Figure legends

Figure 1 –Schematic representation of the NK cell signalling pathway. A) Upon binding of the killing activation receptor (KAR), the ITAM motifs of the intracellular portion are phosphorylated by a member of the Src family. Syk or ZAP70 are activated and further phosphorylate the PI3K complex and activate VAV-1/-2/-3, PLCγ1 and PLCγ2. This will further trigger a cytoskeleton reorganization and affect the calcium flux into the cell, leading to an increase in Ca2+levels. This cascade results in the cytotoxic/cytolytic response of NK cells, with the release of the cytotoxic granules (degranulation) and production of cytokines/chemokines. B) Killing-inhibiting receptors (KIRs), present on the surface of NK cells, may also be engaged by the corresponding inhibiting ligands. These ITIM-bearing receptors are subsequently phosphorylated by Src kinases. This leads to the recruitment of SHP1/2 kinases that terminate the activation signalling emanating from ITAMs, rendering the NK cells inactive. The ultimate activation/inactivation result of these pathways is directly correlated with the number of receptors engaged. Target cells expressing more activation than inactivation ligands will engage more activation receptors resulting in the transduction of more activation signals, triggering the response of NK cells (Koch, Steinle, Watzl & Mandelboim, 2013; Linnartz-Gerlach, Kopatz & Neumann, 2014; Linnartz, Wang & Neumann, 2010; Paul & Lal, 2017; Watzl & Long, 2010).
Figure 2 – NK cell activation/inhibition equilibrium. The relative number of engaged receptor-types dictates the response. A) Normal cells, expressing normal MHC I molecules (inhibitory ligands) do not trigger the cytolytic response of NK cells. B) Stressed cells, such as tumour cells, engaging the same relative number of activating and inhibiting receptors of NK cells, induce a dynamic equilibrium that results in no cytolytic response of NK cells. C) If that equilibrium is tipped over by engaging more activating receptors on NK cells, a cytolytic response will be triggered against the cells presenting the corresponding ligands.
Figure 3 – Structure representation of NKp30. The two antiparallel β-sheets (in yellow) and the two α-helixes (in red) are shown. The two β-sheets are bridged by a disulphide bond between residues Cys39 and Cys108. The stalk domain and the transmembrane α-helix are not represented. Image generated from the published X-ray structure of unbound NKp30 (PDB 3NOI) (Joyce, Tran, Zhuravleva, Jaw, Colonna & Sun, 2011) .
Figure 4 – NKp30 molecular surface with A) B7-H6-contacting residues (red) and N-glycosylation sites (blue) highlighted. Only one possibly N-glycosylated residue (Asp68) is located near the identified binding site, which comprises residues Ile50, Gly51, Ser52, Val53, Leu80, Ser82, Phe85, Leu86, Glu111, Leu113 and Gly114, marked in red; and B) B7-H6-contacting residues differentiated by type of amino acid residue: polar residues (yellow) and hydrophobic residues (cyan). Image generated with PyMOL(Schrodinger, 2015) from the published X-ray structure of unbound NKp30 (PDB ID: 3NOI) (Joyce, Tran, Zhuravleva, Jaw, Colonna & Sun, 2011) . NKp30 structure representation in its unbound (C, E) and B7-H6-bound forms (D, F) are also shown. Isoleucine 50 is labelled as a positioning reference. Conformation differences between the two forms are observed in the distance between residues 52 and 82, labelled in yellow, that differ by about 1 Å, and in the relative positions of arginine 67 and asparagine 68. The loop composed by arginine 67 (blue) and asparagine 68 (orange) seems to bend upon binding. Asparagine 68 moves away from proline 79 (in red) and arginine 67 bends down towards the pocket, approaching valine 53 (not highlighted). Image generated from the published X-ray structures of unbound NKp30 (PDB ID: 3NOI)(Joyce, Tran, Zhuravleva, Jaw, Colonna & Sun, 2011) and B7-H6 bound NKp30 (PDB ID: 3PV6) (Li, Wang & Mariuzza, 2011) .