Abstract

   Immunotherapy has transformed the treatment landscape for non-small
   cell lung cancer (NSCLC), yet achieving lasting benefits remains a
   challenge. The resistance mechanisms to immunotherapy are complex,
   involving interactions between tumor cells and immune cells that are
   not fully understood. Metformin, an FDA-approved diabetes medication,
   shows promise in enhancing immunotherapy efficacy by boosting
   anti-tumor immune responses, although the underlying molecular pathways
   are still being investigated. This study utilized co-culture models of
   cancer and immune cells to explore the effects of combining metformin
   with anti-PD-1/PD-L1 therapies on the anti-tumor immune response in
   LKB1 mutant (LKB1mut) versus wild-type (LKB1wt) NSCLC cells, alongside
   peripheral blood immune cells from NSCLC patients. The transcriptomic
   profiles of LKB1mut and LKB1wt NSCLC cells were characterized via bulk
   RNA sequencing to understand gene expression changes induced by
   metformin. Patients with advanced-stage NSCLC provided peripheral blood
   mononuclear cells (PBMCs) for analysis. The study assessed metformin’s
   impact both alone and in combination with anti-PD-1/PD-L1 agents on
   innate immune pathways. Results indicated that metformin activated the
   cGAS-STING pathway and interferons in PBMCs, enhancing their anti-tumor
   capabilities. Notably, immune cells treated with metformin and
   immunotherapy exhibited synergistic effects, significantly reducing
   colony formation in LKB1mut NSCLC cells. Additionally, monocytes from
   NSCLC patients showed decreased viability of NSCLC cells in co-culture,
   independent of LKB1 status, and shifted towards an anti-tumor M1
   phenotype with combined treatment. These findings were supported by 3D
   co-culture models involving tumor spheroids and patient-derived
   organoids, highlighting a novel biological rationale for using
   metformin alongside immunotherapeutic agents to boost anti-tumor
   activity across various immune cell subsets derived from NSCLC
   patients.

   Subject terms: Translational immunology, Non-small-cell lung cancer,
   Tumour immunology, Cancer immunotherapy

Introduction

   Lung cancer is one of the most frequently diagnosed cancers worldwide
   and is the leading cause of cancer-related mortality [[56]1]. Non-small
   cell lung carcinoma (NSCLC) accounts for the majority of cases (85%)
   and comprises a heterogeneous group that includes lung adenocarcinoma
   (LUAD), lung squamous cell carcinoma (LUSC), and large cell carcinoma
   (LCC) [[57]2]. Although the advent of anti PD-1/PD-L1 drugs has
   revolutionized the management of patients with advanced NSCLC, durable
   benefits are not achieved in all patients [[58]3], and the mechanisms
   of resistance remain largely unknown.

   In LUAD, serine/threonine kinase 11 (STK11), also known as liver kinase
   B1 (LKB1), ranks as the third most commonly mutated gene after TP53 and
   KRAS, with mutations found in up to 33% of primary LUAD cases [[59]4].
   Tumors harboring STK11 mutations exhibit an “immune-cold” phenotype
   characterized by low or absent PD-L1 levels and ineffective T cell
   infiltration. Additionally, STK11/LKB1 inactivation is a major genetic
   driver of primary resistance to anti-PD-1 agents and co-mutation of
   STK11 and KEAP1 genes has been correlated to higher sensitivity to dual
   immunotherapy with anti-CTLA4 and anti-PD-1 agents versus anti-PD1
   alone [[60]5, [61]6]. In this context, our group previously
   demonstrated that stimulators of interferon genes (STING) activation
   identifies NSCLC patients with an immune-responsive phenotype [[62]7].
   Mechanistically, STK11 mutations lead to LKB1 loss, which results in
   the suppression of (STING). Thus, reactivating STING in STK11 mutant
   LUAD has been proposed as a promising approach to support anti-tumor
   immune response [[63]8]. Recently, we also established that cGAS-STING
   activation in peripheral blood mononuclear cells (PBMCs) from lung
   cancer patients, particularly those with NSCLC, may serve as a novel
   predictive biomarker for response to immunotherapy [[64]9].

   Several evidences have suggested that the anti-diabetes drug metformin
   (MET) exerted anti-cancer effects in various cancer types, including
   lung, prostate, and colon [[65]10]. Preclinical data support the role
   of MET as an adjuvant drug in the treatment of lung cancer, both in
   combination with chemotherapy and with targeted drugs and immunotherapy
   [[66]11]. In particular, MET may modulate different components of the
   immune-microenvironment and promote pro-immune features [[67]12];
   however, it is unknown whether MET may enhance the efficacy of
   immunotherapy in particular in STK11 mutant NSCLC.

   Our phase I/II dose-escalation study evaluated the feasibility of
   metformin treatment also in a non-diabetic population of
   cancer-patients and in particular confirmed the acceptable safety
   profile of the combination of metformin and erlotinib as a second-line
   treatment for stage IV NSCLC with wild-type EGFR [[68]13, [69]14],
   making it an ideal partner drug for novel combinations.

   In this study, we hypothesized that MET treatment may impact on
   activation of anti-tumor immune response, including STING activation
   and related pathways, with a direct effect in immune cells. Here, we
   sought to assess the efficacy of combining MET with immunotherapeutic
   agents in LKB1-mutant NSCLC tumors and in various subsets of PBMCs,
   specifically lymphocytes and monocytes isolated from treatment-naive
   NSCLC patients. Finally, we validated the in vitro and ex vivo findings
   in two-dimensional and three-dimensional co-culture models utilizing
   tumor spheroids and patient-derived tumor organoids (PDTOs) to
   elucidate the impact of metformin and anti-PD-1/PD-L1 immunotherapy on
   the interactions between cancer and immune cells.

Materials and methods

Reagents

   Metformin, Atezolizumab, and Pembrolizumab, were purchased from Selleck
   Chemicals (Houston, TX).

Cell lines

   The NSCLC cell lines NCI-H1299 (ATCC Cat#CRL-5803) and NCI-H460 (ATCC
   Cat#HTB-177) were maintained in RPMI-1640 (Sigma-Aldrich) supplemented
   with 10% FBS (Sigma-Aldrich), 1X penicillin–streptomycin
   (Sigma-Aldrich) and 1X Amphotericin B (Sigma-Aldrich) in a
   humidity-controlled environment (37 °C, 5% CO[2]). Cell lines were
   obtained from the American Type Culture Collection (ATCC). Cell lines
   morphology was monitored, and the cell lines were routinely tested for
   mycoplasma using a mycoplasma detection kit (InvivoGen).

RNA-sequencing

   Cell lines and PBMCs were treated with 2 mM Metformin and 10 μg/ml
   Atezolizumab for 48 h. Cells were harvested and RNA was purified using
   an RNeasy Plus Mini kit (QIAGEN). The samples were analysed by Novogene
   Co., Ltd. Detailed RNA-seq procedure is described in the Supplemental
   Material file.

MTT assay

   MTT assay was performed as previously described [[70]15]. Briefly,
   cells were seeded in 96-well plates at a density of 5000 cells/well and
   co-cultured with PBMCs (ratio 1:10). After 48 h of treatment, the
   methylthiazolyldiphenyl-tetrazolium bromide (MTT, Sigma Chemical Co.)
   was added. After 4 h, cells were lysed by addition of DMSO. The number
   of viable cells was determined spectrophotometrically by measuring
   absorbance at 490 nm.

Western blot analysis

   Protein lysates were obtained by homogenization in RIPA buffer [0.1%
   SDS, 0.5% deoxycholate, 1% Nonidet, 100 mmol/L NaCl, 10 mmol/L Tris–HCl
   (pH 7.4), 0.5 mmol/L DTT, 0.5% PMSF, protease inhibitor cocktail
   (Hoffmann-La Roche), and PhosSTOP (Roche Diagnostics)] and clarified by
   centrifugation at 12.450 rpm for 20 min at 4 °C. Protein amounts were
   estimated by a modified Bradford assay (Bio-Rad), resolved by SDS-PAGE
   and electrotransferred onto nitrocellulose membranes (Bio-Rad). After
   blocking the membranes for 90 min at RT, they were incubated overnight
   at 4 °C with primary antibodies, followed by incubation with a
   secondary antibody for 1 h at RT. Proteins were detected with Clarity
   Western ECL Substrate using a ChemiDoc (Bio-Rad). Images were analyzed
   using Image Lab 3.0.1. Primary antibodies for western blot analysis
   against LAMIN A/C (2032, 1:1000), CASPASE-8 (1C12) (9746, 1:1000), and
   β-Actin (8H10D10, 1:2000) was purchased from Cell Signaling.

Patients cohort and PBMCs isolation

   Patients with a diagnosis of NSCLC at baseline before receiving any
   treatment were enrolled. The patients’ characteristics are listed in
   Table [71]1. Representative IHC images of PD-L1 tumor expression levels
   in tissue biopsies are reported in Supplementary Fig. [72]S1. Blood
   samples were collected after obtaining a written informed consent from
   patients in accordance with the Declaration of Helsinki (protocol
   approved by the Ethics Committee of the University of Campania “Luigi
   Vanvitelli” and Naples No. 280 on May 16, 2020). NSCLC patients’ blood
   was collected in BD vacutainer spray-coated K2EDTA tubes as previously
   described [[73]9]. Briefly, PBMCs were isolated by Lymphosep
   (Aurogene). PBMCs were washed with PBS and cultured in RPMI-1640 medium
   (Sigma-Aldrich) supplemented 10% FBS, and 1X penicillin–streptomycin
   (Sigma-Aldrich). After 24 h, the monocytes were mechanically isolated
   by adherence, and lymphocytes were separated by suspension.

Table 1.

   Baseline clinical characteristics of patients.
           Variable         Patients^a (n = 18)
   Age, years (mean ± SD)   70.56 ± 7.21
    Sex (n, %)
     Male                   9 (56.25%)
     Female                 7 (43.75%)
    Clinical stage (n, %)
     III                    2 (12.5%)
     IV                     14 (87.5%)
    PD-L1 expression (n, %)
     < 1%                   9 (56.25%%)
     > 1%                   7 (43.75%%)
   [74]Open in a new tab

   Data are reported as number (n), frequency (%) or mean ± SD: standard
   deviation. ^aPatients race/ethnicity was White/Italian.

Immunohistochemistry (IHC)

   Immunohistochemistry (IHC) was performed on 4-micron sections of
   formalin-fixed, paraffin-embedded (FFPE) tissue samples using Ventana
   platform (Ventana BenchMark ULTRA system), according to the
   manufacturer’s instructions. The following antibody was used:
   anti-PD-L1 (rabbit monoclonal primary antibody, clone SP263, VENTANA).
   The IHC staining were interpreted based on the percentage of tumor
   cells (TCs) exhibiting membranous staining at any intensity and it was
   scored according to the IMpower010 study data [[75]16]:
     * Negative staining: PD-L1 expression on <1% of TCs;
     * Positive staining: PD-L1 expression on ≥1% of TCs;

   The PD-L1 positive cases were stratified based on the percentage of
   positive TCs:
     * PD-L1 expression ranging from 1 to 49% of TCs;
     * PD-L1 expression on ≥50% of TCs.

Colony formation assay

   1000 cells/well were seeded on a 6-well plate. The next day the plates
   were treated with 2 mM Metformin, 10 μg/ml Atezolizumab and 10 μg/ml
   Pembrolizumab. After 10 days, colonies were fixed with 4% PFA and
   stained with 1% crystal violet in 10% methanol.

RNA extraction, cDNA synthesis and qRT-PCR

   Total RNA was obtained using TRIzol (Invitrogen), according to the
   manufacturer’s protocol. Briefly, cells were lysed with TRIzol.
   Chloroform was added and samples were then centrifuged at 12,000 x g
   for 15 min at 4 °C. The aqueous phase was transferred to another tube.
   RNA was precipitated by mixing with isopropyl alcohol for 2 h at
   −80 °C. Samples were centrifuged at 12,000 x g for 10 min at 4 °C and
   washed with 75% ethanol. Total RNA was suspended in RNase-free water,
   0.1 mM EDTA solution and incubated in a heat block at 57.5 °C for
   10 min. The purity and concentration of RNA were determined by OD
   260/280 readings using the Nanodrop 2000 spectrophotometer (Thermo
   Fisher Scientific). cDNA was generated using a SensiFAST cDNA Synthesis
   Kit (Meridian Bioscience). mRNA expression levels were evaluated by
   qRT-PCR with a QuantStudio 7-Flex (Applied Biosystems) using the
   SensiFAST SYBR Hi-ROX Kit (Meridian Bioscience). Relative gene
   expression was determined by normalizing to 18S and calculated by the
   2-ΔΔCt method. The list of primer sequences used for qRT-PCR are shown
   in Supplementary Table [76]1.

Spheroids formation and Immunofluorescence staining

   Tumor spheroids were generated by seeding 8 × 10^4 cells in ultra-low
   attachment (Corning) as previously described [[77]9]. Co-cultures were
   initiated by adding 5 × 10^5 SCLC patients derived immune cells per
   well. Before co-culture, 2.5 µM CellTracker Deep Red dye (Invitrogen)
   was added to PBMC culture medium for 45 min and 5 µM CellTracker Blue
   CMAC (Invitrogen) was added to the spheroid culture medium for 45 min.
   After the staining, the PBMCs were added to spheroids (ratio 1:10) and
   the co-cultures were performed for 48 h. The fluorescence was imaged
   using a Cell Discoverer 7 microscope (ZEISS Olympus, Nikon). The
   quantification of fluorescence intensity was performed with ImageJ
   Software.

Patient-derived tumor organoids generation

   Tumor tissue obtained from a patient with NSCLC was utilized for the
   generation of tumor organoids as previously described [[78]17]. The
   biopsy was processed by cutting it into small pieces of 2-4 mm and
   washed with PBS supplemented with 1% penicillin-streptomycin. The
   minced tissue was incubated with Digestion Medium (5 mg/ml Collagenase
   type II, 100 μg/ml DNase I, 10 μg/ml Y-27632 dihydrochloride in
   Advanced DMEM/F-12, 1X penicillin–streptomycin and 1X Amphotericin B)
   for 1 h at 37 °C on a shaking platform. The digested tissue was
   centrifuged at 200 g for 5 min at RT, washed with AdDMEM/F-12 and
   incubated with trypsin + Y-27632 dihydrochloride (10 μM) at 37 °C on a
   shaking platform. Then, cold adDMEM/F12 + ++ supplemented with 1X
   penicillin–streptomycin, 1X Amphotericin B and 20% FBS were added to
   stop the digestion. The resulting cells were cultured in Matrigel
   droplets of organoid culture medium (composition of organoid medium is
   listed in Supplemental Material).

Flow cytometry analysis

   Tumor-infiltrating lymphocytes were separated by Tumor cell isolation
   kit (Miltenyi Biotec) analyzed by flow cytometry. TILs staining: Live
   Dead Blue (Invitrogen), antibodies used were anti-human CD3 (UCHT1,
   SK7), CD4 (L3T4, SK3), CD8 (SK1), CD56 (NCAM16.2).

Statistical analysis

   Results are expressed as the mean ± SEM or SD. Two-group analyses were
   performed using an unpaired t-test. Three or more groups with one
   independent variable were analyzed using a one-way ANOVA test. Three or
   more groups with two independent variables were analyzed using a
   two-way ANOVA test. The Pearson correlation coefficients and the
   corresponding p-values were calculated using Prism 8 (GraphPad
   Software).

Results

Transcriptomic characterization of LKB1mut and LKB1wt NSCLC cells

   We analyzed the transcriptomic features of LKB1-mutant NSCLC cells
   (H460) compared to LKB1-wild type one (H1299). As shown in Fig. [79]1A
   we found 4752 upregulated and 6234 downregulated genes (fold change
   (FC) > 1.30; p value < 0.05) in LKBmut cells compared to LKB1wt. Given
   the key role of LKB1 loss in the downregulation of AMPK activation, we
   analyzed AMPK signaling pathway (hsa04152) (Fig. [80]1B). RNA-seq data
   confirmed a downregulation of AMPK gene transcription (PRKAG1) and an
   upregulation of its signaling pathway along with mTOR mRNA upregulation
   in LKB1mut cells. Then, we wondered whether the LKB1 loss occurred
   along with a modulation of immune responsiveness in the selected cells.
   Interestingly, both the Toll-like receptor signaling pathway (map04620)
   (Fig. [81]1C) and the cytosolic DNA-sensing pathway (map04623) (Fig.
   [82]1D) showed a strong reduction of key genes transcription involved
   in immune response, such as IFNs, IRFs, TBK1 and cGAS-STING.

Fig. 1. RNA-sequencing of LKB1mut vs LKB1wt NSCLC cell lines.

   [83]Fig. 1
   [84]Open in a new tab

   A DEGs across LKB1mut vs LKB1wt comparisons. Fold Change (FC) > 1.30
   and p value < 0.05. B–D Results of KEGG pathway enrichment analysis.
   The red and green bars represent the upregulated genes and
   downregulated genes, respectively for the differential expression of
   the selected pathways based on their RNA-seq expression. Data are
   presented as Log2FC values, all the listed genes p value is < 0.05.

Effect of metformin on cGAS-STING downstream pathway activation in LKB1mut
and LKB1wt NSCLC cells

   We performed RNAseq and DEGs on both LKB1mut and LKB1wt cell lines
   after treatment with MET. As shown in Fig. [85]2A, AMPK gene expression
   and AMPK signaling pathway (hsa04152) were significantly upregulated in
   LKB1mut cells treated with MET compared to untreated cells.
   Interestingly, treatment with MET in LKB1mut cells induced a
   significant increase of IFNβ, CCL5, CXCL10 and IRF3 related to the
   Cytosolic DNA-sensing pathway (map04623) (Fig. [86]2B). On the
   contrary, MET treatment did not show the same effect on AMPK signaling
   pathway and Cytosolic DNA-sensing pathway (Fig. [87]2C-D) in LKB1wt
   cells, thus indicating that MET activates AMPK and cytosolic DNA
   sensing pathways in LKB1mut NSCLC cells.

Fig. 2. Effect of treatment with metformin on cGAS-STING downstream
activation in LKB1mut NSCLC cell lines.

   [88]Fig. 2
   [89]Open in a new tab

   A, B Results of KEGG pathway enrichment analysis in LKB1mut cells
   treated with MET compared to CTR. Data are presented as Log2FC values,
   all the listed genes p value is < 0.05. C, D Results of KEGG pathway
   enrichment analysis in LKB1wt cells treated with MET compared to CTR.
   The red and green bars represent the upregulated genes and
   downregulated genes, respectively for the differential expression of
   the selected pathways based on their RNA-seq expression. Data are
   presented as Log2FC values; all the listed genes p value is < 0.05. E
   Representative immunofluorescence images, p-IRF3 mean fluorescence
   intensity (MFI) and nuclear localization (green/blue merge) showing
   p-IRF3 (green), MitoTracker (red) and DAPI (blue) in H460 cells treated
   with MET 2 mM (scale bar: 10 μm, Magnification 63X). Each experiment
   was performed in three technical replicates. Data are expressed as
   mean ± SD. ****p < 0.0001; *p < 0.05.

   To further confirm the activation of the DNA sensing pathways induced
   by metformin treatment in LKB1 mutant NSCLC cells, we conducted IF
   staining to assess the subcellular translocation of activated p-IRF3 to
   the nucleus, where it promotes the transcription of IFN genes. As
   illustrated in Fig. [90]2E, we observed a significant translocation of
   p-IRF3 to the nuclear compartment following metformin treatment in H460
   cells compared to the CTR. In contrast, no significant translocation of
   p-IRF3 was detected in H1299 cells, as shown in Supplementary Fig.
   [91]S2. Overall, our results confirm that metformin induces cGAS-STING
   activation to a greater extent in LKB1 mutant NSCLC cells.

Effect of metformin on immune cells derived from naive NSCLC patients

   Given the broad antitumor effects mediated by MET treatment on immune
   cells in different solid tumors, especially increasing tumor
   infiltrating lymphocyte levels and upregulating T-cell inflamed
   expression profile^12, we characterized the effect of MET towards
   immune cells derived from naive NSCLC patients be RNA-seq and DEGs
   (Fig. [92]3A). In particular, as shown in Fig. [93]3B, we found 5181
   upregulated genes and 4266 downregulated genes in PBMCs treated with
   MET compared to CTR. The top down regulated genes (P < 0.05) were VCAN
   (Log2FC = −6,54391), VEGFA (Log2FC = −4,11021), CXCL12
   (Log2FC = −2,36826), CCL5 (Log2FC = −1,152) and CCL20
   (Log2FC = −0,59993) (Fig. [94]3C). VCAN and VEGF play a role in the
   regulation of cell motility, growth and differentiation. CXCL12 and
   CCL5 act as chemoattractant active on T-lymphocytes and monocytes.
   CCL20 is responsible for the chemotaxis of effector/memory T-cells and
   B-cells [[95]18]. Among the most significant upregulated genes in MET
   treated PBMCs compared to CTR we found CXCL5 (Log2FC = 3,284284), CCL18
   (Log2FC = 2,510331), CXCL10 (Log2FC = 2,063503), and STING-gene TMEM173
   (Log2FC = 1,333424). CXCL5 and CCL18 are mainly involved in the
   chemotactic activity of neutrophil and lymphocytes [[96]19]. CXCL10 is
   known to play pleiotropic effects, including stimulation of monocytes,
   natural killer and T-cell migration, and modulation of adhesion
   molecule expression [[97]20]. It is also a key regulator of immune
   response to viral infections [[98]21] and a key cGAS-STING downstream
   chemokine involved in immune response [[99]9]. Finally, innate immune
   activation through STING gene upregulated transcription (TMEM173)
   indicates that PBMCs treated with MET increase features of innate
   immune response.

Fig. 3. Effect of treatment with metformin in NSCLC patients-derived PBMCs.

   [100]Fig. 3
   [101]Open in a new tab

   A DEGs in MET-treated NSCLC patient-derived PBMC vs CTR. B Volcano plot
   of DEGs analysis in PBMCs treated with MET compared to CTR. Data are
   presented as Log2FC values, green and red listed genes p value is <
   0.05. C Most significant DEGs in NSCLC patients-derived PBMCs vs CTR.
   The red and green bars represent the upregulated genes and
   downregulated genes, respectively for the differential expression of
   the selected pathways based on their RNA-seq expression. Data are
   presented as Log2FC values, green and red listed genes p value is <
   0.05. D Dot plot of DEG showing the top significant enriched biological
   process (BP) related to MET treated PBMCs vs CTR. Gene Ratio is defined
   as the ratio between intersection size and query size. The dot sizes
   represent enrichment abundance while the colors represent the adjusted
   p values (padj) (blue to red). E Results of KEGG pathway enrichment
   analysis in PBMCs treated with MET compared to CTR. The red and green
   bars represent the upregulated genes and downregulated genes,
   respectively for the differential expression of the selected pathways
   based on their RNA-seq expression. Data are presented as Log2FC values,
   all the listed genes p value is < 0.05. F Real-time PCR analysis of in
   vitro STING, c-GAS, IFI16, IL1β, CXCL10 and CCL5 mRNA expression in
   patient-derived PBMCs. Changes in mRNA levels were normalized to the
   expression of housekeeping genes (18S). Data are expressed as
   means ± SEM derived from n = 2 technically calculated by the
   comparative method 2 − ∆∆Ct. One-tailed unpaired Student’s t-test with
   CI = 95%. Statistical significance: **p < 0.01; *p < 0.05.

   We then performed Gene Ontology (GO) enrichment analysis of NSCLC
   patients-isolated PBMCs treated with MET compared to untreated CTR.
   Figure [102]3D shows that MET treatment induced a positive regulation
   of gene sets related to transmembrane transporters and ion channels
   activities along with chemokines mediated signaling pathway and humoral
   immune response in PBMCs from NSCLC patients compared to CTR.
   Transmembrane transport of ions represent an essential feature for the
   maintenance of subcellular organelle integrity and homeostasis in
   immune cells [[103]22] and favors the release of inflammatory cytokines
   [[104]23].

   We therefore analyzed DEGs results from RNA-seq data in terms of AMPK,
   Toll-like receptor signaling and cytosolic DNA-sensing pathways
   enrichment in NSCLC patients-derived PBMCs after treatment with MET
   compared to untreated CTR (Fig. [105]3E). Among the AMPK signaling
   pathway (hsa04152), we found that the endoplasmic reticulum (ER)
   protein CAMP Responsive Element Binding Protein 3 Like 1 (CREB2L1) was
   the most upregulated gene along with the TGF-Beta Activated Kinase
   MAP3K7CL, thus indicating an increase of immune activation in PBMCs.
   Regarding the Toll-like receptor signaling pathway (map04620) the most
   enriched gene transcription was the NF-Kappa-B-Activating Protein 502
   (TICAM2), known for its role in facilitating the downstream signaling
   leading to type I interferon induction, along with enrichment of IFNA,
   IL1B and CXCL8. Finally, the analysis of cytosolic DNA-sensing pathway
   (map04623) showed that the most upregulated gene was cGAS as well as
   enrichment of RIPK4 gene which is known for its role in NF-kappa-B
   activation. We further confirmed these findings by RT-qPCR by analyzing
   the expression of cytosolic DNA-sensing pathway and Toll-like receptor
   signaling pathway in PBMCs isolated from NSCLC patients (Fig. [106]3F).
   Interestingly, we found a significant increase of cGAS (p < 0.01),
   IL-1β (p < 0.01) and CCL5 (p < 0.05) in NSCLC patients-derived PBMCs
   after treatment with MET compared to untreated CTR. Overall, these data
   suggest a positive pro-immune effect induced by MET treatment on PBMCs
   of NSCLC patients.

Effect of metformin and anti-PD-L1 combination on lymphocytes antitumor
activity towards NSCLC cells

   We then investigated the effect of MET treatment combined with
   anti-PD-L1 drugs in inducing a functional anti-tumor immune response in
   immune cells towards NSCLC cells based on their LKB1 status. As we
   reported in our previous studies [[107]24], H460 and H1299 cell lines
   show a percentage of PD-L1 of 85%. In Fig. [108]4A, we assessed the
   viability of LKB1mut and LKB1wt NSCLC cells co-cultured with NSCLC
   patients-derived PBMCs pre-treated with MET alone or in combination
   with an anti-PD-L1. Interestingly, we found that LKB1mut NSCLC cell
   line H460 showed a higher reduction of cell viability when co-cultured
   with PBMCs treated with MET alone and combined with anti-PD-L1 compared
   to LKB1wt cells (Fig. [109]4A). After co-culture with immune cells, we
   analyzed the protein expression of NSCLC cells H1299 and H460 in terms
   of cell cycle and apoptosis markers expression to further investigate
   the mechanism of cell death induced by MET-activated immune cells. We
   found that PBMCs treated with MET alone were not able to induce in
   vitro cell death in NSCLC cell lines, whereas the combination of MET
   with anti-PD-1/PD-L1 activated immune cells-mediated cytotoxicity
   through Caspase-8 increase in H1299 and strong Lamin A/C upregulation
   (Fig. [110]4B).

Fig. 4. Effect of metformin and anti-PD-L1 combination on lymphocytes
antitumor activity towards NSCLC cells.

   [111]Fig. 4
   [112]Open in a new tab

   A Cell viability assay of LKB1wt (H1299) and LKB1mut (H460) NSCLC cells
   exposed to NSCLC patient-derived PBMCs pre-treated with metformin (MET)
   and anti-PD-L1 for 48 h. Data are expressed as mean ± SD. At least
   three independent experiments were performed. B Western blotting of
   H460 and H1299 NSCLC cell lines after 48 h of co-culture with NSCLC
   patient-derived PBMCs. β-actin was used to ensure equal loading. C H460
   and H1299 NSCLC cell lines were seeded in a 12-well plate (3000
   cells/well). The cells were preincubated for 48 h with NSCLC
   patient-derived PBMCs pre-treated with drugs for 5 days. Afterward, the
   cells were fixed and stained with 0.5% crystal violet solution. Images
   of representative wells were scanned and are shown. Bar plots of colony
   numbers for H1299 NSCLC cells are shown in blue, while H460 NSCLC cells
   are shown in red. Results are expressed as the mean of three technical
   replicates ± SD. D Real-time PCR analysis of in vitro STING, c-GAS,
   CXCL10 and CCL5 mRNA expression in patient-derived monocytes (n = 11).
   Changes in mRNA levels were normalized to the expression of
   housekeeping genes (18S). Data are expressed as means ± SEM derived
   from n = 2 technically calculated by the comparative method 2 − ∆∆Ct.
   One-tailed unpaired Student’s t-test with CI = 95%. E Cell viability
   assay of NSCLC cell lines cultured with NSCLC patient-isolated
   monocytes treated with MET and/or anti-PD-L1 for 48 h conditioned
   media. Each experiment was performed in triplicate. Data are expressed
   as mean ± SD. ****p < 0.0001; ***p < 0.001; **p < 0.01; *p < 0.05.

   We also investigated the effects of combining MET with
   immunotherapeutic agents on the induction of anti-tumor activity in
   immune cells. As illustrated in Fig. [113]4C-D, a modest decrease in
   colony numbers was observed in H460 cells cultured with PBMCs compared
   to the control group. Treatment with MET alone or in combination with
   anti-PD-L1 significantly diminished the colony-forming capacity as well
   as the immune cell response towards these LKB1-mutant cells. This
   finding suggests that the combination of MET and immunotherapy can
   influence the colony formation ability of LKB1-mutant tumors activating
   immune cells to exhibit anti-tumor activity. Furthermore,
   anti-PD-1/PD-L1 did not affect colony formation, reinforcing the notion
   that LKB1-mutant cells exhibit intrinsic resistance to immunotherapy.
   Interestingly, PBMCs co-cultured with H1299 cells significantly
   impacted colony formation compared to the control. The application of
   MET, anti-PD-L1, anti-PD-1, and their combinations resulted in a
   substantial reduction in colony numbers within H1299 cells. These
   results indicate that the addition of MET to immunotherapeutic regimens
   effectively promotes an immune-mediated anti-tumor response against
   both LKB1-mutant and LKB1-wild type cells. Overall, the co-culture
   findings suggest that MET is capable of activating anti-tumor immune
   cell functions, leading to in vitro cytotoxicity, apoptosis, and
   decreased long-term tumor survival, independent of LKB1 mutational
   status. This implies that patients exhibiting poor responses to
   immunotherapy may benefit from the incorporation of MET into their
   treatment regimen.

   Monocytes play a central role in mediating anti-tumor functions of
   immune responses [[114]25], so we analyzed the effect of MET on
   PBMC-isolated monocytes. PBMCs from naive NSCLC (n = 11) were collected
   and subsequently monocytes were isolated in culture by adherence. After
   treatment with MET and/or anti-PD-L1 for 48 h we analyzed the mRNA
   expression of the cGAS-STING innate immune pathway. Interestingly, both
   STING and cGAS were significantly upregulated after treatment with MET
   alone or combined with anti-PD-L1 in NSCLC patients-derived monocytes
   (p < 0.0001). Then we characterized by RT-PCR the anti-tumoral M1 and
   pro-tumoral M2 phenotypically distinct subpopulations (Fig. [115]4D)
   [[116]26]. As shown in Fig. [117]4D, we found that combination
   treatment of MET and atezolizumab significantly increased the M1
   markers IL-1β, and TNF-α as compared to control and with a higher fold
   increase compared to the M2 markers modulation.

   Finally, in Fig. [118]4E, we evaluated the capacity of pre-treated
   monocytes to mediate an in vitro anti-tumor effect through co-culture
   with NSCLC cells. Notably, the conditioned media from MET-treated
   monocytes significantly reduced cell viability in LKB1-wild type cells.
   In contrast, the combination of MET with immunotherapeutic agents
   demonstrated efficacy against both cell lines, regardless of LKB1
   mutational status. This further corroborates the benefits of utilizing
   MET in conjunction with immunotherapy.

Effect of metformin and anti-PD-L1 combination on immune cells antitumor
activity towards NSCLC tumor spheroids and patients-derived 3D tumor
organoids

   We functionally validated our in vitro findings regarding the
   pro-immune effects of MET on PBMCs derived from NSCLC patients,
   utilizing 3D NSCLC cold tumor spheroids derived from LKB1-mutant cells.
   Following the formation of tumor spheroids, we established a 3D culture
   comprising both tumor spheroids and immune cells derived from NSCLC
   patients, and we monitored the development of the tridimensional
   structure of the spheroids generated from LKB1-mutant NSCLC H460 cells
   over time (Fig. [119]5A). After 48 h of co-culture, we assessed the
   infiltration capacity of immune cells (Red) into the 3D tumors (Blue)
   (Fig. [120]5B). Our results indicated that treatment with MET, both
   alone and in combination with anti-PD-1 or anti-PD-L1 agents,
   significantly enhanced the co-localization of immune cells within the
   tumor spheroids (Fig. [121]5C). Concurrently, there was a notable
   reduction in the viability of the tumor spheroids (Fig. [122]5D). This
   model provides a functional in vitro framework for predicting the
   pro-immune effects induced by MET on LKB1-mutant tumors. Moreover,
   these findings suggest that MET has the potential to convert cold
   tumors into inflamed ones.

Fig. 5. Effect of metformin and anti-PD-L1 combination on immune cells
antitumor activity towards NSCLC LKB1mut 3D tumor spheroids.

   [123]Fig. 5
   [124]Open in a new tab

   A Bright-field images of LKB1mut H460 spheroids plated in culture 6
   days after plating and treated with MET and anti-PD-L1 or anti-PD-1.
   Scale bars, 100 mm. B Representative IF images showing the infiltration
   ability of immune cells in 3D tumor spheroids of H460 cells. 3D tumor
   spheroids were blue stained, and the NSCLC patient-derived PBMCs were
   red stained. The PBMCs were treated with MET and anti-PD-L1 or
   anti-PD-1 for 48 h and added to 3D spheroids to start the co-culture.
   The co-localization of NSCLC cells and PBMCs is reported under the
   merge column. Scale bar 100 µm. C Evaluation of the red/blue intensity
   ratio as a quantitative indication of the colocalization rate of LC
   cells and PBMC. ****p < 0.0001. D Cell viability assay of PDTOs
   co-cultured for 48 h with MET, anti-PD-L1 or anti-PD-1 pre-treated
   NSCLC-patient derived PBMCs. Data are expressed as mean of three
   independent experiments ± SD. ****p < 0.0001.

   Finally, we validated our in vitro findings on the pro-immune effect of
   MET on NSCLC patients derived PBMCs in an ex vivo 3D tumor established
   from NSCLC biopsy. PD-L1 expression levels of tumor tissue biopsy from
   NSCLC patients used for 3D tumor establishment is reported in Fig.
   [125]S2B. After tissue biopsy digestion, we isolated tumor-infiltrating
   lymphocytes (TILs) by magnetic beads conjugation (Fig. [126]6A). Then
   we performed flow cytometry analysis to characterize and quantify the %
   of TILs. As shown in Fig. [127]5A, 90.6% of live TILs after isolation,
   with 4.2% being CD3+ cells and 0.4% being CD56+ cells. Among the CD3+
   gated cells, the majority of cells were cytotoxic T lymphocytes
   (CD8 + ). In Fig. [128]6B we summarized the main subpopulation of TILs
   isolated from NSCLC patient biopsy. Thereafter, we established a 3D
   culture of biopsy isolated tumor and immune cells, and we monitored the
   tridimensional structure formation over time (Fig. [129]6C). Once the
   organoids culture was established, after several passages the immune
   cells could be depleted so we rinsed PDTOs with autologous PBMCs
   pre-treated with MET combined with anti-PD1/PD-L1 drugs. After 48 h of
   co-culture, we tested cell viability of both tumors organoid and
   spheroid. As shown in Fig. [130]6D, MET alone or combined with
   anti-PD-L1 was able to induce a significant reduction in cell viability
   of PDTOs, ultimately confirming the translational potential of
   combination of metformin and anti-PD-1/PD-L1 agents.

Fig. 6. Effect of metformin and anti-PD-L1 combination on immune cells
antitumor activity towards NSCLC patient-derived tumor organoid.

   [131]Fig. 6
   [132]Open in a new tab

   A A graphical summary of the tumor-infiltrating isolation method
   step-by-step. The graphical scheme was produced by the authors using
   the free BioRender platform. B Flow cytometry plots for the analysis of
   live lymphocyte subsets in NSCLC tumor tissue. Cells were first gated
   into CD3+ or CD56 + , then CD4 + T cells (CD3 + CD56 − CD4 + CD8 − )
   and CD8 + T cells (CD3 + CD56 − CD4 − CD8 + ) were identified. Numbers
   adjacent to the outlined areas indicate the percentage of cells in each
   sample. C Bright-field images of NSCLC patient-derived cancer organoids
   plated in disc culture 6 days after plating. Scale bars, 100 mm. D Cell
   viability assay of NSCLC PDTOs treated with MET alone or combined with
   anti-PD-1 or anti-PD-L1 for 48 h. Data are expressed as mean of three
   independent experiments ± SD. ****p < 0.0001.

Discussion

   In recent years, immunotherapy has become one of the major therapy
   options for advanced NSCLC patients that has greatly benefited from
   this approach. According to current selection criteria, however, not
   all patients achieve a durable response to immune checkpoint inhibitors
   [[133]27]. In this context, the development of either primary or
   secondary resistance to immunotherapy highlights a need for alternative
   MET to modulate anti-tumor immune functions. Recently, it has been
   proved that MET combined with PD-1 inhibitor enhanced anti-tumor
   efficacy in STK11 mutant NSCLC through inhibition of STING
   ubiquitination [[134]28], suggesting a positive pro-immune effect of
   MET in patients with poor response to immunotherapy. Previous clinical
   trials have shown a clinical benefit of the addition of MET in NSCLC
   patients undergoing immunotherapy [[135]29, [136]30], but the small
   sample sizes and lack of evaluation of LKB1 mutations may have limited
   their conclusions.

   Our study employed a diverse array of in vitro and ex vivo models,
   including NSCLC cells, patient-derived immune cells, 2D and 3D tumor
   spheroid co-cultures, and PDTOs. We demonstrated that combinations of
   MET with anti-PD-1/PD-L1 effectively stimulate an anti-tumor immune
   response by activating key cytosolic DNA sensors and enhancing
   pro-immune characteristics, particularly in LKB1 mutant NSCLC tumors.
   These results provide new insights into the pro-immune effects of MET,
   suggesting its potential to reactivate immune responses in patients
   with poor immunotherapy outcomes. We propose that MET-treated immune
   cells may increase chemokine secretion, aiding the recruitment of
   additional immune cells to the tumor microenvironment and overcoming
   the immunosuppressive barriers present in LKB1mut tumors.

   However, several limitations must be noted. The results from cellular
   and organoid models may not fully capture the complexities of the in
   vivo tumor microenvironment. Additionally, patient responses to
   immunotherapies can vary significantly due to intrinsic heterogeneity,
   complicating predictions regarding the effectiveness of MET
   combinations across broader populations. Longitudinal studies are
   essential to assess both short-term efficacy and the durability of
   immune responses. These considerations highlight the need for further
   research to validate our findings and refine therapeutic strategies for
   NSCLC.

   In conclusion, our study indicates that the combination of MET with
   anti-PD-1/PD-L1 therapies may serve as a promising therapeutic strategy
   to augment the innate immune response in patients diagnosed with NSCLC,
   especially for those harboring LKB1 mutations which tend to be poorly
   immune responsive By leveraging the unique mechanisms of action
   associated with MET, which appears to modulate key signaling pathways
   and promote pro-immune features, we provide compelling evidence for its
   potential role in reactivating immune responses in a subset of NSCLC
   patients.

Supplementary information

   [137]Supplemental Material^ (2.9MB, docx)
   [138]uncropped gels^ (103.5KB, pdf)

Acknowledgements