To uncover novel ligands for NKARs, we performed a genome-wide loss-of-function genetic screening using a pooled shRNA library targeting 15,000 human genes with multiple sequence-verified constructs ( Figure 1A ). On average, five shRNA designs were used per gene. We selected the K562 cell line as a target due to its expression of numerous ligands for NKARs, making it an ideal model for identifying genes involved in NK cell activation. K562 cells were transduced with the human lentiplex shRNA library (TRC 1.5) at a multiplicity of infection (MOI) ≤1 and selected with puromycin for 4 days. As a control, K562 cells were transduced with a non-targeting shRNA (shCTRL). The transduced cells were then cocultured with freshly isolated NK cells from five healthy donors (HD) at an effector-to-target (E:T) ratio of 15:1, with each donor added separately over three consecutive days. Surviving K562 cells, presumed to have silenced genes essential for NK cell-mediated activation, were selected based on viability and morphology via flow cytometry, and maintained as bulk populations or individual clones ( Figure 1A ).

Functional validation using NK cells from additional HDs and a standard ⁵¹Cr release assay confirmed that K562 cells transduced with the shRNA library were more resistant to NK cell-mediated killing, compared to shCTRL cells ( Figures 1B, C ). The average lytic units (L.U.) at 30% lysis across three replicates (shLib 1-3) were significantly lower for K562-shLib cells (12.6) than for K562-shCTRL cells (17.5) ( Figure 1B , right panel). Similarly, individual clones showed variable levels of resistance, with library-derived clones displaying an average L.U. at 20% of 18, compared to 31 for shCTRL cells ( Figure 1C , right panel). Notably, the 10 clones with L.U. values below the average (<18) were also more resistant to killing by the NKL cell line (25) than the other clones ( Figure 1D ).

To evaluate whether silencing a specific ligand could impair NK cell function, we tested NKL cell recognition of K562 cells with MICA knockdown, a known ligand for NKG2D. K562-shMICA cells exhibited protection levels comparable to the top 10 resistant clones, with a L.U. of 40 versus 20 for K562-shCTRL ( Supplementary Figure S1 ).

DNA sequencing identified the silenced target genes in 15 out of 24 clones ( Supplementary Table S1 ). The remaining 9 clones likely contained multiple shRNA inserts, preventing target identification. Among the identified genes, placenta-specific protein 1 (PLAC1, clone 2.2D2) (24), transcription factor 7 (TCF7, clone 2.1C10) (26), and Pleckstrin Homology Domain-Containing A5 (PLEKHA5, clone 3.2D8) (27) conferred strong resistance to NK cell-mediated lysis by both HD-derived NK cells and the NKL cell line ( Figures 1C, D ). PLAC1 is associated with placental development (24), TCF7 is involved in T cell differentiation (28), and PLEKHA5 has been linked to the suppression of tumor metastasis (29).

Functional validation revealed that silencing PLAC1 significantly impaired NK cell degranulation, as indicated by reduced CD107a expression ( Figure 2A ; Supplementary Figure S2A ). Both K562-shPLAC1 cells and the corresponding library clone (K562-2.2D2) decreased the frequency of CD107a⁺ NK cells by nearly 50% compared to K562-shCTRL cells (0.60 and 0.48, respectively) ( Figure 2A ). In contrast, silencing TCF7 or PLEKHA5 had no significant effect on NK cell degranulation ( Figure 2A ). Based on these results and given its predicted membrane localization (30), we focused on PLAC1 as the most promising target. To further validate its role, PLAC1 was cloned from JEG3 cells and stably expressed in K562 cells ( Figure 2B ; Supplementary Figures S2B, C ).

Overexpression of PLAC1 in K562 cells increased NK cell degranulation by 23%, with the frequency of CD107a⁺ NK cells rising from 45.5% in K562-CTRL cultures to 59% in K562-PLAC1 cultures ( Figure 2C ). Since the overexpression of PLAC1 did not impact the expression of NKAR ligands MICA/B, ULBPs, Nectin-2, PVR, B7-H6, FAS and TRAIL-R2 ( Figure 2D ), the increased degranulation was attributed to a direct effect mediated by PLAC1.

These results highlight PLAC1’s critical role in modulating NK cell function and suggest its potential as a novel ligand for NKARs.

PLAC1 may influence NK cell function by serving as a ligand for NKARs. To test this hypothesis, we examined whether PLAC1 expression affects the binding of NKARs fusion proteins NKG2D, DNAM1, NKp30 and NKp44 to K562 cells overexpressing PLAC1. In this assay, K562 cells were stained with NKARs-Fc fusion proteins, NKG2D-Fc, DNAM1-Fc, NKp30-Fc and NKp44-Fc. Compared with control cells, overexpression of PLAC1 led to significantly increased binding of all fusion proteins tested in K562 cells ( Figure 2E ).

These results suggest that overexpression of PLAC1 results in increased interaction of these cells with the activating receptors NKG2D, DNAM1, NKp30, and NKp44, potentially amplifying the recognition and lysis of tumor cells by NK cells.

PLAC1 is highly expressed during development, particularly in placental tissues, including trophoblast giant cells and the labyrinthine layer derived from the trophoblast lineage (31). According to data from The Cancer Genome Atlas (TCGA), PLAC1 is virtually undetectable in most normal somatic tissues, with the exception of testicular germ cells, skeletal muscle, peritubular cells, and the placenta during gestation ( Figure 3A ; Supplementary Figure S3A ). In contrast, PLAC1 is significantly upregulated in a wide range of tumor types compared to their normal tissue counterparts ( Figure 3B ). Elevated expression was observed in 15 distinct cancers, including bladder urothelial carcinoma (BLCA), breast invasive carcinoma (BRCA), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), cholangiocarcinoma (CHOL), colon adenocarcinoma (COAD), esophageal carcinoma (ESCA), glioblastoma multiforme (GBM), head and neck squamous cell carcinoma (HNSC), kidney renal clear cell carcinoma (KIRC), lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), prostate adenocarcinoma (PRAD), rectum adenocarcinoma (READ), stomach adenocarcinoma (STAD) and thyroid carcinoma (THCA) ( Figure 3B ; Supplementary Table S2 ). Interestingly, the prognostic relevance of PLAC1 expression appears to be highly tumor-specific. Elevated PLAC1 levels were associated with improved overall survival in ESCA, GBM and LUSC, and poorer prognosis in BLCA, CESC, CHOL, COAD, HNSC, KIRC and PRAD ( Figure 3C ; Supplementary Figures S3B, C ). To assess the impact of threshold selection on survival outcomes, we conducted sensitivity analyses using multiple clinically and statistically relevant cut-offs—including the mean, median, and quartiles of PLAC1 expression ( Supplementary Figures S3B ). Although the strength and significance of the associations varied depending on the chosen threshold, consistent trends were observed, particularly in ESCA and LUSC, where the mean expression level reliably stratified survival outcomes ( Supplementary Figures S3B, C ).

These findings underscore the importance of careful threshold selection when utilizing PLAC1 as a prognostic biomarker and suggest that PLAC1 may serve as a context-dependent indicator of clinical outcome, with its prognostic value varying substantially across different tumor types.

To further explore the role of PLAC1 in NK cell activation, we analyzed its correlation with a NK cell gene signature (32) in tumor types where PLAC1 expression affects overall survival of patients. Notably, in tumors where high PLAC1 expression is associated with a favorable prognosis (ESCA, GBM, LUSC), we observed a positive correlation between PLAC1 and the genes XCL1, KLRC1 (NKG2A), KLRC2 (NKG2C) and TKTL1, and a negative correlation with the MLC1 gene ( Figure 4 ). In contrast, these correlations were absent in tumor types where high PLAC1 expression was associated with poorer survival outcomes (BLCA, CESC, CHOL, COAD, HNSC, KIRC and PRAD). These findings further strengthen the association between PLAC1 and NK cell function in specific cancer contexts.

Recent studies have identified Desmoglein 2 (DSG2), a critical component of desmosomes, as a direct interaction partner of PLAC1 (30). Desmosomes are specialized adhesive protein complexes located at intercellular junctions, essential for maintaining tissue mechanical integrity (33). Given PLAC1’s potential role as a ligand for NKARs, we investigated whether DSG2 is expressed on NK cells. According to data from the Human Protein Atlas (HPA), DSG2 is primarily expressed in epithelial cells and trophoblasts (34). Among immune cells, DSG2 is predominantly found on dendritic cells (DC), with no significant expression detected on NK cells (11) ( Supplementary Figure S4A ). To further explore DSG2 expression in NK cells within an oncological context, we analyzed single-cell RNA sequencing (scRNA-seq) data from NK cells isolated from the bone marrow of HD and patients with acute myeloid leukemia (AML) (35). DSG2 was detected in fewer than 1% of total NK cells (approximately 0.3%), with no significant differences between healthy and AML conditions ( Supplementary Figure S4B ). In contrast, well-characterized NK cell receptors, such as NKp46 (NCR1), NKG2D (KLRK1), NKG2A (KLRC1), and DNAM1 (CD226) were expressed in a substantially higher proportion of NK cells, approximately 30%, 10%, 7%, and 1%, respectively ( Supplementary Figure S4B ). Finally, we evaluated the expression of DSG2 on the surface of NK cells purified from HDs. Unlike HeLa cells, which are known to express DSG2, NK cells from all donors analyzed did not express DSG2 ( Supplementary Figure S4C ).

Altogether, these finding indicate that DSG2 is not expressed on the surface of NK cells, neither in healthy conditions nor in cancerous states. Therefore, the interaction between PLAC1 and NK cells likely involves a receptor other than DSG2.

All cell lines were obtained by ATCC and characterized every 6 months by HLA class I typing. Mycoplasma contamination was routinely detected by Mycoplasma Detection kit (Venor-GeM Advance). Cells were maintained in RPMI 1640 medium supplemented with 10% FBS (Gibco), 2 mM glutamine, 100 mg/ml penicillin and 50 mg/ml streptomycin (Euroclone).

Human NK cells were isolated from peripheral blood mononuclear cells (PBMC) of HDs by the RosetteSep NK-cell enrichment mixture method (StemCell Technologies) and Ficoll-Paque Plus (Lympholyte Cedarlane) centrifugation. NK cells with purity greater than 90% were suspended in NK MACS medium (Miltenyi Biotec) supplemented with NK MACS Supplement, AB serum and 500 IU/mL of recombinant human IL-2 (PeproTech) for 18 hours at 37°C.

K562 cells were stable infected with a pooled shRNA human library from the RNAi consortium (TRC 1.5) consisting of over 150000 plasmid-based shRNA constructs targeting 15000 human genes (Sigma). On average, there were five shRNA designs for each gene target. K562 were stably transduced with a multiplicity of infection (MOI) ≤ 1 such that the majority of cells will be transduced with a single shRNA. Transduced cells were selected using 3 µg/ml puromycin for 4 days. As control K562 cell were infected with an shRNA (sgCTR). Transduced cells were cocultured with NK cells isolated from HD at an E:T ratio of 15:1 for 3 days. The surviving cells were FACS sorted for cell viability and morphology and seeded clonally in 96 well plates. The clones obtained were expanded and genomic DNA extracted by standard procedures. Lentiviral particles were generated in HEK293T cells by combining a pLKO.1 plasmid containing shRNA sequences, packaging plasmid pCMV-dR8.74, and envelope plasmid VSV-G/pMD2.G using TransIT-293 transfection reagent (MIRUS Bio LLC, Madison, WI, USA). K562 cells were infected by the spin inoculation method with lentivirus containing a nontarget shRNA control vector (SHC002) or either of MICA, PLAC1, TCF7 and PLEK shRNAs (clone ID: TRCN0000061288, TRCN0000061031, TRCN0000061677, and TRCN0000060846) targeting the indicated genes (Sigma-Aldrich).

Full-length cDNAs encoding human PLAC1 (accession number AF234654.1) was cloned from JEG3 cells in the lentiviral vector pRRL-CMV-PGK-GFP-WPRE (TWEEN) under the control of the CMV promoter. K562 cells were transfected with PLAC1 vector or empty vector using LipofectAMINE 2000 according to the manufacturer’s instructions (Invitrogen Life Technologies). Stably transfected cells were sorted by FACS as GFP expressing cells.

Genomic DNA was isolated with standard procedures and primers designed to flank the shRNA hairpins were used to amplify the shRNA sequences. Hits identified by sequencing the PCR amplicon were matched to database.

Total RNA was extracted using TRIzol Reagent (Invitrogen). First-strand cDNA was synthesized using the SuperScript II First Strand cDNA synthesis kit (Invitrogen). Quantitative real-time PCR (qPCR) reactions were performed using pre-validated TaqMan gene expression assays from Applied Biosystems (Hs06598311_m1 for PLAC1, Hs01556515_m1 for TCF7, Hs01043767_m1 for PLEKHA5, Hs02786624_g1 for GAPDH). Relative gene expression was determined using the 2-ΔΔCt method with ß-actin as endogenous control.

NK cell cytotoxic activity and degranulation assay were performed by a standard 4-hour ⁵¹Cr-release assay and flow cytometric analysis of cell-surface CD107a expression, respectively. In cytotoxicity assay K562 cells were labelled with ⁵¹Cr [Amersham International; 100μCi (3.7 MBq)/1 x 10⁶ cells] and co-cultured (5 x 10³) with NK cells or NKL cell line at different E:T cell ratios, in 96-well plates round bottom in triplicates, and incubated at 37°C. At 4 hours of incubation, 25μL supernatant were removed, and the ⁵¹Cr release was measured with TopCount NXT beta detector (PerkinElmer Life Sciences). The percentage of specific lysis by counts per minute (cpm) was determined as follows: 100 x (mean cpm experimental release – mean cpm spontaneous release)/(mean cpm total release - mean cpm spontaneous release). Specific lysis was converted to lytic units (L.U.) calculated from the curve of the percentage lysis (54) and defined as the number of NK cells required to produce 20% lysis of 10⁶ target cells during the 4 hours of incubation. In degranulation assays, NK cells were co-cultured with K562 target cells at 1:1 ratio for 3 hours, in complete medium in the presence of anti-CD107a (diluted 1:100). During the last 2 hours, GolgiStop (BD Bioscences) was added at 1:500 dilution. Cells were firstly pre-stained with Live/Dead Kit (L/D), stained with anti-CD56, anti-CD3 and then, the expression of CD107a was evaluated in the CD3^-CD56⁺ subset by flow cytometry.

The following antibodies were used: anti-CD3-Alexa-700 (UCHT1), anti-CD56-PE-Cy7 (B159), anti-CD107a-FITC (H4A3), anti-IgG1-FITC (A85-1), anti-IgM-PE (R6-60.2), anti-MICA/B-BV650 (6D4), anti-ULBP2/5/6-BV510 (165903), anti-CD155/PVR-BV605 (SKII.4), anti-CD112/Nectin-2-PE (R2.525), anti-CD95/FAS-BV421 (DX2), anti-CD262/TRAIL/R2-PE (YM366), anti-ULBP1-PE (170818), anti-ULBP3-PE (166510), anti-FAS-BV421 (DX2), anti-B7-H6-BV421 (1A5), anti-DSG2-PE (6D8), purchased from BD Biosciences. For indirect staining, goat F(ab’)2 Fragment anti-mouse IgG FITC (IM1619) was used. All these antibodies were used according to the manufacturers’ protocol. Prior to surface staining, NK cells were pre-stained with Live/Dead™ Fixable Near-IR Dead Cell Stain Kit (Invitrogen). Flow cytometry was performed by using FACSCalibur, FACSCantoII or FACSFortessa X-20 (BD Biosciences) and analyzed by FlowJo Software. Recombinant human NKG2D-Fc, DNAM1-Fc, NKp30-Fc and NKp44-Fc chimera proteins were purchased from R&D. Cells were incubated with fusion proteins for 30 min in ice, washed twice and then incubated with PE-conjugated goat anti-human Fc antibody (Jackson ImmunoResearch Laboratories) for 30 min on ice, and propidium iodide was added to cells before flow analysis.

Equal amounts of protein extracts were resolved on 15% or 10% polyacrylamide gel for the detection of PLAC1 and MICA, respectively, and transferred on nitrocellulose membranes (Amersham Systems, Ge Healthcare Sciences). Filters were blocked with 5% (v/v) non-fat dry milk for 1 hour at room temperature, and then blotted with rabbit anti-human PLAC1 polyclonal antibody (kindly provided by Prof. Zarnani, Sina Biotech Co) or mouse anti-human PLAC1 (G1, Santa Cruz Biotechnology, Cat No 365919), and mouse anti-human MICA monoclonal antibody (Proteintech, Cat No. 66384-1-Ig) to recognize PLAC1 and MICA, respectively. Anti-ERp57, anti-β−actin or anti-GAPDH antibodies were used as loading control. After extensive washing with TBST, filters were incubated with peroxidase-coupled secondary antibody for 1 hour at room temperature. Reactivity was detected with the ECL Western Blotting Detection Kit (Amersham Systems, Ge Healthcare Life Sciences) and the protein bands were quantified using Image J.

Pan-cancer gene expression data and clinical data from 33 tumor types were downloaded from The Cancer Genome Atlas (TCGA, https://portal.gdc.cancer.gov/, accessed April 5^th, 2024) database. The survival curves using log-rank test of Kaplan-Meier overall survival (OS) and the forest-plots indicating hazard-ratio and 95% confidence intervals for different cut-off values for PLAC1 expression were conducted using the survival R package (version 3.4.0) under R version 4.4.1 (https://www.R-project.org/). P values<0.05 were considered significant. The expression of PLAC1 and DSG2 in healthy tissues and cell types was obtained from the Human Protein Atlas (HPA, http://www.proteinatlas.org/, accessed July 2^nd, 2024) database. The Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/, accessed September 6^th, 2024) database was used to obtain normalized scRNA-seq data from NK cells (GSE159624).

For all data, statistical significance was evaluated using GraphPad software. Statistical tests performed are indicated in the figure legends. P values not exceeding 0.05 were considered to be statistically significant.