Comparative Analysis of LFA-1 Activation State and Functional Involvement in Enhanced Cytotoxicity of NK Cell Lines KHYG-1 and NK-92

Background: KHYG-1 and NK-92 are highly cytotoxic IL-2 dependent NK cell lines. NK cytotoxicity is regulated through an array of receptors with activating and inhibitory functions for spontaneous elimination of pathogen infected or tumor cells. Important roles have been described for the leukocyte function-associated antigen (LFA)- 1 adhesion receptor (αLβ2) in NK cytotoxicity. In T cells, the integrin LFA-1 occurs in a non-functional, bent form and requires activation, through divalent cations e.g., for ligand binding. In marked contrast, IL-2 stimulated NK cells could directly engage the LFA-1 ligand intercellular adhesion molecules (ICAM)-1, thereby inducing conjugate formation, granule polarization, degranulation, and tumor cell lysis (Barber DF et al., 2004. JI 173:3653–9). Strikingly, in the NK cell line KHYG-1 granules were constitutively polarized (clustered around the MTOC; Suck G et al., 2006. Int Immunol 18:1347–54) and LFA-1 downstream signaling molecules, the spleen tyrosine kinase (Syk) and extracellular signal-regulated MAP kinase (ERK) constitutively phosphorylated (Suck G et al., 2005. Exp Hem. 33:1160–71). These previous findings prompted us to investigate LFA-1 activation state and functional involvement in cytotoxicity in KHYG-1 and NK- 92.

Methods: Adhesion assays were performed on immobilized ICAMs and % adherence determined in a fluorescence plate reader (Tang XY et al., 2008 JI, 180:4793–804). LFA-1 was activated with Mg/EGTA (5 mM MgCl₂ and 1.5 mM EGTA), monoclonal antibody (mAb) KIM185 (10 mg/mL) or both; LFA-1 mAb, clone MHM24, was used for blocking studies, 1–10 mg/ml and mAb KIM127 for immunoprecipitation; cytotoxicity, conjugate formation, and degranulation (CD107a) were measured by Flow cytometry and cell morphology imaged by phase contrast (Olympus, 1X70) or confocal microscopy (Zeiss, LSM510 Meta/Nikon A1). Excel software was employed for statistical analyses.

Results: In cell binding assays KHYG-1 and NK-92 showed constitutive adhesion to the LFA1-1 ligand ICAM-1, 61% +/− 5.8% SD and 55% +/− 7.5% SD, respectively. However, only 22% NK-92 and 10.5% KHYG-1 cells adhered to the lower affinity ligand ICAM-3 and activation with Mg/EGTA or KIM185 was required to increase binding to 73% in KHYG- 1 and to 62% in NK-92. Immunoprecipitation experiments with the activation reporter mAb KIM127 revealed an activated extended LFA-1 conformation in both cell lines. Together these results suggested an intermediate affinity activation state for the LFA- 1 population in KHYG-1 and NK-92. Despite such similarities, ICAM-1 engagement triggered pronounced cell spreading in KHYG-1, similar to phorbol ester stimulated T cells, but not in NK-92 or primary NK cells. It is conceivable that the capacity to undergo cytoskeletal remodeling in KHYG-1 may be impaired as potentially also reflected in the constitutively polarized granule state. In addition, LFA-1 blocking studies with LFA- 1 antibody MHM24 did not inhibit target conjugate formation in KHYG-1, compared to partial inhibition in NK-92, and cytotoxicity against K562 was only about 40% diminished in KHYG-1, compared to almost complete abrogation (maximum 98%) in NK-92. Nevertheless, overall cytolytic potential against K562 was comparable among the two cell lines, which implicated important functions for other receptors at least in KHYG-1.

Conclusion: Results suggested a constitutively activated intermediate affinity state for LFA-1 in KHYG-1 and NK-92. It is conceivable that a similar activated state occurs in primary NK cells. Interestingly, although cytoskeletal dysfunction was indicated in KHYG-1, cytotoxicity of the cell line was unaffected, rendering it a valuable model for the study of alternative pathways.