**3.3 Ligands of NKG2D-ULBPs**

430 Advances in Cancer Therapy

members of the NKG2 family form disulfide-linked heterodimers with CD94. Each member of the NKG2 family, however, has a unique repeat Alu sequence that functions as a gene promoter. Alu repeats are the putative binding sites for several transcription factors: activator protein-1/3 (AP-1/3), cAMP response element binding/activating transcription factors (CREB/ATF), CCAAT/enhancer binding protein (C/EBP), transcription factor-1 (Sp1), nuclear factor-1/CCAAT transcription factor (NF-1/CTF) and lymphocyte-specific DNA-binding protein-1 (LyF-1), and may contribute to differences in gene regulation among family members (Brostjan et al., 2000; Glienke et al., 1998). The CD94 protein associates with different NKG2 isoforms to form heterodimeric receptors that function either to inhibit or to trigger cytotoxicity of NK cells, depending on the NKG2 isoform. Functionally, therefore, CD94/NKG2 heterodimers are divided into activating (NKG2C, NKG2E/H) or inhibitory (NKG2A/B) isotypes (Glienke et al., 1998). It has also been shown that several cytokines, including interleukin (IL)-12 (Derre et al., 2002), transforming growth factor (TGF)-β (Bertone et al., 1999), IL-15 (Mingari et al., 1998) and IL-10 (Romero et al., 2001), are capable of inducing expression of CD94/NKG2 in human NK or T cells, in which IL-15 and TGF- can upregulate CD94/NKG2A, but not other CD94/NKG2 receptors (Wilhelm et al., 2003). NKG2D is a special activating receptor that is not covalently associated with CD94. NKG2D is downmodulated by TGF-β, but markedly upregulated by IL-15, interferon (IFN)- and IL-12 stimulation in primary NK cells (Dasgupta et al., 2005). IFN- can stimulate the expression of NKG2D, but inhibits the expression of the inhibitory receptor NKG2A, and therefore alters the balance of stimulatory and inhibitory receptors in favor of activation, leading to NK-cell-mediated cytotoxicity. IFN-γ addition exerts the

CD94 is a C-type lectin and is required for the dimerization of the CD94/NKG2 family of receptors, which are expressed on NK cells and T-cell subsets. The CD94 and NKG2 genes are located in the center of the NK-gene complex between the NKRP1 and PRB3 loci in the region containing the STS markers D12S77 and D12S1093. All six genes are closely linked and are situated 100 to 200 kb proximal to the D12S77 STS marker (Sobanov et al., 1999). The CD94 gene is placed in the opposite orientation relative to the NKG2 gene family members (which all have the same transcriptional orientation). CD94 gene expression is regulated by distal and proximal promoters that transcribe unique initial exons specific for each promoter. Both promoters contain elements with IFN-γ-activated and E26 transformationspecific (ETS)-binding sites (called GAS and EBS respectively). The two promoters differentially regulate expression of the CD94 gene in response to treatment with IL-2 or IL-

The two types of NKG2D ligands are MHC class I chain-related proteins (MICA and MICB) and UL16-binding proteins (ULBP) (Sutherland et al., 2001; Bauer et al., 1999). MICA and MICB are located in the MHC complex within the 6q21.3 chromosomal region. The ULBP family contains 10 genes, of which five are expressed (ULBP1–5). ULBPs are located outside the MHC gene complex, in the 6q24.2–25.3 region (Samarakoon et al., 2009). These molecules exhibit highly restricted expression in healthy tissues but are widely expressed on epithelial tumors. In hematological malignancies they are upregulated in nontumor and tumor cells by genotoxic stress (Kim et al., 2006; Venkataraman et al., 2007). Aligned cosmid-derived regions 1.5 kb upstream of the translation start codon (ATG) in MICA and MICB share approximately 90% sequence identity. The 5′-end flanking regions of MICA and

opposite effect (Zhang et al., 2005).

15 (Lieto et al., 2003).

**3.2 Ligands of NKG2D-MIC**

All expressed ULBPs (ULBP1–5) have good conservation of the coding sequence, but the 5′ end flanking regions of these genes have little homology. This situation suggests that ULBPs are regulated specifically in response to different stimuli or pathogens (Lopez-Soto et al., 2006). Mechanisms that regulate the expression of ULBP1 have been described previously. In the core region of the ULBP1 promoter, there are many binding sites for transcription factors, including a canonical TATA box, three GC boxes, one overlapping sequence GC(4)/AP-2, two cytokinin response element (CRE)-like sequences (CRE1 and CRE2) (Zeng et al., 1997; Wajapeyee and Somasundaram, 2003), and one NF-κB site. Transcription of ULBP1 depends strictly on binding of Sp1 and Sp3 to a CRE1 site located in the ULBP1 minimal promoter. The mutation or deletion of this Sp1/Sp3 binding site abolished ULBP1 transcription; Sp3 is the main transcription factor that regulates ULBP1 through the CRE1 site. AP-2α, however, repressed the expression of ULBP1 by binding to the GC(4)/AP-2α site and interfering with the binding of Sp3 and Sp1 to the ULBP1 promoter (Lopez-Soto et al., 2006). It has been shown that heat shock and ionizing radiation treatment can induce the expression of ULBP1/2/3 in human cancer cell lines; HSP 70 is induced by heat shock but not by ionizing radiation (Kim et al., 2006). Human cytomegalovirus (HCMV) can also induce the expression of ULBP1/2/3 proteins that are predominantly localized in the endoplasmic reticulum of infected fibroblasts together with UL16 (Rolle et al., 2003). There is interplay, however, between virus and host cells depending on the viral dose. At low viral dose, ULBP1 or ULBP2 surface expression is completely inhibited compared with ULBP3, while at a higher viral doses cell surface expression of ULBP1 and ULBP2 is delayed (Rolle

Transcription Regulation and Epigenetic

that in other genes.

**3.4.2 Histone modifications of KIR** 

**3.4.3 DNA demethylation of KIR** 

expression (Santourlidis et al., 2002; Chan et al., 2005).

Control of Expression of Natural Killer Cell Receptors and Their Ligands 433

et al., 1997; Wilson et al., 2000; Martin et al., 2000). Moreover, KIR2DL4 is the only KIR gene that lacks the repeat region in intron 1. All KIR genes lack typical promoter features such as TATA and CAAT boxes, so many transcription factor binding sites have been predicted to be present in the upstream regions of KIR, including sites for transcription factors CREB, SP1, ETS, AP-1 and AP-4 (Chan et al., 2003; Santourlidis et al., 2002). None of these elements alone, however, is required for or significantly increases promoter activity, except for a CREB site located outside the core promoter region. To date, the basic KIR promoter appears to either require undefined transcription factors or alternatively only interacts with the basic transcription machinery of RNA polymerase II (Trompeter et al., 2005). It has been suggested previously that a binding site for the transcription factor AML/Runx (Runt Homology Domain Transcription Factors) at –100 bp can be essential for KIR expression (Vilches et al., 2000). However, a further study suggested that AML has a general inhibitory influence on KIR expression in mature NK cells and that AML exerts its repressor function by binding directly to the promoter of the different KIR genes (Trompeter et al., 2005). Recently, it was reported that the transcription factor c-Myc upregulated KIR gene transcription through direct binding to an upstream distal promoter element in peripheral blood NK cells, and that IL-15 promoted this effect (Cichocki et al., 2009). It was suggested that the role of the traditional transcription factors in KIR gene expression is different from

It was found that histone H3 and H4 proteins are substantially acetylated at the Lys9 and Lys14 positions and H4 acetylation at the 5th, 8th, 12th, and 16th positions in both KIR3DL1+ and KIR3DL1– NK cells (Chan et al., 2005). The level of KIR3DL1-associated histone acetylation and methylation was higher in KIR3DL1+ NK cells than in KIR3DL1– NK cells; however, histone H3 methylation at Lys4 (H3K4) is only 2.6-fold higher in KIR3DL1+ NK cells than that in KIR3DL1– NK cells, although the histone H3K4 methylation levels correlated well with the KIR3DL1 promoter to lead to upregulation of KIR3DL1 gene expression, named KIR3DL1-associated histone H3K4 methylation. These findings indicate that histone acetylation and trimethylated modifications, but not histone H3 methylation, are preferentially associated with the transcribed allele in NK cells with monoallelic KIR

KIR expression is increased on T cells with increasing age, and can contribute to age-related diseases (van Bergen et al., 2004). In these cells, the KIR2DL4 promoter is partially demethylated, and dimethylated H3L4 is increased; all other histone modifications are characteristic of an inactive promoter. In comparison, NK cells have a fully demethylated KIR2DL4 promoter and have the full spectrum of histone modifications indicative of active transcription with H3 and H4 acetylation. These findings suggest that an increased T-cell ability to express KIR2DL4 with age is conferred by a selective increase in H3L4 dimethylation and limited DNA demethylation (van Bergen et al., 2004; Li et al., 2008).

Methylation status of CpG islands in NK cells correlates with transcriptional activity of KIR genes. The overall structure of the CpG islands is similar in all expressed KIR, with complete conservation of four CpG dinucleotides upstream of the transcriptional start site, with the exception of KIR2DL4, which shows a highly divergent profile with lower CpG density. In

et al., 2003). Whether this phenomenon is mediated through the ULBP promoter or through transcription factors needs to be investigated further.

Epigenetic modulations are also likely to play a large part in the observed alteration in ULBP expression. Current observations imply that loss of DNA methylation (deficient in DNA methyltransferase cells) is correlated with increased ULBP2 mRNA levels. It has been shown that both demethylation of the ULPB2 promoter, necessary for increased protein levels, and the RAS/MEK signaling pathway, shown to control DNA methylation (Lund et al., 2006), are necessary for maximal protein expression (Sers et al., 2009). In addition, the HDAC inhibitor trichostatin A (TSA) is reported to upregulate ULBP1–3 expression in tumor cells; Sp3 was found to be crucial for activation of the ULBP1 promoter by TSA. One report showed that HDAC3 is recruited to the ULBP1 promoter and acts as a repressor of ULBP expression in epithelial cancer cells. TSA treatment interfered with this interaction and caused the complete release of HDAC3 from the ULBP1–3 promoters (Lopez-Soto et al., 2009).

#### **3.4 KIR receptor family and their ligands**

**(**The KIR family is of particular interest because individual members bind to specific subgroups of HLA allele products, such as HLA-A, HLA-B, HLA-C, and HLA-G, although most KIRs show 90–95% amino acid identity. The high level of homology can facilitate exchange of exons between different KIR loci, by some form of crossing over or gene conversion (Wilson et al., 2000). One study found that due to the polymorphism of KIR genes, two KIR haplotypes segregated at roughly equal frequency in a largely Caucasian population. Group A haplotypes contained seven KIR genes and had KIR2DS4 as the only activating receptor. Group B haplotypes had a greater diversity of KIR genes, had more activating receptors, and were characterized by the KIR2DL2, KIR2DS1, KIR2DS2, KIR2DS3, and KIR2DS5 genes (Uhrberg et al., 1997). KIR can also be further subdivided into inhibitory receptors that carry an inhibitory signal motif within their cytoplasmic domain (KIR2DL and KIR3DL) and into stimulatory receptors (KIR2DS and KIR3DS) that lack this motif. Most of the inhibitory KIRs are specific for the products of HLA class I genes like such as HLA-A/-B and HLA-C (Bellon et al., 1999; Colonna et al., 1993; Dohring et al., 1996). The ligand specificities of many stimulatory KIR are uncertain and might include non-HLA class I ligands. KIR2DL4, however, combines structural and functional features of both stimulatory and inhibitory KIR and is reported to bind to the nonclassical class I protein HLA-G (Chan et al., 2003; Martin et al., 2000; Trompeter et al., 2005; Santourlidis et al., 2002). The expression of KIR appears to be largely independent of each other and the impact of the requirement for inhibition by self class I molecules on the shape of the KIR repertoire is rather subtle.

#### **3.4.1 Transcriptional factors regulation of KIR**

Different groups of 2–9 KIR genes are expressed in NK clones, but the sequences of the KIR promoters are homogeneous in their 5′-untranslated regions (5′ UTR), except for that of KIR2DL4 (Valiante et al., 1997; Wilson et al., 2000). The sequences of KIR genes comprise a continuous loop that extends seamlessly from gene to gene. The repeat of the loop is broken only by a region 14 kb upstream of the KIR2DL4 locus, which displays some unique features and is characterized by L1 repeats. The sequence upstream of KIR2DL4 may be significant because this gene is unique in this group in being expressed in 100% of NK clones (Valiante et al., 1997; Wilson et al., 2000; Martin et al., 2000). Moreover, KIR2DL4 is the only KIR gene that lacks the repeat region in intron 1. All KIR genes lack typical promoter features such as TATA and CAAT boxes, so many transcription factor binding sites have been predicted to be present in the upstream regions of KIR, including sites for transcription factors CREB, SP1, ETS, AP-1 and AP-4 (Chan et al., 2003; Santourlidis et al., 2002). None of these elements alone, however, is required for or significantly increases promoter activity, except for a CREB site located outside the core promoter region. To date, the basic KIR promoter appears to either require undefined transcription factors or alternatively only interacts with the basic transcription machinery of RNA polymerase II (Trompeter et al., 2005). It has been suggested previously that a binding site for the transcription factor AML/Runx (Runt Homology Domain Transcription Factors) at –100 bp can be essential for KIR expression (Vilches et al., 2000). However, a further study suggested that AML has a general inhibitory influence on KIR expression in mature NK cells and that AML exerts its repressor function by binding directly to the promoter of the different KIR genes (Trompeter et al., 2005). Recently, it was reported that the transcription factor c-Myc upregulated KIR gene transcription through direct binding to an upstream distal promoter element in peripheral blood NK cells, and that IL-15 promoted this effect (Cichocki et al., 2009). It was suggested that the role of the traditional transcription factors in KIR gene expression is different from that in other genes.
