of weakened expressions of the antigens. The subgroups cannot be detected when the gene for the weak antigen is inherited together with a normal A or B gene. The subgroups may be detected in the laboratory when weak or unexpected negative results are obtained with the forward grouping and/or anomalous results with the reverse grouping. For subgroups Ael or Bel, the presence of A or B antigens can only be demonstrated by an adsorption and elution technique with the corresponding antibodies. The term weak A covers a large range of reactivity, some bloods giving clear (although weak) results and other bloods giving such weak reactions that detection may prove difficult. The weak A types include A3, Am, Ax, Abantu, Ael, Afinn and Aend. Weak A type A3 gives a characteristic mixed field agglutination pattern when tested with polycolonal anti-A and anti-A,B grouping agents. However, stronger agglutination is detected when using monoclonal blood grouping reagents. Table 8 compares reactions between groups and subtypes. Anti-A1 may or may not be produced, although it is often produced by Ax individuals. Note that type A3 shows mixed field agglutination with anti-A and anti-A,B and that type Ax reacts macroscopically with monoclonal anti-A,B Subgroups of group B are suspected when the expression of the B antigen is weak or cannot be easily detected. Subgroups of B are very rare and are found mainly in populations where the frequency of group B is high as in African and Far Eastern populations. The subgroup cannot be detected if inherited with a normal B allele. The weak B subgroup may be inherited with an A allele giving rise to a normal A, weak B phenotype, ABweak. Acquired-B is caused by the action of enzymes that de-acetylate the group A1 antigen N-acetyl-D-galactosamine to D-galactosamine which is similar to the structure of the group B antigen sugar residue (D-galactose). Some anti-B reagents, especially monoclonal reagents that contain clone ES4, react with the acquired-B phenotype and a group A individual could be incorrectly grouped as group AB. It is important to select anti-B grouping reagents carefully to ensure that they do not react with the acquired-B phenotype. The condition is rare but may be associated with gastrointestinal bacterial disease or caused by bacterial contamination of a blood sample. The individual’s red cells often become polyagglutinable. Healthy adults who do not express a given ABO antigen on their red cells usually have the corresponding antibody in their serum/plasma as a result of stimulation from the environment, such as exposure to certain bacteria or food that may express A-, B- or H-like substances. Additional exposure to the antigen can result in more potent antibody formation. Isoagglutinins that are weak or missing in adults may occur in weak subgroups of A or B, either hypogammaglobulinaemia or agammaglobulinaemia (patients with no, or low levels of serum globulins), twin chimerism, old age or treatment with immunosuppressive drugs, or as the results of a bone marrow/stem cell transplantation. Isoagglutinins are not normally detected in newborn infants but develop after three to six months of life due to exposure to A-like and B-like antigens in the environment. If ABO antibodies are detected in neonatal blood samples, they are usually agglutinating IgG antibodies of maternal origin. Table 9 shows the normal grouping results of a group B newborn and an infant of six months of age. Individuals of phenotypes A2, A2B and weaker subgroups of A may have anti-A1 in their plasma. This antibody will react with group A1 cells. Anti-A1 is usually a “cold-reacting antibody”, which is not of clinical significance. As it seldom reacts above 25°C, it is unlikely to cause transfusion reactions or haemolytic disease of the fetus and newborn (HDFN). It may, however, mask a clinically significant antibody. Anti-A1 occurs naturally in the plasma of about 2% of A2 individuals and 26% of A2B individuals. The antibody occurs more frequently as the strength of the A antigen decreases, therefore weak A (or weak AB) individuals are more likely to have anti-A1 in their serum/plasma than A2 (or A2B) individuals. As individuals of group A1, A1B and B have very little H antigen expressed on their red cells, they sometimes develop anti-H in their plasma. This antibody can be recognised by its strong reaction with O red cells, a weaker reaction with A2 cells and usually a failure to react with A1 or B red cells. Anti-H of this nature, which is formed by individuals who are not H-deficient, is usually a benign autoantibody. Group O serum is not a simple mixture of anti-A and anti-B. It cannot be separated by selective adsorption using either group A or group B cells and is a cross-reacting antibody generally known as anti-A,B. Various theories have been suggested to explain this cross-reactivity (including Wiener’s C theory) and it appears that the anti-A,B produced by group O individuals detects a structure common to both A and B antigens. Of all the blood group systems, the ABO is the most important in transfusion because the isoagglutinins are normally present in the absence of the corresponding antigen. Strong reactions take place when incompatible bloods are mixed with each other, not only in vitro, but also in vivo. Even an initial transfusion of group A blood into a group O or group B patient may be disastrous, because the naturally occurring anti-A in the blood of the recipient would react immediately with the incoming group A cells, activating complement and causing haemolysis of the donor cells. This would lead to an acute haemolytic transfusion reaction which may be fatal. However, whole blood from ‘high titre’ group O donors, which contains immune anti-A and/or -B, may only be transfused into group O recipients (homologous group transfusion). This is because these ‘dangerous’ universal donors have potent isoagglutinins with haemolysing characteristics in their plasma, which may cause severe haemolytic reactions when infused into recipients with A and/or B antigens on their red cells. The risk of transfusing harmful anti-A and anti-B in blood group O whole blood can be reduced by the transfusion of group O red cell concentrates, from which most of the plasma has been removed. In practice, however, it is better to transfuse a patient with blood of the same ABO group (ABO identical) and to conserve stocks of group O blood for group O patients and for emergency use. Some individuals produce potent, high titre anti-A and/or anti-B, consisting of a mixture of IgM and IgG antibodies, with haemolysing characteristics in the presence of complement. This immune anti-A and/or -B in pregnant women can cause ABO HDFN with varying degrees of severity, although the fetus is rarely affected in utero. ABO HDFN typically develops within a few days of birth. See Section 7: Haemolytic diseases, for more information. Certain plant extracts (usually seeds) agglutinate human and animal red cells. Two names have been suggested for these plant agglutinins: phytagglutinins and lectins, the latter term used for those which show red cell antigen specificity. Note that these substances are not antibodies. Lectins are sugar-binding proteins or glycoproteins of non-immunological origin. Some lectins are described in succeeding discussions: The most useful lectin, anti-A1, is extracted from the seeds of Dolichos biflorus: the extract strongly agglutinates A1 and A1B cells; it reacts less strongly with A2 cells and very weakly with A2B cells. The extract can therefore be standardised by dilution as a specific anti-A1 reagent. It also reacts with uncommon red cells that express the polyagglutinin antigens Tn or Cad. Lectin anti-H can be extracted from the seeds of Ulex europaeus or the common European gorse. U. europaeus is invaluable for the classification of group O secretor/ non-secretor saliva (or group O secretor status) and confirming an Oh phenotype. In addition to being expressed on the red cells, A, B and H antigens are also expressed on most other tissues as glycolipids and glycoproteins. Soluble blood group substances of the same ABO group as the red cells may also be found in the serum/plasma and are readily detectable in the saliva and other body fluids of most individuals. The secretor status is controlled by the SE (or FUT2) gene on chromosome 19. SE is a dominant hemizygous gene and is responsible for the secretion of A, B and/or H. Approximately 80% of the general population secrete ABH substances (in the form of water-soluble antigens) in abundance in almost all their body fluids (not found in cerebrospinal fluid). There is no se allele and therefore “se” is used only to indicate the absence of SE. As there is no SE gene product in the absence of the SE gene i.e. in those individuals designated sese, these individuals are “non-secretors” and produce no water-soluble ABH antigens. The ABO group of a secretor may be determined by testing the saliva to determine the presence or absence of A, B and H substance. The remaining 20% of the population are termed non-secretors. Table 10 shows the soluble antigens secreted according to ABO group. The critical unique feature of the ABO and H blood group systems is that unlike other blood group systems, the anti-A and/or anti-B isoagglutinins are invariably present in the serum/plasma of every healthy adult when the corresponding antigen is absent from their red cells. As the ABO and H antigens are widely distributed throughout the body, the ABO group must be considered in organ transplantation. Some organs, e.g. the heart, must be ABO compatible with the recipient. In bone marrow transplantation, ABO incompatibility is acceptable because of the lack of expression of ABO on stem cells, but precautions need to be taken such as removal of the unwanted donor red cells or plasma. Note: Anomalous red cell typing may be seen post transplantation when an ABO incompatible graft was used. The cornerstone of safe blood transfusion practice is to transfuse safe blood of the compatible ABO group. It is critical that the ABO group on all samples, whether from a patient or a donor, is correct, as ABO group mistyping can have fatal consequence Number of antigens: 55 (2019) CD numbers: CD240 The discovery of the Rh groups by Karl Landsteiner and Alexander Wiener in 1940, together with the work of Philip Levine and Rufus Stetson in 1939, heralded the greatest discovery in the blood grouping field since Landsteiner described the ABO system in 1900. In 1939 Levine and Stetson described how the mother of a stillborn fetus suffered a severe haemolytic reaction when transfused with her husband's blood. The mother, who obviously lacked some ‘new’ antigen, must have been immunised by her fetus that expressed this antigen, having inherited the gene encoding it from the father. When the ABO compatible husband’s blood was transfused, the maternal antibody reacted with this same antigen expressed on his red cells. In 1940 Landsteiner and Wiener, having immunised rabbits with the blood of a rhesus monkey (Macaques mulatta), discovered that the resulting antibodies agglutinated not only the monkey red cells but also the red cells of about 85% of the Caucasians tested. Later work, however, showed that the red cell antigens detected by the human-derived antibody and the animal antibody were not identical and in fact belonged to two different blood group systems. The blood group system detected by the human-derived antibodies is now known as Rh (not Rhesus or rhesus) and the antigen is called D. The antigen originally described by Landsteiner and Wiener is designated LW in the Landsteiner-Wiener blood group system (ISBT LW 016). The two systems are serologically, biochemically and genetically different from one another. The locus for the Rh genes (RHD and RHCE) is on chromosome 1 (1p36.11) and is linked to the gene for elliptocytosis. The locus for the LW gene (now called ICAM-4) is on chromosome 19 (19p13.2). It was soon realised that the Rh antibodies produced in humans were not as simple as they had first appeared, and that many sera contained antibodies of more than one specificity. Many