Abstract

   As a member of the proton-sensing GPCR receptors, GPR132 plays a
   crucial role in regulating immune cell functions, but the mechanism by
   which GPR132 affects natural killer (NK) cells has not yet been
   reported. Here, RNA-seq displayed that the expression of GPR132 was
   reduced in activated NK cells, and the proportion of mature NK cells in
   GPR132^−/− mice was substantially increased compared to WT mice, with
   stronger anti-melanoma capabilities. Further investigation indicates
   that GPR132-deficient NK92 cells expressed more GzmB and IFN-γ and
   exhibited stronger cytotoxicity. Mechanically, GPR132 regulates NK cell
   function through the CSK/ZAP70/NF-κB signaling axis. Down-regulation of
   GPR132 weakens the inhibition of NK cell function by lactate, thereby
   enhancing the functional execution of CAR-NK cells against colorectal
   cancer. These results highlight the previously unrecognized role of
   GPR132 in the regulation of NK cell function and that inhibition of
   GPR132 provided an updated insight for NK cell therapy.
     __________________________________________________________________

   GPR132 exhibits an inhibitory effect on the antitumor capabilities of
   NK cells via the CSK/ZAP70/NF-κB signaling axis.

INTRODUCTION

   The G protein–coupled receptor (GPCR) family is the largest family of
   membrane receptors, with more than 800 members ([48]1). GPCRs have a
   seven-transmembrane domain and play a crucial role in regulating
   physiological and pathological processes ([49]2). Around 35% of
   marketed drugs achieve therapeutic effects through GPCRs, making them
   attractive drug targets ([50]3, [51]4). There is a class of GPCRs that
   can sense an extracellular acidic environment and are activated at low
   pH to regulate physiological processes such as cell growth, metabolism,
   and migration. These include GPR4, GPR65 (TDAG8), GPR68 (OGR1), and
   GPR132 (G2A) ([52]5, [53]6). GRR132 is the most abundant proton-sensing
   GPCR in immune cells and is involved in migration, proliferation, and
   differentiation ([54]7–[55]10). It has been reported apoptosis and
   differentiation in acute myeloid leukemia (AML) cells can be inducible
   by activating GPR132 using ONC212 and 8-GL ([56]7, [57]11). In the
   breast cancer microenvironment, GPR132 inhibitors interfere with the
   lactate-GPR132-peroxisome proliferator–activated receptor γ axis and
   encourage tumor-associated macrophages to the M1-phenotype
   transformation ([58]12). Furthermore, antagonists of GPR132 can alter
   macrophage reprogramming within pancreatic islets, thereby improving
   diabetes ([59]13). These studies suggest that GPR132 is a potential
   immunotherapy checkpoint.

   Natural killer (NK) cells are a type of innate immune cell that plays a
   crucial role in controlling malignancy and infections. In contrast to T
   cells, NK cells are able to target MHC (or HLA) molecule-deficient
   tumor cells, induce apoptosis by releasing cytotoxic perforins and
   granzymes, and do not cause immune rejection in adoptive cell therapy
   ([60]14). Therefore, NK cells have great potential in antitumor immune
   cell therapy. Despite this, NK cells are still not widely used in the
   clinical treatment of solid tumors. On the one hand, solid tumors are
   difficult to reach and infiltrate ([61]15), and on the other hand, the
   main obstacle is the hypoxic tumor microenvironment (TME) that produces
   large amounts of lactate through metabolism, leading to NK cell
   depletion ([62]16, [63]17). Recently, most studies have focused on
   using immunostimulants, immune checkpoint blockade (IBD), and chimeric
   antigen receptor (CAR) to enhance NK cell antitumor activity and
   specificity ([64]18).

   In our research, we demonstrated that GPR132 affects NK cell function
   through the Gαs/C-terminal Src kinase (CSK)/ZAP70/nuclear factor κB
   (NF-κB) signaling pathway. In addition, the down-regulation of GPR132
   in NK cells weakens the inhibition by lactate and substantially
   enhances the effectiveness of CAR-NK92 cells against colorectal cancer.
   In summary, our study uncovers the regulatory mechanism of GPR132 on NK
   cells, aiming to reduce the immunosuppression of CAR-NK cells in the
   solid TME by down-regulating GPR132, which offers valuable insights for
   cell therapy.

RESULTS

The expression of GPR132 decreases after NK cell activation

   First, we isolated primary NK cells from human peripheral blood
   mononuclear cells (PBMCs). The treatment group was stimulated with
   ionomycin and phorbol 12-myristate 13-acetate (PMA), and the untreated
   cells were used as control. After 6 hours of activation, the samples
   were collected, and RNA sequencing (RNA-seq) analysis was performed.
   The RNA-seq revealed 3446 differentially expressed genes shared by the
   three donors, most of which were associated with biological processes
   (fig. S1, A and B). The heatmap displays a notable up-regulation of NK
   cell activation markers, including tumor necrosis factor (TNF), CD69,
   and interferon-γ (IFN-γ). In contrast, negative regulatory genes, such
   as TIGIT and CD244, were notably down-regulated ([65]Fig. 1A). In
   addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis
   revealed that NF-κB and other related pathways were enriched following
   NK cell activation (fig. S1C). The volcano map of the differential
   genes showed notable changes in the expression levels of certain GPCRs
   of interest ([66]Fig. 1B). Statistical analysis revealed that the
   GPR132, a member of the proton-sensing GPCR family, was mostly
   significantly down-regulated ([67]Fig. 1C). Furthermore, we stimulated
   the NK92 cell line with ionomycin and PMA; collected the cells at 0, 6,
   and 12 hours to extract RNA; and performed reverse transcription
   polymerase chain reaction (RT-PCR) and Western blot (WB). Consistent
   with the RNA-seq results, the expression of GPR132 decreased profoundly
   in activated NK92 cells ([68]Fig. 1, D and E). Consequently, we
   hypothesized that GPR132 may be related to NK cell function regulation.

Fig. 1. The expression of GPR132 decreased in activated NK cells.

   [69]Fig. 1.
   [70]Open in a new tab

   (A to C) RNA-seq analysis of gene expression in activated NK cells
   compared with a control group (n = 3). (A) Heatmap shows the expression
   variation of NK activation–related markers and inhibition of NK
   activation markers. (B) Volcano plot shows 3446 genes shared by three
   donors. Down-regulated genes, blue; up-regulated genes, yellow;
   interested genes, red. (C) Column chart shows the expression of the
   interested GPCR gene according to FPKM value. (D) RT-PCR to detect the
   GPR132 expression in NK92 cells after being stimulated by ionomycin (1
   μg/ml) and PMA (20 ng/ml) at 0, 6, or 12 hours. β-Actin was used as a
   reference housekeeping gene (n = 3, biological replicates). (E) WB
   determined the GPR132 expression in NK92 cells after being stimulated
   by ionomycin (1 μg/ml) and PMA (20 ng/ml) for 12 hours. β-Tubulin was
   used as a reference housekeeping gene. Data represent means ± SD.
   Analyzed by two-way analysis of variance (ANOVA), one-way ANOVA, or
   unpaired t test (*P < 0.05, **P < 0.01, ***P < 0.001, and ****P <
   0.0001).

Deletion of GPR132 enhances the anti-melanoma function of NK cells in mice

   NK cell proportions in different organs from GPR132^−/− mice and
   wild-type (WT) mice were analyzed by flow cytometry. The results showed
   that the percentage of NK cells in the peripheral blood of GPR132^−/−
   mice increased significantly compared to WT mice, while the proportion
   of T cells remained unchanged (fig. S2). We further observed a
   considerable increase in splenic NK cells of GPR132 deficient mice
   ([71]Fig. 2A), manifested by an increase in the percentage and number
   of CD27^−CD11b^+ NK population ([72]Fig. 2B), which was identified as
   mature NK cells with strong lysis capacity ([73]19). However, there
   were no notable proportion differences in other immune cells present in
   the spleen (fig. S3, A to D). In addition, a moderate increase in NK
   cells was noted in the liver, lung, and bone marrow of GPR132-deficient
   mice (fig. S4, A to D). Afterward, we injected 5 × 10^5 B16-F10 cells
   into WT and GPR132^−/− mice subcutaneously to establish a
   melanoma-bearing model. As per the findings, the tumor volume and
   weight bore with GPR132^−/− mice was notably lower than in WT mice
   ([74]Fig. 2, C, E, and F). In addition, compared to WT mice,
   GPR132-deficient mice exhibited a higher survival rate ([75]Fig. 2D),
   an elevated number of tumor-infiltrating NK cells ([76]Fig. 2G), which
   expressed more granzyme B (GzmB) ([77]Fig. 2H) and IFN-γ ([78]Fig. 2I).
   The results above suggest that GPR132 may affect the maturation of NK
   cells and thus the antitumor ability of mice, but we cannot rule out
   ILC1s from this NK1.1 population, especially in the liver and the lung.

Fig. 2. Knockout of GPR132 enhanced the ability of NK cells to resist
melanoma.

   [79]Fig. 2.
   [80]Open in a new tab

   (A and B) FACS analysis of the NK cells (A) and CD27^−CD11b^+ NK cells
   (B) percentage in the spleen (n = 4). The representative FACS profile
   is on the left, and the column chart is on the right. (C) Tumor volume
   was measured every 2 days, and growth curves were shown (n = 5). (D)
   Kaplan-Meier survival curve visually depicted the overall survival rate
   for WT and GPR132^−/− mice in a 50-day survival study (n = 10). (E to
   I) Melanomas were divested when the tumor volume reached 1500 to 2000
   mm^3, the tumors were photographed (E) and weighted (F) (n = 5), then
   the proportion of NK cells inside the tumors (G) and the expression of
   GzmB (H) and IFN-γ (I) in tumor-infiltrating NK cells (n = 3) were
   analyzed by FACS. (J) FACS analysis of the NK cell proportion in the
   peripheral blood of recipients after receiving WT and GPR132^−/− bone
   marrow cells for 4 weeks. (K) Schematic diagram of melanoma lung
   metastasis experiment. On day 0, 3 × 10^5 B16-F10 cells were injected
   into per mouse via the tail vein. (L) On day 10, FACS analysis was
   performed to determine the NK cell depletion in the peripheral blood of
   anti-NK1.1 or PBS-treated mice. On day 14, the melanoma colony numbers
   were counted in the lungs and the weight of the lungs was measured (n =
   3). (M to O) Photographs of lungs from the PBS and anti-NK1.1–treated
   group are shown at the top (M), and the statistical charts [(N) and
   (O)] are shown at the bottom. Data are shown as means ± SD and were
   analyzed by unpaired t test or log-rank test (*P < 0.05, **P < 0.01,
   ***P < 0.001, and ****P < 0.0001; ns, not significant).

   To further verify whether the inhibitory of GPR132 on NK cell
   homeostasis, we transplanted bone marrow cells from WT and GPR132^−/−
   mice into WT mice irradiated with lethal doses (fig. S5A) and observed
   NK cell proportion and antitumor activity in recipients. Consistent
   with previous results, the mice that received GPR132^−/− bone marrow
   cells exhibited a higher proportion of NK cells in peripheral blood and
   inside the tumor as well as showed stronger resistance to melanoma
   compared to the WT group ([81]Fig. 2J and fig. S5, B and C). Moreover,
   to further explore whether GPR132 deletion could promote the ability of
   NK cells to resist tumor metastasis, we conducted a melanoma lung
   metastasis experiment ([82]Fig. 2K). We found that melanin colonies
   metastasized in the lungs of GPR132^−/− mice were lower than in WT
   mice, and this difference was abolished with the administration of
   anti-NK1.1 ([83]Fig. 2, L to O). The findings reveal that GPR132
   deletion promotes the antitumor ability of mice through NK cells.

The deficiency of GPR132 enhances the cytotoxicity and activation of NK cells

   To further demonstrate the effect of GPR132 on NK function in vitro, we
   isolated splenic NK cells from both WT and GPR132^−/− mice and
   co-incubated with Yac-1 cells for 4 hours at different
   effector-to-target (E:T) ratios. GPR132 deletion resulted in an
   increased killing efficiency of NK cells, as shown in [84]Fig. 3A. In
   addition, the expression of CD107a, GzmB, and IFN-γ in NK cells from
   GPR132^−/− mice increased considerably compared to WT mice ([85]Fig.
   3). To further clarify the effect of GPR132 on NK cells, we designed
   three GPR132–short hairpin RNA (shRNA) sequences and a negative control
   (NC) shRNA (shNC) sequence targeting the NK92 cell for in vitro
   functional validation (fig. S6, A and B). First, we verified that
   shRNA2 was the optimal sequence to knock down the GPR132 gene at
   transcriptional and protein levels by RT-PCR and WB ([86]Fig. 3E and
   fig. S6, C and D) and applied it to subsequent experiments.
   Subsequently, the functionality of GPR132-interfering (SH) and
   non-interfering (NC) NK cells was evaluated after co-incubating with
   K562 cells for 4 hours. Our findings indicate that SH-NK92 cells
   exhibited stronger cytotoxicity against K562 cells at different E:T
   ratios compared to NC-NK92 cells ([87]Fig. 3F and fig. S7) and
   expressed more CD69, CD107a, IFN-γ, and GzmB than NC-NK92 cells
   ([88]Fig. 3, G to J), the similar results have been demonstrated with
   primary human NK cells (fig. S8, A to C). Briefly, these data indicate
   that the knockdown of GPR132 enhances the NK cell function.

Fig. 3. Deficiency of GPR132 improves the cytotoxicity of NK cells.

   [89]Fig. 3.
   [90]Open in a new tab

   (A) The line plots displayed the killing ability of WT and GPR132^−/−
   mice splenic NK cells against Yac-1 cells for 4 hours. The E:T ratios
   are 5:1, 10:1, and 20:1. (B to D) FACS analysis of CD107a, GzmB, and
   IFN-γ expression in splenic NK cells of WT and GPR132^−/− mice after
   coculture with target cells at a 10:1 ratio for 4 hours. (E) NK92 cells
   were infected with Mock, shNC, and shRNA lentivirus for 48 hours and
   then collected to perform an RT-PCR analysis to verify the interference
   efficiency of GPR132. The multiplicity of infection (MOI) is 10. (F)
   Line plots illustrate the killing ability of NK92 cells against K562
   cells at various E:T ratios over a 4-hour duration. NC was the NK92
   cells infected with shRNA-NC lentivirus, and SH was the NK92 cells
   infected with GPR132-shRNA2 lentivirus at MOI = 10. (G to J) After
   coculture with K562 cells at a 5:1 (E:T) ratio for 4 hours, FACS was
   conducted to detect the expression of CD69 (G), CD107a (H), IFN-γ (I),
   and GzmB (J). For (H) and (J), the representative FACS histogram is
   displayed on the left, and the column chart is on the right. The
   experiment was performed at least three times. Data are shown as means
   ± SD. Analyzed by unpaired t test or one-way ANOVA (*P < 0.05, **P <
   0.01, ***P < 0.001, and ****P < 0.0001). MFI, mean fluorescence
   intensity.

Activation of GPR132 inhibits the cytotoxicity and activation of NK cells

   ONC212 has been reported to be a selective agonist of GPR132 ([91]11).
   To further verify that GPR132 has a negative impact on NK function, we
   conducted the following experiments. On the basis of the determination
   of ONC212 concentration in fig. S9, NK92 cells, NC-NK92 cells, and
   SH-NK92 cells were treated with dimethyl sulfoxide (DMSO) or ONC212 (5
   μM) for 12 hours and subsequently cocultured with K562 at different E:T
   ratio for 4 hours to analyze cytotoxicity. Compared to the DMSO group,
   the killing efficiency of ONC212-treated NK92 cells was substantially
   reduced ([92]Fig. 4A and fig. S10). Simultaneously, the expression of
   CD69 and CD107a was also decreased ([93]Fig. 4, B and C). However, this
   effect was eliminated after GPR132 knockdown, demonstrating that ONC212
   specifically activates GPR132 ([94]Fig. 4, A to C, and fig. S10).
   However, this effect was eliminated after GPR132 knockdown,
   demonstrating that ONC212 specifically activates GPR132 ([95]Fig. 4, A
   to C, and fig. S10). In addition, it is well known that a large amount
   of lactate produced by the TME inhibits the function of immune cells
   ([96]20). We investigated whether lactate could have a similar effect
   to ONC212, given that GPR132 belongs to the pH-sensitive GPCR family
   and is supposed to be activated by lactate ([97]12, [98]21). Thus,
   NC-NK92 cells and SH-NK92 cells were treated with 10 mM lactate (HL)
   for 12 hours according to the determination of lactate concentration in
   fig. S11 and then co-incubated with HCT116 and MGC803 for 4 hours at
   different E:T ratios. The bioluminescence imaging displayed that
   lactate-treated SH-NK92 cells still maintained excellent killing
   capacity and higher activation levels than lactate-treated NC-NK92
   cells ([99]Fig. 4, D to G). In contrast with the NC-NK92 cohort,
   lactate-treated SH-NK92 exhibited no significant difference in GzmB and
   IFN-γ expression ([100]Fig. 4, F and G) compared with untreated SH-NK92
   cells. Our findings revealed that activation of GPR132 inhibited the
   function of NK cells, and GPR132 deficiency could weaken the inhibition
   of NK cells by lactate.

Fig. 4. Activation of GPR132 inhibits the function of NK92 cells.

   [101]Fig. 4.
   [102]Open in a new tab

   (A) NK92 cells, NC-NK92 cells, and SH-NK92 cells were treated with DMSO
   or 5 μM ONC212 for 12 hours and cocultured with K562 cells at various
   E:T ratios for 4 hours to detect cytotoxicity. (B and C) Expression of
   CD107a and GzmB in the indicated group was analyzed by flow cytometry
   after coculture with K562 cells at a 5:1 (E:T) ratio for 4 hours. The
   representative FACS histogram is displayed on the left, and the column
   charts are on the right. (D and E) Lysis rate of HCT116-luciferase
   cells (D) and MGC803-luciferase cells (E) was detected by
   bioluminescence imaging after 4 hours cocultured with indicated NK92
   group at different E:T ratios. NC + HL and SH + HL mean NK92 cells were
   treated with 10 mM lactate acid for 12 hours, with untreated NC-NK92
   and SH NK92 cells as a control. (F and G) NC-NK92 and SH-NK92 cells
   were treated with or without 10 mM lactate for 12 hours and cocultured
   with HCT116 at a 5:1 (E:T) ratio for 4 hours, and subsequently, the
   expression of GzmB (F) and IFN-γ (G) were analyzed by flow cytometry.
   The representative FACS profile is displayed on the left, and the MFI
   column is on the right. Data are shown as means ± SD, and the
   experiment was performed three times. The experiment was performed
   three times. Data are shown as means ± SD. Analyzed by one-way ANOVA,
   unpaired t test, or two-way ANOVA (*P < 0.05, **P < 0.01, and ***P <
   0.001).

GPR132 affects the proliferation and anti-apoptotic ability of NK cells

   Previously, we noted that there was a noticeable increase in the
   proportion of NK cells in GPR132 knockout mice. As a result, we carried
   out additional validation using the human NK92 cells. We found that the
   knockdown of GPR132 resulted in a marked increase in the expression
   abundance of the proliferation antigen Ki67 ([103]Fig. 5A).
   Furthermore, we cultured NC-NK92 cells and SH-NK92 cells without human
   recombinant interleukin-2 (hIL-2) to investigate their apoptosis and
   survival. The results revealed that, in comparison to the NC-NK92
   cells, SH-NK92 cells presented reduced apoptosis ([104]Fig. 5B),
   accompanied by a considerable increase in the expression of the
   anti-apoptotic protein Bcl-2 ([105]Fig. 5C). Conversely, ONC212-treated
   NK92 cells displayed lower expression of Ki67 ([106]Fig. 5D), increased
   apoptosis rates, and decreased Bcl-2 expression compared to the
   DMSO-treated NK92 cells ([107]Fig. 5, E and F). These results indicated
   that GPR132 may regulate the proliferation and apoptosis of NK cells,
   and in the absence of IL-2, GPR132 deficiency can enhance the
   anti-apoptotic ability and survival of NK cells.

Fig. 5. GPR132 affects the proliferation and apoptosis of NK92 cells.

   [108]Fig. 5.
   [109]Open in a new tab

   (A to C) NK92 cells were infected with shRNA-NC or shRNA-GPR132
   lentivirus for 48 hours at MOI = 10 and cultured for 3 to 5 days. On
   day 5, the cells were collected to analyze Ki67 expression (A), and on
   day 4, annexin V^+ cell proportion (B) and Bcl-2 expression (C) were
   analyzed by flow cytometry. (D to F) NK92 cells were treated with DMSO
   or ONC212 (5 μM) for 12 hours and cultured for continuous 2 to 5 days.
   FACS analysis was performed for Ki67 expression (D) on day 5, annexin
   V^+ cell percentage (E) on days 2 and 4, and Bcl-2 expression (F) on
   day 4. For analysis of annexin V^+ rate and Bcl-2 expression, NK92
   cells were cultured without supplemented hIL-2. For (A), the
   representative FACS profile is shown at the top, and the column charts
   are at the bottom. For (B) to (F), the representative FACS histogram
   (left) and the column chart (right) are shown. The experiment was
   performed three times. Data are shown as means ± SD. Analyzed by
   unpaired t test or two-way ANOVA (*P < 0.05, **P < 0.01, ***P < 0.001,
   and ****P < 0.0001).

GPR132 regulates NK cell function through the Gαs/CSK/ZAP70/NF-κB signal axis

   The heatmap in [110]Fig. 1 shows that the ZAP70 gene is remarkably
   up-regulated with NK cell activation and enriched among the NF-κB
   pathway (fig. S1C), which is critical for NK function ([111]22,
   [112]23). It has been reported that ZAP70 is regulated by CSK in immune
   cells ([113]24, [114]25). Hence, we assessed the expression of CSK,
   pZAP70, and p-P65 in NK92 cells after down-regulation or activation of
   GPR132. We found that GPR132 positively regulated the expression of CSK
   but negatively regulated pZAP70 and p-P65 expression ([115]Fig. 6, A to
   F, and fig. S12, A to F). To investigate whether CSK regulates the
   expression of pZAP70 and p-P65, we treated NK92 cells and ONC212-NK92
   cells with DMSO or 10 μM CSK degrader, and the flow cytometry analysis
   was performed (fig. S13A). The results showed that the CSK degrader
   relieved the inhibition of pZAP70 and p-P65 expression caused by GPR132
   activation ([116]Fig. 6G and fig. S13, B and C), which indicates that
   CSK inhibits the ZAP70 and NF-κB pathway. We further treated NC-NK92
   and SH-NK92 cells with DMSO or ZAP70 inhibitors to verify that ZAP70
   certainly regulates the NF-κB pathway ([117]Fig. 6H and fig. S14, A and
   B). Moreover, NF-κB inhibitor triggered a significant reduction in the
   expression of GzmB, IFN-γ, and perforin in SH-NK92 cells compared to
   the untreated group ([118]Fig. 6, I to K, and fig. S15, A to D). We
   also demonstrated that GPR132 was positively correlated with the
   expression of Gαs but had no effect on Gαi1/2/3 (fig. S16, A, and B),
   which is consistent with a previous report ([119]7). These results
   indicate that GPR132 regulates NK cell function through the
   Gαs/CSK/ZAP70/NF-κB signaling pathway ([120]Fig. 6L).

Fig. 6. GPR132 inhibits the function of NK cells via the Gαs/CSK/ZAP70/NF-κB
signal axis.

   [121]Fig. 6.
   [122]Open in a new tab

   (A to C) FACS analysis of the CSK, pZAP70, and p-P65 expression in
   NC-NK92 and SH-NK92 cells. (D and F) FACS analysis of the CSK, pZAP70,
   and p-P65 expression in DMSO-NK92 and ONC212-NK92 cells. CSK was
   stained by primary antibody at 4°C for 30 min and then stained with
   AF647 anti-rabbit IgG antibody. (G) DMSO-NK92 and ONC212-NK92 cells
   were treated with DMSO or 10 μM DB-3-291 (CSK degrader) to analyze the
   expression of pZAP70 and p-P65 by flow cytometry. (H) NC-NK92 and
   SH-NK92 cells were treated with DMSO or 5 μM ZAP180013 (ZAP70
   inhibitor) to detect the expression of p-P65 by flow cytometry. (I to
   K) NC-NK92 and SH-NK92 cells were treated with DMSO or 5 μM QNZ (NF-κB
   inhibitor) to detect the expression of GzmB, IFN-γ, and perforin by
   flow cytometry. (L) Graphical abstract shows the summary of the
   signaling mechanism. The experiment was performed three times. Data are
   shown as means ± SD. Analyzed by unpaired t test or one-way ANOVA (*P <
   0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001).

Down-regulation of GPR132 enhances the cytotoxicity and activation of CAR-NK
cells

   According to the methods in our previously published paper ([123]23,
   [124]26), we developed an NKG2D/4-1BBζ-NK cells with a knockdown of
   GPR132 (fig. S17, A and B), which performed considerable GPR132
   interference efficiency at RT-PCR and WB levels (fig. S17, C and D). As
   target cells, colorectal cancer cell lines HCT116 and HCT15 display
   high expression of NKG2D ligand MICA/B (fig. S18). Therefore, we
   incubated Mock-NK92 cells, CAR-NK92 cells, NC-CAR-NK92 cells, and
   SH-CAR-NK92 cells with HCT116 and HCT15 for 4 hours, respectively, with
   the 2.5:1, 5:1, and 10:1 E:T ratio. The data indicated that NK92 cells
   expressing CAR could effectively kill target cells, whereas SH-CAR-NK92
   cells demonstrated enhanced cytotoxicity ([125]Fig. 7, A and B). To
   detect the expression of NK activation–related proteins, CAR-NK92 cells
   were cocultured with HCT116 for 4 hours at a 5:1 (E:T) ratio. The
   results revealed that SH-CAR-NK92 cells expressed higher levels of CD69
   and CD107a than CAR-NK92 cells and NC-CAR-NK92 cells ([126]Fig. 7, C
   and D). In addition, the expression of GzmB and IFN-γ was higher in
   SH-CAR-NK92 cells than in the other groups ([127]Fig. 7, E and F). The
   aforementioned results indicated that GPR132 knockdown increased the
   cytotoxicity capability of CAR-NK92 cells. Furthermore, it was observed
   that the killing capacity of SH-CAR-NK92 cells remained stronger than
   that of NC-CAR-NK92 cells even after a 12-hour treatment with lactate.
   The cytotoxicity of the lactate-treated NC-CAR-NK92 cells was notably
   reduced compared to the untreated group, while in lactate-treated
   SH-CAR-NK92 cells, the reduction was not significant (fig. S19, A and
   B). Briefly, the down-regulation of GPR132 could enhance the CAR-NK
   cell function and could attenuate the inhibition of CAR-NK cell
   function by lactate.

Fig. 7. Down-regulation of GPR132 promotes the antitumor activity of CAR-NK92
cells in vitro.

   [128]Fig. 7.
   [129]Open in a new tab

   To prepare CAR-NK92 cells, 5 × 10^5 NK92 cells were transduced with
   mock, CAR, NC-CAR, and SH-CAR lentiviruses, respectively, at MOI = 10.
   (A and B) Line plots displayed the killing ability of mock-NK92 cells,
   CAR-NK92 cells, NC-CAR-NK92 cells, and SH-CAR-NK92 cells against HCT116
   cells and HCT15 cells at 2.5:1, 5:1, and 10:1 (E:T) ratios for 4 hours.
   (C and D) Expression of CD69 and CD107a in the indicated group was
   performed by flow cytometry after coculture with HCT116 cells at a 5:1
   ratio for 4 hours. The representative FACS histogram (left) and the MFI
   column chart (right) are displayed. (E and F) FACS analysis of GzmB and
   IFN-γ expression in indicated groups after co-culture with HCT116 cells
   at a 5:1 ratio for 4 hours. The representative FACS profile is shown on
   the left, and the MFI column chart is on the right. The experiment was
   performed three times. Data are shown as means ± SD. Analyzed by
   one-way ANOVA (*P < 0.05, **P < 0.01, ***P < 0.001, and ****P <
   0.0001).

Knockdown of GPR132 enhances the antitumor activity of CAR-NK92 cells in vivo

   Next, we further evaluated the ability of GPR132-knockdown CAR-NK cells
   to eradicate colorectal cancer in vivo. Considering that HCT116 cells
   are more sensitive to NKG2D-CAR-NK92 cells than HCT15 cells, HCT116
   cells were selected to establish a colorectal cancer xenograft mouse
   model. A total of 3 × 10^6 HCT116-luciferase cells were injected
   subcutaneously into NSG mice. Mice were randomly divided into four
   groups when tumor volumes reached 60 to 100 mm^3 and received PBS,
   Mock-NK92 cells, NC-CAR-NK92 cells, and SH-CAR-NK92 cells (5 × 10^6
   cells per mouse) through the tail vein, respectively. At the same time,
   hIL-2 (20,000 IU per mouse) was injected intraperitoneally. Weekly
   injected the treatment NK92 cells and hIL-2 and performed in vivo
   imaging system (IVIS) and tumor measurements every few days ([130]Fig.
   8A). The results showed that the SH-CAR-NK92 cells exhibited a marked
   antitumor capacity in vivo compared to Mock-NK92 and NC-CAR-NK92 cells,
   and the volume of tumors from mice treated with SH-CAR-NK92 cells was
   lower than in the other groups ([131]Fig. 8, B to D, and fig. S20, A
   and B). In addition, the mice treated with SH-CAR-NK92 cells showed a
   higher survival rate, as indicated in [132]Fig. 8E. We also evaluated
   the number and activation markers of CAR-NK92 cells inside the tumor.
   The results indicated that the proportion of tumor-infiltrating NK
   cells in the SH-CAR-NK92 group was the highest ([133]Fig. 8F), and the
   expression of CD107a, GzmB, and IFN-γ was remarkably elevated compared
   to other groups ([134]Fig. 8, G to I, and fig. S20, C to E). The
   findings definitely prove that the down-regulation of GPR132 can
   effectively improve the anti-colorectal cancer efficacy of CAR-NK cells
   in vivo.

Fig. 8. Down-regulation of GPR132 enhances the antitumor ability of CAR-NK92
cells in vivo.

   [135]Fig. 8.
   [136]Open in a new tab

   (A) Schematic representation of the animal experimental design. (B)
   IVIS was used to assess the progression of HCT116-luciferase tumors in
   each group on days 0, 13, 19, and 27 (n = 4). (C) On day 27, the total
   bioluminescence flux (photons per second) emitted by tumors of the
   remaining mice in each group was quantified through IVIS. (D) Tumor
   volume was regularly assessed using calipers and presented with line
   plots (n = 4). (E) Overall survival rate for mice in each experimental
   group was depicted by the Kaplan-Meier survival curve (n = 8). (F) FACS
   analysis of the proportion of NK cells (CD45^+CD56^+) inside the tumor
   (n = 3). The representative FACS profile is shown on the left, and the
   column chart is on the right. (G to I) Expression of CD107a, GzmB, and
   IFN-γ in tumor CD45^+CD56^+ cells was determined by FACS (n = 3). Data
   are shown as means ± SD and were analyzed by one-way ANOVA (*P < 0.05,
   **P < 0.01, ***P < 0.001, and ****P < 0.0001).

DISCUSSION

   In recent years, cancer immunotherapy has revolutionized the field of
   oncology, including cytokine therapy, IBD, adoptive cell therapy,
   cancer vaccines, monoclonal antibodies, and CAR therapy ([137]27).
   Among them, the application of therapeutic monoclonal antibodies (mAbs)
   to immunosuppressive “checkpoint” receptors has fundamentally rewritten
   the traditional model of cancer treatment ([138]28). Recently, there
   has been a surge of interest in NK cell adoptive therapies. Unlike T
   cells, NK cells can recognize and respond to cancerous cells without
   prior exposure. In addition, allogeneic NK cells do not trigger immune
   rejection, which strongly promotes the emerging development of NK cell
   adoptive therapy. The combination of NK adoptive cells and ICB therapy
   could achieve remarkable tumor control ([139]29, [140]30). However, the
   majority of clinical patients tolerate immune checkpoint inhibitors
   ([141]31), and the clinical efficacy of adoptive NK cell therapy in
   solid tumors is also unsatisfactory. Therefore, it is essential to
   identify new immune checkpoints and use them to enhance the antitumor
   immune response of NK cells.

   GPCRs are described as very attractive drug targets and play an
   important role in the regulation of NK cell function. It has been
   reported that substance P promotes the expression of perforin and GzmB
   in NK92-MI cells through NK-1R (GPCR) signaling ([142]32). Similarly,
   the activation of GPR55 enhances NK cell activation and IL-12 and TNF-α
   production ([143]33). However, Chang et al. ([144]34) reported that
   GPR56 expression is down-regulated in activated NK cells and inhibited
   cytotoxicity and cytokine production of NK cells. In addition,
   adenosine receptors are expressed in tumor-infiltrating NK cells and
   limit the maturation of NK cells, making it a potential immune
   checkpoint ([145]35, [146]36). Moreover, previous studies in our
   laboratory demonstrated that GPR116 negatively regulates NK cell
   function through the Gαq/HIF1α/NF-κB signaling pathway ([147]23). In
   the present study, RNA-seq results showed that GPR132 was substantially
   down-regulated in activated NK cells, suggesting that GPR132 may
   regulate the NK cell function.

   The proton-sensitive GPCR family is sensitive to low pH and regulates
   cell proliferation and metastasis, immune cell function, inflammation,
   and vascularization ([148]6). Parks et al. ([149]8) found that the
   deletion of GPR132 promoted the migration of macrophages. In addition,
   increased transcription of GPR132 leads to B cell differentiation
   blockage ([150]9). These studies showed that GPR132 plays an important
   role in regulating immune cell function. Therefore, we first performed
   a preliminary study using GPR132^−/− mice. Le et al. ([151]37)
   previously reported that the deletion of GPR132 promotes T cell
   proliferation in aged mice (over 1 year old). Our findings indicated
   that GPR132^−/− mice (6 to 8 weeks old) had a higher proportion of NK
   cells but with no significant difference in other immune cells.
   Accordingly, we performed melanoma lung metastasis and bone marrow
   transplantation experiments, further demonstrating the absence of
   GPR132 promotes the antitumor ability of NK cells. Subsequently, we
   further observed the effects of GPR132 on NK cell function using shRNA
   and ONC212 agonists. The results confirmed that the reduction of GPR132
   levels could boost NK92 cell function, while activation did the
   opposite.

   NF-κB serves as a crucial regulator of cytokine and chemokine
   production in NK cells. In patients with deficiencies in NF-κB
   components [NEMO, inhibitor of nuclear factor κB (IκB), and IκB kinase
   β], NK cells display significant impairments in cytotoxic function and
   IFN-γ production upon recognition of target cells ([152]38, [153]39).
   Our KEGG enrichment showed that the NF-κB signaling pathway was
   significantly enriched in activated NK cells, and the ZAP70 gene is
   among in it, which is known to play a key role in T cell receptor
   signaling and regulates the activation of protein kinase Cθ and NF-κB
   in T cells ([154]40). Hideshima et al. ([155]22) found that
   immunomodulatory drugs can directly activate ZAP-70, thereby leading to
   increased expression of GzmB and activity of NK cells. Our study also
   confirmed that ZAP70 regulates the expression of NF-κB in NK92 cells.
   It is widely known that cyclic adenosine 3′,5′-monophosphate (cAMP) is
   directly produced by G proteins. Within the immune system, cAMP
   inhibits the activity of Lck through the cAMP-dependent protein kinase
   (PKA)/CSK pathway, which further leads to the inhibition of ZAP70
   activation ([156]41). Consistent with the above, our results revealed
   that GPR132 inhibits ZAP70/NF-κB signaling through the Gαs-PKA-CSK
   pathway, further inhibiting the function of NK cells. This suggests
   that GPR132 is a potential immune checkpoint that can be targeted to
   enhance the antitumor response of NK cells.

   Now, CAR-T therapy has revolutionized cancer treatment. Although CAR-T
   cells have shown effectiveness in hematological cancers, they offer
   limited therapeutic benefits for solid tumors. Moreover, CAR-T cells
   may cause graft-versus-host disease, cytokine release syndrome, and
   neurologic toxicities. In contrast, NK cells provide a safer and more
   advantageous CAR-engineering platform ([157]42). Nevertheless, CAR-NK
   cells still face challenges in solid tumors, including target
   selection, CAR engineering, and persistence ([158]43). In addition,
   immunosuppressive TME in solid tumors is another obstacle. Transforming
   growth factor–β is expressed in the majority of TMEs, which inhibits
   the tumor lytic response and causes the expression of eomesodermin
   (EOMES) lost ([159]44), ultimately contributing to NK cell exhaustion.
   TME is also hypoxic, which can down-regulate the expression of
   activating receptors NKG2D and induce the release of HIF1α in
   tumor-infiltrating NK cells, thereby reducing their antitumor response
   and IFN-γ production ([160]17).

   Furthermore, the TME hinders the effectiveness of NK cells by
   metabolizing large amounts of lactate. Because GPR132 is reported to be
   sensitive to lactate in the breast cancer microenvironment ([161]12),
   we treated GPR132-deficient NK92 cells and normal NK92 cells with
   lactate, and the results demonstrated that the GPR132 knockdown can
   lessen lactate-induced inhibition of NK92 cells. Although the evidence
   for GPR132 sensed to low pH is still limited and less reliable than
   other pH-sensing GPCRs, it can be converted to a collaborative pH
   sensor by installing the missing glutamate residues ([162]45). The
   genetically modified NKG2D/4-1BBζ-T cells previously reported in our
   laboratory have enhanced anti-pancreatic and prostate cancer
   capabilities ([163]46–[164]48). In addition, NKG2D/4-1BBζ-T cells have
   been reported to exhibit specific cytotoxic activity against colorectal
   cancer cells ([165]49). On the basis of these findings, we
   down-regulated GPR132 in NKG2D/4-1BBζ-NK92 (CAR-NK92) cells to detect
   its anti-colorectal cancer activity. Compared to traditional CAR-NK
   cells, the GPR132-downregulated CAR-NK cells displayed enhanced
   cytotoxicity against colorectal cancer cells and increased the
   expression of GzmB and IFN-γ. In the treatment of colorectal cancer
   xenografted mice, GPR132-downregulated CAR-NK cells exhibited excellent
   antitumor activity and intratumoral survival. The 4-1BB/CD3ζ
   combination has been prioritized in our CAR-NK design as it has shown
   potent signaling and enhanced cytotoxicity in CAR-NK therapy ([166]23,
   [167]50) as well as a high clinical application rate ([168]51).
   However, DAP10 and DAP12 have also been extensively studied as
   endogenous costimulatory molecules in NK cells ([169]52, [170]53),
   providing advanced insights for our future exploration.

   In summary, our results revealed that the GPR132 receptor regulates NK
   cell function through Gαs/CSK/ZAP70/NF-κB signaling axis. The
   deficiency of GPR132 can attenuate the inhibition of NK cells by
   lactate, consequently increasing the expression of INF-γ and GzmB and
   effectively boosting the anti-colorectal cancer effect of CAR-NK92
   cells. Our study more conclusively proves that GPR132 is a potential
   immune checkpoint and provides an advanced avenue for NK cell therapy.

MATERIALS AND METHODS

Cell preparation and culture

   Human natural killer cell (NK92), chronic myeloid leukemia cell (K562),
   embryonic kidney cell [human embryonic kidney (HEK) 293T], colorectal
   cancer cell (HCT116 and HCT15), gastric cancer cell (MGC803), mouse
   melanoma cell (B16-F10), and mouse lymphoma cell (Yac-1) lines were
   obtained from American Tissue Culture Collection (ATCC, USA). NK92 was
   grown in αMEM medium (Procell, China) supplemented with 12.5% fetal
   bovine serum (FBS; ABW, Uruguay), 12.5% horse serum (Gibco, NY), 0.2 mM
   inositol, 0.1 mM β-mercaptoethanol, 0.02 mM folic acid (Sigma-Aldrich,
   USA), hIL-2 (100 to 200 IU/ml; PeproTech, USA), and 1%
   penicillin-streptomycin (P/S). HEK293T was cultured in Dulbecco’s
   minimum essential medium (Gibco, USA), and K562, HCT116, HCT15, MGC803,
   Yac-1, and B16-F10 were cultured in RPMI 1640 medium (Gibco, USA). The
   culture medium was supplemented with 10% FBS and 1% P/S. Cells were
   grown at 37°C and 5% CO[2]. In addition, HCT116 and HCT15 cells
   transduced luciferase-lentiviral to express firefly luciferase as the
   reporter.

Construction of lentiviral expression plasmid

   Human GPR132 mRNA (GenBank: [171]EU431121.1) was obtained from the
   National Center for Biotechnology Information as a template to design
   shRNA. Three GPR132-shRNA sequences and a NC sequence were randomly
   selected from the website of BLOCK-iT RNAi Designer (listed in table
   S1). At the 3′ end of the shRNA sequence, the loop sequence, respective
   reverse complementary sequence, 6Ts tail, and EcoR І cleavage site were
   added sequentially. Subsequently, inverse complementary sequences were
   generated, and the Xho I sticky and Hpa I sticky were added to the ends
   of the two complementary nucleic acid strands, respectively. To obtain
   the interfering plasmids, the synthesized shRNA sequences were annealed
   and ligated to a pLL3.7 vector containing the U6 promoter and enhanced
   green fluorescent protein. In the CAR-NK study, the most efficient
   shRNA sequence and NC sequence were inserted into the pLL3.7-NKG2D-CAR
   plasmid, which is preserved in our laboratory ([172]46).

Mice

   Global knockout mice GPR132^−/− was developed with CRISPR technology on
   a C57BL/6 background (fig. S21) and gifted by W. Lu. Genotyping was
   performed by PCR using primers (forward 5′-TGGA-CTTCCTCCCCTGATCC-3′;
   reverse 5′- AAACGCAGGTAGTGGTAGCC-3′), and PCR products were sequenced
   to identify the genotype. NOD/SCID/IL2Rγ-chain^−/− (NSG) mice are NOD
   background. All mice were fed, and animal experiments were performed at
   specific pathogen–free conditions. The animal experiments protocols
   were approved by the Institutional Animal Ethics Committee of East
   China Normal University (m20230903).

Bone marrow transplant

   First, WT and GPR132^−/− femurs were harvested and stored in solution A
   (RPMI 1640 + 2% FBS + 1% P/S). Second, the femoral bone marrow cavity
   was flushed with a 1-ml syringe, and the cell suspension was filtered
   with a sterilized cell filter (40 μm). Then, the cells were washed
   three times with B solution (RPMI 1640 + 1% P/S). The cells were
   resuspended and counted after centrifuging at 4°C, 300g, for 5 min.
   Last, 5 × 10^6 bone marrow cells from WT or GPR132^−/− mice were
   injected intravenously through the tail vein into WT recipient mice,
   which were irradiated with lethal doses (9.5 gray of γ-ray). One month
   later, the proportion of NK cells in the peripheral blood of recipient
   mice is analyzed. Moreover, 5 × 10^5 B16-F10 cells were injected
   subcutaneously into recipient mice for further study. A summary diagram
   is shown in fig. S5A.

Lung metastasis model

   B16-F10 melanoma cell line was used to construct a lung metastasis
   model in WT and GPR132^−/− mice. On day 0, each mouse was injected with
   3 × 105 B16-F10 cells via the tail vein. For the depletion of NK cells
   in mice, 200 μg of anti-NK1.1 antibody (108760, BioLegend, USA) was
   diluted with phosphate-buffered saline (PBS) and injected
   intraperitoneally, a total injection volume is 200 μl. The control WT
   and GPR132^−/− mice were injected with PBS. On day 14, the mice were
   euthanized, the lungs were weighed, and the number of colonies of
   melanoma lung metastases was counted.

NK cell isolation

   PBMCs were separated from healthy donor blood, which was provided by
   the Shanghai Blood Center (Shanghai, China). Human primary NK cells
   were enriched from PBMCs by NK cell isolation kit ([173]L00903,
   GenScript, China). Mouse NK cells were isolated from fresh WT or
   GPR132^−/− mice spleen with an NK cell isolation kit (130-115-818,
   Miltenyi, Germany). The cell population was meticulously enriched for
   NK cells, achieving a purity greater than 99%.

RNA sequencing

   Human NK cells isolated from the PBMCs were divided into two groups
   (control and treatment) and incubated in a 24-well plate; the treatment
   group was stimulated by PMA (20 ng/ml) and ionomycin (1 μg/ml). After 6
   hours, cells were collected using TRIzol reagent (Megan, China) and
   flash freezing by liquid nitrogen. RNAs were then extracted
   subsequently, and its integrity was evaluated using the Agilent 2100
   Bioanalyzer (Agilent, USA). The libraries were constructed using the
   Fast RNA-seq Lib Prep Kit V2 (RK20306, ABclonal). The libraries were
   then sequenced using an Illumina NovaSeq 6000 (PE150) platform and
   generated approximately 40 million paired-end reads (Novogene, China).
   The raw read counts were extracted and then normalized by DESeq2.
   R-pheatmap generated an expression heatmap according to cluster
   analysis of the gene FPKM value. The Gene Ontology and KEGG pathway
   enrichment analysis of differentially expressed genes was performed
   with P-clusterProfiler (V3.0.3).

Reverse transcription PCR

   Total RNAs were extracted from NK92 cells and subjected to reverse
   transcription for cDNA synthesis. The specific primer sequences used
   for GPR132, Gαs, Gαi, and β-actin are listed in table S2. The RT-PCR
   procedure included an initial denaturation step at 95°C for 5 min,
   followed by 35 cycles consisting of denaturation at 95°C for 30 s,
   annealing at 60°C for 30 s, and extension at 72°C for 30 s. After
   completing the cycles, the PCR product was loaded onto a 2% agarose gel
   for electrophoresis.

Western blot

   Following the procedures outlined in our previously published papers
   ([174]23, [175]26), we collected cells and lysed them using
   radioimmunoprecipitation assay buffer (Epizyme, China), which was
   supplemented with protease and phosphatase inhibitors. We measured the
   protein concentration with a BCA protein assay kit (Biosharp, China).
   The proteins were then separated on 10% SDS–polyacrylamide gels and
   transferred onto polyvinylidene difluoride membranes (Millipore, USA).
   After a transfer time of 2 hours, the membranes were blocked with 5%
   skim milk and incubated overnight at 4°C with primary antibodies.
   Following this, the membranes were washed three times with TBST and
   subsequently incubated for 30 min with an horseradish
   peroxidase–labeled goat anti-rabbit antibody (Epizyme, China). Last,
   the immunoreactive signals were detected using the ECL detection system
   (Epizyme, China). The primary antibodies used were rabbit anti-GPR132
   (DF4894, Affinity, China) and rabbit anti–β-tubulin (LF204, Epizyme,
   China).

Single-cell suspension acquisition of mouse tissues

   Peripheral blood leukocytes were sampled through the orbital vein,
   followed by erythrocyte lysis, and filtered with a 40-μm filter
   (Corning, USA). The mouse femur and tibia were removed muscle, rinsed
   with a 1 ml syringe of tissue, and followed red blood cell lysis to
   obtain bone marrow cells. For lung and tumor, the tissues were digested
   using a cocktail of collagenase type I (0.1 mg/ml; Sigma-Aldrich, USA)
   and collagenase type IV (1 mg/ml; Sigma-Aldrich, USA), and
   deoxyribonuclease I (3 mg/ml; Sigma-Aldrich, USA) at 37°C for 30 min;
   after, cells were filtered through a 40-μm filter to obtain a
   single-cell suspension. For other tissues, after stripping from mice
   and triturating, red blood cells were lysis, and filtration was
   performed to obtain a single-cell suspension.

Cytotoxicity assays

   Annexin V apoptosis assay for the cytotoxicity of effector cells: 3 ×
   10^4 target cells (Yac-1, K562) were labeled with carboxyfluorescein
   diacetate succinimidyl ester (CFSE) dye and seeded in a 96-well plate.
   Effector cells were incubated at various E:T ratios (mouse splenic NK
   cells: 5:1, 10:1, 20:1; human NK92 cells: 2.5:1, 5:1, 10:1). After 4
   hours of co-incubation, cells were collected and stained with the
   annexin V antibody, and the positive frequency of annexin V under the
   CFSE gate was analyzed by flow cytometry. The killing ability of
   effector cells was quantified after normalizing based on the
   self-apoptosis of target cells.

   Luciferase assay for the cytotoxicity of effector cells: 3 × 10^4
   luciferase-transduced target cells (HCT116, HCT15, and MGC803) were
   seeded in a low-attachment 96-well plate. Effector cells were incubated
   at various E:T ratios (2.5:1, 5:1, and 10:1). After 4 hours of
   co-incubation, cells were collected, and the killing ability of
   effector cells was detected using the Firefly Luciferase Reporter Gene
   Assay Kit (Beyotime Biotech, Shanghai, China). Alternatively, the
   cytotoxicity was analyzed by Living Image software (version 4.4) after
   photographing with bioluminescence imaging.

Flow cytometry analysis

   Cells were washed with PBS supplemented with 2% FBS and 2 mM EDTA
   [fluorescence-activated cell sorting (FACS) buffer]. To stain the
   surface markers, cells were incubated with FACS antibodies for 30 min.
   For intracellular staining of GzmB, IFN-γ, perforin, CSK, ZAP70, and
   NF-κB, cells were fixed and permeabilized using the BD Cytofix/Cytoperm
   Kit (BD Biosciences, USA), according to the instructions. A list of the
   antibodies used is provided in table S3. Data were acquired from either
   the BD LSRFortessa or the Canto II flow cytometer and were analyzed by
   FlowJo software (version 10.8.1).

Mouse models

   To establish a subcutaneous melanoma tumor model, 5 × 10^5 B16-F10
   cells were subcutaneously injected into the right back flank of WT or
   GPR132^−/− mice aged 6 to 8 weeks. The tumor volume was regularly
   measured using calipers, applying the formula V = 1/2 (length ×
   width^2). When the tumor size reached 1500 mm^3, the mice were
   euthanized, and tumor tissues were weighted and collected for the
   preparation of single-cell suspension.

   To evaluate the therapeutic effect of CAR-NK in vivo, 8-week-old female
   NOD/SCID/γ-chain^−/− (NSG) mice were used to establish a colorectal
   cancer xenograft model. A total of 3 × 10^6 HCT116-luciferase cells
   were subcutaneously injected into the right back flank of NSG mice.
   When tumor volume reached approximately 60 to 100 mm^3, the mice were
   randomly divided into four groups (n = 8) and accessed with an IVIS.
   Mice were administered PBS or Mock-NK92, CAR-NK92, and SH-CAR-NK92
   cells intravenously every 7 days, for a total of five treatments.
   During the treatment, 20,000 IU of hIL-2 was injected
   intraperitoneally. Tumor volume was regularly assessed with calipers
   and calculated using the formula V = 1/2(length × width^2).
   Alternatively, tumor size was quantified with total bioluminescence
   flux (photons per second) emitted from the tumors using IVIS. On day
   35, the mice were euthanized, and tumor tissues were collected for
   detection of the proportion and activity of CAR-NK.

Statistical analysis

   GraphPad Prism (version 9.0.0) was used to create the graphs and
   perform the statistical analyses. All data are expressed as the means ±
   SD. The differences between the two groups were assessed using an
   unpaired t test. One-way analysis of variance (ANOVA) was used to
   compare differences among three or more groups, while two-way ANOVA was
   applied when analyzing two or more variable factors. P value of ≤0.05
   was considered statistically significant, with the following
   significance levels: *P < 0.05, **P < 0.01, ***P < 0.001, and ****P <
   0.0001. P value labeled as “ns” indicates not significant.

Acknowledgments