The MICA gene was described in 1994 among a group of genes within the MHC class I region (Bahram et al., 1994). It is a member of the MIC gene family, which consists of 7 members (MICA-MICG). MICA is classified as a non-classical MHC class I gene as opposed to classical MHC class I genes encoding the commonly known Human Leukocyte Antigen (HLA) proteins. Similar to the HLA genes, the MICA gene is highly polymorphic. One hundred alleles have been reported (according to IMGT Release 3.17.0, 2014-07-14), with new alleles being continuously described and added to the database. As such it is significantly less polymorphic than HLA loci A (2,884 alleles), B (3,589), C (2,375), DRB1 (1,540), DQB1 (664), and DPB1 (422). However, MICA is more polymorphic than HLA loci G (50), DRB3 (58), DRB4 (15), DRB5 (20), DRA (7), DQA1 (52), DPA1 (38), and its closely related MICB gene (40) (Robinson et al., 2013).
MICA gene is organized into 7 exons of which exon 5 encodes the transmembrane (TM) region of the MICA molecule. TM encodes the repeat polymorphism (GCT/Ala) and eight types of repeats have been described as A4, A5, A5.1, A6, A7, A8, A9, and A10 (Gambelunghe et al., 2006; Zou et al., 2006). The combinations of extracellular and TM types facilitates the identification of the MICA alleles and reduces the number of potential ambiguous typings in heterozygous individuals. This combination identifies MICA alleles based on polymorphisms in the TM region as well as elsewhere, e.g., MICA*007:01/A4, MICA*008:01/A5.1, etc.

MICA gene encodes 383 amino acids polypeptide. MICA is up-regulated on human endothelial cells by the pro-inflammatory cytokine TNFα. This up-regulation is controlled at the transcriptional level by a master regulatory control element positioned 130-base pair (bp) upstream of the MICA transcription start site integrating input from the NF-kB and heat shock pathways (Lin et al., 2012).

Of the MIC gene family, only MICA and MICB are true genes with protein products while the remaining (MICC-MICG) are pseudogenes.

Due to synonymous substitutions being the only differences among some of the MICA alleles, the 100 MICA alleles encode for only 79 distinct proteins. Two of the MICA alleles are null alleles with no expressed protein products. MICA molecules are considered non-classical MHC class I molecules rather than human leukocyte antigens (HLA) since they are not expressed on the surface of human leukocytes. Nevertheless they are expressed on endothelial cells, dendritic cells, fibroblasts, epithelial cells, and many tumors and serve as targets for both cellular and humoral immune responses (Bahram et al., 1994; Zou and Stastny, 2010). MICA protein at normal states has a low level of expression in epithelial tissues but is upregulated in many tumors and under various stimuli of cellular stress including heat shock proteins (Groh et al., 1996). Similar to MHC class I, MICA molecules have 3 extracellular immunoglobulin-like domains, a transmembrane domain and intracellular cytoplasmic tail. However, unlike class I, MICA is not covalently bound a monomorphic β2 microglobulin and its peptide binding groove is empty and does not present peptides (Figure 3).
Allelic variants of MICA based on a single amino acid substitution at position 129 in the α2 domain have been reported to result in large differences in the affinity of binding to the activating natural killer group 2, member D (NKG2D) (Steinle et al., 2001). MICA alleles with a methionine (M) or valine (V) have been classified as having strong or weak binding affinity for NKG2D, respectively. These variable affinities have been suggested to affect thresholds of NK cell triggering and T cell modulation and consequently influencing clinical phenotypes in autoimmune disorders and malignancies (Amroun et al., 2005; Douik et al., 2009).
MICA molecules exist also in soluble forms (sMICA) encompassing three extracellular domains (Salih et al., 2002). ADAM, a disintegrin and metalloproteinase, is reported to mediate sMICA generation by cleavage of the molecule in the α3 domain close to the papain cleavage site (Figure X), however the precise location of the cleavage site is still unknown (Waldhauer et al., 2008). sMICA are not normally detected in sera of healthy donors and tumor cells are the major source of sMICA generation (Holdenrieder et al., 2006). In addition to patients with malignancies, sMICA is detected in the sera of patients with autoimmune diseases, acute bacterial infections, renal insufficiency, and cholestasis (Holdenrieder et al., 2007). Unlike the surface-bound form of MICA that stimulates immune responses, the secreted soluble counterpart, sMICA abates immune responses primarily by down regulating NKG2D expression which impairs the cellular cytotoxicity of T cells and NK cells against tumor cells (Groh et al., 2002). This may partly explain why higher levels of sMICA were observed in the serum of chronically infected individuals compared to healthy controls and HIV-1 controllers (Nolting et al., 2010). Similarly CMV infection triggers shedding of sMICA (Andresen et al., 2009).

383 amino acids in length (including the 23 amino acids leader sequence that is lost from the 360 amino acid mature protein), 42,915 Da protein, undergoes N-glycoslation as post-translational modification, although not required to interact with NKG2D. MICA has three external (α1, α2, α3), a transmembrane and an intracytoplasmic domains (figure 4).

Co-dominantly and constitutively expressed on cell membranes of human epithelial, endothelial, fibroblasts cells, keratinocytes, monocytes, dendritic cells, thymic medulla, and gastrointestinal epithelial cells but not on the surface of other healthy cells (Zwirner et al., 1999). Activated CD4+ and CD8+ T cells are shown to express MICA. Nuclear factor (NF)-kB regulates MICA expression on activated T lymphocytes by binding to a specific sequence in the long intron 1 of the MICA gene (Molinero et al., 2004). MICA expression is also up-regulated on stressed cells, tumour cells, and pathogen-infected cells (Mistry and OCallaghan, 2007). Its noteworthy that in vitro CMV infection strongly induces expression of MICA (Groh et al., 2001). Similarly, respiratory syncytial virus (RSV) infection of epithelial cells up-regulated cell surface expression of MICA and levels of soluble MICA (Zdrenghea et al., 2012). It is also reported that CMV induced expression on the surface of fibroblasts is skewed towards a common form of MICA (A5.1), which has a nucleotide insertion in exon 5 corresponding to the transmembrane region and no cytoplasmic tail and less activation of the NKG2D pathway (Zou et al., 2005).

Cell surface as a single-pass type 1 membrane protein

Like other MHC Class I molecules, MICA is a highly polymorphic MHC Class I molecule expressed on the cell surface of cells mentioned above. However, unlike other classical MHC Class I molecules, MICA is not involved in antigen presentation because its peptide-binding groove is too narrow to present antigens and is not associated with β2-microglobulin (Groh et al., 1996). MICA is expressed under cellular stress and is a ligand for NKG2D receptor expressed on surface of NK, NKT, CD8+ and TCRγδ T cells (Bauer et al., 1999). NKG2D binding results in up-regulation of MICA and ultimately cytotoxicity and release of IFN? by NKG2D-expressing cells.

Searching the non-redundant protein sequences (nr) database for the protein sequences similar to the MICA reference amino acid sequence (MICA*001) using Blastp (protein-protein BLAST) tool publicly available from the website of The National Center for Biotechnology Information (https://blast.ncbi.nlm.nih.gov/Blast) showed that the closest protein sequences are MICB followed by MHC class I (Wheeler and Bhagwat, 2007). The phylogenetic distance among these 3 sequences as determined by the online ClustalW2 (version CLUSTAL 2.1) Multiple Sequence Alignments tool publicly available at the website of the European Bioinformatics Institute of the European Molecular Biology Laboratory (EMBL-EBI) (http://www.ebi.ac.uk/Tools/msa/clustalw2/) is shown in figure (5) (McWilliam et al., 2013). Aligning MICA *001 and MICB reference amino acid sequence (MICB *001) using the same ClustalW2 tool yields scores of 83.03% sequence identity (Figure 6). Similarly alignment of MICA*001 and HLA class I consensus amino acid sequence yielded an identity score of 26.47% between (Figure 7).

Similar to HLA genes, each allelic variant of MICA is given a distinct name that is officiated by the WHO Nomenclature Committee for Factors of the HLA System. The following are the 100 recognized MICA alleles (according to IMGT Release 3.17.0, 2014-07-14): MICA*001, MICA*002:01, MICA*002:02, MICA*002:03, MICA*002:04, MICA*004, MICA*005, MICA*006, MICA*007:01, MICA*007:02, MICA*007:03, MICA*007:04, MICA*007:05, MICA*007:06, MICA*008:01:01, MICA*008:01:02, MICA*008:02, MICA*008:03, MICA*008:04, MICA*008:05, MICA*009:01, MICA*009:02, MICA*010:01, MICA*010:02, MICA*011, MICA*012:01, MICA*012:02, MICA*012:03, MICA*012:04, MICA*013, MICA*014, MICA*015, MICA*016, MICA*017, MICA*018:01, MICA*018:02, MICA*019, MICA*020, MICA*022, MICA*023, MICA*024, MICA*025, MICA*026, MICA*027, MICA*028, MICA*029, MICA*030, MICA*031, MICA*032, MICA*033, MICA*034, MICA*035, MICA*036, MICA*037, MICA*038, MICA*039, MICA*040, MICA*041, MICA*042, MICA*043, MICA*044, MICA*045, MICA*046, MICA*047, MICA*048, MICA*049, MICA*050, MICA*051, MICA*052, MICA*053, MICA*054, MICA*055, MICA*056, MICA*057, MICA*058, MICA*059, MICA*060, MICA*061, MICA*062, MICA*063N, MICA*064N, MICA*065, MICA*066, MICA*067, MICA*068, MICA*069, MICA*070, MICA*072, MICA*073, MICA*074, MICA*075, MICA*076, MICA*077, MICA*078, MICA*079, MICA*080, MICA*081, MICA*082, MICA*083, MICA*084 (Robinson et al., 2013).