Innate Recognition of Tumor Cells

In the "missing self" hypothesis Karre and colleagues suggested that normal healthy self cells are protected from NK cell attack due to the recognition of a self-antigen [39]. This hypothesis was confirmed by the discovery of the inhibitory NK receptors which recognize self class-I MHC molecules [40]. Thus, it was suggested that once tumor cells reduces the expression of the class-I MHC proteins to evade recognition by CTLs, they also become more susceptible to NK lysis [40, 39]. Today, we know that downregulation of MHC class I proteins per se is not enough to induce NK cytotoxicity and that activating receptors control NK cytotoxicity [41, 5]. In addition, other NK inhibitory receptors recognize non-MHC ligands (see below).

The major human activating NK receptors include the CD16, NKG2D, NKp30, NKp44, NKp46 and NKp80 [5]. Only few tumor ligands were identified for these receptors. CD16 is a low-affinity FcRgIII receptor that binds antibodies and unknown tumor ligands [42]. NKG2D recognize the stress-induced proteins ULBPs, MICA and MICB in human or Rae-1s, MULT and H60 in mice [43] (further discussed below), and NKp80 was recently reported to recognize the mye-loidic antigen AICL [44].

Activation of the CD16 elicits a strong cytotoxic response even on naive NK cells [45]. Several research reports demonstrated the presence of antitumor antibodies in the blood of patients [15]. It is still unknown if these antibodies can provide enough stimulation for the CD16 (or other Fc-receptors). The Antibody Dependent Cellular Cytotoxicity (ADCC) activity, which is mediated by NK cells activation through CD16, can have a significant antitumor activity, and is already used clinically [46].

Only viral ligands were identified for the three NCRs: NKp30, NKp44 and NKp46 receptors [47-49]. The tumor-ligands for the NCRs are unknown; however, it was demonstrated that these receptors specifically recognize various tumors [5053] and that HSPG are involved in tumor recognition [54]. In addition to the direct in vitro data regarding the killing of tumor cells by these receptors, analysis of NK cells derived from acute myeloid leukemia (AML) revealed lower expression of the NKp30 and NKp46 receptors, and therefore reduced activity [55]. Since the unknown tumor ligands for NKp30, NKp44 and NKp46 are expressed in many different tumors [50, 56], it is likely to assume that they are needed to support tumor growth, or could not be easily removed by the tumor cells.

Additional NK receptors are also important for tumor eradication. For example, the DNAX-1 accessory molecule-1 (DNAM-1) was demonstrated to have a major role in NK cytotoxicity and cytokines production [57]. The KIR2DS4 activating receptor was demonstrated to have a tumor-ligand other than MHC class-I [58], and this antigen is yet to be identified. Another example is in the case of defective signaling of the 2B4 in XLP patients, causing inability of NK cells to eliminate EBV-infected cells [59] (further discussed below). Thus, understanding the complex balance of inhibitory, activating and co-stimulatory signals [40, 5] that control NK activity is needed to elucidate the multiple mechanisms by which tumors avoid NK cell attack.

TLR9 is expressed on NK cells, and is activated by non-methylated DNA [12]. Tumors have abnormal DNA-methylation, overall manifesting a general genome-wide hypomethylation, despite local hypermethylation islands. Necrotic and apop-totic tumor cells secret such non-methylated CpGs oligonucleotides and these could affect NK killing. Recently, the possible activation of NK cells through TLR9 was shown in human leukemia as stimulation of NK activity was observed via CpG oligonucleotides [60].

0 0

Post a comment