Combined PD-1 and TIGIT blockade via siRNA enhances anti-leukemic T-cell function and promotes AML cell apoptosis

Abstract

This study introduces a novel siRNA-based dual checkpoint blockade approach using readily available leukoreduction filter-derived T-cells, exploring dual PD-1/TIGIT siRNA-mediated knockdown to enhance anti-leukemic activity against Acute myeloid leukemia (AML) cell lines with distinct molecular characteristics (KG-1, representing minimally differentiated AML-M0; THP-1, representing monocytic AML-M5). T-cells were isolated from leukoreduction filters (98.1% purity), activated with CD3/CD28 Dynabeads and IL-2 to upregulate PD-1 (40.5%) and TIGIT (46.4%), and transfected with siRNA (70.8% efficiency), reducing PD-1 to 11.9% and TIGIT to 19.4%. Transfected T-cells were co-cultured with AML cells at E: T ratios of 1:1, 2:1, and 5:1. Viability, cytotoxicity, IFN-γ and caspase-3 levels, and Bax/Bcl-2 mRNA expression were assessed. Dual knockdown reduced THP-1 viability to 30% and KG-1 to 40% at 5:1, with cytotoxicity increasing to 44.3% (THP-1) and 48.8% (KG-1) (P < 0.01), outperforming single knockdowns (P < 0.01). IFN-γ peaked at 80 pg/mL (THP-1, 5:1, P < 0.0001), caspase-3 rose by 70% (P < 0.0001), Bax increased 3.0-fold, and Bcl-2 decreased 70% (P < 0.001). CD8⁺ T-cells dominated (~ 60% post-transfection). Statistical significance was confirmed for differences between KG-1 and THP-1 responses (P < 0.05), with THP-1 demonstrating greater susceptibility to dual checkpoint blockade. Dual PD-1/TIGIT knockdown additively enhances cytotoxicity, apoptosis, and T-cell activation, with stable CD4⁺/CD8⁺ ratios ensuring effector function. While this approach provides a cost-effective foundation for AML immunotherapy research, clinical translation requires validation in primary AML samples and in vivo models to address the complex bone marrow microenvironment and optimize delivery systems.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-53340-7.

Introduction

Acute myeloid leukemia (AML) is an aggressive hematologic malignancy characterized by clonal expansion of immature myeloid progenitors leading to bone marrow failure^1,2. Despite therapeutic advances, outcomes remain poor, and long-term survival in adults is still unsatisfactory^3–7. A recent national registry study in England reported a median survival of 0.6 years and 5-year overall survival of 18.8% in adults with AML, underscoring the persistent need for more effective therapies⁸.

Immune evasion is a central mechanism underlying AML progression and resistance to therapy. Leukemic blasts impair anti-tumor immunity through defective antigen presentation, immunosuppressive cytokine networks, and induction of inhibitory checkpoint pathways on effector lymphocytes^9–11. T-cell exhaustion, defined by reduced proliferative capacity, diminished cytokine production, and sustained inhibitory receptor expression, is a prominent feature of AML and contributes to ineffective immune surveillance^12,13. Among these inhibitory pathways, programmed cell death protein-1 (PD-1) and T-cell immunoreceptor with Ig and ITIM domains (TIGIT) have emerged as particularly relevant in AML¹⁴. PD-1 signaling suppresses T-cell receptor activation after engagement with PD-L1/PD-L2¹⁵, whereas TIGIT competes with CD226 for binding to poliovirus receptor (PVR), thereby attenuating both T-cell and NK-cell activation^16,17. Recent AML studies have shown that PD-1 and TIGIT are highly co-expressed on bone marrow and peripheral blood CD8 + T cells, PD-1 is broadly elevated in AML T cells, and high TIGIT expression is associated with primary refractory disease and post-transplant relapse, supporting the biological relevance of dual checkpoint targeting¹⁸.

The clinical experience with checkpoint-based therapy in AML has been mixed. In a phase II study of relapsed/refractory AML, azacitidine plus nivolumab produced an overall response rate of 33%, with higher responses in hypomethylating-agent-naïve and salvage 1 patients, but grade 3 to 4 immune-related adverse events still occurred in 11% of patients, showing both activity and clinical constraints of this strategy¹⁹. Beyond T cells, TIGIT also appears to shape the AML microenvironment more broadly, because TIGIT blockade has been shown to repolarize AML-associated TIGIT + M2 leukemia-associated macrophages toward a more inflammatory phenotype²⁰. Together, these findings suggest that single-pathway checkpoint inhibition is unlikely to fully reverse AML-associated immune suppression and that combined targeting of complementary inhibitory axes may be more effective.

RNA interference (RNAi) provides a targeted post-transcriptional approach for silencing immune checkpoint genes at the mRNA level and may therefore offer a cell-intrinsic alternative to systemic antibody blockade^21,22. In parallel, leukoreduction filters represent a practical and underused source of viable immune cells. Prior studies have shown that whole-blood leukoreduction filters can yield large numbers of leukocytes suitable for phenotypic and functional investigations, and that leukocytes isolated from leukoreduction system chambers and filters retain measurable biological activity²³. This makes leukoreduction filters an attractive starting material for ex vivo immune-cell engineering in a translational setting.

In this study, we investigate a novel immunotherapeutic strategy combining T cells isolated from leukoreduction filters with dual siRNA-mediated knockdown of PD-1 and TIGIT. We hypothesize that intracellular silencing of these inhibitory pathways will enhance T-cell cytotoxicity against AML cell lines with distinct molecular characteristics, including KG-1 and THP-1. Activated T cells were transfected with PD-1 and TIGIT siRNAs and co-cultured with leukemic cells at varying effector-to-target ratios. Anti-leukemic activity was evaluated by assessing cell viability, cytotoxicity, IFN-γ production, active caspase-3 levels, Bax/Bcl-2 mRNA expression, and CD4⁺/CD8⁺ T-cell subset distribution.

By integrating checkpoint biology, RNAi technology, and an accessible cellular source, this work aims to establish a proof-of-concept framework for a targeted and potentially safer immunotherapeutic approach in AML.

Materials and methods

Sample collection

Leukoreduction filters were obtained from 10 eligible blood donors (aged 25–50 years) with negative screening results at the Iranian Blood Transfusion Organization, Tehran Province. The study was approved by the Research Ethics Committee of Shariati Hospital, Tehran, Iran (approval number IR.TUMS.HORCSCT.REC.1402.028), and all participants provided written informed consent. All procedures were conducted in accordance with relevant guidelines and regulations. Cell lines (KG-1, passages 15–20; THP-1, passages 18–25) were sourced from the Pasteur Institute’s Cell Bank (Tehran, Iran), a reputable institution under the Ministry of Health, with authenticated cell line catalogs available online. Cell lines were authenticated by short tandem repeat (STR) profiling at the Pasteur Institute and tested negative for mycoplasma contamination.

PBMCs isolation

Peripheral blood mononuclear cells (PBMCs) were isolated from leukoreduction filters using Ficoll-Paque PLUS (GE Healthcare, cat. no. 17-1440-02) density gradient centrifugation. Filters were flushed with 50 mL PBS containing 2 mM EDTA, and the cell suspension was diluted 1:1 with PBS-EDTA, layered over Ficoll-Paque PLUS, and centrifuged (400×g, 30 min, room temperature, no brake). The PBMC layer was collected and washed twice with PBS²⁴.

T-cell isolation

CD3⁺ T-cells were isolated from PBMCs using the MACS CD3 MicroBeads Human Positive Selection Kit (Miltenyi Biotec, cat. no. 130-050-101). PBMCs were incubated with CD3 MicroBeads (20 μL per 10⁷ cells) for 15 min at 4 °C, washed, and passed through an LS column in a MACS Separator. Magnetically labeled CD3⁺ T-cells were eluted, washed with MACS Buffer, assessed for viability via trypan blue exclusion, and plated at 2 × 10⁶ cells/well in six-well dishes.

Flow cytometry analysis

T-cell purity and checkpoint expression were evaluated using flow cytometry with antibodies against CD3 (clone UCHT1), PD-1 (clone EH12.1), TIGIT (clone 741182), CD4-FITC, and CD8-PE (all from BD Biosciences). Cells were stained for 30 min at 4 °C in the dark, then washed and analyzed on a BD FACSLyric Flow Cytometer. Data were analyzed using FlowJo (version 10.7.1, BD Biosciences). Gating involved singlets (FSC-A vs. FSC-H), live cells (DAPI-negative), and CD3⁺ populations, with PD-1, TIGIT, and CD4/CD8 expression assessed within CD3⁺ subsets.

Induction of PD-1 and TIGIT on T-cells

To assess the knockdown efficiency of PD-1 and TIGIT on isolated T-cells, we induced the expression of PD-1 and TIGIT via Dynabeads Human T-Activator CD3/CD28. Briefly, these cells were cultured in RPMI 1640 medium (Gibco, cat. no. 11875093) supplemented with 10% heat-inactivated fetal bovine serum (FBS, Gibco, cat. no. 10270106), 2 mM L-glutamine (Gibco, cat. no. 25030081), 100 U/mL penicillin, and 100 μg/mL streptomycin (Gibco, cat. no. 15140122). Cells were activated using Dynabeads Human T-Activator CD3/CD28 (Thermo Fisher Scientific, cat. no. 11131D) at a 1:1 bead-to-cell ratio and 100 U/mL recombinant human IL-2 (PeproTech, cat. no. 200-02) at 37 °C in a humidified atmosphere with 5% CO₂. After activation, T-cells were analyzed again by flow cytometry for PD-1 and TIGIT expression as described above. T cell exhaustion and the upregulation of inhibitory receptors, such as PD-1 and TIGIT, is a gradual process that can occur over days to weeks. Significant increases in PD-1 and TIGIT expression can be observed within 3 to 7 days of continuous stimulation^25,26.

siRNA design and transfection

siRNAs targeting PD-1 (PDCD1, NM_005018) and TIGIT (TIGIT, NM_173799) were designed using BLOCK-iT RNAi Designer (Thermo Fisher Scientific) and siDESIGN Center (Dharmacon) to minimize off-target effects and avoid secondary structures in the target site while ensuring asymmetry and energy valley within the siRNA sequences. To reduce potential off-target effects, siRNAs were designed with low seed complement frequencies (SCFs) and avoided complementarity to 3′ untranslated regions of genes. Four groups of siRNA transfections were performed: (1) PD-1 siRNA, (2) TIGIT siRNA, (3) PD-1 + TIGIT siRNA, and (4) Scramble siRNA (control). siRNAs were obtained from Metabion (sequences available upon request for reproducibility).

All experiments included two control groups: (1) untreated T-cells (no activation/siRNA), and (2) scramble siRNA-transfected activated T-cells. Activated T-cells were transfected with siRNAs (50 nM final concentration) using Lipofectamine RNAiMAX Transfection Reagent (Invitrogen, cat. no. 13778150) according to the manufacturer’s protocol. Briefly, for each well of a 24-well plate, 1 × 10⁵ activated T-cells were resuspended in 500 μL of Opti-MEM medium (Gibco, cat. no. 31985070). siRNA-lipid complexes were prepared and added to the cells, resulting in a final siRNA concentration of 50 nM. Cells were incubated for 48 h at 37 °C, 5% CO₂. Transfection efficiency was assessed using a delivery check (FAM-labeled control siRNA, Thermo Fisher Scientific, cat. no. AM4620) and analyzed by flow cytometry, utilizing a 488 nm laser for FAM detection.

Confirmation of PD-1 and TIGIT inhibition via flow cytometry and qRT-PCR

Expression of PD-1 and TIGIT on T-cells was assessed 24 h post-transfection. Cells (5 × 10⁵) were stained with anti-PD-1-PE-Cy7 (clone EH12.1) and anti-TIGIT-BV421 (clone 741182) antibodies (BD Biosciences) for 30 min at 4 °C in the dark. Samples were analyzed on a BD FACSCanto II flow cytometer, acquiring at least 10,000 events per sample. Data were analyzed using FlowJo software version 10.7.1, calculating the mean fluorescence intensity (MFI) and percentage of positive cells for each marker. To confirm siRNA knockdown efficiency at the mRNA level, quantitative reverse transcription PCR (qRT-PCR) was performed 48 h post-transfection. Total RNA was extracted using TRIzol reagent (Invitrogen), and cDNA was synthesized using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). qRT-PCR was performed using SYBR Green PCR Master Mix (Applied Biosystems) with primers designed for the 5’ regions of PD-1 and TIGIT mRNAs to ensure accurate detection of knockdown. Relative expression was normalized to β-actin and calculated using the 2^-ΔΔCt method.

Co-culture experiments

KG-1 and THP-1 cells were labeled with CFSE (Invitrogen, cat. no. C34554) and cultured in RPMI 1640 with 10% FBS and antibiotics. For all E: T ratio experiments, T-cells from the same donor were used to minimize inter-donor variability. Transfected T-cells were co-cultured with CFSE-labeled leukemic cells at effector-to-target (E: T) ratios of 1:1, 2:1, and 5:1 in 96-well plates for 6 h (for cytotoxicity) or 24 h (for viability, IFN-γ, caspase-3, and qRT-PCR assays) at 37 °C, 5% CO₂. Five groups were tested: (1) untreated control T-cells, (2) PD-1 siRNA T-cells (P group), (3) TIGIT siRNA T-cells (T group), (4) PD-1 + TIGIT siRNA T-cells (P + T group), and (5) scramble siRNA T-cells (control). Experiments were performed in triplicate.

Metabolic activity and cytotoxicity assays

MTT assay

Metabolic activity was assessed using the MTT Cell Proliferation Assay Kit (Abcam, cat. no. ab211091). After 24 h of co-culture at E: T ratios of 1:1, 2:1, and 5:1, 10 μL of MTT reagent was added to 100 μL of the co-culture, and the mixture was incubated for 4 h at 37 °C. This was followed by the addition of 100 μL of MTT solvent and 15 min of shaking. Absorbance was measured at 570 nm, and viability was calculated relative to controls. Each sample was assayed in triplicate, and the experiment was repeated 3 times.

Cytotoxicity assays

Co-cultured cells (E: T ratios of 1:1, 2:1, and 5:1) were stained with propidium iodide (PI, 50 μg/mL) after 6 h and analyzed by flow cytometry. CFSE⁺ leukemic cells were gated, and PI positivity was quantified to assess cell death. Each sample was assayed in triplicate, and the experiment was repeated 3 times.

IFN-γ and active caspase-3 assays

Supernatants from 24-h co-cultures (E: T ratios 2:1 and 5:1) were analyzed for IFN-γ and active caspase-3 using the Human IFN-γ ELISA Kit (R&D Systems, cat. no. DY285B) and a Human Active Caspase-3 ELISA Kit, respectively. Assays were performed according to the manufacturer’s instructions, with optical density measured at 450 nm and background subtraction at 570 nm applied to ensure accurate experimental results. Concentrations were determined from standard curves. Each sample was assayed in triplicate, and the experiment was repeated 3 times.

qRT-PCR for bax and Bcl-2 mRNA expression

Total RNA was extracted from co-cultured cells (E: T ratios 2:1 and 5:1) after 24 h using TRIzol (Invitrogen). cDNA was synthesized using a reverse transcription kit (Thermo Fisher Scientific), and qRT-PCR was performed with primers for Bax and Bcl-2 using SYBR Green (Applied Biosystems). Primer sequences were designed for 5′ regions of target mRNAs: Bax forward: 5′-GCAGATCATGAAGACAGGGGC-3′, reverse: 5′-CCGATGCGCTTGAGACACTC-3′; Bcl-2 forward: 5′-GCGCGTGTAAATTGCCGAGA-3′, reverse: 5′-GCAGGAATCCCAACCGGAGA-3′. Relative expression was normalized to β-actin and calculated using the 2^-ΔΔCt method^27,28.

Statistical analysis

Experiments were conducted in triplicate using T-cells from the same donor for each set of E: T ratio experiments to minimize inter-donor variability. Data are expressed as mean ± SEM. Statistical analyses were performed using GraphPad Prism (version 9.0.0). Normality was tested with the Shapiro–Wilk test. Normally distributed data were analyzed using one-way ANOVA (Tukey’s post-hoc test) or two-way ANOVA (Sidak’s test); non-normal data were analyzed using the Kruskal–Wallis test with Dunn’s post-hoc test. Statistical significance between KG-1 and THP-1 responses was assessed using unpaired t-tests. P < 0.05 was considered significant (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

Results

T-cell isolation, activation, and transfection efficiency

T-cells were isolated from peripheral blood mononuclear cells (PBMCs) obtained from leukoreduction filters using the MACS CD3 MicroBeads Human Positive Selection Kit, achieving a high purity of 98.1% as confirmed by flow cytometry with CD3-FITC staining (Fig. 1b). The untreated control exhibited minimal background fluorescence at 0.041% (Fig. 1a). Flow cytometry gating involved selecting singlets (based on FSC-A vs. FSC-H), live cells (DAPI-negative), and CD3⁺ populations to ensure the accurate identification of T-cells.

Fig. 1.

Flow cytometric analysis of isolated T cells showing CD3 expression and PD-1 expression on T cells under different conditions. (a) Unstained control showing background fluorescence (0.041%). (b) CD3-FITC staining demonstrated 98.1% purity of isolated T cells. SSC-A (Side Scatter-Area) vs. FITC-A (Fluorescein isothiocyanate-Area) plots are shown. (c) Unstained control for baseline PD-1 expression (0.14%). (d) Baseline PD-1 expression (15.3%) on untreated T cells. (e) Unstained control for activated T cells (0.012%). (f) Increased PD-1 expression (40.5%) following T cell activation with CD3/CD28 Dynabeads and IL-2. (g) Unstained control for siRNA-treated T cells (0.072%). (h) Reduced PD-1 expression (11.9%) after siRNA-mediated knockdown. SSC-A (Side Scatter-Area) vs. FITC-A/PE-A plots are shown, with percentages indicating positive populations. Antibodies used for staining: CD3-FITC (clone UCHT1) and PD-1-PE-Cy7 (clone EH12.1, BD Biosciences). Each sample was assayed in triplicate, and the experiment was repeated 3 times.

Baseline checkpoint expression in untreated T-cells was assessed to establish reference values. Baseline PD-1 expression on untreated T-cells was 15.3% (mean fluorescence intensity, MFI: 234 ± 18), with an unstained control showing 0.14% background (MFI: 12 ± 3; Fig. 1c, d). Similarly, baseline TIGIT expression was 19.3% (MFI: 287 ± 22), with an unstained control at 0.027% (MFI: 8 ± 2; Fig. 2a, b).

Fig. 2.

Flow cytometric analysis of TIGIT expression on T cells under different conditions. (a) Unstained control for baseline TIGIT expression (0.027%). (b) Baseline TIGIT expression (19.3%) on untreated T cells. (c) Unstained control for activated T cells (0.012%). (d) Increased TIGIT expression (46.4%) following T cell activation with CD3/CD28 Dynabeads and IL-2. (e) Unstained control for siRNA-treated T cells. (f) Reduced TIGIT expression (19.4%) after siRNA-mediated knockdown. SSC-A vs. APC-A plots are shown, with percentages indicating positive populations. Antibodies used for staining: TIGIT-APC (clone 741182, BD Biosciences). Each sample was assayed in triplicate, and the experiment was repeated 3 times.

To induce an exhausted phenotype representative of AML microenvironment conditions, T-cells were activated using Dynabeads Human T-Activator CD3/CD28 (at a 1:1 bead-to-cell ratio) and 100 U/mL recombinant human IL-2 for 3 days. Post-activation, PD-1 expression increased to 40.5% (MFI: 589 ± 34, P < 0.001 vs. baseline), with an unstained control at 0.012% (MFI: 5 ± 1; Fig. 1e, f), and TIGIT expression rose to 46.4% (MFI: 672 ± 41, P < 0.001 vs. baseline), with an unstained control at 0.012% (MFI: 4 ± 1; Fig. 2c, d).

Following 48 h of siRNA transfection (50 nM) using Lipofectamine RNAiMAX, both flow cytometry and qRT-PCR confirmed effective knockdown. Flow cytometry analysis showed PD-1 expression decreased to 11.9% (MFI: 178 ± 15, P < 0.001 vs. activated), with an unstained control at 0.072% (MFI: 9 ± 2; Fig. 1g, h), and TIGIT expression was reduced to 19.4% (MFI: 263 ± 19, P < 0.001 vs. activated), with an unstained control at 0% (MFI: 0 ± 0; Fig. 2e, f). qRT-PCR validation confirmed mRNA-level knockdown with PD-1 mRNA reduced by 78% ± 6% and TIGIT mRNA reduced by 72% ± 8% compared to scramble controls (P < 0.001 for both).

Transfection efficiency was assessed using FAM-labeled control siRNA, showing 70.8% FAM-positive T-cells (MFI: 845 ± 53) compared to 0.015% in untransfected controls (MFI: 10 ± 3; Fig. 3a, b). T-cell viability post-transfection was assessed by trypan blue exclusion, confirming > 90% viability across all groups (untreated: 95.2 ± 1.8%, activated: 93.5 ± 2.3%, scramble: 92.3 ± 2.1%, PD-1: 91.8 ± 2.0%, TIGIT: 93.1 ± 1.9%, PD-1 + TIGIT: 90.5 ± 2.2%).

Fig. 3.

Flow cytometric analysis of FAM-labeled siRNA transfection efficiency and CD8/CD4 ratio pre- and post-transfection. (a) Un-transfected control cells showed background fluorescence (0.015%), while (b) T-cells showed 70.8% transfection efficiency. Histograms display cell count versus FAM fluorescence intensity, with percentages indicating FAM-positive populations. (c) Unstained, and (d) CD4-FITC/CD8-PE pre-transfection (CD4⁺ 64.5%, CD8⁺ 46.7%). (e) CD8/CD4 ratio pre- and post-transfection, showing stable proportions (CD8⁺  ~ 60% post-transfection). Each sample was assayed in triplicate, and the experiment was repeated 3 times.

Flow cytometry analysis of CD4⁺/CD8⁺ ratios across different treatment conditions revealed stable proportions of T-cell subsets. Pre-transfection, untreated T-cells showed CD4⁺ T-cells at 64.5% (MFI: 512 ± 29) and CD8⁺ T-cells at 46.7% (MFI: 398 ± 25; Fig. 3d). Post-transfection, the ratio shifted slightly, with CD8⁺ T-cells dominating at 60.2 ± 3.5% (MFI: 467 ± 31) and CD4⁺ T-cells at 39.8 ± 3.2% (MFI: 423 ± 27; Fig. 3e). This stability indicates that siRNA treatment did not disrupt the balance of T-cell subsets, thereby preserving the effector potential of CD8⁺ cytotoxic T-cells, which is critical for anti-leukemic activity.

Impact on leukemic cell viability across effector-to-target ratios

The MTT assay was used to evaluate the metabolic activity of KG-1 and THP-1 AML cell lines after 24 h of co-culture with transfected T-cells at effector-to-target (E: T) ratios of 1:1, 2:1, and 5:1. Five groups were tested: untreated control, scramble siRNA (control), PD-1 siRNA, TIGIT siRNA, and dual PD-1 + TIGIT siRNA.

At a 1:1 E: T ratio, dual PD-1 + TIGIT knockdown significantly reduced THP-1 viability to 50.2 ± 4.5% (P < 0.01) compared to scramble (92.1 ± 3.8%) and untreated controls (100 ± 2.9%), while KG-1 viability decreased to 60.8 ± 5.1% (P < 0.01) from 94.3 ± 4.2% (scramble) and 100 ± 3.1% (untreated) (Fig. 4a, b). Single knockdowns showed intermediate effects: PD-1 siRNA reduced THP-1 viability to 70.5 ± 5.3% (P < 0.05) and KG-1 to 78.9 ± 4.8% (P < 0.05), while TIGIT siRNA reduced THP-1 to 68.3 ± 4.9% (P < 0.05) and KG-1 to 76.4 ± 5.0% (P < 0.05) compared to scramble.

Fig. 4.

Cell viability analysis following PD-1 and TIGIT knockdown in leukemia cell lines. (a) THP-1 and (b) KG-1 cells at the ratio of 1:1, (c) THP-1 and (d) KG-1 cells at the ratio of 2:1, (e) THP-1 and (f) KG-1 cells at the ratio of 5:1, were treated with scramble control siRNA (Scramble), PD-1 siRNA (PD-1), TIGIT siRNA (TIGIT), or combination of both siRNAs (PD-1 + TIGIT) compared to untreated control after 24 h of co-culture. Cell viability was assessed by MTT assay. Groups are color-coded as follows: scramble, PD-1 siRNA, TIGIT siRNA, and PD-1 + TIGIT siRNA. Data are presented as mean percentage viability ± SEM of three independent tests. Statistical significance was determined using p-values (P < 0.05). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Each sample was assayed in triplicate, and the experiment was repeated 3 times.

At a 2:1 E: T ratio, dual knockdown further decreased THP-1 viability to 40.1 ± 3.7% (P < 0.001) and KG-1 to 50.6 ± 4.3% (P < 0.01) compared to scramble (88.7 ± 3.5% for THP-1, 91.2 ± 4.0% for KG-1) (Fig. 4c, d). PD-1 siRNA alone reduced THP-1 viability to 58.4 ± 4.1% (P < 0.01) and KG-1 to 65.7 ± 4.6% (P < 0.01), while TIGIT siRNA reduced THP-1 to 60.2 ± 4.4% (P < 0.01) and KG-1 to 68.1 ± 4.9% (P < 0.01).

At the 5:1 E: T ratio, the effect was most pronounced, with dual knockdown reducing THP-1 viability to 30.4 ± 3.2% (P < 0.0001) and KG-1 to 40.9 ± 3.8% (P < 0.0001) compared to scramble (85.3 ± 3.1% for THP-1, 89.5 ± 3.6% for KG-1) (Fig. 4e, f). Single knockdowns at this ratio resulted in THP-1 viability of 45.7 ± 3.9% (PD-1, P < 0.001) and 48.2 ± 4.0% (TIGIT, P < 0.001) and KG-1 viability of 55.3 ± 4.2% (PD-1, P < 0.001) and 57.8 ± 4.4% (TIGIT, P < 0.001). Statistical analysis confirmed significant differences between THP-1 and KG-1 responses across all E: T ratios (P < 0.05), with THP-1 consistently showing greater susceptibility to dual checkpoint blockade. The dose-dependent reduction in viability highlights the enhanced anti-leukemic activity of dual checkpoint inhibition, with dual knockdowns consistently outperforming single knockdowns.

Enhanced T-cell-mediated cytotoxicity against leukemic cells

T-cell-mediated cytotoxicity was assessed by flow cytometry using propidium iodide (PI, 50 μg/mL) staining after 6 h of co-culture at E: T ratios of 1:1, 2:1, and 5:1. CFSE-labeled leukemic cells (KG-1 and THP-1) were gated to distinguish them from T-cells, with PI positivity indicating cell death. The results, summarized in Fig. 5, demonstrate a dose-dependent increase in cytotoxicity across all E: T ratios. At a 1:1 E: T ratio, dual PD-1 + TIGIT siRNA significantly increased cytotoxicity against THP-1 (14.0%, P < 0.01) and KG-1 (15.7%, P < 0.01) compared to scramble controls (9.5% and 8.7%, respectively). Single knockdowns showed modest but significant improvements over scramble at this ratio. At a 2:1 E: T ratio, dual knockdown further enhanced cytotoxicity, reaching 28.2% for THP-1 (P < 0.001) and 38.8% for KG-1 (P < 0.001), outperforming both PD-1 siRNA (18.2% for THP-1; 16.2% for KG-1) and TIGIT siRNA (14.0% for THP-1; 12.5% for KG-1). The most pronounced effects were observed at a 5:1 E: T ratio, where dual knockdown peaked at 44.3% for THP-1 (P < 0.0001) and 48.8% for KG-1 (P < 0.0001) compared to scramble controls (5.7% and 5.1%). While single PD-1 and TIGIT knockdowns also significantly increased cytotoxicity at the 5:1 ratio, the dual knockdown consistently exhibited the most pronounced effect, highlighting the additive and complementary impact of targeting both checkpoints simultaneously.

Fig. 5.

Enhanced T-cell-mediated cytotoxicity against AML cell lines at different effector-to-target (E:T) ratios. T cells were transfected with Scramble siRNA (control), PD-1 siRNA, TIGIT siRNA, or dual PD-1 + TIGIT siRNA (Double), and co-cultured with CFSE-labeled AML cells for 6 h. Cytotoxicity was assessed by propidium iodide (PI) staining of CFSE⁺ leukemic cells using flow cytometry. (a–c) Cytotoxicity against THP-1 cells at E:T ratios of 1:1, 2:1, and 5:1, respectively. (d–f) Cytotoxicity against KG-1 cells at E:T ratios of 1:1, 2:1, and 5:1, respectively. Data are presented as mean ± SEM from triplicate experiments. Significant differences compared to the Scramble control are indicated (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

Active caspase-3 levels and IFN-γ production in co-cultures

To evaluate the mechanisms underlying the enhanced cytotoxicity, we measured active caspase-3 levels and IFN-γ production in co-culture supernatants after 24 hours at E: T ratios of 2:1 and 5:1 using ELISA (Figure 6). At a 2:1 E: T ratio, active caspase-3 levels in THP-1 co-cultures with dual siRNA-treated T-cells increased to 62.3 ± 5.1% (P < 0.0001) compared to scramble (40.1 ± 3.8%), PD-1 siRNA (48.7 ± 4.2%), and TIGIT siRNA (46.4 ± 4.0%) (Figure 6a). A similar pattern was observed for KG-1 cells, where caspase-3 levels rose to 58.9 ± 4.8% with dual knockdown versus 39.2 ± 3.6% with scramble (Figure 6b). At a 5:1 E: T ratio, THP-1 caspase-3 levels increased to 68.4 ± 5.3% with dual knockdown (P < 0.0001) compared to scramble (38.5 ± 3.5%) (Figure 6c), and KG-1 levels reached 64.7 ± 5.0% (P < 0.0001) (Figure 6d).

Fig. 6.

Active caspase-3 levels and IFN-γ production following PD-1 and TIGIT knockdown in leukemia cell lines. (a) THP-1 and (b) KG-1 cells at the ratio of 2:1, (c) THP-1 and (d) KG-1 cells at the ratio of 5:1, were treated with scramble control siRNA (Scramble), PD-1 siRNA (PD-1), TIGIT siRNA (TIGIT), or a combination of both siRNAs (PD1 + TIGIT). Active caspase-3 levels in culture supernatants were measured by ELISA after 24 h of co-culture compared to untreated controls. (e) THP-1 and (f) KG-1 cells at the ratio of 2:1, (g) THP-1 and (h) KG-1 cells at the ratio of 5:1, were treated with scramble control siRNA (Scramble), PD-1 siRNA (PD1), TIGIT siRNA (TIGIT), or a combination of both siRNAs (PD1 + TIGIT). IFN-γ levels in culture supernatants were measured by ELISA after 24 h of co-culture compared to untreated controls. Statistical significance was determined using p-values (P < 0.05). Each data point is expressed as the mean ± SEM of three independent tests. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Each sample was assayed in triplicate, and the experiment was repeated 3 times.

IFN-γ production mirrored the cytotoxicity trends. At a 2:1 E: T ratio, THP-1 co-cultures with dual knockdown T-cells produced 60.5 ± 4.7 pg/mL IFN-γ (P < 0.0001), significantly exceeding scramble (20.3 ± 2.1 pg/mL), PD-1 siRNA (35.8 ± 3.4 pg/mL), and TIGIT siRNA (32.4 ± 3.2 pg/mL) (Figure 6e). For KG-1 co-cultures at the same ratio, IFN-γ reached 50.2 ± 4.3 pg/mL with dual knockdown (P < 0.001) (Figure 6f). At the 5:1 ratio, dual knockdown induced peak IFN-γ production of 80.1 ± 5.5 pg/mL in THP-1 co-cultures (P < 0.0001) (Figure 6g) and 70.4 ± 5.2 pg/mL in KG-1 co-cultures (P < 0.0001) (Figure 6h). THP-1 cells consistently exhibited higher IFN-γ levels and caspase-3 activity than KG-1 cells (P < 0.05), reflecting their differential immune responsiveness. A strong positive correlation was confirmed between IFN-γ production and cytotoxicity in both THP-1 (r = 0.91, P < 0.001) and KG-1 (r = 0.87, P < 0.001) co-cultures at a 5:1 E: T ratio (Supplementary Figure S1).

Bax and Bcl-2 mRNA expression in leukemic cells

To further investigate the apoptotic mechanisms, qRT-PCR was performed to measure Bax (pro-apoptotic) and Bcl-2 (anti-apoptotic) mRNA expression in co-cultured AML cells after 24 h at E: T ratios of 2:1 and 5:1. Relative expression was normalized to β-actin using the 2^-ΔΔCt method.

At a 2:1 E: T ratio, Bax expression in THP-1 cells increased 2.5-fold with dual PD-1 + TIGIT knockdown (P < 0.01) compared to scramble, while PD-1 siRNA and TIGIT siRNA alone resulted in 1.8-fold (P < 0.05) and 1.6-fold (P < 0.05) increases, respectively (Fig. 7a). For KG-1 cells, Bax expression rose 2.0-fold with dual knockdown (P < 0.01), compared to 1.5-fold (PD-1, P < 0.05) and 1.4-fold (TIGIT, P < 0.05) with single knockdowns (Fig. 7c).

Fig. 7.

Bax and Bcl-2 mRNA expression levels following PD-1 and TIGIT knockdown in leukemia cell lines. qRT-PCR was performed to examine the BAX mRNA expression levels in leukemia cell lines after siRNA transfection for 24 h. (a) THP-1 and (b) KG-1 cells at the ratio of 2:1, (c) THP-1 and (d) KG-1 cells at the ratio of 5:1, were treated with scramble control siRNA (Scramble), PD-1 siRNA (PD-1), TIGIT siRNA (TIGIT), or a combination of both siRNAs (PD1 + TIGIT). qRT-PCR was performed to examine the Bcl-2 mRNA expression levels in leukemia cell lines after siRNA transfection for 24 h. (e) THP-1 and (f) KG-1 cells at the ratio of 2:1, (g) THP-1 and (h) KG-1 cells at the ratio of 5:1, were treated with scramble control siRNA (Scramble), PD-1 siRNA (PD-1), TIGIT siRNA (TIGIT), or a combination of both siRNAs (PD1 + TIGIT). Statistical significance was determined using p-values (P < 0.05). Each data point is expressed as mean ± SEM of three independent tests. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Each sample was assayed in triplicate, and the experiment was repeated 3 times.

At a 5:1 E: T ratio, Bax expression in THP-1 cells increased 3.0-fold with dual knockdown (P < 0.001), compared to 2.1-fold (PD-1, P < 0.01) and 1.9-fold (TIGIT, P < 0.01) with single knockdowns (Fig. 7b). For KG-1 cells, Bax expression rose 2.5-fold with dual knockdown (P < 0.001), compared to 1.8-fold (PD-1, P < 0.01) and 1.7-fold (TIGIT, P < 0.01) (Fig. 7d).

Conversely, Bcl-2 expression decreased significantly with dual knockdown. At a 2:1 E: T ratio, Bcl-2 expression in THP-1 cells was reduced by 60% (P < 0.01) compared to scramble, while PD-1 siRNA and TIGIT siRNA alone resulted in 40% (P < 0.05) and 35% (P < 0.05) reductions, respectively (Fig. 7e). For KG-1 cells, Bcl-2 expression decreased by 50% with dual knockdown (P < 0.01), compared to 30% (PD-1, P < 0.05) and 28% (TIGIT, P < 0.05) with single knockdowns (Fig. 7g).

At a 5:1 E: T ratio, Bcl-2 expression in THP-1 cells decreased by 70% with dual knockdown (P < 0.001), compared to 50% (PD-1, P < 0.01) and 45% (TIGIT, P < 0.01) with single knockdowns (Fig. 7f). For KG-1 cells, Bcl-2 expression was reduced by 60% with dual knockdown (P < 0.001), compared to 40% (PD-1, P < 0.01) and 38% (TIGIT, P < 0.01) (Fig. 7h). The Bax/Bcl-2 ratio, a key indicator of apoptotic propensity, increased dramatically with dual knockdown (e.g., 8.3-fold in THP-1 at a 5:1 ratio), confirming a substantial pro-apoptotic shift in leukemic cells following PD-1/TIGIT inhibition.

Discussion

This study shows that siRNA-mediated dual knockdown of PD-1 and TIGIT in leukoreduction-filter-derived CD3⁺ T cells enhances anti-leukemic activity against KG-1 and THP-1 AML cells, supporting the idea that AML-associated T-cell dysfunction can be reversed by coordinated checkpoint suppression. PD-1 and TIGIT coexpression has been documented in AML bone marrow T cells, where it is associated with a more immunosuppressive microenvironment and impaired T-cell function, which makes their combined targeting biologically plausible¹⁸.

Our data extend this concept by showing that dual checkpoint silencing increased cytotoxicity, promoted leukemic apoptosis, and restored IFN-γ production while preserving T-cell viability and CD4⁺/CD8⁺ balance. This pattern is consistent with the broader biology of T-cell exhaustion, in which sustained inhibitory receptor signaling limits proliferation, cytotoxicity, and cytokine output, while dual checkpoint targeting can restore antitumor effector function more effectively than single-pathway inhibition^18,29. The observed superiority of dual PD-1/TIGIT silencing over single knockdown is also mechanistically coherent. PD-1 directly suppresses T-cell activation through inhibitory signaling that dampens PI3K-AKT and related downstream pathways, whereas TIGIT inhibits T-cell activity through SHIP1-associated signaling and repression of PI3K, MAPK, and NF-κB programs³⁰, so simultaneous blockade is expected to remove complementary brakes on T-cell function³¹.

Current AML immunotherapy has largely focused on checkpoint blockade, especially in combination with hypomethylating agents, but clinical benefit has remained variable and the field is still supported mainly by early-phase studies rather than definitive randomized evidence¹⁹. In relapsed/refractory AML, azacitidine plus nivolumab produced encouraging activity in a phase II study, yet the broader clinical experience has not established checkpoint blockade as a consistently effective stand-alone strategy in AML¹⁹. This limited efficacy is biologically consistent with the fact that AML T cells frequently show coordinated PD-1 and TIGIT co-expression, and this exhausted phenotype has been linked to impaired cytokine production and functional suppression¹⁸. Accordingly, dual checkpoint suppression is a rational next step, because TIGIT has emerged as an additional inhibitory axis in AML and may complement PD-1 targeting more effectively than single-pathway inhibition³².

The greater sensitivity of THP-1 relative to KG-1 in our assays likely reflects intrinsic biological heterogeneity between AML subtypes rather than a uniform checkpoint dependency across the disease. This subtype-specific response is important because AML is not a single biological entity, and therapeutic susceptibility can vary substantially with lineage, differentiation state, apoptotic threshold, and immunophenotype³³. A practical strength of this study is the use of leukoreduction filters as the T-cell source. Leukoreduction-system chambers and leukocyte reduction filters have been described as accessible, economical, and workable starting material for cellular research and manufacturing, which supports the feasibility of this platform for autologous immune-cell manipulation³⁴.

Several limitations temper direct clinical interpretation. First, the study was restricted to AML cell lines, and primary AML patient samples were not tested, so the findings cannot yet be assumed to generalize across the full molecular and microenvironmental heterogeneity of patient disease. Second, the in vitro siRNA transfection strategy is not clinically optimized, because unprotected siRNA is vulnerable to nuclease degradation and rapid clearance, and in vivo application will likely require chemically stabilized constructs and delivery systems such as lipid nanoparticles or other targeted carriers^35–37. Third, durability of knockdown was not examined beyond the short experimental window, so persistence of checkpoint suppression remains unknown. That issue matters because translational RNAi therapy depends not only on initial transfection efficiency, but also on sustained intracellular exposure, acceptable pharmacokinetics, and repeatable delivery³⁵. Fourth, immune functional characterization was incomplete. TIM-3 and LAG-3 were not assessed, and T-cell proliferation and persistence assays were not included, even though these markers commonly participate in AML-associated exhaustion and may identify compensatory inhibitory pathways after PD-1/TIGIT inhibition³⁸. Fifth, no mechanistic pathway analysis was performed, so downstream signaling changes in NF-κB, STAT, or PI3K-AKT were not directly measured. This limits the ability to link checkpoint silencing to the apoptotic shift observed in AML cells, and future studies should include pathway-level readouts to define how dual checkpoint suppression alters immune-effector signaling³⁰. Sixth, off-target effects were not exhaustively profiled. This is a meaningful limitation because siRNA seed-mediated off-target silencing is a recognized problem in RNAi biology, and transcriptomic or proteomic validation would strengthen confidence in target specificity and safety³⁹. Taken together, these findings support dual PD-1/TIGIT knockdown as a strong proof-of-concept for AML immunotherapy, but the current work remains preclinical and hypothesis-generating. The next step is validation in primary AML samples and in vivo models using clinically realistic delivery systems, together with broader immune phenotyping, pathway analysis, and safety profiling.

Conclusion

Dual siRNA-mediated knockdown of PD-1 and TIGIT in leukoreduction-filter-derived T cells produced robust checkpoint silencing, restored effector cytokine release, and enhanced cytotoxicity against AML cell lines. These effects were associated with preserved T-cell viability and a favorable balance of effector function over exhaustion, indicating that coordinated checkpoint suppression can meaningfully reprogram anti-leukemic immunity. At the same time, the work remains limited by the absence of primary patient validation, short-term in vitro delivery, incomplete exhaustion phenotyping, and a lack of pathway-level mechanistic analysis. Therefore, dual PD-1/TIGIT knockdown should be viewed as a promising preclinical platform that now requires validation in patient-derived AML systems, durable and targeted in vivo delivery, and deeper immune-mechanistic characterization before clinical translation can be considered.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

S.A: Methodology, Investigation, Visualization, Writing-Original draft preparation, Data Curation, Software, Visualization, Resources, Formal analysis, Writing—Reviewing and Editing; M.N: Methodology, Investigation, Visualization, Writing-Original draft preparation, Data Curation, Software, Visualization, Resources, Formal analysis, Writing—Reviewing and Editing; M.V: Methodology, Investigation, Visualization, Writing-Original draft preparation, Data Curation, Software, Visualization, Resources, Formal analysis, Writing—Reviewing and Editing; S.A.K: Funding acquisition, Conceptualization, Project Supervision; M.A: Funding acquisition, Conceptualization, Project Supervision.

Data availability

Our study involved a limited number of gene expression measurements across a small set of samples and flow cytometry experiments. All data are fully presented within the manuscript and supplementary files. Given the limited dataset, deposition in a public repository is not applicable. Nevertheless, all raw data, including flow cytometry FCS files and gene expression measurements, are available from the corresponding author upon reasonable request. (Ahmadvand.mohamad64@yahoo.com)

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval and consent to participate

The Institutional Ethics Committee of Shariati Hospital, Tehran, Iran, approved the study with the approval number IR.TUMS.HORCSCT.REC.1402.028. All participants were provided written informed consent during the study. All methods were performed in accordance with the relevant guidelines and regulations.