自然殺手細胞類免疫球蛋白受體基因群的基因多樣性與結構複雜性：人類泛基因體組裝序列的詳盡分析

Abstract

殺手細胞免疫球蛋白樣受體（killer-cell immunoglobulin-like receptor, KIR）基因複合體是人類基因組中多型性最強、結構最複雜的區域之一，對自然殺手（NK）細胞的功能具有關鍵影響。雖然 KIR 多樣性在感染狀況、癌症進展以及移植相容性方面的重要性早已獲得肯定，但之前由於技術上遭遇的困難，始終缺乏對 KIR 區域的高解析度定型方法。本文利用Human Pangenome Reference Consortium (HPRC) 所產生的高品質且定相（haplotype-resolved）基因體組裝序列，針對 47 位來自多元族群的個體（共94條單套染色體）進行KIR多樣性的完整描繪。 為達此目標，我開發並應用了名為「Structural KIR annoTator (SKIRT)」的生物資訊流程，將IPD-KIR資料庫裡完整的KIR等位基因序列對齊至組裝後的基因體contigs，以解析基因層級的拷貝數、等位基因差異，以及結構重排情形。透過廣泛的生物資訊分析與實驗室驗證（包含使用特異性引子對的聚合酶連鎖反應 PCR-SSP、定量 PCR(qPCR)，以及Sanger定序確認新穎斷點），SKIRT能夠達到七位數字等位基因的高精度解析。研究結果顯示，本方法鑑定出 570 筆新發現的KIR等位基因、多種基因拷貝數變異（CNVs），以及例如KIR2DS4/3DL1、KIR2DL3/2DP1等不預期的基因融合（fusion），凸顯該區域能產生新型受體構造的潛力。值得注意的是，許多新發現的等位基因帶有改變胺基酸序列的變異（nonsynonymous）或調控序列異常，顯示它們可能影響NK細胞的功能。 比較分析也顯示，SKIRT無論在結構變異偵測或等位基因分型層面皆優於現今存在的其他分析工具，成為KIR基因定型的新標竿。此發現一方面擴充了對人類免疫多樣性的認知，另一方面亦彰顯了「組裝式分析」如何有效解析基因組中的“暗區域”。從臨床角度來看，更完整的 KIR 單套基因體與等位基因變異資訊，有助於更精確地進行免疫基因體學研究，包括疾病關聯分析及移植免疫領域。

The killer-cell immunoglobulin-like receptor (KIR) gene complex is among the most polymorphic and structurally complex regions of the human genome, controlling key aspects of natural killer (NK) cell function. Despite the established importance of KIR diversity in infection outcomes, cancer progression, and transplant compatibility, systematic high-resolution genotyping of the KIR region has long been hindered by technical challenges. In this study, I leverage high-fidelity haplotype-resolved genome assemblies from the Human Pangenome Reference Consortium (HPRC) to comprehensively map KIR variation in 47 diverse individuals (94 haplotypes). To accomplish this, I developed and applied the Structural KIR annoTator (SKIRT) pipeline, which aligns complete KIR allele sequences from the IPD-KIR database to assembled genome contigs, resolving gene-level copy number, allele-level differences, and structural rearrangements. Through an extensive bioinformatic and laboratory validation process—including PCR with sequence-specific primers (PCR-SSP), quantitative PCR (qPCR), and Sanger sequencing of novel junctions—SKIRT achieved up to seven-digit allele resolution. The method revealed 570 novel KIR alleles, numerous copy number variants (CNVs), and unexpected gene fusions such as KIR2DS4/3DL1 and KIR2DL3/2DP1, underscoring the region’s capacity for generating new receptor configurations. Many of these newly discovered alleles harbor nonsynonymous substitutions or altered regulatory sequences, suggesting a potential impact on NK cell function. Comparative analysis showed that SKIRT outperforms existing pipelines in structural variant detection and allele calling, setting a benchmark for KIR genotyping. These findings expand our fundamental understanding of human immune diversity and underscore how assembly-based approaches can illuminate genomic “dark regions.” From a clinical perspective, this enhanced knowledge of KIR haplotypes and allele-level variation paves the way for more precise immunogenetic research, including disease association studies and transplantation immunology.