Cell cultivation and handling

Culturing the MFC mouse GC cell line (catalog number: MZ-0120, Ningbo Mingzhou Biotechnology Co., Ltd.) was performed in RPMI-1640 medium (catalog number: 11875093, Gibco, USA) enriched with 10% fetal bovine serum (FBS) to ensure optimal cell growth. Through lentiviral vector-mediated transduction, the cells were engineered to consistently produce a green fluorescent protein (GFP) and a fluorescent marker, as well as a specific T cell receptor (TCR) targeting GFP in CD8+ T cells (source: HuiZhi Biological Technology Co., Ltd., Suzhou, Jiangsu, China) (Baldominos et al. 2022). The transduced cells were maintained in a T cell growth medium supplemented with IL-2 (catalog number: A1048501, Gibco, USA) to sustain their activity and proliferation (Baldominos et al. 2022).

Mouse CD8+ T cells lacking programmed cell death protein 1 (PD-1) were generated utilizing CRISPR/Cas9 technology. The PD1-sgRNA: 5'-CAGCTTGTCCAACTGGTCGGAGG-3' (PAM: NGG) was integrated into the Lenti-CRISPR v2 vector housing the Streptococcus pyogenes Cas9 nuclease gene (source: Hanheng Biotechnology Co., Ltd., Shanghai). Application of lentiviral Lenti-CRISPR v2 vector resulted in cell transduction, and the PD1-knockout CD8+ T cells were achieved through the CRISPR/Cas9 editing system. The cells transfected with sgRNA plasmids and donor sequences were selected employing 4 μg/mL puromycin (catalog number: HY-K1057, MCE, USA). The PD1 knockout in CD8+ T cells was authenticated by Western blot analysis (Fig. S1) (Zeng et al. 2023; Rahimi et al. 2019).

Western blot

Extraction of cellular total protein was conducted by employing RIPA lysis buffer (P0013C, Bioteke, Shanghai, China) containing PMSF. Upon ice incubation lasting 30 min, the samples were centrifuged at 4 °C and 8000g for 10 min to accumulate the supernatant fluid. The total protein content was measured with the BCA assay kit (Catalog No: 23227, ThermoFisher, USA). After being dissolved with 2 × SDS loading buffer, the protein sample of 50 μg was boiled for 5 min and then underwent electrophoresis on an SDS-PAGE gel. To transfer the proteins, a PVDF membrane was employed, followed by blocking with 5% skim milk for 1 h at room temperature. Next, primary antibodies PD1 (1:1500, NBP1-77276, Novusbio) and GAPDH (1:1000, NB100-56875, Novusbio) were diluted and incubated with the membrane overnight at 4 °C. The membrane was rinsed with TBST on three occasions, each wash lasting 10 min, followed by a 1-h incubation with the HRP-conjugated goat anti-rabbit IgG H&L secondary antibody (ab97051, 1:2000 dilution, Abcam, based in Cambridge, UK). Post TBST rinsing steps, the membrane was situated on a clear glass panel. A suitable amount of Solution A and Solution B from the ECL fluorescence testing kit (Catalog No: abs920, Abiotech (Shanghai) Biotechnology Co., Ltd, Shanghai, China) were blended in the dark and dropped onto the membrane. The membrane was then photographed through the Bio-Rad imaging system (BIO-RAD, USA). Analysis was carried out using Quantity One v4.6.2 software, where the protein content was indicated by the ratio of grayscale values between the protein band and the GAPDH band. Iterated thrice, the experiment yielded the average results (Liu et al. 2019).

Experimental animals

Wild-type BALB/c mice and nude mice (catalog numbers 211, 401, obtained from Beijing Weitonglihua Experimental Animal Technology Co., Ltd.), at the age of 6 weeks, were accommodated in an SPF-level animal facility where the humidity levels were kept at 60–65% and the temperature maintained within the range of 22–25 °C. After acclimatization for one week, experiments were initiated with a prior health assessment of the mice. This experimental protocol and animal handling procedures were authorized by the Animal Ethics Committee of the First Affiliated Hospital of China Medical University.

PD1−/− CD8+ T cell cytotoxicity experiment

MFC GC cell lines expressing GFP and red fluorescent protein (mCherry) were mixed in a 1:1 ratio. Approximately 1 × 106 GC cells were subsequently administered subcutaneously into the right abdominal region of the mice. Within 7 days after tumor establishment, PD1−/− CD8+ T cells were intravenously injected via the tail vein, with approximately 2 × 107 T cells per mouse. After 3 weeks, the tumor tissue was harvested, and the cytotoxicity of T cells against cancer cells was quantitatively assessed through flow cytometry. Additionally, the infiltration of T cells in tumor tissue was assessed through immunofluorescence experiments (Baldominos et al. 2022).

Immunofluorescence

Treatment of tumor tissue sections involved deparaffinization and hydration steps. The tissue samples were subjected to antigen retrieval in sodium citrate buffer for 5 min, followed by an hour-long exposure to 5% goat serum to block non-specific antibody binding. Application of the primary antibody was succeeded by an overnight incubation at 4°C, followed by exposure to the secondary antibody for 1 h at room temperature in a light-protected setting. The slices were treated with a nuclear marker, DAPI, for 5 min under light protection and analyzed utilizing an Olympus BX51 fluorescence microscope (Olympus Corp, Tokyo, Japan). The following antibodies were used: rabbit anti-CD3 (1:100, ab231775, Abcam, Cambridge, UK), CD4 (1:50, ab288724, Abcam, Cambridge, UK), CD8 (1:200, NBP1-49045, novusbio), CD45 (1:200, #70257, CST), Ki67 (1:200, ab15580, Abcam, Cambridge, UK), GFP (1:500, ab183734, Abcam, Cambridge, UK), and the secondary antibodies: Goat anti-rabbit IgG (Alexa Fluor 488) (1:500, ab150077, Abcam, Cambridge, UK), Goat anti-rabbit IgG (Alexa Fluor 405) (1:800, ab175652, Abcam, Cambridge, UK).

The assessment of tumor hypoxia involved administering Pimonidazole (75mg/kg body weight) via intraperitoneal injection 30–60 min before euthanasia, followed by staining using Hypoxyprobe® (Catalog number: H10498, Invitrogen, USA) as directed by the manufacturer's guidelines (Ko et al. 2020).

Flow cytometry

After collecting tumor tissue, it was minced and relocated into a 10 mL tube. Then, a solution consisting of 5 mL of digesting mixture (400U/mL Collagenase IV, 6.8U/mL Hyaluronidase, 20μg/mL DNase I, and 10% FBS in HBSS) was introduced, and the samples underwent an incubation process utilizing a rotator set at 37 degrees Celsius for a period of 20 min. Once digested, the specimens underwent filtration via a 100 μm cell strainer to remove impurities. The single-cell suspension was blocked with 1:100 diluted TruStain FcX (#101319 Biolegend) in flow buffer (2 mM EDTA, 0.1% BSA, PBS). Finally, the samples were analyzed using a CytoFLEX flow cytometer (Beckman) based on the fluorescence emitted by the fluorescent proteins, and FlowJo v.10 software was employed for data analysis purposes (Baldominos et al. 2022).

Assessment of cancer cell proliferation capacity

To evaluate the in vivo proliferation of cancer cells, mice received intraperitoneal injections of EdU at a dose of 10 mg/Kg daily for the final 5–8 days of tumor development. As described above, tumor tissue was collected and subjected to immunofluorescence staining. By adhering to the directions stipulated by the manufacturer, the immunofluorescence BCK-EdU647 staining kit (BaseClick, Sigma) was used to stain EdU in the tumor tissue, allowing for the assessment of cancer cell proliferation (Baldominos et al. 2022).

Contribution of QCCs in PD1−/− CD8+ T cell therapy

To investigate the involvement of QCCs in PD1−/− CD8+ T cell therapy, we constructed MFC tumor cells labeled with mVenus-p27K as a quiescent reporter. Immunofluorescence staining of mVenus-p27K_High cells was performed with Ki67 to validate their proliferative status. To further confirm the cell cycle arrest of mVenus-p27K_High cells, EdU experiments were conducted. Additionally, QCCs were marked by introducing tdTomato-p27K into GFP+ and miRFP670+ GC cells. Subcutaneous injection of a mixture of GFP+ and miRFP670+ labeled GC cells was performed in mice, with miRFP670 (far-red) used as a non-immunogenic control. Immunofluorescence and flow cytometry techniques were utilized to evaluate the cytotoxic impact of PD1−/− CD8+ T cell therapy on GFP+ tumor cells, as well as the changes in QCCs before and after treatment in GFP+ and miRFP670+ tumor cells, and the relative proportion of QCCs in the two tumor cell types (Baldominos et al. 2022).

Tumor initiation assay

To assess the tumor-initiating potential of QCCs, mCherry + mVenus-p27K_High (QCCs) and mCherry + mVenus-p27K_Neg cancer cells were sorted from MFC tumor tissue using FACS and then transplanted into the right abdominal areas of BALB/c mice (1000 cells). The tumor-initiating potential was evaluated, and T cell infiltration in tumor tissue was assessed using immunofluorescence experiments. The antibodies that were utilized are listed below: rabbit anti-CD3 (1:100, ab231775, Abcam, Cambridge, UK), anti-CD4 (1:50, ab288724, Abcam, Cambridge, UK), anti-CD8 (1:200, NBP1-49045, novusbio), anti-CD45 (1:200, #70257, CST), goat anti-rabbit IgG (Alexa Fluor 488) (1:500, ab150077, Abcam, Cambridge, UK), and goat anti-rabbit IgG (Alexa Fluor 405) (1:800, ab175652, Abcam, Cambridge, UK) (Baldominos et al. 2022).

Single-cell sequencing of intragroup and intergroup regions of QCCs

The samples included in the sequencing comprised two regions of tumor tissue: p27K_High (n = 2) and p27K_Neg (n = 2). Initially, the tumor tissue from these regions was washed in cold PBS. Subsequently, the tissue was digested using 1mg/mL of collagenase (Sigma-Aldrich, USA, catalog number: C2674) and exposed to 37 °C for a time span of 10 min. Then, the tissue was incubated with trypsin/EDTA (Gibco, USA, catalog number: 25200072) at 37 °C for 5 min to generate a single-cell suspension. The single-cell suspension was captured using the C1 single-cell auto-preparation system (Fluidigm, Inc., South San Francisco, CA, USA). Once captured, the cells underwent lysis within the chip, releasing mRNA, which was subsequently reverse transcribed, generating cDNA. The cDNA, post-lysis and reverse transcription underwent pre-amplification in microfluidic chips for subsequent sequencing. The amplified cDNA was used to construct libraries for single-cell sequencing utilizing the HiSeq 4000 Illumina platform (parameters: paired-end reads, read length of 2 × 75 bp, approximately 20,000 reads per cell) (Keefe et al. 2023).

The Seurat software package was used for scRNA-seq data analysis. Firstly, data quality control was performed based on the criteria of nFeature_RNA > 500 & nCount_RNA > 3000 & nCount_RNA < 20,000 and percent.mt < 5. Subsequently, the LogNormalize function was applied for normalization. Next, applying the RunPCA function allowed for the execution of Principal Component Analysis (PCA) on a subset of 2000 genes known for their significant variability, and important principal components (PCs) were identified for UMAP clustering analysis through the utilization of the JackStrawPlot and ElbowPlot functions. Utilization of the FindAllMarkers function facilitated the identification of distinct marker genes associated with each cell cluster, while cell classification was achieved through consultation of the CellMarker database. Finally, scatter plots and violin plots were produced to visually represent the expression of marker genes across various cell clusters using the FeaturePlot and VlnPlot functions (Zhang et al. 2019).

Transcriptome sequencing of QCCs

RNA-seq was performed on QCCs (mVenus-p27K_High; n = 6) and non-quiescent tumor cells (mVenus-p27K_Neg; n = 6). Total RNA was extracted using Trizol reagent (15596026 Invitrogen, Car, Cal, USA), and analysis of the RNA samples' concentration and purity involved the use of a Nanodrop 2000 spectrophotometer (1011U, Nanodrop, USA). Total RNA samples that fulfilled the specified conditions were employed in the ensuing assays: RNA Integrity Number (RIN) ≥ 7.0 and 28S:18S ratio ≥ 1.5.

The generation and sequencing of libraries were carried out at CapitalBio Technology, located in Beijing, China. 5 μg of RNA was allocated for each sample. Briefly, Ribo-Zero™ Magnetic Kit (MRZE706, Epicentre Technologies, Madison, Wisconsin, USA) was used to remove ribosomal RNA (rRNA) from the total RNA. Utilization of NEB Next Ultra RNA Library Prep Kit (#E7775, NEB, USA) enabled the creation of libraries for Illumina sequencing. The RNA fragments were initially fragmented into around 300 base pair (bp) long segments using the first-strand synthesis reaction buffer (5 ×) by NEB Next. The reverse transcriptase primer and random primers were utilized for the synthesis of the first cDNA strand, whereas the second cDNA strand was synthesized in the reaction buffer containing dUTP Mix (10 ×). The cDNA fragments underwent end-repair, which entailed adding poly-A tails as well as sequencing adaptors. After the successful ligation of the Illumina sequencing adaptors, a strand-specific library was generated by digesting the second cDNA strand with USER Enzyme (#M5508, NEB, USA). The library DNA was amplified, purified, and enriched by PCR. The assessment of the library involved the application of Agilent 2100 and quantification was achieved using the KAPA Library Quantification Kit (KK4844, KAPA Biosystems). Ultimately, the NextSeqCN500 (Illumina) sequencer was utilized for paired-end sequencing.

Quality assessment of the sequencing data's raw paired-end reads was executed using FastQC software version 0.11.8. The preprocessing of unprocessed data was performed using Cutadapt software v1.18, which entailed the elimination of Illumina sequencing adapters and poly(A) tails. Perl script was used to discard reads with more than 5% N content. The extraction operation was carried out through the utilization of FASTX Toolkit software 0.0.13, focusing on reads with a base quality greater than 20 and comprising 70% of the total bases. The paired-end sequences underwent repair with the aid of BBMap software. Finally, the hisat2 software (0.7.12) was applied to align the filtered high-quality read fragments with the reference genome (Landolt et al. 2016).

Selection of differentially expressed genes (DEGs)

The analysis of gene differential expression on sequencing data was executed through the "limma" package in R software. The false discovery rate (FDR) approach was utilized to correct the diverse p-values. The threshold for selecting DEGs was set as FDR < 0.05 and |logFC|> 1 to identify significantly DEGs. Utilization of the "pheatmap" and "ggplot2" packages in R facilitated the generation of the heat map and volcano plot, respectively. The comparison between the two groups of data was analyzed using the Wilcox. test. The analyses conducted in this research were carried out utilizing R version 4.2.1 (R Foundation for Statistical Computing) (Wu et al. 2022).

GO and KEGG functional enrichment analysis

GO and KEGG enrichment analysis was performed on DEGs via the ClusterProfiler package in R language. The enrichment outcomes from KEGG were visualized using the gglot2 package. Differential gene enrichment was analyzed to identify significantly enriched cellular functions and signaling pathways at a critical level of P less than 0.05 (Yuan et al. 2018).

Gene selection

Leveraging the outcomes of differential analysis from transcriptome sequencing data, the protein produced by the top 100 genes showcasing upregulation was submitted to the String database (https://string-db.org) to investigate protein–protein interaction (PPI). The exportation of the PPI network and its corresponding node files facilitated the generation of a graphical representation showcasing node statistics via R software. Selected as the central genes in the network, those genes exhibiting the greatest number of connections were subject to detailed analysis (Liu et al. 2020).

RT-qPCR

TRIzol (15596026, ThermoFisher, USA) was employed to extract total RNA from cells, with subsequent measurement of RNA concentration and purity performed using the Nanodrop 2000 spectrophotometer (ThermoFisher, USA). The mRNA was converted to cDNA via reverse transcription with the PrimeScript RT Reagent Kit (Takara Code: RR047A, Takara, Japan) as directed by the manufacturer's guidelines and recommendations. TaKaRa produced gene-specific primers through synthesis (Table S1). The real-time fluorescence quantitative PCR procedure, performed through the 7500 Fast Real-Time PCR system (4351106, ThermoFisher, USA), was executed under defined reaction conditions: pre-denaturation at 95 °C for 10 min, denaturation at 95 °C for 10 s, annealing at 60 °C for 20 s, extension at 72 °C for 34 s, for a total of 40 cycles. The relative transcription levels of the target gene were calculated using the comparative Ct method (2−ΔΔCt method), with GAPDH as the reference gene. ΔΔCt was calculated as ΔCt experimental group—ΔCt control group, where ΔCt = Ct (target gene)—Ct (reference gene). The relative transcription levels of the target gene mRNA were expressed as 2−ΔΔCt (Sun et al. 2020). The experiment was reiterated three times in total.

Construction of Lentiviral Vectors

Acquisition of the lentiviral interference vector pSIH1-H1-copGFP (sh-, interference vector, catalog: SI501A-1, System Biosciences, USA) facilitated the construction of a lentivirus-based HIF1A gene interference vector. The lentiviral particles carrying the vector were introduced into HEK-293T cells (catalog number: iCell-h237, Sibang Biotech Co., Ltd., Shanghai, China) using the lentivirus packaging kit (catalog number: A35684CN, Invitrogen, USA). The lentivirus, with a titer of 1 × 108 TU/ml, was extracted from the cell supernatant after 48 h. The sh-NC sequence was AGGCTACAATGATCAGACTAAT, while the sh-HIF1A-1 sequence was GCCACTTTGAATCAAAGAAAT, and the sh-HIF1A-2 sequence was GCCGCTCAATTTATGAATATT (Jiang et al. 2017).

Dual-Luciferase reporter assay

To investigate the impact of HIF1A on the transcriptional activity of ALDOA, LDHA, and HK2 promoters, sh-NC and sh-HIF1A were separately co-transfected into human embryonic kidney HEK293T cells with dual-luciferase reporter gene vectors containing the promoter sequences of ALDOA, LDHA, and HK2, along with their mutant binding sites. Lipofectamine 2000 reagent (Catalog Number: 11668019, Thermo Fisher, USA) was used for co-transfection. The cells were harvested and lysed after 48 h, followed by the assessment of luciferase reporter gene expression employing a luciferase assay kit (K801-200, BioVision, USA). The detection of the luciferase reporter genes was performed using the Dual-Luciferase Reporter Assay System (Promega, Madison, WI, USA). Employing Renilla luciferase as the internal reference gene, the comparison of the activation levels of the target reporter genes was achieved by dividing the relative light units (RLUs) of firefly luciferase by the RLUs of Renilla luciferase (Taniue et al. 2016).

ChIP experiment

Assessing the enrichment of HIF1A in the promoter sequences of ALDOA, LDHA, and HK2 genes was accomplished through the ChIP assay. The ChIP analysis was executed employing the ChIP Kit (KT101-02) provided by Sai Cheng Biotechnology Co., Ltd. in Guangzhou, China. The following steps were carried out: initially, cells were cultured until they reached 70–80% confluency, followed by fixation with 1% formaldehyde at ambient temperature for 10 min to cross-link DNA and proteins within the cells. After cross-linking was completed, DNA fragmentation in cells was achieved through sonication to create random DNA fragments of suitable sizes. Sonication was performed for 10 s with a 10-s pause, repeated 15 times. Subsequent to centrifugation at 13,000 rpm at a temperature of 4 °C, the resulting supernatant was segregated into two tubes. One tube was treated with a rabbit IgG negative control antibody (ab172730, 1:100 dilution, Abcam, UK), while the other tube was incubated with a rabbit-specific antibody specific to the target protein HIF1A (1:100 dilution, Catalog #36169, CST), both maintained at 4 °C for an overnight period. Utilizing Protein Agarose/Sepharose beads, we precipitated the DNA–protein complexes, then performed a brief centrifugation to separate the supernatant and wash away any nonspecific complexes. Later on, the complexes were de-crosslinked at 65 °C for an extended period, and the DNA sections were cleansed through phenol/chloroform extraction for further analysis by qPCR (Table S2) to detect the promoter regions of the ALDOA, LDHA, and HK2 genes (Nelson et al. 2006).

Co-culture of QCCs with T cells

CFSE (Catalog Number: C34554, ThermoFisher, USA) was diluted in a 1:1000 ratio and carefully blended with the T cell mixture. CFSE-labeled T cells underwent co-culture with QCCs in which HIF1A had been silenced. In the co-culture system, 10μg/mL of the immune checkpoint inhibitor anti-PD1 (ab214421, Abcam, Cambridge, UK) was added, with anti-IgG (ab172730, Abcam, Cambridge, UK) as a control. Flow cytometry was employed to measure CFSE density in each group (Zha et al. 2022).

T-cell proliferation and activity assay

T-cells were collected and labeled with a quantitative fluorescent antibody against CD8_FITC (NBP1-49045F, novusbio) to identify T-cell subgroups. Cells were then fixed and permeabilized before staining with antibodies against Ki-67_PE (NB500-170PE, novusbio) and IFN-γ_PerCP (FAB485C, novusbio). Lastly, the samples were subjected to analysis with a BD FACS Canto II flow cytometer (BD Immunocytometry Systems) to assess T-cell proliferation and activity.

Flow cytometry is used to detect cell apoptosis

The cells were amassed in a flow cytometry tube, subjected to centrifugation, and the liquid above them was removed. Three rounds of cold PBS washes were performed on the cells, with subsequent discarding of the supernatant post-centrifugation. The preparation of the Annexin-V-FITC/PI staining solution involved mixing Annexin-V-FITC, PI, and HEPES buffer in a 1:2:50 ratio, as instructed by the Annexin-V-FITC Cell Apoptosis Detection Kit (C1065, Biyun Biotechnology Company, Shanghai, China). Upon resuspension in 100 μL of staining solution, each 1 × 106 cells underwent gentle mixing and subsequent incubation at room temperature for 15 min. Then, 1 mL of HEPES buffer was added and mixed gently. Fluorescence was measured with a flow cytometer to assess cell apoptosis. The excitation wavelength was set at 488 nm, while detection wavelengths of 525 nm for FITC and 620 nm for PI fluorescence were employed in the evaluation process. The experiment was independently repeated three times. Apoptosis Index (AI): AI = (number of apoptotic cells) / (number of apoptotic cells multiplied by the number of normal cells) (Zhu et al. 2017).

Construction of an In vivo GC model

QCCs were prepared in cell suspension with a density of 2 × 107/mL for sh-NC and sh-HIF1A groups, respectively. Into the left axilla of BALB/c nude mice, 0.2 mL of the cell suspension was injected subcutaneously to establish the subcutaneous xenograft model of GC. Six nude mice per group were used, and after inoculation, the mice were relocated to a SPF animal housing facility. The progression of the tumor was monitored, and observations were documented at intervals of 4, 8, 12, 16, 20, and 24 after inoculation. On day 25, mice in each group were euthanized via cervical dislocation, followed by excision and weighing of the tumor tissues.

In addition, a volume of 0.2 mL of QCCs-lucifer cell suspension was administered into the gastric wall of BALB/c mice using a 1 mL syringe to establish the GC orthotopic tumor model. Tumor size was monitored using bioluminescence imaging. After 4 weeks, cervical dislocation was performed to euthanize the mice in every group, and inguinal lymph nodes were collected to observe lymph node metastasis. Tumor tissues were processed into single-cell suspensions, and T cell content was analyzed by flow cytometry (Miao et al. 2023).

Immunohistochemistry

GC in situ tumor tissue was immersed in a 10% formalin solution for fixation and subjected to two rounds of deparaffinization using xylene for 10 min each time. Hydration was performed using a gradient of ethanol concentrations (100%, 95%, 75%, 50%), followed by 10-min incubation at room temperature. After adding 0.01mol/L citrate buffer, microwave retrieval was carried out for 20 min, and then 1 drop of 5% goat serum was added. Incubation at room temperature was continued for 5 min, followed by removal of the serum. Incubation at 4 °C overnight with rabbit anti-CD8 primary antibody (NBP2-29475, novusbio) commenced. Incubation at 37 °C for 1 h was conducted, followed by the addition of biotinylated goat anti-rabbit secondary antibody (1:500, ab150077, Abcam, USA) for a 30-min incubation at the same temperature. Then, a freshly prepared DAB chromogenic solution (Catalog No: DA1015, Tianjin Solabao Technology Co., Ltd, China) was added, and the staining process lasted for 1–2 min. Counterstaining was done with hematoxylin (Catalog No: G1080, Tianjin Solabao Technology Co., Ltd, China) for 1 min, followed by dehydration and transparency using a neutral mounting medium. Five representative high-power fields were chosen at random for observation under an optical microscope (Zong et al. 2021).

Statistical analysis

The data for this study were analyzed using SPSS software (version 21.0, IBM, United States). Mean ± SD was used to represent quantitative data. Two data sets were compared using the Unpaired Student's t-test, assuming a normal distribution. Employing one-way ANOVA, a comparative analysis of multiple groups was conducted, with Tukey's post hoc test utilized for subsequent analysis.