Rare gems: null phenotypes of blood groups

For most of the 30 defined blood group systems, there is at least one associated biological function. When searching for specific serological phenotypes, these “functional phenotypes” may prove to be instrumental for donor screening. A typical candidate is an erythrocyte population lacking a functional protein belonging to a particular blood group; identification of such a blood group can be made by the lack of the protein’s function in the red cell membrane. Some of these phenotypes are exceedingly rare, including the null phenotype of the Kidd (JK) blood group system Jk(a-b-). These serological gems deserve careful attention from the clinical as well as the scientific perspective.

In this issue of Blood Transfusion, Deelert et al.¹ report using the urea lysis assay, known for 27 years^2,3, to screen more than 20,000 blood samples. The current study appears to be the first published random screen in a large set of donors for identifying Jk(a-b-) donors using the urea lysis assay. Five Jk(a-b-) donors have been identified, a significant number for any rare donor program⁴. The European Frozen Blood Bank of the Council of Europe, for instance, currently lists a total of 12 Jk(a-b-) units.

The authors did not hesitate to document the results immediately after completion of their study. In Thailand and other Asian countries, where Jk(a-b-) donors are more frequent, their approach deserves application to an even larger number of donors, as it does outside of Asia where reliable estimates for frequency of Jk(a-b-) carriers are unavailable. The assay is remarkably inexpensive, welcome in Thailand and elsewhere, similar to many serological techniques which are often very inexpensive relative to the benefits they provide for patients.

With a random screen, the phenotype frequency in the population can be estimated. Assuming that 5 of 20,613 donors were found (excluding the possibility of more than one donation by the same donor), a lower estimate for the Jk(a-b-) phenotype in the donor population served by the Bangkok Blood Center is 1 in 4,122 (1 in 1, 844 to 1 in 10,463; 95% confidence interval, Poisson distribution). It is more difficult to estimate the allele frequency without knowing the possible consanguinity in the population and whether any of the 5 Jk(a-b-) donors were related⁵. One in every 33 donors, however, may carry a JK^null allele in heterozygous form in the Thai population.

Null phenotypes are very rare for at least nine of the 30 blood group systems (Table I)⁶. Besides the JK system, the most recently described RhAG system and – I like to suggest – the Rhesus system are among these nine systems with very rare null phenotypes, if RhCE and RhD are considered as one functional group. For most of these nine systems the null phenotypes are apparently disadvantageous, associated with diseases or incapacitating conditions, either obligatory or under stress factors, and so remain rare in the populations.

Table I.

Clinical consequences of null phenotypes

Clinical relevance	Blood group system	Prevalence of null phenotype	Symptoms or clinical benefit, remarks
Always symptomatic or associated with disease	Diego	Very rare	Severe haemolytic anemia
	RHAG	Very rare	Haemolytic anemia, often compensated
	Kx	Very rare	McLeod syndrome: acanthocytosis, neurological symptoms
	GLOB	Very rare	Repeated fetal loss in some individuals, resistance to parvovirus B19
	I	Very rare	Congenital cataracts in some alleles
	Ch/Rg	Very rare	Systemic lupus erythematosus
Symptomatic under stress conditions	Colton	Very rare	Impaired urine concentrating ability
	Kidd	Very rare	Impaired urine concentrating ability
Never symptomatic, no advantage known	Rhesus	Lack of both Rhesus proteins, RhD and RhCE, very rare	Function unknown. Rarity may hint to a biological relevance of the structures missing in null phenotype.
	Kell, Yt, Scianna, Dombrock, LW, H, Cromer, Knops, Indian, OK, JMH, GIL	Rare	Rarity may hint to a biological relevance of the structure missing in null phenotypes and to a unrecognized clinical disadvantage
	ABO, Lewis, P, Raph, Xg	Frequent	Major unrecognized disadvantage unlikely, but significant biological relevance still possible
Never symptomatic, advantageous under certain conditions	Lutheran	Rare	Reduced thrombosis in sickle cell disease possible, acanthocytosis in “inhibitor”-type
	MNS and Gerbich	Frequent in affected populations	Resistance to some Plasmodium falciparum, mild elliptocytosis in “Leach”-type of Gerbich
	Duffy	Frequent in affected populations	Resistance to Plasmodium vivax

Modified from Flegel and Wagner.⁶

Why then does Jk(a-b-) ever occur and apparently in higher numbers in tropical Asia^7,8, if lack of the JK protein is so disadvantageous? We do not have an obvious explanation. However, the driving force that ultimately results in the null phenotype is often the heterozygous status. As a functional protein variant is exposed and its suitability becomes crucial within a given environment, it may prove to be more advantageous for the carrier than the null phenotype itself. The environmental challenges vary, but parasites, notably malaria, are notorious in affecting variation in many populations. Once utilized as an escape from a pathogen, one of the spontaneously occurring variant proteins can prove to be superior in physiological function.

To speculate further, a population with a greater number of rare variants that can induce null phenotypes may also have a greater rate of variant but functionally active alleles. This prediction can be explored experimentally. For instance, nucleotide sequencing is clearly sensitive enough to detect heterozygous alleles, whether these are functional or null alleles.

It may prove interesting to study heterozygous carriers with only one functional allele, as their second chromosome harbors a null allele. These carriers can be considered functionally hemizygous, like D positive individuals who carry a functional RHD allele on one chromosome and the RHD deletion on the other. Here, the authors’ approach fails, as they point out in their publication: the urea lysis test is unsuitable for detection of functionally hemizygous carriers of the JK protein. However there is a solution, as genotyping for the known or at least the more common JK^null alleles can easily detect individuals with only one functional allele.

Determining the molecular bases of the 5 Jk(a-b-) donors will be an important step, as the authors have suggested. Twelve JK^null alleles are published^9,10. Will these known alleles, two of which were reported to be more frequent in the Polynesian and Finnish populations¹¹, be found in the Thai donors? How prevalent is the common JK(IVS5-1G>A) allele in Thailand? Do the alleles in the Thai population eventually match the alleles in Indian populations, for whom a surprisingly high prevalence of the Jk(a-b-) phenotype has been reported^7,12, but any follow up has been lacking for more than a decade?

Protein variants become exposed on erythrocytes as a consequence of heterozygosity with a null allele, for instance serendipitously as an escape from a pathogen. Still, some of these protein variants are likely to express their primary, and possibly improved, functions in tissues other than red blood cells. In the case of the JK protein, the most sensitive tissue to functional variants may well be the kidney. The physiological function and response to medication in hemizygous carriers with or without protein variants is amenable to clinical study.