Monalizumab efficacy correlates with HLA-E surface expression and NK cell activity in head and neck squamous carcinoma cell lines

NKG2A, an inhibitory receptor expressed on NK cells and T cells, leads to immune evasion by binding to HLA-E expressed on cancer cells. Here, we investigated the relationship between HLA-E surface expression on head and neck squamous cell carcinoma (HNSCC) cell lines and the efficacy of monalizumab, an NKG2A inhibitor, in promoting NK cell activity. Six HNSCC cell lines were used as target cells. After exposure to IFN- γ, HLA-E surface expression on HNSCC cell lines was measured by flow cytometry. Peripheral blood mononuclear cells (PBMCs) from healthy donors and isolated NK cells were used as effector cells. NK cells were stimulated by treatment with IL-2 and IL-15 for 5 days, and NK cell-induced cytotoxicity was analyzed by CD107a degranulation and 51Cr release assays. We confirmed that HLA-E expression was increased by IFN-γ secreted by NK cells and that HLA-E expression was different for each cell line upon exposure to IFN-γ. Cell lines with high HLA-E expression showed stronger inhibition of NK cell cytotoxicity, and efficacy of monalizumab was high. Combination with cetuximab increased the efficacy of monalizumab. In addition, stimulation of isolated NK cells with IL-2 and IL-15 increased the efficacy of monalizumab, even in the HLA-E low groups. Monalizumab efficacy was correlated with HLA-E surface expression and was enhanced when NK cell activity was increased by cetuximab or cytokines. These results suggest that monalizumab may be potent against HLA-E-positive tumors and that monalizumab efficacy could be improved by promoting NK cell activity.


Introduction
Patients with head and neck squamous carcinoma (HNSCC) have a poor prognosis (Johnson et al. 2020). Although therapies that modulate immune checkpoint molecules on T cells, such as programmed death protein-1 and programmed death ligand-1, show promising clinical responses for some cancer types (Herbst et al. 2014;Powles et al. 2014), the response rate to T-cell-based immunotherapy is generally low (Chen et al. 2018;Gotwals et al. 2017). Therefore, research focusing on other types of immune cells or immune checkpoint molecules is needed to develop new treatment strategies.
NKG2A, which dimerizes with CD94, is an inhibitory receptor expressed on NK cells and T cells that binds to HLA-E, a non-classical HLA class I molecule containing leader peptides of other HLA class I molecules (Iwaszko and Bogunia-Kubik 2011;Leibson 1998). Although HLA-E expression is often weak or absent on the cell surface of tumor cells, it is increased by IFN-γ from immune cells and protects against NK cell-induced lysis (Lo Monaco et al. 2011;Nguyen et al. 2009). Levels of HLA-E can increase within the TME due to antitumor responses, and the overexpression of HLA-E in several types of solid tumors is associated with poor prognosis (Benevolo et al. 2011;de Kruijf et al. 2010;Gooden et al. 2011;Morinaga et al. 2022). HLA-E plays a more predominant role in inhibiting NK cells than other HLA class I molecules (Sheffer et al. 2021) and also inhibits antibody-dependent cellular cytotoxicity (ADCC) (Ehlers et al. 2021;Levy et al. 2009;Ward et al. 2004). Several studies show that NKG2A blockade promotes the antitumor immunity of T cells and NK cells (André et al. 2018;Ducoin et al. 2022;Kamiya et al. 2019;Ruggeri et al. 2016;Zaghi et al. 2019). Monalizumab, a novel IgG4 monoclonal antibody (mAb) targeting NKG2A, was recently developed (van Hall et al. 2019). A preclinical study shows that monalizumab inhibits tumor progression by increasing NK cell and T cell activity, and clinical trials of monalizumab in combination with cetuximab are ongoing (André et al. 2018;Cohen et al. 2020;Galot et al. 2021;Herbst et al. 2022;Salomé et al. 2022). However, factors predicting tumor response to monalizumab are not yet understood. In particular, it is unclear whether the level of HLA-E expression on tumor cells predicts response to monalizumab therapy.
The objective of this study was to evaluate the association between HLA-E surface expression on HNSCC cell lines and the efficacy of monalizumab on NK cells. We also investigated the correlation between NK cell activity and monalizumab efficacy after stimulation of NK cells by cytokines ex vivo.

Cell lines and cell culture
The human HNSCC cell lines SNU-1041, SNU-1066, SNU-1076 (purchased from the Korean Cell Line Bank, Seoul, Korea) were cultured in Roswell Park Memorial Institute (RPMI) 1640 medium, and Detroit-562 and FaDu (purchased from the American Type Culture Collection, Manassas, Virginia, USA) were cultured in Minimum Essential Media containing 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin. Recombinant human IFN-γ (R&D Systems, Minneapolis, MN, USA) was used to increase HLA-E surface expression.

Establishment of β2M KO SNU-1041 cells by CRISPR
Guide RNA, scrambled control, and GFP-puro donor in beta 2 microglobulin (β2M) Human Gene Knockout CRISPR Kit (ORIGENE, Rockville, MD, USA) was used to establish SNU-1041 MHC-I knockout (KO) cells. Lipofectamine and PLUS™ reagent (Invitrogen, Carlsbad, CA, USA) were used as transfection reagents, and transfection was performed according to the manufacturer's instructions. Forty-eight hours after transfection, one-tenth of cells were split and grown again for 3 days, which was repeated for 3 weeks. β2M KO SNU-1041 cells were selected by puromycin (2 μg/ ml), and non-expression of MHC-I was verified by flow cytometry.

Preparation of peripheral blood NK cells, NK cells isolated from peripheral blood mononuclear cells, and NK cell conditioned medium
Using Ficoll density-gradient sedimentation, peripheral blood mononuclear cells (PBMCs) from healthy donors were isolated using leukoreduction system chambers. PBMCs were activated in RPMI 1640 medium supplemented with 10% FBS and 1 ng/ml rhIL-15 (PeproTech, Rocky Hill, NJ, USA) for 5 days.
For the 51 chromium (Cr) release assay and preparation of NK cell conditioned medium (CdM), NK cells were isolated from PBMCs using MACS columns and an NK cell isolation kit (Miltenyi Biotec, Bergisch Gladbach, Germany). Purified NK cells were cultured in NK MACS® Medium (Miltenyi Biotec) supplemented with 10% FBS, 500 IU/ml rhIL-2, and 10 ng rhIL-15. The purity of isolated NK cells was confirmed by flow cytometry. Purified NK cells were cultured in RPMI with 500 IU/ml rhIL-2 for 3 days to generate CdM. Cell culture supernatants were harvested by centrifugation and passed through a 0.45-µm Acrodisc syringe filter (Pall Corporation, Port Washington, NY, USA). IFN-γ mAb (NIB42; Thermo Fisher Scientific, Waltham, MA, USA) was used to neutralize IFN-γ in CdM.

Reagents and antibodies
To evaluate the efficacy of monalizumab to promote NK cell activity and ADCC, monalizumab and cetuximab were used. Monalizumab (purchased from Creative Biolabs, Shirley, NY, USA) is a humanized IgG4 mAb that targets NKG2A.
Cetuximab is a chimeric IgG1 mAb targeting EGFR that induces ADCC.

Analysis of cell surface protein expression using flow cytometry
Cell surface expression of HLA-E was measured by flow cytometry. Cells were incubated with or without 10 ng/ml IFN-γ for 24 h and stained with fluorochrome-conjugated HLA-E antibody or isotype control for 20 min at 4 °C. FACS Calibur (BD Biosciences) was used for data acquisition, and data were analyzed using FlowJo software (LCC, Ashland, OR, USA). mAb staining was measured as geometric mean fluorescence intensity (gMFI). We defined the difference between HLA-E-stained cancer cells and isotype control antibody-stained cancer cells as delta (Δ)gMFI.

CD107a degranulation assay
CD107a degranulation assay was performed using PBMCs or isolated NK cells from healthy donors with an effector-totarget (E:T) ratio of 1:1 using a previously described method (Park et al. 2020). PBMCs (2 × 10 5 ) or isolated NK cells were incubated with HNSCC cells for 1 h at 37 °C in U-bottom 96-well plates. Next, GolgiStop solution was added followed by incubation for 3 h. All cells were transferred to a FACS tube, washed with FACS buffer, and stained with fluorescently labeled mAbs. Data were acquired and analyzed using a FACS Canto II (BD Biosciences) and FlowJo software.

51
Cr release assay 51 Cr release assay methodology is described in our previous study (Park et al. 2019). Target cells were labeled with 50 μCi for 1 h at 37 °C, and then 2.5 × 10 5 cells were washed three times with growth media. Target cells were cultured with 10 μg/ml of the indicated antibodies, seeded at 5 × 10 3 cells/well into U-bottom 96-well plates, and co-incubated with effector cells at a 10:1 E:T ratio for 4 h at 37 °C. Following co-culture, 75 μl supernatant from each well was transferred to 96-well PP 1.2-ml cluster tubes for gamma counting. To calculate radioactivity, a WIZARD2 Automatic Gamma Counter (PerkinElmer, Waltham, MA, USA) was used.

Statistical analysis
GraphPad Prism v7.0 (GraphPad Software, San Diego, CA, USA) was used to analyze data, which is reported as mean ± standard deviation (SD). Two-tailed paired Student's t tests were used to compare groups, and a two-sided P < 0.05 was considered statistically significant.

Surface expression of HLA-E in HNSCC cell lines
We measured the surface expression of HLA-E in HNSCC cell lines using flow cytometry. In normal conditions, all cell lines showed low HLA-E surface expression except for MHC-I KO cells. However, when IFN-γ was added, an increase in HLA-E expression was observed in all cell lines (Fig. 1A). Depending on the HLA-E expression level after IFN-γ treatment, cell lines with a change in mean fluorescence intensity (ΔMFI) ≥ 20 were classified into the HLA-E high group (SNU-1041, SNU-1066, and FaDu), whereas cell lines with ΔMFI < 20 were classified into the HLA-E negative or low group (SNU-1041 MHC-I KO, SNU-1076, and Detroit-562). There were no significant correlations between the HLA-E mRNA expression or genotype of each cell line and its level of HLA-E surface expression ( Supplementary  Fig. S1).
To confirm that HLA-E surface expression was increased due to IFN-γ secreted from NK cells, we prepared NK cell CdM. When SNU-1041 cells were cultured in CdM, the surface expression of HLA-E increased to a level similar to that following IFN-γ treatment, whereas neutralization with IFN-γ mAb attenuated this increase (Fig. 1B).

Correlation between HLA-E surface expression in HNSCC cell lines and monalizumab efficacy
To investigate the correlation between HLA-E surface expression in HNSCC cell lines and monalizumab efficacy, we measured the cytotoxicity of primary NK cells in PBMCs by CD107a degranulation assay. PBMCs were activated with a low concentration of IL-15 (1 ng/ml) for 3 days, and NK cells were identified by flow cytometry ( Supplementary  Fig. S2). There were no statistically significant effects of monalizumab among HNSCC cell lines in the negative or low HLA-E group either under control conditions or after IFN-γ treatment ( Fig. 2A). However, among HNSCC cell lines in the HLA-E high group, monalizumab showed significant effects, with even larger effects when monalizumab was combined with cetuximab (Fig. 2B).
The level of HLA-E surface expression was negatively correlated with NK cell cytotoxicity, including ADCC induced by cetuximab (Fig. 3A, B). In addition, the recovery of cytotoxic NK cells by monalizumab was correlated with HLA-E surface expression (Fig. 3C). Taken together, these results suggest that HLA-E surface expression in HNSCC cell lines predicts the efficacy of monalizumab in promoting NK cell activity.

Ex vivo stimulation of NK cells increases the efficacy of monalizumab
To measure the efficacy of monalizumab upon stimulation of NK cells with a cytokine, we performed a CD107a degranulation assay after isolating NK cells from PBMCs.
When the assay was performed with NK cells incubated in growth medium for 1 day, the effect of monalizumab was not significant in any HNSCC cell line, even when combined with cetuximab ( Supplementary Fig. S3). By contrast, after stimulation with IL-2 (500 IU/ml) and IL-15 (10 ng/ml) for 5 days, effects of monalizumab alone and monalizumab combined with cetuximab were observed for SNU-1041 cells with high HLA-E surface expression as well as SNU-1076 cells with negative or low HLA-E surface expression (Fig. 4A). The effect of monalizumab alone was comparable to that of cetuximab in SNU-1076 cells and was higher than that of cetuximab in SNU-1041 cells.
We also performed a 51 Cr release assay to confirm whether ex vivo stimulation with a cytokine improves the effect of monalizumab. As with the CD107a degranulation assay results, effects of monalizumab alone and monalizumab in combination with cetuximab were observed in all HNSCC cell lines including those in the HLA-E negative A Surface expression of HLA-E on HNSCC cell lines with or without treatment with IFN-γ for 24 h was measured by flow cytometry. B SNU-1041 and SNU-1041 MHC-I KO cells were exposed to NT, IFN-γ, NK cell CdM, or NK cell CdM with anti-IFN-γ mAb, and HLA-E surface expression was measured after 24 h by flow cytometry. Data are shown as mean ± SD. ΔgMFI difference in geometric mean fluorescence intensity, NT no treatment, CdM conditioned medium Fig. 2 Increased NK cell cytotoxicity after monalizumab treatment of HNSCC cell lines. The cytotoxicity of primary NK cells was measured by CD107a degranulation assay after co-culture of PBMCs and tumor cells with an E:T ratio of 1:1. PBMCs from healthy donors were used for effector cells and were activated by IL-15 (1 ng/ml) for 3 days before co-culture. NK cells in PBMCs were identified using flow cytometry after fluorescence staining. The experiment was conducted by dividing the control and IFN-γ pretreated groups. IFN-γ pretreated groups were treated with IFN-γ for 24 h. The effects of monalizumab alone and monalizumab in combination with cetuximab were measured. HNSCC cell lines were treated with monalizumab (gray), cetuximab (red), or monalizumab plus cetuximab (blue) or were not treated (NT, black). A CD107a degranulation assay results for SNU-1041 MHC-IKO, SNU-1076, and Detroit-562, which had negative or low surface HLA-E expression. B CD107a degranulation assay results for SNU-1041, SNU-1066, and FaDu, which had high surface HLA-E expression. Data are shown as mean ± SD. *P < 0.05, ***P < 0.001, ****P < 0.0001. Statistical significance was determined by two-tailed paired Student's t tests. neg/lo negative or low, hi high, Ctrl control, NT no treatment, Mo monalizumab, Ctx cetuximab, ns not significant 1 3

Fig. 3
Correlations between surface expression of HLA-E and monalizumab efficacy. A NK cell cytotoxicity (measured by the percentage of CD107a + NK cells) in HNSCC cell lines with or without treatment with IFN-γ (left) and correlation between HLA-E surface expression and NK cell cytotoxicity (right).
B NK cell cytotoxicity when treated with cetuximab in HNSCC cell lines with or without treatment with IFN-γ (left) and correlation between HLA-E surface expression and NK cell cytotoxicity (right).
C Correlation between HLA-E surface expression and increased NK cell cytotoxicity when treated with monalizumab alone (left) or in combination with cetuximab (right). NT no treatment, Ctx cetuximab or low group, although larger effects were observed in the HLA-E high group (Fig. 4B, C). These results suggest that stimulation of NK cells by IL-2 and IL-15 enhances the efficacy of monalizumab.

Discussion
We found a correlation between HLA-E surface expression on HNSCC cell lines and monalizumab efficacy in promoting NK cell activity. We also showed that combination of cetuximab and stimulation of NK cells with IL-2 and IL-15 improved the efficacy of monalizumab, even in cell lines with low HLA-E surface expression (Fig. 5). These findings suggest that the antitumor effect of monalizumab is stronger Fig. 4 Ex vivo stimulation of NK cells increased the efficacy of monalizumab. To determine whether stimulation with NK cells increases the efficacy of monalizumab, NK cells were isolated from PBMCs. Subsequently, NK cells were stimulated with IL-2 (500 IU/ ml) and IL-15 (10 ng/ml) for 5 days. HNSCC cells were treated with IFN-γ for 24 h. A Results of CD107a degranulation assay for SNU-1041 MHC-KO (HLA-E negative), SNU-1076 (HLA-E low), and SNU-1041 (HLA-E high) cell lines. 51 Cr release assay was performed using stimulated NK cells. HNSCC cells were treated with IFN-γ or not for 24 h before the assay and labeled with 51 Cr. NK cells and tumor cells were co-cultured for 4 h at an E:T ratio of 10:1. Results of 51 Cr release assay for B HLA-E negative or low cell lines and C HLA-E high cell lines. Data are shown as mean ± SD. **P < 0.01, ****P < 0.0001. Statistical significance was determined by two-tailed paired Student's t tests. neg/lo negative or low, hi high, Ctrl control, NT no treatment, Mo monalizumab, Ctx cetuximab, ns not significant in tumors with higher HLA-E expression and depends on NK cell activity.
By measuring levels of HLA-E surface expression on HNSCC cell lines, we showed that HLA-E expression was increased by IFN-γ secreted from NK cells. The surface expression of HLA-E requires binding to the MHC-I leader sequence, and HLA-E surface expression and affinity for NKG2A vary depending on the type of leader peptide that is bound (Battin et al. 2022;Matsunami et al. 2002). Antibodies that bind HLA-E leader peptide complexes enhance NK cell cytotoxicity (Li et al. 2022). HLA-E surface expression is also affected by tapasin, TAP, and β 2 M, which are involved in movement to the cell surface (Braud et al. 1998;Lee et al. 1998;Lo Monaco et al. 2008). We found that the surface expression of HLA-E in HNSCC cell lines increased upon exposure to IFN-γ, consistent with previous studies (Derré et al. 2006;Nguyen et al. 2009), likely due to activation of the STAT-1 pathway by IFN-γ (Lee and Benveniste 1996;Zhou 2009). As HLA-E expression in the TME is also increased by IFN-γ, expression of IFN-γ and STAT-1 pathway-related genes in the TME may also be affected. Furthermore, HLA-E surface expression is high in acute leukemia patients with the genotype HLA-E*01:03 (Lauterbach et al. 2015;Xu et al. 2019), suggesting that the genotype of HLA-E could affect HLA-E surface expression. However, in the present study, neither the surface expression of HLA-E nor the level of HLA-E mRNA was significantly associated with genotype. Together, these findings suggest that the surface expression of HLA-E is affected by complex interactions among multiple factors, which requires clarification through future studies.
Using CD107a degranulation assay, we observed correlations between HLA-E expression and monalizumab efficacy, with stronger correlations observed when Fig. 5 Schematic representation of a model representing monalizumab efficacy depending on HLA-E expression and NK cell activity. HLA-E high cells strongly inhibit NK cell cytotoxicity, and have higher efficacy of monalizumab than HLA-E low cells. Increasing NK cell activity by stimulating NK cells with cytokines or inducing ADCC with IgG1 mAb such as cetuximab can improve the efficacy of monalizumab. In particular, stimulation of NK cells using IL-2 and IL-15 enhances monalizumab efficacy even in HLA-E low cells monalizumab was combined with cetuximab. The higher efficacy of monalizumab, when used in combination with cetuximab, may be due to an increase in NK cell activity through the CD16 receptor, which also mediates ADCC (Capuano et al. 2021). Monalizumab has been demonstrated to have potency in combination with other treatments (van Hall et al. 2019), and clinical trials of monalizumab combination therapies are ongoing (NCT02643550, NCT02671435, NCT05221840). Our results suggest that HLA-E surface expression is a tumor biomarker for monalizumab efficacy and that the use of monalizumab in combination with other cancer therapies that promote NK cell or T cell activity could be more effective. INTER-LINK-1 phase 3 trial (NCT04590963) which compared cetuximab with/without monalizumab in HNSCC had been terminated because of futility issue. We speculate that better outcomes could be obtained when they apply molecular enrichment based on HLA-E-positivity in patient selection.
In unstimulated NK cells, even the combination of monalizumab and cetuximab did not show efficacy, whereas stimulation of NK cells with IL-2 and IL-15 increased the efficacy of monalizumab in both CD107a degranulation and 51 Cr release assays, even within the HLA-E low group. In vitro stimulation of NK cells with IL-2 and/or IL-15 increases NK cell-activating receptors and NK cell cytotoxicity (Croxatto et al. 2017;Horng et al. 2007;Lehmann et al. 2001;Sanchez-Correa et al. 2017;Wang et al. 2000). Also, in vitro stimulation increases the ratio of NKG2A + NK cells (André et al. 2018;Hromadnikova et al. 2013), and NKG2A + NK cells may induce more cytotoxicity (Mahaweni et al. 2018). Taken together, these findings suggest that the efficacy of monalizumab increases when NK cell activity is promoted by cytokine treatment and that NK cell activity in the TME is important. In addition, considering that adoptive NK cell transfer is an attractive therapeutic candidate for solid tumors and that NKG2A is an interfering factor (Cichocki and Miller 2019), these findings imply that adoptive NK cell therapy may exert greater antitumor activity when combined with monalizumab after ex vivo stimulation.
Our study has several limitations. As our study employed only in vitro experiments, verification of our results using in vivo experiments is required. Our experiment did not take into account the infiltration of NK cells into the TME or interactions with other immune cells that may affect the efficacy of monalizumab. Furthermore, it is necessary to determine the major factors that affect HLA-E surface expression in vivo. However, this study is meaningful in that it first identifies a potential biomarker and effective therapeutic strategy involving monalizumab.
In conclusion, our study reveals that HLA-E surface expression on HNSCC cell lines is correlated with monalizumab efficacy and that increasing NK cell activity by combined treatment with cetuximab or cytokines enhances the efficacy of monalizumab. Our findings suggest that monalizumab has potency against tumors with higher HLA-E surface expression and that the efficacy of monalizumab is improved when NK cell activity is increased by other treatments.