Exeley Inc.

Bennett, Ashwini

Boyapati, Ray K.

Hong, Frank S.

IMMUNOHEMATOLOGY, Volume 31, Number 4, 2015 163
Suspected acute hemolytic transfusion 
reaction mediated by anti-Dia
A. Bennett, R.K. Boyapati, and F.S. Hong
Case report
Anti-Dia can mediate hemolytic disease of the fetus and newborn, 
but it is unclear if it can cause hemolytic transfusion reactions 
(HTRs). To date, there has only been one report of a possible 
immediate HTR attributed to anti-Dia. Our case report details an 
immediate HTR due to anti-Dia in a patient with pre-existing liver 
failure. This reaction triggered multi-organ failure, and the patient 
subsequently died. This case also highlights the importance of 
considering HTRs even when routine antibody screening has 
been unremarkable, particularly when electronic crossmatch is 
used, because of the potential for an alloantibody against a low-
prevalence antigen. Immunohematology 2015;31:163–165.
Key Words: Anti-Dia, transfusion reaction, low prevalence, 
alloantibodies
A 59-year-old Chilean woman, 18 years post-orthotopic 
liver transplant for primary biliary cirrhosis, was admitted to 
hospital with primary biliary cirrhosis recurrence manifest 
with decompensated liver failure (ascites, encephalopathy, 
jaundice, renal impairment, and coagulopathy), pancytopenia, 
and fever, for which she was empirically started on ceftriaxone 
and ampicillin. Her other medications included spironolactone, 
furosemide, mycophenolate mofetil, tacrolimus, lactulose, and 
ursodeoxycholic acid; azathioprine had recently been ceased 
because of pancytopenia.
Because of significant anemia (hemoglobin 6.2 g/dL), 
blood transfusion was arranged. Group and screen procedures 
revealed that the patient was blood group A, D+; no 
alloantibodies were identified on a standard three-cell 
screening panel (Abtectcell III 0.8%; CSL Biotherapies, 
Melbourne, Australia) using column agglutination technology 
(Diamed-ID gel card method; Diamed, Cressier, Switzerland). 
One unit of group A, D+, red blood cells (RBCs) was issued 
using electronic crossmatch.
Less than 1 hour into the RBC transfusion, the patient 
developed chills, fever (39.0°C), tachycardia (115 beats/min), 
hypertension (blood pressure 160/70 mmHg) and a non-
blanching purpuric rash over her trunk and legs, prompting 
immediate cessation of the transfusion. The fever subsequently 
resolved. Twelve hours later, the rash had worsened (fulminant 
retiform purpura) and was newly associated with tongue 
swelling. The patient was transferred to the intensive care 
unit for intubation, nebulized adrenaline, inotropic support, 
and continuous venovenous hemodialysis because of the 
development of anuria. Coagulation parameters worsened, 
including hypofibrinogenemia refractory to cryoprecipitate 
and fresh frozen plasma transfusions. Pertinent investigation 
results are listed in Table 1. Creatinine kinase peaked at 20,325 
U/L (reference interval 20–180 U/L) and haptoglobin fell to 
<10 mg/dL (reference interval 50–330 mg/dL). As part of the 
evaluation of the suspected transfusion reaction, the direct 
antiglobulin test (DAT) was performed post-transfusion using 
the gel card method (Diamed) with a reaction score of 2 (scale 
0–4) in a pre-transfusion sample but 3 (scale 0–4) in a post-
transfusion sample.
Results
Following the transfusion reaction, an indirect antiglobulin 
test (IAT) crossmatch between donor and recipient was found 
to be incompatible (reaction score 12 [scale 0–12]), suggestive 
of an alloantibody in the patient serum. Testing of a pre-
transfusion sample with a 3-cell antibody screening panel and 
an 11-cell antibody identification panel (Phenocell B 0.8%; 
CSL Biotherapies) was negative, suggesting an alloantibody 
against a low-prevalence antigen. Further testing of the pre-
transfusion sample with a different 11-cell antigen panel 
(Phenocell C 0.8%, CSL Biotherapies) revealed the presence of 
possible anti-Dia. This result was confirmed by the Australian 
Red Cross Blood Service Red Cell Reference Laboratory, 
which also performed phenotyping of the recipient (using pre-
transfusion sample) and donor. The recipient’s phenotype was 
group A; DCce; K–; Fy(a–b+); Jk(a+b–); M+N+S–s+; Di(a–). 
Anti-Dia was detected in pre- and post-transfusion samples 
by multiple methods, including saline 22°C and 37°C, low-
ionic IAT, polyethylene glycol-IAT, low-ionic-strength saline/
Coombs card (Diamed), and enzymes. No other alloantibodies 
were detected. Recipient DAT was positive because of IgG, as 
detected by tube, score 3 (scale 0–4) and Bio-Rad DAT card, 
score 10 (scale 0–12) (DC-Screening II, Diamed). The donor’s 
phenotype was group A, DcE; K–k+; Fy(a+b–); Jk(a+b+); 
M+N+S–s+; Di(a+b+); DAT was negative.
164 IMMUNOHEMATOLOGY, Volume 31, Number 4, 2015
The patient received five further RBC transfusions, of which 
the initial two were released following electronic crossmatch, 
with the remaining three units released after IAT crossmatch 
once anti-Dia was suspected. Of the two units released 
following electronic crossmatch, one donor was confirmed to 
be Di(a–). The other donor did not return so phenotyping was 
not possible. It was also not possible to retrospectively perform 
a crossmatch of this donor’s RBC unit with patient plasma due 
to donor sample unavailability. Donor typing had not been 
performed on the three IAT-crossmatched units because their 
compatibility indicated that they were antigen negative. None 
of these RBC transfusions caused any adverse effects. Despite 
active management, the patient continued to deteriorate and 
died 6 days following the implicated transfusion. The cause 
of death was thought to be underlying liver failure, with the 
hemolytic transfusion reaction (HTR) as an acute precipitant 
of multi-organ failure.
Discussion
This episode most likely represents an immediate 
acute HTR due to anti-Dia, based on clinical presentation 
and laboratory parameters such as increasing lactate 
dehydrogenase, falling haptoglobin, and positive DAT. 
Reticulocyte count remained inappropriately low due to bone 
marrow suppression secondary to liver failure and recent 
azathioprine. The hemolytic reaction was further complicated 
by progressive disseminated intravascular coagulation and 
worsening renal failure, which were likely contributed to by 
underlying liver failure. As anti-Dia was the only antibody 
present, we concluded that the HTR was triggered by anti-Dia. 
Other explanations for the reaction were considered but 
thought less likely. For example, sepsis was unlikely given 
that multiple blood cultures from the patient and returned 
donor unit were negative for bacterial growth. Polymerase 
chain reaction for cytomegalovirus, varicella zoster virus, and 
herpes simplex virus were negative. Epstein–Barr virus IgM 
was negative. Anaphylactic reaction was also unlikely given 
the lack of improvement to adrenaline, the patient’s normal 
serum IgA levels, and the fact that prior transfusions were not 
associated with reactions.
This patient may have been sensitized to Dia during 
pregnancy [noting her partner is also of South American 
origin, and therefore may be Di(a+)] or previous blood 
A. Bennett et al.
Table 1. Investigation results 
Day 0* Day 1 post-transfusion Day 2 post-transfusion Day 3 post-transfusion Reference interval
Hemoglobin (g/dL) 6.5 6.3 8.1 6.8 11.5–16.5
Lactate dehydrogenase 
(IU/L)
166 751 1693 — 135–214
Haptoglobin (mg/dL) 50† <10 <10 — 50–330
Reticulocytes (× 109/L) 20‡ 49 55 — 20–100
Direct antiglobulin test 
(IgG only)
2 3 — — Scale 0–4
Prothrombin time (s) 29 33 38 35 11–15
Activated partial 
thromboplastin time (s)
43 73 55 48 22–38
Fibrinogen Clauss (g/L) 1.7 1.0 <0.8 <0.8 2.0–4.0
Creatine kinase (U/L) 46§ 13,073 20,325 6673 20–180
Creatinine (mg/dL) 4.0 5.6 3.2¶ 2.0¶ 0.7–1.2
Bilirubin (mg/dL) 35.1 36.8 33.4 29.2 <1.1
Blood film Hypochromia with target 
cells, acanthocytes, and 
occasional fragments
Hyposplenism, mild 
polychromasia
Target cells, nucleated 
red blood cells, 
crenated cells, and 
schistocytes
Numerous nucleated 
red blood cells, 
polychromasia, target 
cells, fragments
—
RBC = red blood cell.
*Pre-transfusion sample on day of reaction, except where otherwise indicated.
†Day minus 19.
‡Day minus 20.
§Day minus 3.
¶Patient on continuous venovenous hemodialysis.
IMMUNOHEMATOLOGY, Volume 31, Number 4, 2015 165
Hemolytic reaction due to anti-Dia
transfusions (e.g., as part of the prior liver transplantation 
procedure). Given the known pattern of Dia antigen expression 
in the body, it is unlikely that the transplanted liver in itself 
was a stimulus. There was no history of hemolytic disease 
of the fetus and newborn (HDFN) complicating the patient’s 
prior pregnancies.
The Diego blood group was first discovered in 1955, 
during investigations into a case of HDFN.1,2 Dia is mostly 
found in Mongolian populations, 36 percent of South 
American Indians, approximately 12 percent of Japanese 
and Chinese individuals, and a smaller proportion of Polish 
people.3–6 Most Caucasians are Di(a–b+). The Diego protein 
(also known as the SLC4A1 protein) is expressed on RBCs 
and in renal tubules, where it plays an important role in RBC 
membrane stability and anion exchange. Anti-Dia is known 
to mediate HDFN.4,6–8 To our knowledge, there has only been 
one other report of an immediate HTR attributed to anti-Dia.9 
In that case, the patient developed fevers, flank pain, and 
hemoglobinuria, which self-resolved. There has also been one 
case report of a delayed transfusion reaction due to multiple 
alloantibodies (including antibodies to E, Jka, Dia, Fyb, and 
S), although in that case, the clinical presentation could not 
be attributed to anti-Dia alone, because of the presence of the 
other alloantibodies.10
Because the standard commercial RBC antibody testing 
panels in Australia are prepared from donor RBCs, and 
the donors are predominantly Caucasian, antibodies to 
low-prevalence antigens such as Dia may not be detected 
during routine screening. In combination with electronic 
crossmatching, this can expose patients to risks of hemolysis 
from undetected alloantibodies. It is important, therefore, for 
clinicians and laboratory scientists to consider the possibility 
of HTR even when no antibodies are detected on screening 
and the RBCs are issued using electronic crossmatch.
Acknowledgments
The authors would like to thank the staff at the Australian 
Red Cross Blood Service Red Cell Reference Laboratory for 
their assistance in evaluating the implicated RBC antibody.
References
 1. Levine P, Koch E, McGee R, et al. Rare human isoagglutinins 
and their identification. Am J Clin Path 1954;24:292–304.
 2. Levine P, Robinson E, Layrisse M, et al. The Diego blood factor. 
Nature 1956;177:40–1.
 3. Dean L. Blood Groups and Red Cell Antigens [Internet]. 
Bethesda (MD): National Center for Biotechnology Information 
(US); 2005. Chapter 11, The Diego blood group. Available from: 
http://www.ncbi.nlm.nih.gov/books/NBK2273/
 4. Kusnierz-Alejska G, Bochenek S. Haemolytic disease of the 
newborn due to anti-Dia and incidence of the Dia antigen in 
Poland. Vox Sang 1992;62:124–6.
 5. Komatsu F, Hasegawa K, Yanagisawa Y, et al. Prevalence of 
Diego blood group Dia antigen in Mongolians: comparison 
with that in Japanese. Transfus Apheresis Sci 2004;30: 
119–24.
 6. Lenkiewicz B, Zupanska B. The first example of anti-Diego(b) 
found in a Polish woman with the Di(a+b–) phenotype and 
haemolytic disease of the newborn not requiring treatment. 
Transfus Med 2003;13:161–3.
 7. Alves de Lima LM, Berthier ME, Sad WE, et al. Characterization 
of an anti-Dia antibody causing haemolytic disease in a 
newborn infant. Transfusion 1982;22:246–7.
 8. Ting JY, Ma ES, Wong KY. A case of severe haemolytic disease 
of the newborn due to anti-Di(a) antibody. Hong Kong Med J 
2004;10:347–9.
 9. Hinckley ME, Huestis DW. An immediate haemolytic 
transfusion reaction apparently caused by anti-Dia. Rev Fr 
Transfus Immunohematol 1979;22:581–5.
 10. Yamane K, Yagihashi A, Sasaki M, et al. A delayed haemolytic 
transfusion reaction (DHTR) with multiple alloantibodies 
(Anti-E, Jka, Dia, Fyb, and S) induced by E-antigen 
negative, crossmatch-compatible blood. Immunopharmacol 
Immunotoxicol 1998;20:531–9.
Ashwini Bennett, MBBS, Haematologist (corresponding author), 
Haematology Department, Monash Medical Centre, 246 Clayton 
Road, Clayton, Victoria 3168, Australia, ashwini.bennett@gmail.
com; Ray K. Boyapati, MBBS, Clinical Research Fellow, Department 
of Gastroenterology, Western General Hospital, Edinburgh, 
Scotland; and Frank S. Hong, MBBS, Haematologist, Austin 
Pathology, Austin Hospital, Heidelberg, Victoria, Australia.


Suspected acute hemolytic transfusion reaction mediated by anti-Dia

Erratum

Barron, Christina

IMMUNOHEMATOLOGY, Volume 30, Number 4, 2014 153
Allogeneic red blood cell adsorption for 
removal of warm autoantibody
C. Barron
Adsorption studies are usually required to confirm or rule out 
the presence of underlying alloantibodies in samples containing 
warm autoantibody. Allogeneic adsorptions are necessary if the 
patient has been recently transfused. Most commonly, allogeneic 
adsorptions are performed using a trio of phenotyped reagent 
red blood cells to rule out clinically significant alloantibodies to 
common antigens. The adsorbing cells may be used untreated 
or treated with enzymes or with ZZAP before adsorption. 
Adsorption may also be performed using enhancement such 
as low-ionic strength saline or polyethylene glycol added to 
the mixture. Multiple adsorptions may be necessary to remove 
strongly reactive autoantibodies. Allogeneic adsorptions 
will not detect alloantibodies to high-prevalence antigens. 
Immunohematology 2014;30:153–155.
Key Words: AIHA, adsorption, autoantibody, alloantibody, 
allogeneic RBCs
The serum from patients with warm autoimmune 
hemolytic anemia (WAIHA) typically demonstrates broad 
reactivity with all reagent red blood cells (RBCs). This 
reactivity usually necessitates the use of incompatible blood for 
transfusion. When blood is incompatible as a result of an RBC 
autoantibody, transfused RBCs will survive at a rate similar 
to that of the patient’s own RBCs.1 However, autoantibody 
reactivity may mask the reactions of underlying clinically 
significant alloantibodies to common antigens, which can 
cause rapid RBC destruction. The rate of alloimmunization in 
patients with warm autoantibodies varies among investigators. 
Reported alloimmunization rates range from 17.5 to 38 percent, 
with an average rate of 32 percent.2–7 Issitt et al.4 found that 
47 percent of alloantibody specificities identified after 
adsorption were actually partially adsorbed autoantibodies 
with mimicking specificities. Mimicking autoantibody 
specificities were found most often when autologous adsorption 
was used to remove autoantibody reactivity. Identification of 
alloantibody specificities underlying a warm autoantibody is 
critical for a safe and effective transfusion.
Methods used to determine the presence of an alloantibody 
underlying an autoantibody include serum dilution, autologous 
adsorption, and allogeneic adsorption. The dilution screening 
method uses a 1:5 dilution of serum in phosphate-buffered 
saline. The diluted serum is tested to determine whether the 
autoantibody has been removed and whether alloantibody 
is present.8 Although this may be useful as a screening tool, 
this method has been shown to be unreliable in consistently 
detecting alloantibody underlying autoantibody.9 Autologous 
adsorption, which uses the patient’s own RBCs to remove 
the autoantibody, is a preferred method. Using autologous 
RBCs ensures that only autoantibodies are removed from the 
serum, leaving intact any alloantibody the patient may have 
produced, including antibodies to high-prevalence antigens. 
However, autologous adsorptions may only be performed 
when the patient has not been recently transfused or pregnant, 
Reagents/Supplies
Treatment 
Method Reagents Supplies
Papain • 1% cysteine-activated 
papain
• Isotonic 
saline
• Allogeneic 
RBCs
• Glycine max 
var. soja
• Pipettes
• 37°C water 
bath
• Large-bore 
test tubes
• Calibrated 
centrifuge
Ficin • 1% ficin
ZZAP • 1% cysteine-activated 
papain or 1% ficin
• 0.2 M DTT
• Commercial LISS 
enhancement reagent
• 20% PEG or commercial 
PEG enhancement
DTT = dithiothreitol; RBCs = red blood cells; LISS = low-ionic-strength 
saline; PEG = polyethylene glycol.
Procedural Steps
Preparation of RBCs for adsorption
• Select reagent RBCs for use as adsorbing cells.
• Treat the adsorbing cells with the desired enzyme or ZZAP.
Adsorption
• Add 1 to 2 volumes of treated adsorbing cells to 1 volume of 
patient serum or plasma and 1 volume of enhancement reagent (if 
applicable).
• Incubate for 30–60 minutes at 37°C without enhancement, or 15 
minutes at 37°C with LISS or PEG enhancement.
• Harvest the adsorbed serum or plasma and test for the presence of 
underlying alloantibodies.
RBCs = red blood cells.
S
E
R
O
L
O
G
IC
 M
E
T
H
O
D
 R
E
V
IE
W
154 IMMUNOHEMATOLOGY, Volume 30, Number 4, 2014
C. Barron
commonly defined as having no transfusions or pregnancy in 
the prior 3 months. Alloantibody may be removed from the 
test system if autologous adsorptions are performed using 
a sample from a recently transfused patient. Laine et al.10 
showed that even small amounts of transfused RBCs are 
capable of removing all alloantibody reactivity. Allogeneic 
RBC adsorption is the method of choice for removal of warm 
autoantibody reactivity in patients with recent transfusions. 
Note that the term “serum” is used to denote the patient sample 
throughout this article; however, either serum or plasma may 
be used for testing.
Principle
Allogeneic adsorptions are used to remove broadly reactive 
autoantibody from a patient’s serum to allow for detection of 
underlying clinically significant alloantibodies to common 
antigens. The adsorbing cells will remove autoantibody along 
with any additional antibody specificities to antigens that are 
present on the RBC. When performing allogeneic adsorptions, 
the goal is to eliminate or confirm the presence of those 
alloantibodies directed at the common antigens D, C, E, c, e, 
K, S, s, Fya, Fyb, Jka, and Jkb.
Traditional allogeneic adsorptions use a set of three 
adsorbing cells with known phenotypes: R1R1, R2R2, and rr. 
At least one of the adsorbing cells should be negative for K and 
Jka or Jkb. In addition, at least one of the reagent cells should 
also lack S or s and Fya or Fyb unless the cells are pretreated 
to denature the antigens. A separate aliquot of patient serum 
is incubated with each adsorbing cell. Once all autoantibody 
reactivity is removed, the three adsorbed serums are tested to 
eliminate or prove the presence of any underlying alloantibody. 
Each aliquot of adsorbed serum contains alloantibody 
corresponding to those antigens for which the adsorbing cell 
was negative.
If a patient’s RBC phenotype is known, it may be possible 
to adsorb using a single phenotypically similar reagent RBC. 
For instance, if a patient is only negative for E, K, S, and Fya, 
one reagent RBC negative for those antigens can be used for 
adsorption. After adsorption, the adsorbed serum can be 
tested to confirm or eliminate the presence of anti-E, -K, -S, 
and/or -Fya.
Various methods are used for performing the adsorption. 
Treatment of the adsorbing cells using enzymes or ZZAP may 
enhance the uptake of the autoantibody.9 These cell treat- 
ments will also destroy some antigens, which may be 
advantageous when selecting cells for adsorption and testing 
the adsorbed serum. Rarely, autoantibodies may demonstrate 
specificity corresponding to an antigen that is destroyed by 
enzymes (e.g., anti-Ge2) or ZZAP (e.g., anti-LWa or Kell system 
antibodies). In these cases, adsorption with untreated reagent 
RBCs is necessary.
The addition of enhancement media to the adsorbing 
mixture has also become commonplace and can be used in 
place of RBC pretreatment. The addition of low-ionic-strength 
saline (LISS) or polyethylene glycol (PEG) reduces the 
processing time and may decrease the number of adsorptions 
required to remove the autoantibody.9,11,12 However, some 
investigators caution that some underlying alloantibodies may 
not be detected when PEG adsorptions are performed.13,14
Indications
Allogeneic adsorptions should be used when a patient has 
been recently transfused or when autologous cells are of limited 
supply. Adsorption eliminates the reactivity of the warm 
autoantibody, allowing for the identification of alloantibodies 
in the adsorbed serum.
Procedure
Select cells for adsorption. If the patient’s RBC phenotype is 
known, select a phenotypically similar reagent RBC that lacks 
the same common RBC antigens as the patient’s RBCs. If the 
patient’s RBC phenotype is unknown, select a trio of reagent 
RBCs: R1R1, R2R2, and rr. Additionally, at least one cell of the 
set should be negative for K and Jka or Jkb so that all common 
alloantibodies can be detected using the adsorbed serum. If 
adsorptions will be performed using untreated reagent RBCs, 
at least one cell should be negative for S or s and Fya or Fyb as 
well.15–18
If desired, treat the adsorbing cells with enzymes (ficin 
or papain) or ZZAP (ficin or papain and 0.2 M DTT). This 
pretreatment will typically enhance the adsorbing process.
Following pretreatment, wash the adsorbing cells, 
centrifuge, and remove all excess saline before use. Perform 
the adsorption by mixing one to two volumes of allogeneic 
RBCs with one volume of patient serum at 37°C for 30 to 60 
minutes. 
Another option is to perform the adsorption by mixing 
one volume of patient serum and one volume of LISS or PEG19 
with one to two volumes of allogeneic reagent RBCs at 37°C 
for 15 minutes.15–18
After the adsorption, centrifuge the mixture and harvest 
the adsorbed serum. Test the adsorbed serum to determine 
whether adsorption is complete. This may be accomplished 
IMMUNOHEMATOLOGY, Volume 30, Number 4, 2014 155
Allogeneic RBC adsorption for autoantibody removal
by testing the adsorbed serum against the adsorbing cells 
or by testing the adsorbed serum against a reagent RBC 
phenotypically similar to the adsorbing cell. If necessary, 
repeat the adsorption using a new aliquot of adsorbing cells.
Once adsorption is complete, test the adsorbed serum 
against reagent RBCs to rule out the presence of underlying 
alloantibodies. For instance, if the adsorbing cell is R1R1, S–, 
K– Jk(a–), then the following alloantibodies can be ruled out 
in that aliquot of adsorbed serum; anti-E, -c, -S, -K, and -Jka. 
Note: Test the adsorbed serum from the LISS or PEG 
enhancements using four drops of the serum/enhancement 
reagent mixture.
Limitations
Allogeneic adsorption will not rule out the presence of 
alloantibodies to high-prevalence antigens. The adsorbing cells 
are presumed to be positive for all high-prevalence antigens; 
therefore, an alloantibody to a high-prevalence antigen would 
be adsorbed along with the autoantibody.
When a patient’s serum contains a very potent warm 
autoantibody, multiple adsorptions are needed to remove the 
broad reactivity. Generally, for each grade of reactivity, at least 
one adsorption will be necessary (a serum sample with an 
autoantibody that reacts 2+ will require a minimum of two 
adsorptions). Multiple adsorptions have the potential to dilute 
the serum. Increasing the ratio of adsorbing cells to serum 
may enhance the removal of the autoantibody.
The use of PEG adsorption may cause weakening or loss 
of antibody reactivity in some samples. The use of six drops of 
the serum/PEG mixture in antibody identification tests may 
minimize this effect.
Quality Control
To determine whether the adsorbing cells are adequately 
treated with enzymes, test the cells with Glycine max var. soja 
lectin. Enzyme-treated RBCs should demonstrate positive 
reactivity with G. max var. soja and untreated RBCs should 
be nonreactive.
References
 1. Petz LD. A physician’s guide to transfusion in autoimmune 
haemolytic anaemia. Br J Haematol 2004;124:712–16.
 2. Wallhermfechtel MA, Pohl BA, Chaplin H. Alloimmunization 
in patients with warm autoantibodies. Transfusion 1984;24: 
482–5.
 3. Laine ML, Beattie KM. Frequency of alloantibodies 
accompanying autoantibodies. Transfusion 1985;25:545–6.
 4. Issitt PD, Combs MR, Bumgarner DJ, Allen J, Kirkland 
A, Melroy-Carawan H. Studies of antibodies in the sera of 
patients who have made red cell autoantibodies. Transfusion 
1996;36:481–6.
 5. Laine EP, Leger RM, Arndt PA, Calhoun L, Garratty G, 
Petz LD. In vitro studies of the impact of transfusion on the 
detection of alloantibodies after autoadsorption. Transfusion 
2000;40:1384–7.
 6. Young PP, Uzieblo A, Trulock E, Lublin DM, Goodnough LT. 
Autoantibody formation after alloimmunization: are blood 
transfusions a risk factor for autoimmune hemolytic anemia? 
Transfusion 2004;44:67–72.
 7. Maley M, Bruce DG, Babb RG, Wells AW, Williams M. The 
incidence of red cell alloantibodies underlying panreactive 
warm autoantibodies. Immunohematology 2005;21:122–5.
 8. Øyen R, Angeles ML. A simple screening method to evaluate 
the presence of alloantibodies with concomitant warm 
autoantibodies. Immunohematology 1995;11:85–7.
 9. Leger RM, Garratty G. Evaluation of methods for detecting 
alloantibodies underlying warm autoantibodies. Transfusion 
1999;39:11–16.
 10. Laine EP, Leger RM, Arndt PA, Calhoun L, Garratty G, 
Petz LD. In vitro studies of the impact of transfusion on the 
detection of alloantibodies after autoadsorption. Transfusion 
2000;40:1384–7.
 11. Chiaroni J, Tousinssi M, Mazet M, De Micco P, Ferrera V. 
Adsorption of autoantibodies in the presence of LISS to detect 
alloantibodies underlying warm autoantibodies. Transfusion 
2003;43:651–5.
 12. Cheng CK, Wong M, Lee AW. PEG adsorption of autoantibodies 
and detection of alloantibodies in warm autoimmune hemolytic 
anemia. Transfusion 2001;41:13–17.
 13. Barron CL, Brown MB. The use of polyethylene glycol (PEG) 
to enhance adsorption of autoantibodies. Immunohematology 
1997;13:119–22.
 14. Judd WJ, Dake L. PEG adsorption of autoantibody causes loss 
of concomitant alloantibody. Immunohematology 2001;17: 
82–5.
 15. Judd J, Johnson ST, Storry JR. Judd’s methods in 
immunohematology. 3rd ed. Bethesda, MD: American 
Association of Blood Banks, 2008;469–72.
 16. Fung MK. Technical manual. 18th ed. Chapter 17: Positive DAT 
and immune-mediated hemolysis. Bethesda, MD: American 
Association of Blood Banks, 2014;425–51.
 17. Beatie K, Crawford M, Martin J, et al. Immunohematology 
methods. Rockville, MD: American Red Cross, 1993;47–51.
 18. Fung MK. Technical manual. 18th ed. Method 4–9: Adsorbing 
warm-reactive autoantibodies using allogeneic red cells. 
Bethesda, MD: American Association of Blood Banks, 2014.
 19. Fung MK. Technical manual. 18th ed. Method 4–10: 
Polyethylene glycol adsorption procedure. Bethesda, MD: 
American Association of Blood Banks, 2014.
Christina Barron, MT(ASCP)SBB, Manager, Immunohematology 
Reference Laboratory, American Red Cross, Missouri-Illinois 
Region, 4050 Lindell, St. Louis, MO 63108.


Allogeneic red blood cell adsorption for removal of warm autoantibody

Billingsley, Katrina L.

Posadas, Jeff B.

Moulds, Joann M.

Gaur, Lakshmi K.

IMMUNOHEMATOLOGY, Volume 29, Number 4, 2013 145
A novel JK null allele associated with typing 
discrepancies among African Americans
K.L. Billingsley, J.B. Posadas, J.M. Moulds, and L.K. Gaur
repOrt
The Jknull (Jk-3) phenotype, attributable to null or silenced 
alleles, has predominantly been found in persons of Polynesian 
descent. With the increased use of molecular genotyping, many 
new silencing mutations have been identified in persons of other 
ethnic backgrounds. To date, only two JK null alleles have been 
reported in African Americans, JK*01N.04 and JK*01N.05. 
A comparative study was undertaken to determine whether 
JK mutations were present in the regional African American 
population. Results of donor genotyping were compared with 
previously recorded results of serologic tests, and discrepant 
results were investigated. Although the two previously identified 
polymorphisms were not detected in the discrepant samples, a 
novel allele (191G>A) was identified and was assigned the ISBT 
number JK*02N.09. This study illustrates a limitation of using 
single-nucleotide polymorphisms for prediction of blood group 
antigens. Immunohematology 2013;29:145–148.
Key Words: blood group, null allele, Jk antigens, Jk:–3
The Kidd blood group system was first described in 1951 
when anti-Jka was found to cause hemolytic disease of the 
fetus and newborn (HDFN).1 Anti-Jkb was reported in 1953, 
also as a case of HDFN,2 and in 1959, anti-Jk3 was described 
in a posttransfusion patient whose red blood cells (RBCs) 
typed Jk(a–b–).3 Since this time, many examples of anti-Jka 
and anti-Jkb have been reported. Owing to the rapid decrease 
in antibody titer, these clinically significant antibodies can be 
difficult to detect and are frequently associated with severe 
delayed hemolytic transfusion reactions. In fact, fatal delayed 
hemolytic transfusion reactions caused by anti-Jka and anti-
Jkb remain among those most reported to the U.S. Food and 
Drug Administration.4
The Kidd glycoprotein is found on RBCs as well as on 
endothelial cells of the vasa recta, the vascular supply of the 
renal medulla. It is an integral protein of the RBC, spanning 
the membrane ten times. The glycoprotein functions as a 
urea transporter for the cells and helps maintain osmotic 
stability. Individuals lacking the glycoprotein are known to 
have decreased urine concentrating ability, but they have no 
known disease associations or RBC abnormalities. The Kidd 
glycoprotein consists of 389 amino acids and is encoded by 
the SLC14A1 gene located on chromosome 18. Eleven exons 
are distributed across 30 kb of DNA. As exons 1 through 3 
are not translated, the mature protein is encoded by exons 4 
through 11.5 The SLC14A1 gene has two major codominant 
alleles, JK*A (JK*01) and JK*B (JK*02), which result from a 
single-nucleotide polymorphism (SNP), 838A>G (rs1058396).
To date, 14 polymorphisms causing JKnull phenotypes 
have been reported: 6 on the JK*A allele and 8 on the JK*B 
allele (Table 1). Of the known polymorphisms, only two 
have been associated with individuals of African American 
ethnicity (JK*01N.04 and JK*01N.05).6,7 Investigation into 
samples whose genotype results for JK*A and JK*B differed 
from results previously obtained with serologic testing led to 
the identification of an additional mutation that resulted in the 
lack of expression of the JK protein.9
Materials and Methods
Sample Selection
After informed consent was obtained, whole-blood 
samples from African American blood donors and patients 
from the tristate service area (Louisiana, Texas, and Arkansas) 
were obtained for testing. Genomic DNA was extracted and 
prepared using either a manual or an automated method 
(Gentra PureGene or QIAcube, Qiagen, Valencia, CA).
Serologic Studies
All of the discrepant donor samples were tested either on 
multiple donations or, if confirmation testing was performed 
on the same sample, with multiple antisera. Of the three 
donors with genotype or phenotype discrepancies, donor 
one was tested on two different donations using a polyclonal 
reagent (Anti-Jkb, Immucor, Norcross, GA). Donor two was 
tested twice on the same donation: once with a polyclonal 
reagent (anti-Jkb, Immucor) and once with an in-house human 
source of anti-Jkb. Finally, donor three was tested twice, first 
with a polyclonal serum (anti-Jkb Immucor) and then with a 
monoclonal reagent (Anti-Jkb, Ortho Clinical Diagnostics, 
Raritan, NJ). No weakened expression was observed with 
any of the antisera; in fact, all serologic results were negative. 
As for the patient, because his discrepancy was discovered 
on a retrospective genotype and phenotype comparison of 
146 IMMUNOHEMATOLOGY, Volume 29, Number 4, 2013
results from 250 patients with sickle cell disease, testing was 
not confirmed by additional serology. The patient’s complete 
phenotype was performed as part of the immunohematology 
reference laboratory (IRL) serologic workup that led to the 
identification of alloanti-Jkb.
Molecular Assays
RBC blood group genotyping was performed using a 
microarray kit (HEA BeadChip assay version 1.2, BioArray 
Solutions, LLC, Warren, NJ) according to the manufacturer’s 
directions. This microarray detects the 838G>A polymorphism 
corresponding to JK*A or JK*B, as well as 23 additional 
polymorphisms associated with 34 RBC antigens. Genotype 
results were compared with previous serologic typing results 
for concordance. For samples with discordant results, the 
genotype was confirmed using another kit (Red Cell EZ Type 
KDK assay, GTI Diagnostics, Milwaukee, WI), according to the 
manufacturer’s directions. Samples that remained discordant 
were sent to Puget Sound Blood Center for DNA sequencing. 
The coding regions of the JK genes (exons 4–10) along with 
the promoter region6 were enzymatically amplified in a total 
volume of 50 µL with 100 ng of DNA, 1.25 units of  polymerase 
(Taq, 1× Green GoTaq Flexi Buffer, Promega, Madison, WI), 
1.5 mM MgCl2, 0.125 mM deoxyribonucleoside diphosphate, 
and 100 ng each forward and reverse primer. Polymerase chain 
reaction conditions included denaturation for 2 minutes at 
95°C, followed by 35 cycles of 95°C for 30 seconds, 55°C for 30 
seconds, and 72°C for 1 minute, and then a final elongation for 
5 minutes at 72°C. Polymerase chain reaction products were 
purified with a kit (QIAquick PCR purification kit, Qiagen) and 
sequenced on an automated system (ABI Prism 3100 Applied 
Biosystems, Foster City, CA). The amplified products were 
sequenced (ABI Prism 3100 Applied Biosystems), they were 
aligned to the reference JK allele, and the electropherograms 
were scanned for SNPs. The sequences were aligned in FASTA 
format using a software program (BLAST, http://blast.ncbi.
nlm.nih.gov/).
Results
Data were collected from January 1, 2009, through 
September 30, 2010. Identified during the 21-month study 
period were 1079 genotyped donors who had been previously 
typed for Jka and 1235 genotyped donors who had been 
previously typed for Jkb by serologic methods. On comparison 
of serologic results with molecular typing results, three 
samples were found to be discordant. All three were 
heterozygous for the 838G>A SNP and were predicted to be 
Jk(a+b+); however, the samples typed Jk(b–) by serologic 
methods.
DNA sequencing of the three samples indicated that all 
had wild-type promoter regions and confirmed heterozygosity 
at 838G>A. However, a 191G>A (R64Q) polymorphism 
was also present in the heterozygous state (Fig. 1). Given 
the serologic phenotype in the three donors, it is most likely 
that the 191G>A polymorphism is carried on the JK*B 
allele. Further, this missense SNP is proposed to result in 
no or undetectable levels of Jkb on the RBC surface. Further 
Table 1. International Society for Blood Transfusion terminology for Kidd blood group alleles*
Phenotype Allele name Nucleotide change Intron/exon Amino acid change Reference number
JK:–3 or Jk(a–b–) JK*01N.01 Exons 4 and 5 deleted 4 and 5 Initiation Met absent
JK:–3 or Jk(a–b–) JK*01N.02 202C>T 5 Gln68Stop
JK:–3 or Jk(a–b–) JK*01N.03 582C>G 7 Tyr194Stop
JK:–3 or Jk(a–b–) JK*01N.04 956C>T 10 Thr319Met 6
JK:–3 or Jk(a–b–) JK*01N.05 561C>A 7 Tyr187Stop 7
JK:–3 or Jk(a–b–) JK*01N.06 IVS5–1g>a Intron 5 Exon 6 skipped; in frame
JK:–3 or Jk(a–b–) JK*02N.01 IVS5–1g>a Intron 5 Exon 6 skipped; in frame 8
JK:–3 or Jk(a–b–) JK*02N.02 IVS5–1g>c Intron 5 Exon 6 skipped; in frame
JK:–3 or Jk(a–b–) JK*02N.03 222C>A 5 Asn74Lys
JK:–3 or Jk(a–b–) JK*02N.04 IVS7+1g>t Intron 7 Exon 7 skipped; frameshift➞Leu223Stop
JK:–3 or Jk(a–b–) JK*02N.05 723delA 8 Frameshift➞Ile262Stop
JK:–3 or Jk(a–b–) JK*02N.06 871T>C 9 Ser291Pro 8
JK:–3 or Jk(a–b–) JK*02N.07 896G>A 9 Gly299Glu
JK:–3 or Jk(a–b–) JK*02N.08 956C>T 10 Thr319Met
JK:–3 or Jk(a–b–) JK*02N.09 191G>A 4 Arg64Gln 9, this study
*Reference allele JK*02 encodes JK2, JK3. Source: http://www.isbtweb.org/working-parties/red-cell-immunogenetics-and-blood-group-terminology/blood-
group-terminology/.
K.L. Billingsley et al.
IMMUNOHEMATOLOGY, Volume 29, Number 4, 2013 147
investigation, including cDNA analysis and protein modeling, 
would be necessary to demonstrate these interpretations. The 
allele has been assigned the ISBT name JK*02N.09. 
A retrospective review of patient testing also yielded a 
discrepancy between serology and molecular typing. Of the 
250 patient records reviewed, one patient genotyped JK*A/
JK*B, but phenotyped Jk(b–) and had produced an alloanti-
Jkb. The DNA sample was sequenced, and the patient was 
confirmed to carry the 191G>A substitution.
This previously unpublished SNP has thus far only been 
detected in Asians (Gaur LK, unpublished data) using a primer 
set JK-E3-FP1: 5′ACAGCCCCACTATGGTTAGA3′ and JK-
E4-RP1: 5′CTGGCTGAGCAAGAGGGC3′ that encompasses 
codons 10 to 111 (872 bp); so this is the first report of this 
polymorphism in African Americans. 
Discussion
Although no longer in its infancy, the role of molecular 
technologies in the blood bank environment continues 
to evolve. With the advent of high-throughput molecular 
assays that are often more economical and efficient than 
routine serology, donor screening has been revolutionized.10 
Although some microarray assays have incorporated JK null 
allele detection (IDCore+, Progenika, Derio, Spain), these are 
usually limited to the more common Polynesian and Finnish 
alleles, JK*02N.01 and JK*02N.06, respectively.9 Two JK nulls 
(JK*01N.04 and JK*01N.05) have previously been found in 
African Americans, but their occurrence is rare.6,7 However, 
this new allele, JK*02N.09, occurred in 1 of 412 donors tested. 
As shown by this study, antigen screening that uses only SNP 
analysis may not detect blood group mutations caused by 
silencing mutations and may lead to erroneous testing results. 
Of note, this method of donor screening will not increase 
the risk to alloimmunized patients, as a unit would appear 
to be positive for the antigen and, therefore, not selected for 
transfusion to a patient with a known antibody. However, 
there is a risk of mistyping patients, and if a patient’s genotype 
will be used for prophylactic antigen matching, as in the case 
of sickle cell disease or warm autoimmune hemolytic anemia, 
limitations of molecular assays must be considered.
Given the prevalence of the 191G>A SNP in the study 
population, a comparison of previously typed patients with 
sickle cell disease was conducted, and indeed the new JK 
allele was detected in one patient who had produced alloanti-
Jkb. This knowledge has led to a change in LifeShare Blood 
Centers protocol, such that any African American patient who 
genotypes as JK*A/JK*B will have the phenotype confirmed 
by serology whenever possible.
African Americans are known to have silencing mutations 
in other blood group genes, e.g., FY and GYPB, and detection of 
these has been incorporated into existing microarrays. Perhaps 
as molecular assays continue to evolve, detection of additional 
mutations leading to JK nulls will be incorporated as well. 
More data as to the frequency of the 191G>A polymorphism 
in African Americans in other geographic locations may be of 
interest. Additionally, given that anti-Jka and anti-Jkb can be 
weakly reactive as well as show a dosage effect, those donors 
used for reagent RBC panels should be confirmed as truly 
homozygous for the JK*A or JK*B gene. As with many other 
blood group systems, what once appeared to be a simple blood 
group system with two common antigens and one rare null 
type has increasingly become more complex.
Acknowledgment
This work was supported in part by funds from the U.S. 
Armed Forces (USAMRMC-BAA05-01) to L.K.G.
JK null alleles in African Americans
Fig. 1 Electropherogram of the DNA sequence for exon 4 of the JK 
gene showing the silencing mutation 191G>A.
C C N
191G>A, R64Q
G G
C C N G G
148 IMMUNOHEMATOLOGY, Volume 29, Number 4, 2013
References
 1. Allen FH Jr., Diamond LK, Niedziela B. A new blood-group 
antigen (letter). Nature 1951;167:482.
 2. Plaut G, Ikin EW, Mourant AE, Sanger R, Race RR. A new 
blood-group antibody, anti-Jkb. Nature 1953;171:431.
 3. Pinkerton FJ, Mermod LE, Liles BA, Jack JA Jr., Noades J. The 
phenotype Jk(a–b–) in the Kidd blood group system. Vox Sang 
1959;4:155–60.
 4. U.S. Food and Drug Administration. Fatalities Reported 
to FDA Following Blood Collection and Transfusion: 
Annual Summary for Fiscal Year 2008. Available at: http://
www.fda.gov/BiologicsBloodVaccines/SafetyAvailability/
ReportaProblem/TransfusionDonationFatalities/ucm113649.
htm. Accessed October 25, 2013.
 5. Reid ME, Lomas-Francis C. The blood group antigen factsbook. 
2nd ed. San Diego, CA: Academic Press, 2004.
 6. Wester ES, Johnson ST, Copeland T, et al. Erythroid urea 
transporter deficiency due to novel JK null alleles. Transfusion 
2008;48:365–72.
 7. Horn T, Castilho L, Moulds JM, et al. A novel JKA allele, 
nt561C>A, associated with silencing of Kidd expression. 
Transfusion 2012;52:1092–6.
 8. Posadas J, Shnyreva M, Gaur P, et al. Distribution of JK silent 
alleles in Asian American and Pacific Islander populations 
(abstract). Transfusion 2010;50(Suppl):143A.
 9. Billingsley K, Posasdas JB, Gaur LK, Moulds JM. Identification 
of a new JK silencing mutation in the African American 
population (abstract). Transfusion 2011;51(Suppl):150A.
 10. BeadChip molecular immunohematology. Moulds JM, Ness 
PM, Sloan SR, eds. New York, NY: Springer, 2010.
Katrina L. Billingsley, MT(ASCP)SBB (corresponding author), 
Supervisor, Clinical Immunogenetics, Scientific Support Services, 
LifeShare Blood Centers, 8910 Linwood Avenue, Shreveport, LA 
71106; Jeff B. Posadas, BS, Research Technologist II, Puget Sound 
Blood Center, Seattle, WA; Joann M. Moulds, PhD, MT(ASCP)SBB, 
Director, Scientific Support Services, LifeShare Blood Centers, 
Shreveport, LA; and Lakshmi K. Gaur, PhD, Senior Scientist, Puget 
Sound Blood Center, Seattle, WA.
K.L. Billingsley et al.


A novel JK null allele associated with typing discrepancies among African Americans

Curvers, Joyce

Scharnhorst, Volkher

de Haas, Masja

Warnier-Wandel, Loes

van de Kerkhof, Daan

IMMUNOHEMATOLOGY, Volume 28, Number 3, 2012 85
Blood group genotyping in a multitrauma 
patient: a case report
J. Curvers, V. Scharnhorst, M. de Haas, L. Warnier-Wandel, and D. van de Kerkhof
Currently DNA-based analysis of blood groups is mainly used to 
improve transfusion safety by reducing alloantibody formation 
in multiply transfused patients and by monitoring pregnancies at 
risk for hemolytic disease of the fetus and newborn. We present a 
case in which genotyping was performed after massive transfusion 
with unmatched group O, D– blood in a trauma setting. Our 
patient was genotyped as O1A1 and predicted to be D–, and we 
therefore transfused group A, D– red blood cell concentrates. 
This case demonstrates how the use of blood group genotyping 
in an acute setting can lead to a decrease in the unnecessary 
use of group O, D– blood products. Immunohematology 
2012;28:85–7.
Key Words: genotyping, serology, massive transfusion, 
unmatched transfusion
Case Report
Our transfusion laboratory received an emergency 
request for six unmatched group O, D– red blood cell (RBC) 
concentrates for a patient. These units were to be transfused 
not inside the hospital but at the scene of a car accident (at a 
peripheral highway), where the patient was bleeding heavily 
while trapped in a car wreck. Again, another four unmatched 
RBC units were ordered together with five group AB, D– 
fresh-frozen plasma (FFP) units and one randomly selected 
unit of platelets obtained from five donors. All products were 
leukocyte reduced, with less than 106/L white blood cells per 
unit. We had not received blood for crossmatching.
When released from the car wreck, the 62-year-old white 
female patient arrived at the hospital emergency department 
and was admitted to the operating room immediately. The 
patient suffered from complex high-impact trauma. Her car 
had slid underneath a truck, and the patient had been trapped 
for at least 1.5 hours. She had multiple fractures, in the nose, 
left wrist, left humerus, bilateral femur, and ramus superior 
and inferior, and she had a pneumothorax. Also, the patient 
exhibited various superficial wounds and a head injury that 
was closed using staples. Besides damage-control surgery, 
additional acute surgery was necessary to implant a stent 
because of dissection of the descending aorta. During this 
procedure the patient was in need of more blood products, and 
at that time the transfusion laboratory received the first sample 
for crossmatching. Serologically it was no longer possible to 
determine a blood group owing to the number of transfused 
RBC units. Even a mixed-field reaction was hardly visible 
(Fig. 1). Also, a reverse blood group typing was not possible as 
a result of the number of FFP units she had received. After the 
surgical procedure, the patient was stable and was transferred 
to the intensive care unit with a hemoglobin concentration of 
8.8 g/dL (5.5 mmol/L). Three additional FFP units and two 
five-donor platelet concentrates were transfused because 
of low levels of fibrinogen (0.87 g/L) and platelets (76/nL), 
respectively. Altogether, our patient received 18 units of group 
O, D– leukocyte-reduced packed cells, 12 units of FFP (blood 
Case RepoRt
Fig. 1. Blood group typing cards (Ortho Clinical Diagnostics, Inc., 
Raritan, NJ) from our patient’s blood upon arrival at the hospital, 
showing a mixed-field reaction with anti-A (left lane). The mixed-
field reaction was only visible after addition of 10x the erythrocyte 
volume used in the usual procedure. No reaction with the A1 cells 
and B cells (Ortho Clinical Diagnostics, Inc.) is visible.
Case XXXXXXX
86 IMMUNOHEMATOLOGY, Volume 28, Number 3, 2012
J. Curvers et al.
group AB), and 6 units of leukocyte-reduced five-donor platelet 
concentrates (group O, D–) during her intensive care stay.
Unfortunately, we were unable to obtain information 
regarding the blood group of our patient (even after contacting 
the patient’s general practitioner and the local blood bank). 
Therefore, we chose to determine her blood group by means 
of DNA analysis to avoid further (unnecessary) use of group 
O, D– RBC products. We collected venous blood samples 
(anticoagulated in K3-EDTA) and sent them to the laboratories 
of the Sanquin Blood Supply Foundation. ABO genotyping 
was performed using a sequence-specific primer polymerase 
chain reaction (SSP-PCR).1 RHD genotyping was performed 
using a multiplex reverse transcriptase (RT) PCR method.2 
Our patient’s genotype was determined to be O1A1, and she 
was predicted to be D–.
Fortunately, the patient appeared to no longer need blood 
products. She stayed in the intensive care unit for another 20 
days, during which time only one unit of RBCs was transfused 
(blood group A, D–). At that time the patient’s blood sample 
was serologically typed as group A, D– (without mixed-field 
reactions), and the screening for irregular antibodies was 
negative (using a three-cell screening panel, Ortho Clinical 
Diagnostics, Inc.). She was then moved to the surgery 
department for necrotomy of wounds and was transferred to a 
nursing home 2 months after the car accident.
Results
The chosen SSP-PCR is based on the presence of O1, 
O2, B, or A2 genes (the presence of glucosyltransferases 
on chromosome 9q34-2) in DNA isolated from peripheral 
blood samples.1 A PCR product is produced only when the 
specific chosen primer binds to and amplifies the sequence 
of interest. Eight mixtures of four allele-specific primer sets 
are used. The PCR products are transferred to an ethidium-
bromide gel, and patterns are read and interpreted manually. 
Extrapolating results of eight different incubations (O1, O2, B, 
A2, non-O1, non-O2, non-B, and non-A2) led to the genotype. 
Therefore, this assay cannot be used for ABO variant analysis. 
In our case, blood group A1 was predicted as determined by 
exclusion of the other genotypes present. This phenotype has a 
frequency of 33 percent in the white population. The presence 
of an A2 blood group was excluded, as was the presence of a 
B or Bombay (Oh) phenotype. The frequency of other type A 
(or B) blood groups is very rare in the white population, and 
would not have led to a mixed-field reaction of the patient’s 
RBCs with the anti-A serum (Fig. 1).
We were not able to confirm the blood group by reverse 
blood typing because of the amount of FFP received. However, 
after several days, serum anti-B was found but no anti-A; this 
is consistent with the ABO genotype we found.
The multiplex PCR for genotyping the Rh blood group 
system is based on single-nucleotide polymorphisms (SNPs) 
using six different primer sets that cover exons 3, 4, 5, 6, 7, and 
9 of RHD and is able to identify most RHD variants.2 Based on 
the genotyping results, our patient was predicted to be D– , 
and for our white patient, the frequency of RHDψ or a variant 
D is very low. We did not confirm genotyping with buccal cells 
because the clinical consequences of a false-negative result are 
far less severe than those of a false-positive result. Serologically, 
on increasing volumes of RBCs to observe mixed-field 
reaction in our column technique, the D phenotype remained 
negative (Fig. 1). When blood group serology was repeated 
before transfusion 2 weeks after the accident, no mixed-field 
agglutination was observed, although transfused RBCs were 
likely to have been present. This was consistent with a D– 
phenotype and showed concordance between genotype and 
phenotype results. Again, using a three-cell screening panel, 
we found no irregular erythrocyte antibodies.
Of course, interpretation of genotyping results should be 
performed carefully because ethnic differences (e.g., presence 
of RHDψ, or the DAR variant) and silencing alleles (e.g., GATA 
box mutations) exist that could lead to discrepancies between 
genotype and phenotype.3,4 When the patient is of a certain 
ethnicity with a higher risk of ABO variants, sequencing may 
be an option, although costly and time-consuming.
Discussion
Recently, excellent reviews5,6 have been published 
regarding genotyping of the blood donor population and 
its possible drawbacks. However, in transfusion practice, 
genotyping is not commonly performed on patient samples. 
The Dutch national guideline currently states that a blood 
group can only be confirmed when it has been established 
by testing two different samples, irrespective of the method 
used.7 One method may be genotyping.
From 2001, leukocytes are depleted from all blood products 
within the Netherlands by means of filtration, resulting in the 
presence of less than 1 × 106 white blood cells per unit. In case 
of massive transfusion of RBCs and platelet or plasma units, 
the load of donor white blood cells in the patient remains low 
(undetectable when compared with the patient’s white blood 
cell count, on average 5 × 109/L), hence genotyping on DNA 
from the patient’s white blood cells is achievable. Our patient 
IMMUNOHEMATOLOGY, Volume 28, Number 3, 2012 87
Genotyping in massive transfusion
had a white blood cell count of 4.5 leukocytes/nL at the time of 
DNA analysis. Of these, a maximum of 4 percent (0.16/nL: 16 
blood products times 106 white blood cells) could potentially 
originate from the received RBC units. To circumvent these 
limitations or when white blood cell counts in peripheral blood 
are too low to perform genotyping, other material can be used, 
such as nails or buccal epithelial cells (with potential errors 
in bone marrow transplant patients).3,4,8 Therefore, leukocyte 
counting for the patient is essential before performing these 
analyses, and care should be taken in case of neutropenic or 
lymphopenic patients.
Ideally, all transfusions should be closely monitored, 
especially transfusions of unmatched RBC units, to detect acute 
transfusion reactions in case of transfusion of incompatible 
blood group(s). Unfortunately the clinical practice shows that 
unmatched units are transfused in emergency situations, in 
which close monitoring of transfusion can be difficult. Because 
transfused RBCs remain in circulation for several weeks and 
can cause mixed-field agglutination when antigen typing 
is performed, genotyping can determine the blood group of 
trauma patients who have received repeated unmatched RBC 
units. 
Conclusions
As has previously been suggested, blood group 
genotyping can be helpful in acute situations, when serologic 
(hemagglutination with antisera) blood group typing is 
not possible or yields inconclusive results (i.e., mixed-field 
agglutination, discrepant results).6,8 The success of genotyping 
in cases of emergency depends on the speed of processing and 
the turnaround time. Improvements in these will eventually 
lead to fewer unmatched transfusions. The blood transfusion 
laboratory should, however, always be aware of the limitations 
of genotyping (vs. serologic determination of blood groups). In 
a trauma center, genotyping a patient’s sample to determine 
ABO and D might be a suitable alternative to transfusing 
group O, D– RBCs, thus reducing the unnecessary use of such 
blood products. 
References
 1. Gassner C, Schmarda A, Nussbaumer W, Schönitzer D. ABO 
glycosyltransferase genotyping by polymerase chain reaction 
using sequence-specific primers. Blood 1996;88:1852–6.
 2. Maaskant-van Wijk PA, Faas BH, de Ruijter JA, et al. Genotyping 
of RHD by multiplex polymerase chain reaction analysis of six 
RHD-specific exons. Transfusion 1998;38:1015–21.
 3. Castilho L, Rios M, Bianco C, et al. DNA-based typing of blood 
groups for the management of multiply-transfused sickle cell 
disease patients. Transfusion 2002;42:232–8.
 4. Tax MGHM, van der Schoot CE, van Doorn R, Douglas-
Berger L, van Rhenen DJ, Maaskant-vanWijk PA. RHC and 
RHc genotyping in different ethnic groups. Transfusion 2002; 
42:634–44.
 5. Veldhuisen B, van der Schoot CE, de Haas M. Blood group 
genotyping: from patient to high-throughput donor screening. 
Vox Sang 2009;97:198–206.
 6. Avent NO. Large-scale blood group genotyping: clinical 
implications. Br J Haematol 2009;144:3–13.
 7. CBO (Centraal Begeleidings Orgaan) consensus. Richtlijn 
bloedtransfusie (Blood transfusion directive) 2004. 
Kwaliteitsintituut voor de Gezondheidszorg (Dutch Institute 
for Healthcare Improvement). Utrecht, The Netherlands.
 8. Rozman P, Dovc T, Gassner C. Differentiation of autologous 
ABO, RHD, RHCE, KEL, JK, and FY blood group genotypes 
by analysis of peripheral blood samples of patients who 
have recently received multiple transfusions. Transfusion 
2000;40:936–42.
Joyce Curvers, PhD (corresponding author), Resident in Clinical 
Chemistry, Volkher Scharnhorst, PhD, Resident in Clinical 
Chemistry, Clinical Laboratory Department of the Catharina 
Hospital, Michelangelolaan 2, 5623 EJ Eindhoven, The Netherlands; 
Masja de Haas, MD, PhD, Head of the Diagnostic and Research 
Laboratory of Immunohematology, Sanquin Blood Supply 
Foundation, Amsterdam, The Netherlands; Loes Warnier-Wandel, 
BSc, Head of the Immunohematology Department of the Blood Bank 
Southwest Region, Sanquin Blood Supply Foundation, Dordrecht, 
The Netherlands; and Daan van de Kerkhof, PhD, Resident in Clinical 
Chemistry, Clinical Laboratory Department of the Catharina 
Hospital, Michelangelolaan 2, 5623 EJ Eindhoven, The Netherlands.


Blood group genotyping in a multitrauma patient: a case report

Tsimba-Chitsva, Farai

Bishop, Susanne

Kezeor, Kelly

88 IMMUNOHEMATOLOGY, Volume 28, Number 3, 2012
Warm autoadsorption with enzyme-treated red 
blood cells
F. Tsimba-Chitsva, S. Bishop, and K. Kezeor
Patients demonstrating warm autoantibody specificity present 
serologic challenges for laboratory staff performing antibody 
identification in the blood bank. Autoantibody can be removed 
from plasma or serum by adsorption onto autologous red blood 
cells (RBCs) provided the patient has not been transfused in the 
previous 3 months. The adsorption process can be enhanced by 
enzyme pretreatment of autologous RBCs. Immunohematology 
2012;28:88–90.
Key Words: warm autoadsorption, ficin, papain, autologous
Principle
Performing pretransfusion testing in the presence of 
warm autoantibodies optimally reactive at 37°C presents 
serologic challenges when attempting to determine the 
presence of underlying clinically significant alloantibodies.1 
Warm-reactive autoantibodies may mask the presence of other 
antibodies because of their agglutination with all cells tested, 
including autologous red blood cells (RBCs). To determine the 
presence of clinically significant antibodies, the removal of 
autoantibody may be achieved by adsorption of patient plasma 
or serum with autologous RBCs.2–4 Allogeneic RBCs may be 
used for adsorption if the patient has been recently transfused 
or when the quantity of autologous RBCs is insufficient for 
autoadsorption.2
Indications
As circulating autologous RBCs are coated with 
autoantibody, antigen sites on the autologous cells may 
become blocked with immunoglobulin. To efficiently achieve 
autologous adsorption of warm-reactive autoantibodies, 
autoantibody must be dissociated from the RBC before 
the adsorption process.2 Autoantibody dissociation can be 
accomplished by several methods, including partial heat elution 
at 45°C, gentle heat elution at 56°C, a combination of 0.2 M 
dithiothreitol (DTT) and papain or ficin treatment, acid-EDTA 
treatment, ZZAP (a mixture of cysteine-activated proteolytic 
papain and DTT) treatment, and chloroquine diphosphate 
(CDP) treatment.5 (For specific procedures concerning 
these methods and reagents, see the applicable procedural 
reference documents or manufacturer’s package insert.) After 
dissociation of the autoantibody from the autologous RBCs, the 
antigen sites should be available for binding the autoantibody 
from the plasma or serum during the adsorption process.
To further enhance the adsorption process, autologous 
RBCs can be treated with proteolytic enzymes such as ficin or 
papain. This additional treatment enhances antibody uptake 
onto the autologous cells by cleaving some of the glycoprotein 
chains extending from the cell membrane, leading to improved 
accessibility of the antigens not removed by the dissociation 
method.4
Procedure
For papain or ficin (1%) treatment of RBCs, one volume 
of 1% cysteine-activated papain or 1% ficin is added to two 
volumes of packed RBCs. For example, if 2 mL of packed 
RBCs is treated, 1 mL of ficin or papain is added. The mixture 
S
E
R
O
L
O
G
IC
 M
E
T
H
O
D
 R
E
V
IE
W
Reagents/Supplies
Enzyme 
Treatment 
Method Reagents Supplies
Papain 1% 
cysteine-
activated 
papain
• Isotonic 
saline
• Patient 
autologous 
red blood 
cells
• 1 mL graduated pipettes
• 37°C water bath
• Large bore test tubes
• Calibrated centrifugeFicin 1% ficin
Procedural Steps
Treatment 
of Cells
• Mix enzyme with packed red cells
• Incubate mixture
• Wash mixture to remove enzyme
• Centrifuge to pack the red cells
• Remove supernatant
Adsorption • Mix enzyme-treated cells with patient plasma/serum
• Incubate  the mixture
• Centrifuge the mixture
• Remove plasma/serum
IMMUNOHEMATOLOGY, Volume 28, Number 3, 2012 89
Warm autoadsorptions with enzyme-treated RBCs
is incubated at 37°C for a predetermined time to allow 
for the most effective enzyme treatment. If commercially 
prepared enzyme is used, the manufacturer’s insert is used 
to determine the incubation time. If enzyme is prepared in-
house, qualification studies are performed to determine the 
appropriate treatment time.6
The enzyme-treated RBCs are washed with copious 
amounts of isotonic saline a minimum of three times, or until 
the supernatant saline is clear, to ensure complete removal of 
enzyme. The last wash is centrifuged a minimum of 5 minutes 
at 900 to 1000× g to completely pack the RBCs.7 This step 
is critical to achieve proper packing of RBCs and removal of 
maximum amounts of saline. To avoid plasma or serum dilution 
during adsorption, as much saline as possible is carefully 
removed after the final wash. This can be accomplished with 
suction using a small-bore pipette or filter paper.4
To perform adsorption, one volume of enzyme-treated 
RBCs is mixed with an equal volume of the patient’s plasma or 
serum. To enhance antibody uptake, the proportion of RBCs to 
plasma or serum can be increased. Adsorption is more effective 
if the area of contact between the RBCs and plasma or serum 
is large; use of a large-bore test tube (16 × 100 mm) increases 
the surface area and allows for optimum plasma or serum and 
RBC contact. The mixture is incubated at 37°C for 10 to 60 
minutes.4 The mixture is then centrifuged and the plasma or 
serum is carefully removed. If the harvested adsorbed plasma 
or serum will be used for additional adsorptions, disturbing 
the packed RBCs to achieve maximum plasma removal is 
acceptable. If, however, the harvested adsorbed plasma or 
serum will be used for antibody detection, the packed RBCs 
should not be disturbed in an attempt to recover additional 
plasma.
The number of adsorptions required to exhaust 
autoantibody from the plasma or serum is proportionally 
related to the strength of its  reactivity in plasma or serum; 
however, in some instances, additional adsorptions may be 
required to remove the antibody.7–9 If the original plasma or 
serum reactivity is 1+ with the indirect antiglobulin test (IAT), 
the first aliquot of adsorbed plasma or serum should be tested 
against group O reagent RBCs or direct antiglobulin test (DAT) 
negative autologous RBCs, if available. If reactivity persists, 
the adsorption procedure should be repeated, as needed, using 
a fresh volume of enzyme-treated RBCs. Once reactivity is 
abolished, the adsorbed plasma or serum is tested against the 
appropriate group O reagent panel RBCs to exclude underlying 
alloantibodies.
Limitations
A recently transfused patient (within the last 3 months) 
is not a candidate for autologous adsorption. Circulating 
transfused RBCs may adsorb clinically significant plasma or 
serum alloantibodies.2 It has been demonstrated that very small 
amounts of antigen-positive RBC are capable of eliminating 
alloantibody reactivity from the plasma or serum.2,10
Large volumes of autologous RBCs from the patient are 
needed to perform autoadsorption. In the case of a low patient 
hematocrit, it may be difficult to obtain a sufficient volume 
of autologous cells for testing without exacerbating clinical 
symptoms of anemia.
Cell treatment may be unsuccessful in stripping 
autoantibody from the autologous, coated RBCs. As a result, 
adsorption may be less efficient in removing autoantibody.
Enzyme treatment of RBCs will destroy antigen sites on 
the RBCs. Enzymes remove sialic acid from the RBCs and will 
destroy some antigens (-M, -N, -S, -s, -Fya, -Fyb, -Ena, -Ge, 
-JMH, -Ch/Rg, -Inb). Autoantibodies specific for enzyme-
sensitive antigens will not be removed by this method.4
Potential dilution of adsorbed plasma or serum must be 
considered a consequence of preparation of enzyme-treated 
cells. If the saline is not properly removed after the final wash, 
the residual saline will dilute the plasma or serum when 
adsorptions are performed. The more adsorptions performed, 
the greater the dilution factor. As a result, diluted antibody 
may become undetectable in the adsorbed plasma or serum.
Enzyme treatment of autologous RBCs can cause 
hemolysis. If the RBCs are hemolyzed by enzyme treatment, 
there may be an insufficient quantity of RBCs remaining to 
perform adsorptions.
Acknowledgment
The authors thank Kerry Burright-Hittner, MT(ASCP)
SBB, Director IRL, American Red Cross, Midwest Region, for 
providing technical assistance and expertise, as well as for her 
critical reading and editing of the manuscript.
References
 1. Leger RM, Garratty G. Evaluation of methods for detecting 
alloantibodies underlying warm autoantibodies. Transfusion 
1999; 39:11–16.
 2. Leger, RM. Positive DAT and immune-mediated hemolysis. 
In: Roback JD (ed). Technical manual. 17th ed. Bethesda, MD: 
American Association of Blood Banks, 2011:505–6.
S
E
R
O
L
O
G
IC
 M
E
T
H
O
D
 R
E
V
IE
W
90 IMMUNOHEMATOLOGY, Volume 28, Number 3, 2012
F. Tsimba-Chitsva et al.
 3. Morel PA, Bergren MO, Frank BA. A simple method for the 
detection of alloantibody in the presence of autoantibody 
[abstract]. Transfusion 1978;18:388.
 4. Judd J, Johnson ST, Storry JR. Judd’s methods in 
immunohematology. 3rd ed. Bethesda, MD: American 
Association of Blood Banks, 2008:45, 457–9.
 5. Beattie K, Crawford M, Martin, J, et al. Immunohematology 
methods. Rockville, MD: American Red Cross, 1993:44-1, 48-
1, 51-1.
 6. Method 3-5-1: Preparing ficin enzyme stock, 1% W/V; Method 
3-5-2: Preparing papain enzyme stock, 1% W/V; Method 
3-5-3: Standardizing enzyme procedures. In: Roback JD 
(ed). Technical manual. 17th ed. Bethesda, MD: American 
Association of Blood Banks, 2011:902–3.
 7. Method 4-8: Adsorbing warm-reactive autoantibodies using 
autologous red cells. In: Roback JD (ed). Technical manual. 
17th ed. Bethesda, MD: American Association of Blood Banks, 
2011:924–5.
Notice to Readers
Immunohematology is printed on acid-free paper.
For information concerning Immunohematology or the 
Immunohematology Methods and Procedures manual, 
contact us by e-mail at immuno@redcross.org
For information concerning the National Reference 
Laboratory for Blood Group Serology, including the American 
Rare Donor Program, contact Sandra Nance, by phone at 
(215) 451-4362, by fax at (215) 451-2538, or by e-mail at 
Sandra.Nance@redcross.org
 8. Garratty G. What do you do when all units are incompatible? 
43. Instituto Mexicano del Seguro Social, 2005:S107–11.
 9. Method 3-12: Adsorption procedure. In: Roback JD (ed). 
Technical manual. 17th ed. Bethesda, MD: American 
Association of Blood Banks, 2011:914.
 10. Garratty G, Telen MJ, Petz LD. Red cell antigens as functional 
molecules and obstacles to transfusion. Hematology AmSoc 
Hematol Educ Program 2002:445–50.
Farai Tsimba-Chitsva, MLS(ASCP)SBBCM, Reference Technologist, 
Susanne Bishop, MLS(ASCP)SBBCM, Reference Technologist, 
and Kelly Kezeor, MT(ASCP), (corresponding author) Reference 
Technologist, Midwest Region American Red Cross Reference 
Laboratory. American Red Cross, 3838 Dewey Avenue, Omaha, NE 
68105.
S
E
R
O
L
O
G
IC
 M
E
T
H
O
D
 R
E
V
IE
W


Warm autoadsorption with enzyme-treated red blood cells

Johnson, Nanette C.

68 IMMUNOHEMATOLOGY, Volume 27, Number 2, 2011
The XG blood group system is best known for its contributions 
to the fields of genetics and chromosome mapping. This system 
comprises two antigens, Xga and CD99, that are not antithetical 
but that demonstrate a unique phenotypic relationship. XG is 
located on the tip of the short arm of the X chromosome with 
exons 1 to 3 present in the pseudoautosomal region of the X 
(and Y) chromosome(s) and exons 4 to 10 located only on the 
X chromosome. Xga demonstrates a clear X-linked pattern of 
inheritance. MIC2, the gene encoding the CD99 antigen, is found 
in the pseudoautosomal region of both the X and Y chromosomes. 
Anti-Xga is comparatively rare, and only two examples of anti-
CD99 have ever been identified. Alloanti-Xga is considered 
clinically insignificant; only one example of autoanti-Xga has been 
reported, but it resulted in severe hemolytic anemia. Insufficient 
data exist to determine the clinical significance of anti-CD99. 
Linkage of XG to several X-borne genes encoding inherited 
disorders has been demonstrated. CD99 is an adhesion molecule, 
and high levels are associated with some types of cancer. 
Immunohematology 2011;27:68–71.
Key Words: Xga, X-linked, CD99, XG, MIC2, PBDX, 12E7, 
Lyonization, qualitative polymorphism, pseudoautosomal
History
The XG blood group system is not revered for its 
antigens and antibodies, but is instead respected for its 
contributions to the fields of genetics and chromosome 
mapping. XG is the first, and so far the only, blood group 
system to be assigned to the X chromosome.1 There are 
only two antigens in the system: Xga and CD99. Xga may be 
expressed only on red blood cells (RBCs), whereas CD99 is 
expressed on many tissue cells. Anti-Xga can be naturally 
occurring and is not considered clinically significant. Only 
two examples of alloanti-CD99 have ever been reported; its 
clinical significance is unknown.2
The first example of anti-Xga was reported in 1961 in 
the serum of a man who had been multiply transfused for 
severe nosebleeds as a result of familial telangiectasia.3 
Shortly thereafter, in 1962, Mann et al. characterized Xga 
as originating from a locus on the X chromosome.1 Xga is 
named after the X chromosome and Grand Rapids, where 
the first-reported patient to have the antibody (Mr. And) was 
treated.4 It was later discovered that PBDX, a gene found to 
span the pseudoautosomal boundary on the X chromosome, 
is the gene that encodes Xga (see Molecular Basis section).5 
The function of the Xga protein is not known.6
Two decades after the first example of anti-Xga was 
reported, Goodfellow and Tippett observed a mouse 
monoclonal antibody, 12E7, which recognized another 
determinant that appeared to be controlled by a gene also 
found on the X chromosome.7 This gene, MIC2, is located 
in the pseudoautosomal region of both the X and the Y 
chromosomes and encodes CD99. CD99 is an adhesion 
molecule,5,8 and high levels are associated with some 
types of cancer. Xga and CD99 enjoy a unique phenotypic 
relationship, which will be discussed in further detail in 
the Genetics and Inheritance section. Fourteen years later, 
in 1995, Uchikawa et al. reported the only examples of 
alloanti-CD99, detected in two unrelated, healthy Japanese 
blood donors.2 CD99, became part of the XG blood group 
system officially in 2000 when it was confirmed that MIC2 
and XG are adjacent, homologous genes.4
Because of its location in the pseudoautosomal boundary 
region of the sex chromosome, study of the inheritance of 
Xga has helped to define the mechanisms responsible for 
Turner, Klinefelter, and other syndromes resulting from an 
abnormal number of sex chromosomes. Sex chromosome 
aneuploidies are commonly associated with infertility.
Nomenclature
The XG blood group system has been assigned the 
ISBT symbol and number XG and 012, respectively. The 
ISBT terminology for the two antigens in the system, Xga 
and CD99, is listed in Table 1. No antithetical counterpart 
to the Xga antigen has been discovered, so the failure of 
RBCs to react with anti-Xga defines the silent allele, Xg, or 
an absence of Xga.
Genetics and Inheritance
Xga is inherited as an X-linked, dominant trait, and 
the pattern of inheritance is as expected for an X-borne 
characteristic.1,8 If Xga is present, it is expressed on RBCs. 
XG: the forgotten blood group system
N.C. Johnson
Review
Table 1. Nomenclature
Xga CD99
ISBT symbol XG1 XG2
ISBT number 012001 012002
IMMUNOHEMATOLOGY, Volume 27, Number 2, 2011 69
XG blood group system: a review
The Xg(a+) phenotype is found in 88.7 percent of females 
and 65.6 percent of males. (Table 2.) Females can have a 
single (Xga/Xg) or a double dose (Xga/Xga) of Xga. Having 
only one X chromosome, males can be hemizygous for Xga 
(Xga/Y) or be Xg/Y and not express the antigen at all.6
Lyonization is a process by which one X chromosome 
is inactivated early in somatic cell development in XX 
females.9 The process is random in each cell, ensuring 
that all genes on the two X chromosomes are found at the 
phenotypic level. By studying RBCs of Xga/Xg women it can 
be shown that all cells carry Xga, proving that Xga is not 
subject to inactivation or Lyonization. XG was the first X 
chromosome locus proven to escape this process.10 It was 
later shown that MIC2 is not subject to Lyonization either.11
New Guineans, Australian aborigines, Navajo Indians, 
and Sardinians are reported to have the highest prevalence 
of Xga. On the other hand, native Taiwanese; Chinese in 
Singapore, Hong Kong, Taiwan, and Hakka; and Malays in 
Singapore are observed to have the lowest incidence.3,12–14 
Interestingly, Asian populations with a higher prevalence of 
Xg(a–) phenotypes do not have a higher incidence of anti-
Xga.6
Using monoclonal antibodies, it has been shown that 
CD99 is present on all tissue cells tested15; however, the 
expression of the adhesion molecule on RBCs is unique. 
CD99 is expressed at different levels on RBCs, and this 
expression is directly related to the presence or absence of 
Xga .16 In females, the Xg(a–) phenotype is always associated 
with low expression of CD99. (Table 3.) Curiously, among 
Xg(a–) males, 74 percent are high expressors of CD99.17 
To explain this quantitative polymorphism, Goodfellow 
and Tippett proposed the existence of a locus on the Y 
chromosome, YG, analogous to XG on the X chromosome. 
Like XG, YG would have two alleles, Yga and Yg. In Xg(a–) 
males, the presence of Yga would result in high expression 
of CD99, and the absence of Yga (Yg) would result in low 
expression of CD99.7,16
The presence of YG has been substantiated by family 
studies,17 but to date, there has been no Yga antigen 
identified. It is important to note that MIC2 is responsible 
for expression of CD99 on all tissue cells, but XG and YG 
create the CD99 quantitative polymorphism observed on 
RBCs.8
Molecular Basis
Sex chromosomes have two genetically distinct regions: 
sex chromosome–specific sequences and pseudoautosomal 
sequences. Pseudoautosomal refers to a structurally 
homologous region on sex chromosomes, the function 
of which is to facilitate pairing during male meiosis.8 XG 
spans the pseudoautosomal boundary between the two 
regions of the X chromosome at Xp22.3; exons 1 to 3 are 
located in the pseudoautosomal region (PAR) and exons 4 
to 10 are found in the sex chromosome–specific region.4–6 
MIC2, the closest neighbor to XG, is located in the PAR at 
position Xp22.2. An identical copy of MIC2 is found in the 
PAR region of the Y chromosome at Yp11.2.4 Almost half 
(48%) of the predicted amino acid sequences of XG and 
MIC2 are identical.18
Biochemistry
Immunoblotting of immunoprecipitates confirmed that 
CD99 and Xga antigens are located on different structures, 
later confirmed to be sialoglycoproteins.18,19 CD99 and Xga 
are associated in the membrane, possibly in the form of a 
heterodimer.7,18
The RBC membrane molecule expressing Xga was first 
identified in 1989 by Herron and Smith, by immunoblotting 
and electropheresis.19 Several years later, Petty and Tippett 
used immunoblotting and immunoprecipitation to show that 
Xga is carried on a broad band of apparent molecular weight 
24.5 to 29.5 kDa, consisting of a darkly stained component 
at 24.5 kDa and a more diffusely stained component 
between 26.5 and 29.5 kDa.18 No bands are observed with 
protease-treated cells, and the apparent molecular weight 
of Xga is decreased after treatment with neuraminidase, 
suggesting that Xga resides on a sialoglycoprotein.
Table 2. Phenotype prevalence and genotype frequencies
Male (%) Female (%)
Phenotype
Xg(a+) 65.6 88.7
Xg(a–) 34.4 11.3
Genotype
Xga/Xga 0.434
Xga/Xg 0.450
Xg/Xg 0.116
Xga/Xg 0.656
Xg/Xg 0.344
Table 3. Relationship of Xga, Yga, and CD99
Females Xg(a+) CD99 high
Xg(a+w) CD99 high
Xg(a–) CD99 low
Males Xg(a+) CD99 high
Xg(a–) Yga CD99 high
Xg(a–) Yg CD99 low
70 IMMUNOHEMATOLOGY, Volume 27, Number 2, 2011
N.C. Johnson
Xga is sensitive to treatment with ficin, papain, trypsin, 
bromelin, α-chymotrypsin, and pronase, and is resistant to 
sialidase, 0.2 M dithiothreitol (DTT), neuraminidase, and 
2-aminoethylisothiouronium bromide.4,13,20 The effects of 
acid are unknown. The expression of the antigen decreases 
as the RBC ages and it has been shown by indirect 
radioimmunoassay to have an in vivo half-life of 47 days.21
Early experiments using a microcomplement fixation 
method and indirect radioimmunoassay suggested that 
Xga might be present on other cell lines and tissues,22 but 
subsequent testing was unable to duplicate these results. 
XG is expressed on hematopoietic tissues8 and is strongly 
expressed as mRNA in fibroblasts.20
Apart from humans, the only animals found to have 
Xga on their RBCs are one species of gibbons. Of 52 gibbons 
tested, 30 percent of males and 53 percent of females were 
Xg(a+). Chimpanzees (67); gorillas (2); orangutans (20); 
another species of gibbons (5); various monkeys, including 
60 baboons; and a few nonprimates, including mice and 
dogs, were all Xg(a–).3,4
Monoclonal anti-CD99 identify a protein of apparent 
molecular weight of 32 kDa.5,8 Immunoblotting and 
immunoprecipitate studies suggest the CD99 protein is 
composed of two polypeptides, a 32-kDa surface compo-
nent, and a 30-kDa intracellular polypeptide.18 CD99 
is sensitive to treatment with trypsin, α-chymotrypsin, 
ficin, papain, and pronase15 and, like Xga, is carried on a 
sialoglycoprotein. CD99 is resistant to 0.2 M DTT and is 
usually resistant to neuraminidase treatment; results vary 
with sialidase treatment.
CD99 is present on all human cell types and tissues 
tested.15 High levels of CD99 are a tumor marker in Ewing 
sarcoma, some neuroectodermal tumors, lymphoblastic 
lymphoma, and acute lymphoblastic leukemia.8 CD99 was 
not detected on the RBCs or peripheral blood lymphocytes 
of 10 gibbons, regardless of their Xga phenotype. CD99 
was detected on RBCs and fibroblasts of chimpanzees and 
gorillas, but not of orangutans or any of the other mammals 
tested.23
Antibodies in the System
Anti-Xga is comparatively rare; most immunohema-
tologists will never see anti-Xga in their careers. When 
identified, it is usually found as a lone antibody specificity. 
Some examples of anti-Xga are naturally occurring, yet 
are more frequently IgG than IgM.4 Optimal methods for 
identification include room temperature incubation (even 
when they are IgG in composition3), indirect antiglobulin 
test (IAT), and capillary testing.4 One IAT-reactive anti-
Xga was shown to have IgG1 and IgG2 subclasses.24 Some 
examples of anti-Xga have been shown to fix complement.1 
The only reported example of autoanti-Xga was found in 
a pregnant woman. The antibody appears to have caused 
significant hemolytic anemia, but the possibility that the 
hemolytic anemia was exacerbated by the pregnancy was 
not ruled out.25
As a result of the sex-related frequency differences of 
Xga, it is expected that men would make more examples of 
anti-Xga. Although this is true, it is striking that greater 
than 85 percent of antibody makers are men when the 
incidence of the Xg(a–) phenotype is only 23 percent higher 
in men than in women. It has been postulated, although 
not studied, that this higher than expected prevalence 
of anti-Xga in men may be associated with the expression 
of CD99.6
The first example of anti-CD99 ever made was a mouse 
monoclonal antibody called 12E7, which was generated to 
screen for antigen differences in human acute lymphocytic 
leukemia.26 The only two examples of anti-CD99 ever 
detected were found in two unrelated, healthy Japanese 
blood donors. One antibody was reported to react with 
variable strength with different RBC samples. The second 
blood donor had a history of pregnancy. One of the two 
siblings of the second blood donor was determined to be 
CD99 negative also.2 The antibodies were determined to 
be IgG in nature and reacted optimally by the IAT.5 It is 
not known whether the antibodies were capable of binding 
complement.
Clinical Significance
Alloanti-Xga has never been reported to cause 
hemolytic disease of the fetus or newborn or a hemolytic 
transfusion reaction.4 One patient received six units of 
antigen-positive blood with no signs of a reaction.27 A 
chromium survival study on another patient showed normal 
survival of Xg(a+) RBCs and no posttransfusion increase 
in antibody strength.28 One male Japanese patient had a 
slight temperature elevation and exhibited urticaria when 
transfused with Xg(a+) blood. Subsequent transfusions 
with Xg(a–) blood were tolerated with no symptoms.29
The one example of autoanti-Xga reported was 
identified in a pregnant female. The antibody was IgG in 
nature, activated complement, and caused severe hemolytic 
anemia. The infant was Xg(a+) and demonstrated a strongly 
positive direct antiglobulin test at birth, and anti-Xga was 
eluted from the infant’s cells. The clinical course was 
remarkable only for a mild rise in bilirubin after birth.25
Because the only two examples of anti-CD99 were 
found in healthy blood donors, information on the clinical 
significance of this antibody is not known.
IMMUNOHEMATOLOGY, Volume 27, Number 2, 2011 71
XG blood group system: a review
XG is linked to genes responsible for several genetic 
disorders.3,4 Ichthyosis is a disorder in which the skin 
resembles scales on a fish. Ocular albinism is unique in that 
only the eyes lack pigment. Retinoschisis is a sex-linked 
form of macular degeneration affecting only men. High 
levels of CD99 have been reported in Ewing sarcoma, some 
neuroectodermal tumors, lymphoblastic lymphoma, and 
acute lymphoblastic leukemia.8
Summary and Future Perspectives
The XG blood group system is most important for the 
insight it has provided into the study and understanding of 
meiotic errors that lead to sex chromosome aneuploidies. As 
the roles of Xga and CD99 become more clearly delineated, 
knowledge of the forgotten blood group system may aid in 
the development of cures for certain genetic disorders and 
types of cancers.
References
 1. Mann JD, Cahan A, Gelb AG, et al. A sex-linked blood group. 
Lancet 1962;1:8–10.
 2. Uchikawa M, Tsuneyama H, Tadokoro K, Juji T, Yamada 
M, Maeda Y. An alloantibody to 12E7 antigen detected in 2 
healthy donors [abstract]. Transfusion 1995;35:23S.
 3. Race RR, Sanger R. The Xg blood groups. In: Blood groups 
in man. 6th ed. Oxford, PA: Blackwell Scientific Publication, 
1975:578–635.
 4. Reid ME, Lomas-Francis C. The blood group antigen 
factsbook. 2nd ed. San Diego, CA: Academic Press, 2004: 
338–43.
 5. Ellis N, Tippett P, Petty A, et al. PBDX is the XG blood group 
gene. Nat Genet 1994;8:285–90.
 6. Issitt PD, Anstee DJ. Applied blood group serology. 4th ed. 
Durham, NC: Montgomery Scientific Publications, 1998.
 7. Daniels G. Xg blood group system. In: Human blood groups. 
2nd ed. Malden, MA: Blackwell Science, 2002:374–86.
 8. Tippett P, Ellis NA. The Xg blood group system: a review. 
Transfus Med Rev 1998;12:233–57.
 9. Lyon MF. Gene action in the X-chromosome of the mouse 
(Mus musculus L.). Nature 1961;190:372–3.
 10. Lawler SD, Sanger R. Xg blood-groups and clonal-origin 
theory of chronic myeloid leukaemia. Lancet 1970;1:584–5.
 11. Goodfellow P, Pym B, Mohandas T, Shapiro LJ. The cell 
surface antigen locus, MIC2X, escapes X-inactivation. Am J 
Hum Genet 1984;36:777–82.
 12. Dewey WJ, Mann JD. Xg blood group frequencies in some 
further populations. J Med Genet 1967;4:12–15.
 13. Siniscalco M, Filippi G, Latte B, et al. Failure to detect 
linkage between Xg and other X-borne loci in Sardinians. 
Ann Hum Genet 1966;29:231–52.
 14. Simmons RT. X-linked blood groups, Xg, in Australian 
Aborigines and New Guineans. Nature 1970;227:1363.
 15. Goodfellow P. Expression of the 12E7 antigen is controlled 
independently by genes on the human X and Y chromosomes. 
Differentiation 1983;23(Suppl):S35–9.
 16. Goodfellow PN, Tippett P. A human quantitative poly-
morphism related to Xg blood groups. Nature 1981;289: 
404–5.
 17. Tippett P, Shaw M-A, Green CA, Daniels GL. The 12E7 red 
cell quantitative polymorphism: control by the Y-borne 
locus, Yg. Ann Hum Genet 1986;50(Pt 4):339–47.
 18. Petty AC, Tippett P. Investigation of the biochemical 
relationship of the blood group antigens Xga and CD99 
(12E7 antigen) on red cells. Vox Sang 1995;69:231–5.
 19. Herron R, Smith GA. Identification and immunochemical 
characterization of the human erythrocyte membrane 
glycoproteins that carry the Xga antigen. Biochem J 1989; 
262:369–71.
 20. Ellis NA, Ye T-Z, Patton S, German J, Goodfellow PN, Weller 
P. Cloning of PBDX, and MIC2-related gene that spans the 
pseudoautosomal boundary on chromosome Xp. Nat Genet 
1994;6:394–400.
 21. Campana T, Szabo P, Piomelli S, Siniscalco M. The Xga 
antigen on red cells and fibroblasts. Cytogenet Cell Genet 
1978;22:524–6.
 22. Fellous M, Bengtsson B, Finnegan D, Bodmer WF. Expression 
of the Xga antigen on cells in culture and its segregation in 
somatic cell hybrids. Ann Hum Genet 1974;37:421–30.
 23. Shaw MA, Tippet P, Delhanty JD, Andrews M, Goodfellow 
P. Expression of Xg and the 12E7 antigen in primates. J 
Immunogenet 1985;12:115–18.
 24. Devenish A, Burslem MF, Morris R, Contreras M. Serologic 
characteristics of a further example of anti-Xga in North 
London blood donors. Transfusion 1986;26:426–7.
 25. Yokoyama M, McCoy JE Jr. Further studies on auto-anti-
Xga antibody. Vox Sang 1967;13:15–17.
 26. Levy R, Dilley J, Fox RI, Warnke R. A human thymus-
leukemia antigen defined by hybridoma monoclonal 
antibodies. Proc Natl Acad Sci U S A 1979;76:6552–6.
 27. Cook IA, Polley MJ, Mollison PL. A second example of anti-
Xga. Lancet 1963;1:857–9.
 28. Sausais L, Krevans JR, Townes AS. Characteristics of a third 
example of anti-Xga [abstract]. Transfusion 1964;4:312.
 29. Azar PM, Saji H, Yamanaka R, Hosoi T. Anti-Xga suspected 
of causing a transfusion reaction. Transfusion 1982;22: 
340–1.
Nanette C. Johnson, MT(ASCP)SBB, Director, Immunohematol-
ogy Reference Laboratory, Greater Chesapeake and Potomac 
Region, 4700 Mt. Hope Drive, Baltimore, MD 21215.


XG: the forgotten blood group system

Clay, Mary E.

Schuller, Randy M.

Bachowski, Gary J.

Granulocyte serology: current concepts and clinical signifcance

Smart, Elizabeth A.

Storry, Jill R.

IMMUNOHEMATOLOGY, Volume 26, Number 3, 2010124
The OK blood group system: a review
E.A. Smart and J.R. Storry
History
 The very high-incidence RBC antigen Oka was first de-
scribed by Morel and Hamilton1 in 1979, after the detection 
of clinically significant antibodies in the serum of a Japa-
nese woman (Mrs. S.Ko.G.) who had been admitted to the 
hospital with severe gastrointestinal bleeding. On the ba-
sis of extensive serologic tests, it was shown that the anti-
body was distinct from other antibodies to high-incidence 
or “public” antigens. The antibody was named anti-Oka and 
the antigen Oka after the name of the patient who lacked 
the antigen and developed the antibody. The patient had 
a history of previous blood transfusions but had not been 
pregnant. The patient and her family were from a small is-
land in the Inland Sea off the coast of Honshu in Japan. 
Family studies showed that Oka is inherited. The proband’s 
parents were consanguineous (cousins), and two siblings 
typed Ok(a–) and one sibling typed Ok(a+). Apart from the 
two siblings, no other Ok(a–) individuals were found when 
testing blood samples from more than 400 people living 
on the same island. No Ok(a–) samples were found when 
testing blood samples from a further 870 Japanese indi-
viduals, 1378 Mexican-Americans, 9053 Caucasians, 261 
Asians, 911 Blacks, and 1570 blood donors of undesignated 
ethnicity. Anti-Oka was not detected in the serum of Mrs. 
S.Ko.G.’s compatible sister although she had five children, 
two of whom tested Ok(a+).
 The Ok(a–) phenotype is extremely rare. To date, it has 
been identified in eight families in Japan.2 However, two 
other variants have been identified: anti-Oka was identi-
fied in the plasma of a woman of Iranian background. Her 
RBCs gave variable reactivity with a panel of anti-Oka, indi-
cating that her phenotype was different from that of previ-
ously identified Ok(a–) individuals.3 Second, an antibody 
in the plasma of a prenatal Hispanic woman reacted with 
all RBCs except for three samples of Ok(a–) RBCs. How-
ever, her RBCs were reactive with all anti-Oka tested, again 
indicating a variant phenotype.4 In both cases, molecular 
analysis revealed point mutations in the gene that were 
different from Oka (discussed in more detail later), and 
both have been assigned provisional numbers by the ISBT 
Working Party on Red Cell Immunogenetics and Blood 
Group Terminology.
Nomenclature
There are three RBC antigens in the OK blood group sys-
tem (ISBT 024), Oka (OK1), and the provisionally assigned 
antigens OK2 and OK3. The antigens are located on CD147 
encoded by the gene BSG on chromosome 19p13.3 (Tables 
1 and 2).5,6.
Genetics and Inheritance
 In the original family studies1 two siblings were Ok(a–), 
and one was Ok(a+), showing that Ok is an inherited char-
acteristic on the basis of the serologic tests. Further studies 
by Williams et al.,7 reported in 1988, showed that the Ok de-
terminant is controlled by a single gene OK and that the OK 
locus is on chromosome 19, in the region 19pter-p13.2. Us-
ing the monoclonal antibody TRA-1-85, which they showed 
was specific for the Ok-bearing glycoprotein, these authors 
identified that Oka was carried not only on human RBCs but 
also on leukocytes and on many immortalized human cell 
lines of different tissue origin. Spring et al.6 later demon-
strated that Oka antigen was carried on the M6 leukocyte 
activation antigen (now known as basigin; other synonyms 
EMMPRIN, CD147, TCSF) and identified the point muta-
tion responsible for the Ok(a–) phenotype (see below). 
Using immunohistochemical and Western blotting tech-
niques, the Ok glycoprotein was identified on a wide range 
of tissues.
Molecular Basis
 The gene encoding basigin (BSG) consists of eight exons 
and encodes a type I single-pass transmembrane glycopro-
tein of 269 amino acids. The leader sequence of 21 amino 
acids is cleaved from the mature membrane-bound protein, 
which consists of a glycosylated amino-terminal extracel-
lular domain of 187 amino acids that is arranged into two 
Ig-like subunits and a transmembrane/cytoplasmic domain 
of 40 amino acids. Sequencing of cDNA extracted from three 
Ok(a–) individuals showed a unique missense mutation 274 
G>A in exon 3, which encodes a Glu92Lys substitution in 
the amino-terminal Ig-like subunit of the glycoprotein.6 This 
substitution was shown by transfection studies in murine 
NS-0 cells to be solely responsible for the Ok(a–) phenotype. 
A silent mutation in exon 4 (384T>C) was also identified.
Table 1. The OK blood group system
ISBT No.
System 
name
System 
symbol
Gene 
name
Chromo-
somal 
location CD number
024 OK OK BSG 19p13.3 CD147
Table 2. Blood group antigens within the OK system
Antigen number OK1 OK2* OK3*
Antigen name Oka OKGV OKVM
*Provisionally assigned by the ISBT Working Party on Red Cell 
Immunogenetics and Blood Group Terminology (2010 meeting 
report in preparation).
Review
IMMUNOHEMATOLOGY, Volume 26, Number 3, 2010 125
 The OK:-2 phenotype described in 2003 was shown 
to arise from a nucleotide substitution in exon 2 of BSG: 
176G>T, which encoded the amino acid change Gly59Val.3 
This mutation was found only in the proposita, together 
with two additional silent single-nucleotide exchanges; 
195C>T (Asp65) and 234G>C (Ser78). Interestingly, an ad-
ditional silent transition, 327T>C (Ala108), was identified 
in one of the control samples sequenced in this study. Crew 
and her colleagues also determined the molecular basis of 
the OK:-3 phenotype; sequence analysis of BSG revealed 
another mutation very close by in exon 2, 178G>A, causing 
a Val60Met change in basigin.4
Biochemistry
 As mentioned above, the OK blood group antigens are 
located on basigin (CD147), which is an immunoglobulin 
superfamily glycoprotein having an apparent molecular 
weight on SDS-PAGE of 35,000 to 68,000. It is widely dis-
tributed in human tissues and is present on leukocytes and 
RBCs. Spring and colleagues used the monoclonal antibody 
MA103 to immunopurify the OK glycoprotein.6 MA103 
defines an epitope on the OK glycoprotein but reacts with 
Ok(a–) RBCs. They found that the N-terminal 30 amino ac-
ids were identical to those of the M6 glycoprotein and went 
on to confirm that M6 was the carrier protein for the Oka 
antigen. M6 glycoprotein was also known as the M6 leuko-
cyte activation antigen, basigin (BSG), or the extracellular 
matrix metalloproteinase inducer, EMMPRIN.8
 Extensive studies have been performed on the tissue 
distribution and the role of CD147 in functions such as cell 
adhesion and tumor invasion. Basigin induces the produc-
tion of matrix metalloproteinases in normal fibroblasts as 
well as in tumor cells. It also induces expression of vascular 
endothelial growth factor and hyaluronan and thus stimu-
lates angiogenesis in tumor tissue.9 Basigin has been shown 
to regulate lymphocyte responsiveness, monocarboxylate 
transporter expression, and spermatogenesis, and more re-
cently research has revealed its role as a signaling receptor 
for the extracellular cyclophilin family of proteins.10 It has 
also been reported that basigin facilitates HIV-1 infection 
by an interaction with the virus-associated cyclophilin A on 
the target cell.11
Antibodies in the System
  The first reported anti-Oka (Mrs. S.Ko.G.) was 
detected in a patient who had a history of previous blood 
transfusions, but no history of pregnancy. The anti-Oka was 
determined to be an IgG antibody using the 0.01 M DTT 
test (which differentiates between IgM and IgG antibod-
ies) and was also shown to be monospecific. The titer was 
128 using the antiglobulin technique, and the antibody was 
reactive against both cord and adult RBCs. One other ex-
ample of anti-Oka has been found in Japan.2
 Both OK2 and OK3 antigens were identified by the 
presence of an anti-Oka-like specificity in the plasma of 
both probands, although further examples have not been 
described.
 Despite the paucity of human polyclonal antibodies to 
antigens on CD147, there are monoclonal antibodies that 
detect the protein. The monoclonal anti-Oka TRA-1-185 is 
an IgG1 monoclonal antibody that reacts by the antiglobu-
lin test with all RBC samples tested except Ok(a–) RBCs. 
TRA-1-185 was prepared using spleen cells from a mouse 
immunized with a human teratocarcinoma cell line.
 The monoclonal antibody MA10312 and a number of 
other monoclonal antibodies2 have been produced; these 
antibodies recognize an epitope on the OK glycoprotein and 
react with CD147 but do not have Oka specificity.
 Both human anti-Oka and the monoclonal antibody 
TRA-1-85 react with RBCs that have been pretreated with 
AET, chymotrypsin, neuraminidase, papain, pronase, and 
trypsin.13
 Examples of human Oka antibodies are rare. Three 
novel monoclonal Oka antibodies have been generated that 
may be useful reagents in the laboratory for typing patients’ 
RBCs and for screening donors to find the very rare Ok(a–) 
blood.14
Clinical Significance
 The first example of anti-Oka was clearly shown to be 
of clinical significance. RBC survival studies showed that 
when 10 mL of Ok(a+) whole blood labeled with 51Cr was in-
jected into the patient, only 10 percent of the Ok(a+) RBCs 
survived 3 hours later. The percentage of donor RBCs sur-
viving was reduced to 2 percent at 6 hours.1 In addition, a 
human macrophage assay gave values that correlated with 
antibodies known to cause significant shortening of RBC 
survival.
 There have been no reports of anti-Oka causing hemo-
lytic transfusion reactions or hemolytic disease of the fetus 
and newborn, perhaps not surprising given the rarity of 
the phenotype.
Summary
 The OK blood group system is rarely encountered in 
transfusion medicine because of the high incidence of its 
three antigens and the sparse examples of the correspond-
ing antibodies. The carrier protein basigin, however, is 
widely distributed throughout the body and is functionally 
very important as a signaling molecule in a variety of 
normal and pathogenic processes.
References
1. Morel PA, Hamilton HB. Oka: an erythrocytic antigen of 
high frequency. Vox Sang 1979;36:182–5.
2. Daniels G. Human blood groups. 2nd ed. Oxford, UK: 
Blackwell Scientific, 2002.
3. Karamatic Crew V, Daniels G, Poole J. A new vari-
ant in the Ok blood group system. Transfus Med 
2003;13(Suppl 1):32.
Review of OK system
IMMUNOHEMATOLOGY, Volume 26, Number 3, 2010126
E.A. Smart and J.R. Storry
4. Karamatic Crew V, Thomas R, Gillen B, Poole J, and 
Daniels G. A novel variant in the Ok blood group sys-
tem. Transfus Med 2006;16(Suppl 1):41.
5. Kaname T, Miyauchi T, Kuwano A, Matsuda Y, Mura-
matsu T, Kajii T. Mapping basigin (BSG), a member of 
the immunoglobulin superfamily, to 19p13.3. Cytogenet 
Cell Genet 1993;64:195–7.
6. Spring FA, Holmes CH, Simpson KL, et al. The Oka 
blood group antigen is a marker for the M6 leukocyte 
activation antigen, the human homolog of OX-47 anti-
gen, basigin and neurothelin, an immunoglobulin su-
perfamily molecule that is widely expressed in human 
cells and tissues. Eur J Immunol 1997;27:891–7.
7. Williams BP, Daniels GL, Pym B, et al. Biochemical and 
genetic analysis of the Oka blood group antigen. Immu-
nogenetics 1988;27:322–9.
8. Biswas C, Zhang Y, DeCastro R, et al. The human tumor 
cell-derived collagenase stimulatory factor (renamed 
EMMPRIN) is a member of the immunoglobulin super-
family. Cancer Res 1995;55:434–9.
9. Nabeshima K, Iwasaki H, Koga K, Hojo H, Suzumiya 
J, Kikuchi M. Emmprin (basigin/CD147): matrix me-
talloproteinase modulator and multifunctional cell rec-
ognition molecule that plays a critical role in cancer 
progression. Pathol Int 2006;56:359–67.
10. Yurchenko V, Constant S, Bukrinsky M. Dealing with 
the family: CD147 interactions with cyclophilins. Im-
munology 2006;117:301–9.
11. Pushkarsky T, Zybarth G, Dubrovsky L, et al. CD147 
facilitates HIV-1 infection by interacting with virus-
associated cyclophilin A. Proc Natl Acad Sci U S A 
2001;98:6360–5.
12. Mattes MJ, Cairncross JG, Old LJ, Lloyd KO. Monoclo-
nal antibodies to three widely distributed human cell 
surface antigens. Hybridoma 1983;2:253–64.
13. Reid ME, Lomas-Francis C. The blood group antigen 
factsbook. 2nd ed. San Diego, CA: Academic Press, 
2003.
14. Tian MH, Halverson GR. Characterization of three 
novel monoclonal anti-Oka. Immunohematology 
2009;25:174–8.
Elizabeth A. Smart, Dip. Med. Tech. SA (Blood Trans-
fusion), Durban, South Africa, and Jill R. Storry, PhD, 
FIBMS (corresponding author), Clinical Immunology and 
Transfusion Medicine, Regional and University Laborato-
ries, Klinikgatan 21, SE-22185, Lund, Sweden.
Free Classified Ads and Announcements  
Immunohematology will publish classified ads and announcements (SBB schools, meetings, symposia, etc.) 
without charge.  Deadlines for receipt of these items are as follows:
Deadlines
1st week in January for the March issue
1st week in April for the June issue
1st week in July for the September issue
1st week in October for the December issue
E-mail or fax these items to immuno@usa.redcross.org or (215) 451-2538.
Immunohematology is on the Web!
www.redcross.org/immunohematology
For more information, send an e-mail message 
to immuno@usa.redcross.org
Important Notice About Manuscripts
for
Immunohematology
Please e-mail all manuscripts for consideration to  
immuno@usa.redcross.org


The OK blood group system: a review

Meny, Geralyn M.

Introduction to the 25th anniversary issues

Not a member yet