Long Noncoding RNA lncNDEPD1 Regulates PD-1 Expression via miR-3619-5p in CD8⁺ T Cells

CD8⁺ T cells are an essential component of the adaptive immune response that can rapidly expand and differentiate into cytotoxic effector T cells to control and eliminate pathogens upon stimulation. Along with its activation, programmed cell death protein 1 (PD-1) is upregulated and prevents the excessive activation of T cells (1). PD-1 expression is tightly modulated and remains high during persistent stimulation, such as in anti-tumor responses (2, 3). Although PD-1 itself does not impair the effector function of CD8⁺ T cells, the engagement of programmed cell death protein ligand 1 (PD-L1) with PD-1 can lead to the dysfunction of host cells (4). Hence, T cell–based therapies, including chimeric Ag receptor (CAR)-T cell therapy, have unsatisfactory effects on the treatment of solid tumors (5, 6). To improve the anti-tumor effects of T cells, anti–PD-1 Abs have been used in therapy for patients with solid tumors. Such combinations have yielded satisfactory clinical efficacies (7–9).

Various factors, including transcription factors (10), epigenetic components (2, 11), and noncoding RNA (ncRNA) (12–14), have been reported to affect PD-1 expression. Transcription factors, including Eomes (15), NFATc1 (16), NF-κB (17), T-bet (18), and Notch1 (19, 20), have been well characterized in the regulation of PD-1 expression. Among ncRNAs, microRNAs (miRNAs) have been reported to regulate PD-1 expression through translational repression and mRNA destabilization (21, 22). Long ncRNAs (lncRNAs) also play important roles in regulating gene expression by affecting chromatin modification, transcription, and posttranscriptional processing (23–25). It has been reported that lncRNAs are widely expressed in various immune cells and can integrate distinct signal regulators to fine-tune gene expression (26, 27). However, whether lncRNAs regulate PD-1 expression remains unclear.

In this study, we profiled the lncRNAs expressed in activated CD8⁺ T cells, explored the regulatory roles of lncRNAs on PD-1 expression, and determined the underlying mechanisms of these processes. In addition, we explored the feasibility of lncRNA modification to enhance CAR-T cell function against PD-L1–expressing solid tumors.

Peripheral blood samples from 21 patients with non–small cell lung cancer (NSCLC) were collected from The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China. Buffy coats from healthy donors were obtained from the Central Blood Bank of Zhengzhou, Zhengzhou, China. The collection and usage of samples from patients and healthy donors were approved by the Ethics Committee of The First Hospital of Zhengzhou University, Zhengzhou, China, and informed consent was obtained from all patients.

PBMCs were isolated through Ficoll-Hypaque density gradient centrifugation as previously described (28). Primary CD8⁺ T cells were further purified via MACS. The T cells were then activated for 48 h with an activation/expansion kit (Miltenyi Biotec, Bergisch Gladbach, Germany) and IL-2 (100 U/ml, R&D Systems).

After 48 h of activation, PD-1^highCD69⁺CD8⁺ T cells and PD-1⁻CD69⁺CD8⁺ T cells were isolated via FACS. Total RNA was extracted using TRIzol reagent (Invitrogen) following the manufacturer’s instructions. Quality control was performed and the certified RNAs were sequenced using a HiSeq PE151 sequencing platform (Illumina).

To detect the competitive interaction between lncNDEPD1 and PDCD1 with miR-3619-5p, we performed a luciferase reporter assay. miRNAs that target both lncNDEPD1 and PDCD1 were predicted using TargetScan (http://www.targetscan.org/) and miRBase (http://www.mirbase.org). The sequences of wild-type (WT) and mutated lncNDEPD1-3′-untranslated region (UTR) and PDCD1-3′-UTR (lncNDEPD1-WT/lncNDEPD1-mutant [MUT], PDCD1-WT/PDCD1-MUT) were synthesized and inserted into the pGL3 luciferase reporter plasmid. Then, WT-3′-UTR or MUT-3′-UTR vectors were cotransfected with the predicted miR-3619-5p mimics or miR-NC into HEK293T cells using Lipofectamine 3000, as described previously (29). Firefly and Renilla luciferase activities were measured using the Dual-Glo luciferase assay system (Promega) 48 h after transfection, and the relative luciferase activity was normalized to Renilla luciferase activity according to the manufacturer’s instructions.

Total RNA was extracted with TRIzol reagent (Invitrogen), and cDNA was prepared using a PrimeScript RT reagent kit with genomic DNA (gDNA) Eraser (Takara) according to the manufacturer’s instructions. Quantitative real-time PCR (qRT-PCR) was performed using SYBR Premix Ex Taq II (Roche) on a Bio-Rad iQ5 iCycler (Bio-Rad). For the reverse transcription of miRNAs, the miScript II RT kit (Qiagen) was used. The relative expression levels of lncRNAs and mRNAs were normalized to that of GAPDH, and miRNAs were normalized to U6 small nuclear RNA. The sequences of the primers used for qRT-PCR were as follows: GAPDH, forward, 5′-ACAACTTTGGTATCGTGGAAGG-3′, reverse, 5′-GCCATCACGCCACAGTTTC-3′; lncNDEPD1, forward, 5′-CGTGATGATGGAGAAGGTCA-3′, reverse, 5′-TACCCATAACGTTGCCCACT-3′; miR-3619-5p, forward, 5′-TCAGCAGGCAGGCTGGTGCAGC-3′, reverse, 5′-CTAGCAAAATAGGCTGTCCC-3′; PDCD1, forward, 5′-TACCCATAACGTTGCCCACT-3′, reverse, 5′-TTTAGCACGAAGCTCTCCGAT-3′.

To determine the presence of surface markers, the cells were washed in FACS buffer (PBS with 2% FBS) and incubated with CD69-FITC (catalog no. 310904, clone FN50, BioLegend), CD8-allophycocyanin-Cy7 (catalog no. 344714, clone SK1, BioLegend), 7-aminoactinomycin D-PerCP (catalog no. 420404, BioLegend), PD-1-PE-Cy7 (catalog no. 329918, clone EH12.2H7, BioLegend), or PD-L1-PE (catalog no. 374512, clone MIH3, BioLegend) Abs for 30 min at 4°C in the dark. The cells were then subjected to FACS analysis. For intracellular staining, CD8⁺ T cells were stimulated for 6 h with PMA (1 mg/ml, Sigma-Aldrich) and ionomycin (50 ng/ml, Sigma-Aldrich) in the presence of brefeldin A solution (5 mg/ml, BioLegend) at 37°C in a 5% CO₂ atmosphere. The cells were fixed, permeabilized, and stained with IL-2-allophycocyanin (catalog no. 500310, clone MQ1-17H12, BioLegend), IFN-γ-allophycocyanin (catalog no. 502512, clone 4S.B3, BioLegend), granzyme B-PE (catalog no. 372208, clone QA16A02, BioLegend) for 30 min at 4°C in the dark. Isotype IgG served as the negative control for cell surface and intracellular staining. Cells were subsequently collected on a BD FACSCanto II flow cytometer (BD Biosciences) and analyzed using FlowJo v7 software (Tree Star).

To detect the expression of lncNDEPD1, fluorescence in situ hybridization (FISH) probes of lncNDEPD1 were designed and synthesized by Ribobio (Guangzhou, China). CD8⁺ T cells from healthy donors were washed with PBS and fixed in 4% formaldehyde for 10 min on slides. The fixed cells were permeabilized with PBS containing 0.5% Triton X-100 at 4°C for 5 min, washed with PBS three times, and prehybridized at 37°C for 30 min. Hybridization was carried out using anti-lncNDEPD1, anti-18S, and anti-U6 oligodeoxynucleotide probes at 37°C overnight. After incubation, the cells were rinsed with saline-sodium citrate buffer and PBS and were counterstained with DAPI. The images were captured using a fluorescence microscope (IX71, Olympus).

Chromatin immunoprecipitation (ChIP) assays were carried out using the SimpleChIP enzymatic ChIP kit (Cell Signaling Technology) following the manufacturer’s instructions. Briefly, CD8⁺ T cells were cross-linked using 1% formaldehyde. Then, the cross-linked chromatin DNA was sonicated into 150- to 900-bp fragments. Disrupted chromatin fragments were immunoprecipitated with the corresponding Abs against the protein of interest (anti-Notch1 Ab, Cell Signaling Technology). A positive control (anti-histone H3 Ab) and negative control (anti-IgG Ab) were also detected. The precipitated DNA was purified using spin columns and analyzed via qPCR.

An RNA-binding protein immunoprecipitation (RIP) assay on Ago2 was performed using a Magna RIP kit (Millipore) according to the manufacturer’s instructions. Briefly, CD8⁺ T cells were lysed using RIP lysis buffer and incubated with anti-Ago2 Abs (Proteintech). The coprecipitated RNAs were isolated and quantified to determine the expression levels of lncNDEPD1, PDCD1, and miR-3619-5p via qRT-PCR. Samples incubated with anti-IgG Ab (Cell Signaling Technology) were used as negative controls. Anti-EZH2 Ab (Millipore) was used as the positive control.

To generate entire DNA sequences of CARs targeting mesothelin (MSLN) and short hairpin RNA (shRNA) targeting lncNDEPD1, the nucleotide sequence of scFv-anti-human MSLN and specific shRNA sequences of lncNDEPD1 were synthesized and fused to the transmembrane (TM) domains of CD28 and CD3ζ. The generated sequence of MSLN-ScFv-Hinge-TM-CD28-CD3ζ, MSLN-shNC-ScFv-Hinge-TM-CD28-CD3ζ, and MSLN-shLNC-ScFv-Hinge-TM-CD28-CD3ζ were subcloned into pCDH-EF1-GFP lentiviral vectors using XhoI and EcoRI restriction sites. Then, the CAR-expressing vectors and packaging vector were transferred into the HEK293T retroviral packaging cells. Lentivirus-containing supernatants were collected after 48–72 h. T cells were transduced with lentiviral particles containing human anti-MSLN scFv (name MSLN-CAR-T) and lncNDEPD1-directed shRNA or a scrambled shRNA control, namely MSLN-CAR-T-SHNC and MSLN-CAR-T-SHLNC, respectively. The transfection efficiency and phenotypic characteristics of CAR-T cells were detected via FACS analysis of GFP expression 72 h later.

The lentiviral vector expressing MSLN-specific CAR and shRNA was constructed as previously reported (11). The shRNA sequences of lncNDEPD1 were as follows: lncNDEPD1 shRNA-1 (target sequence: 5′-GTTCTGTATGAGACCACTTGGATCCGCAAAGTGGGCAACGTTATGAGAATAACGTTGCCCACTTTGCTTTTTGAATTCTTCGATCTGCTTTTT-3′) and lncNDEPD1 shRNA-2 (target sequence: 5′-GTTCTGTATGAGACCACTTGGATCCGCCAGACTCCTGACTAGAAGAGATTCTAGTCAGGAGTCTGGCTTTTTGAATTCTTCGATCTGCTTTTT-3′). The small interfering RNA (siRNA) sequence of Notch1 was as follows: siNotch1 (5′-CAGGGAGCAUGUGUAACAUTTAUGUUACACAUGCUCCCUGTT-3′). T cells were transfected using Lipofectamine 3000 reagent according to the manufacturer’s protocol (Thermo Fisher Scientific, L300015).

CAR-T cells were cocultured with the lung cancer cell line H322 at a 1:1 ratio in a 96-well culture plate. After 24 h of coculture, the release of IFN-γ and IL-2 into the medium was determined using ELISA as described previously (30).

To generate a lung cancer xenograft mouse model, female NOD-SCID mice aged 6–8 wk were purchased from Beijing Vital River Laboratory Animal Technology. FFluc-expressing H322 cells (1 × 10⁶) in 100 μl of PBS were injected s.c. into the mice. Five days later, the tumor-bearing mice were injected once with CAR-T cells or PBS. Tumor growth was monitored using an in vivo imaging system (IVIS Lumina Series III; PerkinElmer) on days 3, 6, 9, and 14 after CAR-T cell infusion. The animal study protocol was approved by the Review Board of The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China.

Experiments were performed in triplicate. Statistical analyses were conducted using GraphPad Prism 7 and SPSS Statistics Program (GraphPad Software, La Jolla, CA). The data are presented as the mean ± SEM. Statistical significance was set at p < 0.05. The t test and ANOVA were used to compare the differences among groups. Pearson’s coefficient correlation was used to analyze the correlation between two variables. Statistical significance is indicated as follows: *p < 0.05, ** p < 0.01, ***p < 0.001.

Upon stimulation with anti-CD3/CD28 beads, T cells had upregulated PD-1 expression (Fig. 1A). However, there were fractions of activated T cells (CD69⁺) that were negative for PD-1 expression (Fig. 1B). PD-1 transcription was confirmed in T cells that were negative for PD-1 or expressed high levels of PD-1. As shown in (Fig. 1C, the mRNA levels of PD-1 were in agreement with the protein production levels, indicating that PD-1 transcription was differentially regulated in activated T cells. In addition to transcription factors, lncRNAs represent a more flexible approach in regulating gene expression. Hence, we analyzed the alterations in lncRNA levels in activated T cells. RNA sequencing data showed that several lncRNAs demonstrated differential expression between T cells expressing high or low levels of PD-1 (Fig. 1D). Among them, NONHSAT068957.2 (hereafter referred to as lncNDEPD1) was confirmed to be overexpressed in PD-1^high T cells (Fig. 1E). Further assays showed that lncNDEPD1 was upregulated in active T cells, especially in PD-1^high subsets (p < 0.05; (Fig. 1F). FISH assays demonstrated that lncNDEPD1 was mainly located in the cytoplasm of T cells (Fig. 1G). In addition, FISH assays also demonstrated that lncNDEPD1 was gradually upregulated during T cell activation (Fig. 1H). Taken together, these data indicate that lncNDEPD1 is differentially expressed in activated CD8⁺ T cells with different PD-1 expression levels.

lncRNAs can stabilize mRNAs to promote gene expression. As lncNDEPD1 and PD-1 mRNA showed concerted expression, we speculated whether lncNDEPD1 was involved in PD-1 expression. In CD8⁺ T cells from patients with NSCLC, the expression of lncNDEPD1 was positively correlated with PD-1 levels (p < 0.0001; (Fig. 2A). Then, T cells were transfected with siRNAs specific for lncNDEPD1 and activated. lncNDEPD1-specific siRNAs markedly downregulated the transcription of lncNDEPD1 and PD-1 (p < 0.01; (Fig. 2B, 2C). In addition, the surface expression of PD-1 protein was downregulated by siRNAs against lncNDEPD1 (p < 0.05; (Fig. 2D, 2E). These results suggest that lncNDEPD1 modulates the mRNA expression of PD-1.

Cytoplasmic lncRNAs have been reported to regulate mRNA levels by acting as decoys to sequester miRNAs (31). Using the TargetScan and miRBase algorithms, five miRNAs were predicted to target both lncNDEPD1 and PD-1. Of these miRNAs, miR-3619-5p was markedly downregulated in PD-1^highCD8⁺ T cell subsets (Fig. 3A). Moreover, miR-3619-5p was increased in T cells transfected with silncNDEPD1 (p < 0.01; (Fig. 3B). Thus, we focused on miR-3619-5p and explored its interaction with lncNDEPD1 and PD-1. Naive CD8⁺ T cells were transfected with synthesized miR-3619-5p (miR-3619-5p mimics) or a scrambled miRNA (miR-NC as a control). RT-PCR demonstrated that miRNA-3619-5p significantly decreased lncNDEPD1 and PD-1 expression at the mRNA level (p < 0.01; (Fig. 3C). In addition, the surface expression of PD-1 protein was downregulated by miR-3619-5p mimics (p < 0.01; (Fig. 3D, 3E). To further confirm the specific binding of miR-3619-5p to lncNDEPD1 and PDCD1, we performed a RIP assay. The results showed that lncNDEPD1, PDCD1, and miR-3619-5p were predominantly enriched in the group treated with anti-Ago2 Abs (Fig. 3F). As shown in (Fig. 3G, there were complementary sequences among miR-3619-5p, lncNDEPD1, and PDCD1. To test the direct interaction among these three molecules, pmirGLO luciferase reporter vectors expressing lncNDEPD1 with the WT (WT-lncNDEPD1-3′-UTR) or a mutated 3′-UTR (MUT-lncNDEPD1-3′-UTR), and PDCD1 with the WT (WT-PDCD1-3′-UTR) or a MUT 3′-UTR (MUT-PDCD1-3′-UTR) were constructed. These vectors were then cotransfected with miR-3619 or miR-NC into HEK293 cells. miR-3619-5p overexpression reduced the luciferase activity of pmirGLO-lncNDEPD1-WT and pmirGLO-PDCD1-WT (Fig. 3H). In contrast, miR-3619-5p had little impact on the luciferase activity of pmirGLO-lncNDEPD1-MUT and pmirGLO-PDCD1-MUT (p < 0.01; (Fig. 3I). Taken together, these data suggest that lncNDEPD1 can upregulate PD-1 expression by competing with miR-3619-5p.

To assess whether lncNDEPD1 and miR-3619-5p affected T cell function, T cells were transfected with siRNAs specific for lncNDEPD1 or miR-3619-5p-mimics when cocultured with A549 cells (non–PD-L1 overexpression). The results showed that lncNDEPD1 knockdown had little effect on T cell function (Fig. 4A). Then, we generated the PD-L1–overexpressing A549 cell lines and detected the overexpression efficiency (Supplemental Fig. 1). We found that knockdown of lncNDEPD1 rescued the inhibition of T cells function by PD-L1, when cocultured with A549 cells with PD-L1 overexpression) (p < 0.05; (Fig. 4B), indicating that lncNDEPD1 affects T cell function by regulating PD-1 expression. Similarly, miR-3619-5p overexpression had little impact on T cell activation in the absence of PD-L1 (Fig. 4C), but remarkably it enhanced the cytotoxic potential of T cells when PD-L1 was present (p < 0.05; (Fig. 4D). These results indicate that miR-3619-5p and lncNDEPD1 affect T cell activity by regulating PD-1/PD-L1 axis–mediated immunosuppression.

Notch1 has been reported to regulate PD-1 expression in mouse T cells (20). We wondered whether Notch1 contributes to the differential expression of PD-1 in human T cells. As shown in (Fig. 5A, Notch1 expression was comparable in T cells with different PD-1 expression levels. We knocked down the Notch1 expression using siRNA (Supplemental Fig. 2A). PD-1 was downregulated at both the mRNA and protein levels upon siRNA-mediated knockdown of Notch1 (Fig. 5C–E). The same results were observed using Notch1 inhibitor (Supplemental Fig. 2B). lncNDEPD1 and miR-3619-5p were also altered by Notch1 inhibitor or knockdown (Fig. 5B, 5C). As the potential binding domains of Notch1 existed in the promoters of both PDCD1 and lncNDEPD1 (Fig. 5F, 5G), it was not clear whether Notch1 directly regulated PDCD1 transcription. Therefore, we performed a ChIP assay. The results showed that the regulatory subunits of RBP-Jκ were physically associated with the lncNDEPD1 promoter (p < 0.01; (Fig. 5I) but not with the PDCD1 promoter (p < 0.01; (Fig. 5H), indicating that Notch1 directly regulates lncNDEPD1 expression, which further regulates PD-1 expression by competing for miR-3619-5p.

To confirm the regulatory effects of lncNDEPD1 on PD-1 expression and T cell function, we employed CAR-T cells for in vitro and in vivo assays. We sorted the CD8⁺ T cells (Supplemental Fig. 3A) and generated MSLN-directed CAR-T cells expressing scramble or lncNDEPD1-targeted shRNAs (Fig. 6A, 6C). MSLN is a tumor differentiation Ag that was significantly overexpressed in multiple solid tumors and has emerged as an important target of immunotherapies. T cells transduced with lentiviral particles containing human anti-MSLN scFv, named MSLN-CAR-T, were transduced with anti-MSLN scFv and lncNDEPD1-dericted shRNA or a scrambled shRNA control named MSLN-CAR-T-SHNC or MSLN-CAR-T-SHLNC, respectively. To prevent off-target effects, we first examined the MSLN level in CAR-T cells. The results showed no expression of MSLN in CAR-T cells (Supplemental Fig. 3B). Compared with scrambled shRNA-expressing CAR-T cells, shlncNDEPD1-expressing T cells showed markedly reduced PD-1 expression (p < 0.0001; (Fig. 6B). Then, CAR-T cells were coincubated with H322 cells, which expressed high levels of PD-L1 (Supplemental Fig. 4A). In the coincubation assay, shlncNDEPD1 CAR-T cells demonstrated elevated cytotoxicity in H322 cells (Fig. 6D). In the supernatants of CAR-T cells coincubated with tumor cells at a 10:1 ratio, the expression of effector cytokines was increased in the shlncNDEPD1 group (Fig. 6E). Next, we examined the effect of lncNDEPD1 on the anti-tumor effects and PD-1 expression of CAR-T cells. In comparison with the scrambled group, shlncNDEPD1-expressing CAR-T cells demonstrated enhanced tumor inhibition on day 9 (Fig. 6F–H). Additional tumor growth curves (flux values) strengthen the reproducibility of this experiment (Supplemental Fig. 4B). Moreover, shlncNDEPD1-expressing CAR-T cells maintained low levels of lncNDEPD1 in vivo (Fig. 6I). Further analysis showed that shlncNDEPD1-expressing CAR-T cells expanded more efficiently but had lower PD-1 expression (Fig. 6J–L). These data collectively indicate that lncNDEPD1 promotes PD-1 expression and immunosuppression in T cells in vivo.

It remains unclear whether lncRNAs regulate PD-1 expression in T cells. In this study, we found that the transcription of lncNDEPD1 was higher in PD-1^highCD69⁺CD8⁺ T cells than in their PD-1⁻ counterparts. Knockdown of lncNDEPD1 significantly inhibited PD-1 expression (p < 0.001). Moreover, we observed that the blockade of Notch1 signaling could inhibit lncNDEPD1 transcription, further abrogating PD-1 expression. Mechanistically, Notch1 activation directly upregulated lncNDEPD1 transcription by binding to the promoter of lncNDEPD1, which further upregulated PD-1 expression at posttranscriptional levels in vitro and in vivo.

Technological advances in RNA sequencing have shed light on the role of ncRNAs in the immune system. There is growing evidence showing that ncRNAs regulate the differentiation, activation, and proliferation of T cells through transcriptional, posttranscriptional, and epigenetic modification (32, 33). lncRNAs have distinct functions based on their unique subcellular distribution patterns. Cytoplasmic lncRNAs can regulate mRNA levels by acting as decoys to sequester miRNAs and affect posttranscriptional modulation (34). miRNAs are highly conserved small noncoding RNAs that regulate gene expression by binding to the 3′-UTR of their target mRNA (35). Both lncRNAs and miRNAs have been shown to participate in the development and differentiation of T cells (36, 37). Thus, we speculated that ncRNAs regulate PD-1 expression. As expected, we found that a novel lncRNA, lncNDEPD1, could regulate PD-1 expression. lncNDEPD1 was robustly upregulated in the PD-1^highCD8⁺ T cell subset upon activation. Moreover, lncNDEPD1 expression was significantly downregulated along with PD-1 when the Notch signaling pathway was blocked. The results of the ChIP assay showed that Notch1 directly regulated lncNDEPD1 expression instead of PD-1. Thus, we speculated that lncNDEPD1 may be an important regulator of PD-1 expression. Bioinformatics analysis demonstrated that both lncNDEPD1 and PD-1 were targets of miR-3619-5p. Concordantly, we demonstrated that the expression levels of lncNDEPD1 and PD-1 were inversely correlated with miR-3619-5p levels in activated T cells. Further studies showed that miR-3619-5p overexpression reduced the expression of both lncNDEPD1 and PD-1. RIP and luciferase reporter assays further supported that lncNDEPD1 and PD-1 mRNA were able to bind to miR-3619-5p. Taken together, our data suggest that lncNDEPD1 can upregulate PD-1 expression by competitively sponging miR-3619-5p.

The Notch pathway is an evolutionarily conserved signaling pathway that plays a critical role in the development and function of T cells. When activated, cleavage of the Notch intracellular domain (NICD) migrates into the nuclear complex with the transcription factor RBP-Jk. Then, the NICD/RBP-Jk complex upregulates expression of target genes of Notch signaling (38, 39). In human T cells, Notch1 can regulate T cell effector functions at the transcriptional level (40, 41), enhance differentiation toward short-lived effector cells, and maintain memory T cell subsets (42, 43). In mouse T cells, Notch1 upregulates PD-1 expression by directly enhancing PDCD1 transcription (20). However, we noticed that Notch1 does not directly regulate PDCD1 transcription in human T cells. Instead, Notch1 affected PD-1 expression at the posttranscriptional level by modulating lncNDEPD1 expression. As mentioned above, Notch1 is critical for the optimal activation of human T cells. Direct blockade of Notch1 not only inhibits PD-1 expression but also reduces the effector functions of T cells. Therefore, the identification of lncNDEPD1 provides an opportunity to decrease PD-1 expression but not T cell activation. Our data showed that lncNDEPD1 was involved in PD-1 mRNA stabilization but not T cell activation. Targeting lncNDEPD1 can specifically downregulate PD-1 expression without affecting T cell function.

CAR-T cell therapy has emerged as a promising strategy for treating patients with cancer. The effects of PD-1 on CAR-T cells have been investigated using pharmacological agents, Abs, and molecular methods. In our study, knockdown of lncNDEPD1 significantly decreased PD-1 expression and reversed the PD-L1–dependent suppression of T cell function. We generated CAR-T cells expressing anti-lncNDEPD1 shRNAs to knock down lncNDEPD1 expression. In the presence of PD-L1, lncNDEPD1 knockdown enhanced cytokine production and the cytotoxic function of CAR-T cells. Adoptive transfer of lncNDEPD1-knockdown CAR-T cells effectively delayed tumor growth in a mouse model.

In summary, this study demonstrated that Notch1 directly promotes lncNDEPD1 transcription and that lncNDEPD1 upregulates PD-1 expression by sponging miR-3619-5p, which destabilizes PD-1 mRNA. In addition, lncNDEPD1 knockdown can downregulate PD-1 and enhance CAR-T cell therapy efficacy against PD-L1–expressing tumor cells. Our study suggests that lncNDEPD1 may be a potential target for decreasing PD-1 expression in T cells.

We are grateful for the sample collection performed by the cancer center at The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China. Also, we thank Editage for English language editing.

The authors have no financial conflicts of interest.