An uncommon MALady: is the AnWj puzzle complete?

In this issue of Blood, Tilley and colleagues report the discovery of the gene underlying the inherited AnWj-negative phenotype.¹ They identified a 6646-base-pair deletion at the MAL locus in 8 individuals with the AnWj-negative phenotype but not in family members with the AnWj-positive phenotype. Anti-Mal agglutinated AnWj-positive red blood cells (RBCs) but not AnWj-negative RBCs, which was confirmed by flow cytometric analysis. They further demonstrated that overexpression of the Mal protein resulted in expression of the AnWj antigen, which was inhibited by competitive antibody binding assays.

However, a number of unanswered questions remain in the enigmatic AnWj story. Human blood groups are currently defined as inherited polymorphisms of the erythrocyte surface that have, via transfusion or pregnancy, stimulated the production of alloantibodies. The first blood group system described was ABO in 1900, and thereafter discovery of the subsequent 46 blood group systems described to date has mirrored advances in detection techniques, both in the clinical laboratory and in advances in our ability to tease out the complexities of the erythrocyte genome and proteome. Guidelines for approving a new system are currently under revision by the International Society of Blood Transfusion’s Working Party (WP) for Red Cell Immunogenetics and Blood Group Terminology (https://www.isbtweb.org/isbt-working-parties/rcibgt.html) following debate surrounding the recent approval of CD36 as a novel human blood group.² CD36 is barely detectable on mature RBCs but highly expressed on more immature erythroid cells in the bone marrow. For the first time, the guidelines will state that the antigen(s) in the system must be found on erythrocytes. This requirement is new since it was always taken for granted that blood group antigens would be present on RBCs. Based on the intrinsic low expression of some surface proteins and, in some cases, the waning of protein expression during erythropoiesis, the WP has agreed to consider an antigen that is difficult to detect by standard hemagglutination if it can be detected by alternative methods (eg, flow cytometry or mass spectrometry) and evaluate each antigen candidate given the whole body of evidence presented. The WP will consider candidates detected only on reticulocytes or other earlier erythroid cell stages and the potential clinical consequences thereof (eg, fetal anemia).

Here, Tilley et al have demonstrated inheritance of the phenotype in 2 unrelated families and have demonstrated the independence of MAL from other loci. Although there is plenty of historical and current information regarding allo-anti-AnWj, the evidence for the Mal protein on early erythroid cells remains elusive. Neither Mal protein nor AnWj antigen was detected on the immortalized cell line BEL-A even if the MAL transcript was present. The BEL-A cell line is immortalized at the pro- to early basophilic erythroblast stage,³ and maybe Mal is produced at a later stage of differentiation. However, Mal was also consistently undetectable in primary erythroblasts throughout all stages of differentiation, suggesting that ex vivo erythropoiesis conditions do not support the expression of Mal on the membrane surface. Tilley et al showed also a bimodal expression pattern in mature RBCs, indicating that only a proportion of cells express this antigen. Thus, several questions remain unanswered regarding the expression of this antigen in erythroid cells. What is the expression profile of the antigen in bone marrow erythroblasts? Additionally, why is it that only a subset of RBCs expresses it?

Mal is a lipoprotein, and perhaps the AnWj epitope is not always accessible to detection by our routine methods. Although this theory is speculative, it prompts the question as to what burden of proof is required to ensure that a very rare finding is in fact the true basis of a rare phenotype. Mal is the third candidate identified as the potential bearer of the AnWj antigen. CD44 was first suggested as a candidate in 1995,⁴ and more recently a missense variant in SMYD1 identified in 6 subjects of Middle Eastern origin with AnWj-negative phenotype led to the claim that Smyd1 was responsible for AnWj antigen expression.⁵ Although no molecular changes were identified in CD44, the latter has proven to be interesting and illustrates the risk of red herrings when investigating such rare variants: both SMYD1 and MAL are located on chromosome 2, on either side of the centromere, and the SMYD1 variant previously identified was shown to be in linkage disequilibrium with the MAL deletion. Mal protein is highly expressed in T lymphocytes, and it is involved in their differentiation and activation via Lck transport to the plasma membrane.^6,7 This expression is consistent with the reported detection of AnWj antigen in the Jurkat T-cell line. Intriguingly, the absence of this protein in individuals with the hereditary AnWj-negative phenotype does not result in any apparent hematological disorders in the subjects. Extensive characterization of lymphocytes from these individuals would be clearly relevant in the understanding of the lymphoid function of this protein.

Because blood group antigens are inherited traits, they are generally not altered throughout an individual’s lifetime. However, rarely, an acquired blood phenotype spontaneously occurs even in the absence of previous transfusion or transplantation, resulting from apparent antigen loss at the RBC surface. Changes in blood phenotypes have been reported in the course of hematologic malignancies and appear to be predominantly clonal.⁸ The acquired AnWj-negative phenotype occurs typically in the setting of an underlying lymphoproliferative disorders.⁹ It has been proposed that this phenotype is caused by a transient decreased antigen expression, with a concomitant appearance of anti-AnWj.⁹ The pathogenesis of the downregulation of AnWj in this setting remains unclear; however, in several cases successful treatment of the underlying disease has resulted in the reappearance of the antigen and disappearance of the antibody. Tilley and colleagues showed that Mal protein is absent in RBCs of these subjects, despite the absence of any mutation in the gene, consistent with our current understanding of acquired antigen loss, where epigenetic factors such as altered methylation in the promoter region have been suggested as the cause.¹⁰ In summary, although the experimental proof presented by Tilley et al that identifies MAL as the gene underlying AnWj antigen expression is convincing, unanswered questions still remain surrounding this enigmatic phenotype.

Conflict-of-interest disclosure: The authors declare no competing financial interests.