Test Catalog


Molecular Analysis for Determining Blood Group Phenotypes

March 2012

Editor’s note: This article describes a subset of data from a larger study. The entire article was published in Immunohematology in 2011 (Immunohematology 2011;27:12-19)1


Blood transfusion has become part and parcel of supportive care with many benefits to patients needing improved oxygen-carrying capacity. However, there are intrinsic risks to the transfusion recipient, including transfusion-transmissible diseases (TTD) caused by donor viruses, and parasites and bacterial contaminants of blood products. In addition, there is a risk of alloimmunization (immune response to donor antigens) due to donor-recipient antigen phenotype disparities. Despite this risk, routine selection of red blood cells (RBCs) for blood transfusion has largely been restricted to matching for the major antigens, ABO and D, an approach that is considered safe and cost-effective in most situations. An exception to this practice includes chronic transfusion recipients,  whose management sometimes dictates extended matching for minor antigens because of existing preformed alloantibodies.

Figure 1

The ability to perform extended phenotype-serology testing on all donors has been limited due to 
several factors:

  1. Reagent antisera for rare and minor antigens are increasingly unavailable and costly.
  2. There is a lack of high-throughput platforms for extended antigen-phenotype characterization by hemagglutination.
  3. The added complexities of inventory management would be impractical for many blood banks.

While agglutination tests have been the gold standard for identification of blood group antigens for many years, the reliability of serology testing can be hampered in certain instances. Unreliable results can occur when patients have had recent transfusions (eg, massively transfused patients) and in patients demonstrating a positive direct antiglobulin test due to in vivo sensitization. The reliability of serology can also be compromised in certain phenotypic profiles, such as when there is simultaneous expression of low-incidence antigens that alter the expression of certain clinically significant antigens. For example, red cells that express low-incident antigens such as TSEN (TSEN RBC+) are known to alter the expression of the S antigen in the MNS blood group system, which can result in false-negative reactions. The presence of certain variant antigens (weak D alleles, Del alleles) can also cause false-negative reactions depending on the method of testing or the source of the reagent antisera.

Recently, DNA-based testing for red cell phenotype analysis has been developed, offering many advantages over current serology or hemagglutination platforms. This testing has broad implications for the future of transfusion medicine. Unlike serology-based methods, molecular analysis for determining blood group phenotypes directly sequences alleles for genes that code RBC surface antigens2. Some advantages of molecular-based methods for determining blood group phenotypes include automation with higher throughput; elimination of variation caused by serology platforms, including their inherent limitations on sensitivity and specificity of weakly expressed RBC antigens; and reduction of human errors due to inexperience and lack of competence. Molecular methods can also mitigate the cost of reagent antisera and problems associated with supply shortages.

DNA-based methods, such as BioArray, provide extended phenotype profiles at high throughput and opportunities to explore extended antigen matching of clinically significant antigens of the Rh, Kell, Duffy, and MNS blood group systems. The BioArray Solutions HEA (human erythrocyte antigen) BeadChip platform combines microparticle or bead chemistry with semiconductor processing, molecular biology for multiplex nucleic acid, and protein analysis to test for blood group polymorphism. Most blood group antigens are associated with single-nucleotide polymorphisms (SNP) with the exception of Rh, Duffy, and MNS blood group systems, which are associated with multiple SNPs. This technology allows for a direct genotype-to-phenotype prediction based on single- or multiple-nucleotide polymorphisms.

The Potential of Molecular-Based Testing: A Study from Mayo Clinic Rochester Transfusion Medicine

Study Design

The purpose of our initial study1 was to determine the degree of patient and donor matching by comparing the phenotypic distribution of Mayo Clinic blood donors, based on molecular analysis, with the published Rh, Kell, Kidd, Duffy, and MNS phenotypes of several ethnic groups. To review the complete study, see Badjie et al: Red blood cell phenotype matching for various ethnic groups, published in Immunohematology 20111. In this article, we focus on our findings in the Caucasian population.

After approval from the Mayo Clinic Institutional Review Board, molecular testing results of 1,000 blood donors were analyzed. Molecular testing was limited to blood group A and O donors to maximize blood product inventory. Likewise, because Rh- (negative) donors were selectively tested for inventory purposes, initial data was disproportionately skewed to a higher percentage (30%) of Rh- donors. The normal distribution in the general Caucasian population is 85% Rh+ (positive) and 15% Rh-. To correct for this bias, a subsample of 800 was randomly selected from the 1,000 samples and stratified by Rh, such that the resulting distribution would match the normal distribution. These 800 samples were evaluated for the phenotype distribution of Rh, Kell, Kidd, Duffy, and MNS, and the observed phenotypic distributions were compared to the distribution of the published Caucasian phenotypes.

Genomic DNA from peripheral blood samples of donors was extracted in the Mayo Clinic Tissue Typing Laboratory using Genom-6 DNA extraction machine. Molecular testing for 28 antigens from 11 blood group systems: Rh, Kell, Kidd, Duffy, MNS, Lutheran, Diego, Colton, Dombrock, Landsteiner-Wiener (LW), and Scianna, was performed using the BioArray (Immucor, Norcross, GA) BeadChip HEA. After analysis, the results of each donor antigen and phenotype profile for the most common antigens in the Rh, Kell, Kidd, Duffy, and MNS blood systems were collected and evaluated.

Blood System Antigen Mayo Clinic Donors (%) Published Data for Caucasians2,3 (%)
Rh D 68.2 85.0
  C 54.0 68.0
  E 25.0 29.0
  c 84.0 80.0
  e 97.0 98.0
Kell K 9.6 9.0
  k 99.4 99.8
Kidd Jka 81.6 77.0
  Jkb 70.4 74.0
Duffy Fya 64.8 66.0
  Fyb 82.0 83.0
MNS M 77.4 78.0
  N 71.4 72.0
  S 48.2 55.0
  s 92.0 89.0
Lutheran Lua 5.6 8.0
  Lub 100 99.8
Diego Dia 0.4 0.0
  Dib 99.8 100
Colton Coa 99.6 99.9
  Cob 9.0 10.0
Dombrock Doa 59.2 67.0
  Dob 85.6 82.0
  Joa 100 100
  Hy 100 100
LW LWa 100 100
  LWb 0.6 <1
Scianna Sc1 100 99.0
  Sc2 0.8 <1

Table 1. Antigen Frequencies of Rh, Kell, Kidd, Duffy, MNS, Lutheran, Diego, Colton, Dombrock, LW, and Scianna Blood Group Systems (n=1000)

Study Results

A summary of Mayo Clinic blood donor antigen frequencies, categorized by Rh, Kell, Kidd, Duffy, MNS, Lutheran, Diego, Colton, Dombrock, LW, and Scianna compared with those published for Caucasians is shown in Table 1. The frequencies of D and C antigens in Mayo Clinic donors are slightly lower than in the Caucasian population2,3 since the initial selection process of donors slightly favored a majority of Rh- donors.

In the Kell blood group system, the frequency of K and k are comparable to published frequency for Caucasians4. Likewise, the frequencies of the Kidd, Duffy, and MNS system antigens are also comparable to published data. There are also no significant differences between Mayo Clinic donors and the published antigen frequencies in the Caucasian population for the following blood group systems: Lutheran, LW, Diego, Dombrock, Colton, and Scianna.

A statistical summary of Rh phenotype frequencies of the Mayo Clinic blood donors from the 800 subsamples compared to published frequencies for Caucasians is shown in Table 2. With the exception of the DCCee phenotype (21.4%) showing a statistically significant difference (p<0.0001), the distribution of the Rh phenotypes in our donor pool is comparable to that of the general Caucasian population.5

Phenotype Mayo Clinic Donors (%) Published Data for Caucasians5 (%)
DCCee 21.4 16.0†
DCcEe 11.2 14.0*
DccEE 2.9 3.0*
Dccee 2.1 1.5*
DCcee 35.0 32.0*
DccEe 12.4 13.0*
dccee 13.7 15.0*
dCcee 0.5 0.4*
dccEe 0.8 0.2*

Table 2. Statistical Summary for the Rh Blood Group System (n=800)
p-values compare the percentage of each phenotype with the Mayo Clinic donor percentage
*p-value >0.01 (not significant)
†0.001 > p-value >0.0001

Phenotypic distribution for the Kell, Kidd, and Duffy blood group systems is shown in Table 3. The phenotype distribution in the Kell and Kidd systems is comparable between our donors and the published data for the Caucasians. However, for the Duffy blood group system, the distribution of Fy(a+b+) (44.5%), Fy(a+b–) (20.9%), and Fy(a–b–) (0.1%) phenotypes among our donors demonstrate statistical difference (p<0.0001) when compared to the published distribution in Caucasians.4

Phenotype Mayo Clinic Donors (%) Published Data for Caucasians4 (%)
K- k+ 90.4 91.0
K+ k- 0.4 0.2
K+ k+ 9.2 8.8
JK (a+ b-) 27.5 26.3
JK (a- b+) 20.9 23.4
JK (a+ b+) 51.6 50.3
FY (a+ b-) 20.9 17.0
FY (a- b+) 34.5 34.0
FY (a+ b+) 44.5 49.0
FY (a- b-) 0.1 0.0

Table 3. Statistical Summary for the Kell, Kidd, and Duffy Blood Group Systems (n=800)
p-values compare the percentage of each phenotype with the Mayo Clinic donor percentage
*p-value >0.01 (not significant)
†p-value <0.0001

A review of the results for the MNS blood group system (see Table 4) shows that the most common MNS phenotypes in our donors, MNs (23.5%),  MNSs (21.2%), Ns (15.9%), and MSs (15.0%) are comparable to the published data for the general Caucasian population.5

Phenotype Mayo Clinic Donors (%) Published Data for Caucasians5 (%)
MNSs 21.2 24.0*
MNS 3.1 4.0*
MNs 23.5 22.0*
MSs 15.0 14.0*
MS 6.8 6.0*
Ms 10.0 8.0*
NSs 4.2 6.0*
NS 0.3 1.0*
Ns 15.9 15.0*

Table 4. Statistical Summary for MNS Blood Group System (n=800)
p-values compare the percentage of each phenotype with the Mayo Clinic donor percentage
*p-value >0.01 (not significant)

Practice Implications

Mayo Clinic Rochester blood donors are predominantly Caucasian and our study demonstrates that, in most cases, phenotypes of this population accurately match the published phenotypic data. With that knowledge, we would potentially be able to identify extended phenotype profiles on all blood donors and create a database from dual donor-patient molecular analysis for comprehensive phenotypic profiling and matching to decrease alloimmunization due to donor-recipient antigen disparity. Certain patient populations, such as transplant patients, obstetric patients, and patients who rely on chronic blood transfusion could potentially benefit from individualized RBCs through extended-antigen matching. Use of extended-antigen matched blood could mitigate alloimmunization from blood transfusion due to the inherent disparity between donor-recipient antigen profiles.

One could assume that in most cases of Caucasian donor-recipient blood transfusion, an appropriate extended match could be provided. Additionally, molecular-based methods overcome the inherent limitations of serology platforms and enable transfusion medicine laboratories to quickly determine extended phenotypes of blood product recipients. In doing so, a recipient-donor match can quickly be determined and extended antigen-matched blood–beyond ABO and D–can be provided when necessary.


Molecular and genotype analyses offer many potential advantages over current hemagglutination methods. The characterization of donor phenotypes is a process that offers the feasibility of developing a computerized database of donor phenotypes, as well as maintaining an inventory of select donor phenotypic profiles that closely match certain patients. This information could result in faster turnaround time to obtain a blood recipient’s extended phenotype and provide compatible blood when a patient presents with unexpected antibodies.

Authored by: Karafa Badjie and Dr. James Stubbs


  1. Badjie KSW, Tauscher C, Van Buskirk C, et al: Red blood cell phenotype matching for various ethnic groups. Immunohematology 2011;27(1):12−19
  2. Roback JD, Combs MR, Grossman BJ, Hillyer CD: Technical Manual. Sixteenth edition. Bethesda, MD, AABB; 2008
  3. Reid ME, Lomas-Francis C: The Blood Group Antigen FactsBook. Second edition, New York, NY, Elsevier Academic Press, 2004
  4. Reid MR, Lomas-Francis C: Blood Group Antigens & Antibodies: A Guide to Clinical Relevance & Technical Tips, New York, NY,  Star Bright Books, 2007
  5. Harmening D: Modern Blood Banking and Transfusion Practices. Second edition Philadelphia, PA, FA Davis Company,1989,  pp 78−102

Back to top

Tell Us What You Think