ABO Blood Group Antigen Changes in Acute Myeloid Leukemia and No Significant Association With RUNX1 and GATA2 Somatic Variants

In Hwa Jeong; Ji Young Seo; Sooin Choi; Hyun-Young Kim; Duck Cho

doi:10.3343/alm.2023.43.6.635

Dear Editor,

A or B antigen loss in the ABO blood group system is observed in patients with hematologic malignancies [1, 2]. As A or B antigen expression relies on the transcriptional regulation of ABO, regulatory regions and related transcription factors have been studied to understand the molecular basis of these weak phenotypes. The runt-related transcription factor 1 (RUNX1) and GATA-binding protein 2 (GATA2) genes encode transcription factors that bind to specific ABO motifs [3]. The +5.8-kb site in intron 1 of ABO includes one recognition motif for the RUNX1 protein and two motifs for the GATA2 protein [4]. Associations with somatic variants of genes such as RUNX1 and GATA2 have been studied; however, their relationship has not yet been validated in clinical samples [4, 5]. We expand on the previous discoveries of A or B antigen loss in AML patients and examine whether they are correlated with RUNX1 and GATA2 somatic variants. The study was approved by Institutional Review Board of Samsung Medical Center (2022-08-094).

We retrospectively analyzed acute leukemia patients between June 2018 and July 2021. Serologic typing was performed using Galileo Neo (Immucor, Norcross, GA, USA) and QWALYS-3 (DIAGAST, Loos, France). The variants were screened using a targeted next-generation sequencing panel for 36 AML-related genes with genomic DNA extracted from bone marrow aspirates. Sequencing libraries were prepared using custom probes by IDT xGen (Integrated DNA Technologies, Coralville, IA, USA) and sequenced on NextSeq 550Dx (Illumina, San Diego, CA, USA).

Among 264 newly diagnosed acute leukemia patients, 27.3% (72/264) had blood type O and 72.7% (192/264) had blood type A, B, or AB. Among the 192 non-O acute leukemia patients, 37.0% (71/192) had ALL and 63.0% (121/192) had AML. Weak antigen expression was not observed in any ALL patient but was observed in 6.6% (8/121; patients with blood type A, N=6; B, N=0; AB, N=2) of the 121 AML patients. Among the AML patients with weak antigens at diagnosis, weak A antigen was found in all patients (100%), whereas weak B antigen was observed in one patient (12.5%). Antigen recovery was investigated in the eight AML patients with weak antigen expression (Table 1). After treatment, blood group antigen recovery was observed in all patients, except one, and the time to antigen recovery varied from one to five months. Antigen expression decrease was dominant in the A antigen. Among these eight AML patients, four were tested for the presence of disease-associated somatic variants, which revealed that two patients harbored RUNX1 or GATA2 variant (Table 2). Among the 113 AML patients without blood group antigen loss, 81 patients were tested for somatic variants, and RUNX1 or GATA2 variants were detected in 14. There was no significant association between ABO antigen changes and the presence of RUNX1 and GATA2 variants (P=0.159).

There were more patients with A antigen loss than with B antigen loss, which is interesting considering that the percentages of A and B blood types in the Korean ethnic population are 34% and 27%, respectively [6]. Examination of the antigen recovery time and agglutination strength change revealed that red blood cell (RBC) antigen recovers after induction or consolidation chemotherapy. Based on transient A or B antigen loss or weak expression, we hypothesize that the putative somatic aberrations that triggered the hematologic malignancies also affected the RBCs, hence contributing to weak ABO antigen expression.

Mechanistically, ABO antigen changes are largely due to hypermethylation of the ABO promoter region, leading to loss of protein expression [7 -9]. However, the exact molecular mechanism of antigen decrease, especially in patients with hematologic disease, has not been fully elucidated. Hayakawa, et al. [5] investigated the mechanism underlying A antigen reduction in an MDS patient and identified a frameshift variant of RUNX1 caused by a two-nucleotide deletion in exon 4 as a potential reason for A antigen loss on RBCs. They also speculated that GATA2 somatic variants may also be associated with A antigen loss in MDS patients [4]. However, in our cohort, RUNX1 or GATA2 somatic variants were not associated with weak A antigen expression in AML patients.

This study is subject to limitations as it focused solely on the analysis of acute leukemia patients. Further investigation is required to comprehend the correlation between decreased RBC antigen expression and somatic variants. Nevertheless, the examination of clinical samples led to the retrieval of results that contradict previous experimental studies. These findings facilitate a better understanding of the relationship between the expression of ABO blood group antigens and AML.

ACKNOWLEDGEMENTS

None.

Notes

AUTHOR CONTRIBUTIONS

Cho D and Choi S designed the study, Seo JY collected data, Jeong IH wrote the first draft, and Cho D and Kim HY revised and supervised the study. All authors read and approved the final manuscript.

CONFLICTS OF INTEREST

None declared.

REFERENCES

1. Schenkel-Brunner H. 2000. Human blood groups. Springer;Wien:

2. Daniels G. 2013. Human blood groups. Wiley-Blackwell;West Sussex:

3. Kominato Y, Sano R, Takahashi Y, Hayakawa A, Ogasawara K. 2020; Human ABO gene transcriptional regulation. Transfusion. 60:860–9. DOI: 10.1111/trf.15760. PMID: 32216153. PMCID: PMC7187371.

4. Hayakawa A, Sano R, Takahashi Y, Okawa T, Kubo R, Harada M, et al. 2022; Reduction of blood group A antigen on erythrocytes in a patient with myelodysplastic syndrome harboring somatic mutations in RUNX1 and GATA2. Transfusion. 62:469–80. DOI: 10.1111/trf.16766. PMID: 34918362.

5. Hayakawa A, Sano R, Takahashi Y, Kubo R, Harada M, Omata M, et al. 2020; RUNX1 mutation in a patient with myelodysplastic syndrome and decreased erythrocyte expression of blood group A antigen. Transfusion. 60:184–96. DOI: 10.1111/trf.15628. PMID: 31840280.

6. Choi SY, Kim SI, Cho HI. 1984; Study on gene frequencies of blood groups in Koreans. Korean J Hematol. 19:63–75.

7. Orlow I, Lacombe L, Pellicer I, Rabbani F, Delgado R, Zhang ZF, et al. 1998; Genotypic and phenotypic characterization of the histoblood group ABO(H) in primary bladder tumors. Int J Cancer. 75:819–24. DOI: 10.1002/(SICI)1097-0215(19980316)75:6<819::AID-IJC1>3.0.CO;2-Y.

8. Iwamoto S, Withers DA, Handa K, Hakomori S. 1999; Deletion of A-antigen in a human cancer cell line is associated with reduced promoter activity of CBF/NF-Y binding region, and possibly with enhanced DNA methylation of A transferase promoter. Glycoconj J. 16:659–66. DOI: 10.1023/A:1007085202379. PMID: 10972144.

9. Kominato Y, Hata Y, Takizawa H, Tsuchiya T, Tsukada J, Yamamoto F. 1999; Expression of human histo-blood group ABO genes is dependent upon DNA methylation of the promoter region. J Biol Chem. 274:37240–50. DOI: 10.1074/jbc.274.52.37240. PMID: 10601288.

10. Li MM, Datto M, Duncavage EJ, Kulkarni S, Lindeman NI, Roy S, et al. 2017; Standards and guidelines for the interpretation and reporting of sequence variants in cancer: a joint consensus recommendation of the Association for Molecular Pathology, American Society of Clinical Oncology, and College of American Pathologists. J Mol Diagn. 19:4–23. DOI: 10.1016/j.jmoldx.2016.10.002. PMID: 27993330. PMCID: PMC5707196.

Table 1

List of the eight AML patients showing change in agglutination strengths during follow-up

Case number	Diagnosis	Sex/age (yr)	Blood type	RBC blood typing by serologic method								BMT	Somatic variant test

				At diagnosis	During follow-up (months)

					1	1.5	2	2.5	3	5	12
1	AML-MRC	M/63	A	Trace	Trace	Trace	4+	4+				Yes	Yes
2	t-AML	F/58	A	2+	3+	N/T	4+					No	No
3	AML	M/78	A	2+	1+	Trace	Trace	N/T	Trace	4+		No	Yes
4	AML	M/75	A	Trace	Trace							No	Yes
5	AML	M/49	A	2+	4+	4+	4+	N/T	4+	N/T	4+	Yes	Yes
6	AML-MRC	M/48	A	3+	3+	3+	3+	4+				Yes	No
7*	AML	M/71	AB	2+	1+	4+	4+	N/T	4+	N/T	4+	Yes	No
8	AML	M/18	AB	1+/1+	2+/4+	3+/4+	4+/4+	4+/4+	4+/4+	4+/4+		Yes	No

Agglutination grading from “negative” to “3+” on automated analyzers and observation of “mixed field” agglutination on manual typing were defined as weak antigen expression. Blood types of allogenic peripheral blood stem cell transplantation (aPBSCT) donors did not affect agglutination strength as the antigens recovered before aPBSCT in all eight patients. The patient marked with an asterisk (*) only exhibited weak A antigen. Expression. Case number. 4 was not followed up because of referral to another hospital.

Abbreviations: AML-MRC, AML with myelodysplasia-related changes; BMT, bone marrow transplantation; N/T, not tested; t-AML, therapy-related AML; RBC, red blood cell; M, male; F, female.

Table 2

RUNX1 and GATA2 somatic variants detected in AML patients with weak and normal ABO group antigens

Group Patients tested, N	Weak ABO antigen (N=8) 4		Normal ABO antigen (N=113) 81

Gene	RUNX1 (N=1)	GATA2 (N=1)	RUNX1 (N=9)	GATA2 (N=5)
Variant description	c.1274C>T; p.P425L	c.1085G>A; p.R362Q	c.1274C>T; p.P425L c.497G>A; p.R166Q c.1129_1130del; p.Y377fs222 c.611G>A; p.R204Q c.406_433del; p.N136Efs7 c.445G>C; p.A149P c.1415T>C; p.L472P c.508+4_508+5ins15; p.? c.508+4_508+5ins21; p.? c.439_440insGAG; p.N146_A147insG	c.1219A>T; p.S407C c.1121G>T; p.G374V c.1085G>A; p.R362Q c.1114G>A; p.A372T c.1127A>G; p.Y376C

Only tier I and II variants according to the 2017 Association for Molecular Pathology/American Society of Clinical Oncology/College of American Pathologists guideline [10] detected in the RUNX1 and GATA2 genes are listed. Fisher’s exact test showed no significant difference between the weak ABO antigen and normal antigen groups (P=0.159). The two variants marked with asterisk (*) were detected in the same patient. National Center for Biotechnology Information IDs: NM_001754.4 (RUNX1) and NM_032638.4 (GATA2).