Prediction of immunogenicity of Rh antigens using in silico analysis of binding to human leukocyte antigen peptide : In silico analysis of HLA class II restricted RBC peptide binding = 인실리코 분석을 통한 조직적합성항원(Human Leukocyte Antigen, HLA) 펩타이드 결합 기반 Rh 항원의 면역원성 예측

다국어 초록 (Multilingual Abstract)

Background: Red blood cell (RBC) alloimmunization remains a major complication of blood transfusion, particularly in patients requiring repeated transfusions. Although human leukocyte antigen (HLA) class II molecules are known to play a central role in antigen presentation to CD4⁺ T cells, the mechanistic association between HLA polymorphisms and the immunogenicity of blood group antigens has not been fully elucidated. Previous studies have predominantly focused on predicting immunization events in cancer immunotherapy, but not blood group antigens.

Aim: This study aimed to investigate whether HLA peptide binding predictions could explain the immunogenicity of major blood group antigens, with a particular focus on the Rh blood group system, using an in silico approach.

Materials and Methods: This study performed in silico binding analysis of Rh antigens and representative HLA class I and class II alleles using the NetMHCpan-4.1 (HLA class I) and NetMHCIIpan-4.1 (HLA class II) algorithms. Strong binding regions (“hotspots”) were identified based on predicted binding affinity and percentile rank. HLA allele frequency data across different ethnic populations were incorporated to evaluate population-specific hotspot distributions.

Results: A pilot analysis was conducted using representative HLA alleles and six Rh system antigens, including normal RhD, weak D variants, and RhCE. Binding predictions were initially assessed for HLA class I molecules and subsequently expanded to include HLA class II loci, such as HLA-DRB, -DQA-DQB, and -DPA-DPB. HLA class I binding analysis demonstrated scattered and nonsignificant strong binding regions across Rh antigens. In contrast, HLA class II analysis revealed distinct and clustered hotspot distributions that varied by HLA locus. Both RhD and RhCE antigens exhibited multiple strong binding hotspots for HLA-DRB, -DQA-DQB, and -DPA-DPB molecules. Notably, a unique hotspot shift was observed in RHD*01W.1, corresponding to a p.Val270Gly substitution, which differed from RHD*01.01, RHD*01W.2, and RHD*01W.3. Hotspots were distributed across exofacial, transmembrane, and intracellular domains of the Rh proteins. The distribution of hotspots differed across HLA loci and ethnic groups.

Conclusions: This in silico approach provides novel insights into understanding and managing alloimmunization events. These findings highlight the potential clinical utility of HLA peptide binding prediction in assessing alloimmunization risk and guiding transfusion strategies, particularly in patients with multiple alloantibodies. Taken together, the results of this study indicate that the proposed in silico approach offers a novel conceptual framework for understanding blood group antigen immunogenicity beyond conventional serologic observations.

목차 (Table of Contents)

• ABSTRACT i
• CONTENTS iii
• LIST OF TABLES v
• LIST OF FIGURES vi
• INTRODUCTION 1
• MATERIALS AND METHODS 6
• 1. Blood group antigen selection 6
• 2. Pilot study 7
• 3. Selecting an in silico binding prediction algorithm 7
• 4. HLA allele selection 8
• 5. Statistical analysis 8
• RESULTS 10
• 1. Pilot study of Rh antigen binding to HLA class I 10
• 2. Pilot study of Rh antigen binding to a part of HLA class II 12
• 3. Hotspot analysis of Rh antigens and HLA class II alleles, with emphasis on the p.Val270Gly substitution in RHD*01W.1 14
• 4. Predicted core amino acids (hotspot regions) of Rh antigens that bind to HLA class II peptides 20
• 5. Analysis of HLA-DRB1 hotspot regions among the four ethnic groups 25
• 6. HLA-DRB1 hotspot regions among Korean and Japanese populations 36
• DISCUSSION 40
• REFERENCES 49
• KOREAN ABSTRACT 54