Possible Polyphyletic Origin of Major Histocompatibility Complex Class I Chain-Related Gene A (MICA) Alleles Mun-kit Choy, Maude E. Phipps Allied Health Sciences Department, Faculty of Medicine, University of Malaya, 50603 Kuala Lumpur, Malaysia Received: 29 March 2002 / Accepted: 30 December 2002 Abstract. Phylogenetic relationships among 23 nonhuman primate (NHP) major histocompatibility complex class I chain-related gene (MIC) sequences, 54 confirmed human MICA alleles, and 16 human MICE alleles were constructed with methods of se- quence analysis. Topology of the phylogenetic tree showed separation between NHP MICs and human MICs. For human MICs, the topology indicated monophyly for the MICB alleles, while MICA alleles were separated into two lineages, LI and LII. Of these, LI MICA alleles shared a common ancestry withgorilla(Ggo)MIC.Oneconservativeaminoacid difference and two nonconservative amino acid dif- ferences in the a3 domain were found between the MICA lineages. The nonconservative amino acid differences might imply structural and functional differences. Transmembrane (TM) trinucleotide-re- peat variants were found to be specific to the MICA lineages such as A4, A9, and A10 to LI and A5 to LII. Variants such as A5.1 and A6 were commonly found in both MICA lineages. Based on these anal- yses, we postulate a polyphyletic origin for MICA alleles and their division into two lineages, LI and LII. As such, there would be 30 alleles in LI and 24 alleles in LII, thereby reducing the current level of polymorphism that exists, based on a presumed monophyletic origin. The lower degree of polymor- phism in MICA would then be in line with the rest of the human major histocompatibility complex non- classical class I genes. Key words: MICA — Alleles — Phylogenetics — Polyphyly — Transmembrane trinucleotide re- peat Introduction Major histocompatibility complex (MHC) class I chain-related genes (MIC) and Perth beta block transcript 11 (PERB11), which are now widely ac- cepted to be the same loci, were identified separately by Bahram et al. (1994) and Leelayuwat et al. (1994). The human MIC gene family consists of seven members, MICA to MICG (Bahram 2001). Only MICA and MICB are functionally expressed; MICC, MICD, MICE, MICF, and MICG are pseudogenes. MICA and MICB products share 90% amino acid homology but their sequences show only 19, 25, and 35% similarity to the extracellular domains (a1, a2, and a3) of other MHC class I genes (Stephens et al. 1999; Zhang et al. 2000). MICA acts as a ligand for natural killer (NK) cells, cd Tcells,and ab CD8 + Tcells,whichexpressa common activating NK cell receptor, NKG2D, co- localized with DNAX-activation protein (DAP) 10, a transmembrane signaling adapter protein (Collins et al. 2002). MICA (11.7 kb) is located within the human MHC gene complex on chromosome 6, ap- proximately 46.4 kb centromeric to human leukocyte antigen (HLA)-B (Stephens 2001). The domain J Mol Evol (2003) 57:38–43 DOI: 10.1007/s00239-002-2444-8 Correspondence to: Assoc. Prof. Dr. Maude E. Phipps; email: phipps@ummc.edu.my