Tongue squamous cell carcinoma exosomal miR-122-5p promotes the conversion from fibroblasts to inflammatory cancer-associated fibroblasts

Abstract

Objectives

The aim of this study is to investigate whether tongue squamous cell carcinoma (TSCC)-derived exosomal microRNAs (miRNAs) promote the conversion to tumor-associated fibroblasts (CAFs) and to elucidate the underlying molecular mechanisms.

Methods

First, we isolated and characterised exosomes from the CAL27 tongue squamous cell carcinoma cell line. These exosomes were then co-cultured with human oral mucosal fibroblasts (hOMFs). We evaluated the expression of inflammatory cancer-associated fibroblasts (iCAFs)-related markers using Western blotting (WB), quantitative real-time PCR (qRT-PCR) and ELISA. We then conducted miRNA sequencing of the exosomes to identify the top five highly expressed miRNAs. WB was then employed to assess which of the corresponding miRNA mimics most effectively promotes the transformation of hOMFs into iCAFs. Finally, we predicted the potential target genes of the selected miRNA through bioinformatics analysis and validated them using a dual-luciferase reporter assay.

Results

Under transmission electron microscopy, the isolated exosomes exhibited a characteristic cup-shaped morphology with a double-membrane structure. Nanoparticle tracking analysis (NTA) revealed a size distribution ranging from 30 to 150 nm. Western blotting confirmed positive expression of the exosomal markers CD9, CD81, HSP90 and TSG101, while microRNA (miRNA) sequencing identified the most highly expressed miRNAs as miR-21-5p, miR-122-5p, miR-148a-3p, miR-143-3p and let-7i-5p. Of these, miR-122-5p induced the most significant changes in the expression levels of iCAF-related markers. Bioinformatic prediction combined with dual-luciferase reporter assays confirmed that GNPDA1 is a direct target of miR-122-5p.

Conclusion

CAL27-exo may facilitate the conversion of hOMFs into iCAFs by targeting glucosamine-6-phosphate deaminase 1 (GNPDA1) via exosomal miR-122-5p.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12672-026-04754-2.

Introduction

Oral squamous cell carcinoma (OSCC) is a highly malignant condition and the most prevalent malignancy of the head and neck region [1]. TSCC constitutes 40–50% of all oral cancers and is more prone to invasion and metastasis compared to other types of head and neck squamous cell carcinoma (HNSCC). The 5-year overall survival rate for patients with advanced TSCC is only 30–40% [2, 3]. Therefore, it is imperative that the molecular mechanisms of TSCC are elucidated and effective therapeutic strategies are developed.

Paget et al. proposed the “seed and soil” doctrine, highlighting the significance of the TME [4]. CAFs are a vital component of the TME [5]. There exists notable heterogeneity among CAFs, including inflammatory cancer-associated fibroblasts (iCAFs), myogenic cancer-associated fibroblasts (myCAFs) and antigen-presenting cancer-associated fibroblasts (apCAFs) found in pancreatic ductal adenocarcinoma [6–9]. CAFs are among the most critical cell subtypes of the tumour microenvironment (TME). As research in this area continues to expand, their role as potent drivers of tumour progression is becoming increasingly recognised [10–12]. Therefore, elucidating the mechanisms governing CAF generation is essential, as this may have important therapeutic implications.

Exosomes are extracellular vesicles characterized by a lipid bilayer membrane, ranging from 30 to 150 nm in diameter [13, 14]. They play a crucial role in tumor development [15], containing various biomolecules, including proteins, lipids, DNAs, mRNAs, and micro-RNAs (miRNAs) [16]. miRNAs are 17–24 nucleotide-long RNAs that mediate post-transcriptional gene silencing [17]. Exosomal miRNAs mediate intercellular communication within the tumor microenvironment, influencing processes such as tumor progression and metastasis [18, 19]. Wang et al. found that miR-122-5p inhibits CHMP3 through the MAPK signaling pathway, promoting invasiveness and epithelial-mesenchymal transition (EMT) in triple-negative breast cancer [20]. Ding et al. demonstrated that exosomal miR-105-5p, which is found in highly metastatic breast cancer cells, promotes the transformation of normal fibroblasts into cancer-associated fibroblasts (CAFs) by downregulating LATS2 and activating NF-κB signalling. This process also enhances epithelial-mesenchymal transition (EMT) in cancer cells [21]. The study by Khaloozadeh et al. demonstrated that exosomal miR-224, which is derived from colorectal cancer (CRC), promotes stromal reprogramming and angiogenesis by targeting PHLPP1 and PHLPP2. This activates Akt signalling in fibroblasts [22]. To further validate this transformation process, Yin et al. developed highly sensitive and specific tetrahedral DNA sensors for real-time imaging of exosomal miR-1247-3p derived from the highly metastatic human hepatocellular carcinoma cell line, HCCLM3, as well as FAP mRNA in fibroblasts. This approach provided visual evidence of tumor-derived exosome-induced conversion of fibroblasts into CAFs [23]. It has been reported that tumor-derived exosomal miRNAs promote the transformation of fibroblasts into cancer-associated fibroblasts (CAFs). Kewitz-Hempel et al. demonstrated that melanoma-derived extracellular vesicles (EVs) deliver miR-92b-3p to normal human dermal fibroblasts (NHDFs), promoting their conversion into cancer-associated fibroblasts (CAFs), including the inflammatory subtype (iCAFs) and myogenic cancer-associated fibroblasts (myCAFs), by targeting PTEN [24]. In lung adenocarcinoma, tumour-derived exosomal miR-1290, which is induced by COX-2, promotes cancer-associated fibroblast (CAF) activation and tumour progression via the CUL3/Nrf2 pathway. Notably, the activated CAFs in this study predominantly exhibited a myCAF phenotype [25]. However, it remains unclear whether TSCC-derived exosomal miRNAs influence the conversion of fibroblasts into CAFs and, if so, which CAF subtypes are induced and by what molecular mechanisms. Further investigation of this important question is warranted.

This study demonstrates that exosomes derived from tongue squamous cell carcinoma (TSCC) can transform human oral mucosal fibroblasts (hOMFs) into inflammatory cancer-associated fibroblasts (iCAFs). Further mechanistic investigations reveal that exosomal miR-122-5p plays a critical role in this process by targeting and regulating GNPDA1. This research aims to elucidate early mechanisms of CAF activation in TSCC, offering potential targets to disrupt the tumor-promoting microenvironment.

Methods

Culture of TSCC Cell Line and hOMFs

The CAL-27 TSCC cell line was obtained from Guangzhou Cellcook Biotech Co. Ltd., and hOMFs were sourced from Shenzhen Otwobiotech Co. respectively. Ltd. CAL-27 cells and hOMFs were both cultured under the same conditions: high-glucose DMEM (VivaCell, China) supplemented with 10% foetal bovine serum (VivaCell, China), at 37 °C in a 5% CO₂ atmosphere. Both cell types were passaged at a split ratio of 1:3. The hOMF medium was changed every two days, and only passages 3–6 were used for subsequent experiments.

Acquisition and identification of TSCC-derived exosomes

Once the cell density reached approximately 50%, the cells were washed three times with phosphate-buffered saline (PBS), then cultured in a media containing 10% exosome-free fetal bovine serum (VivaCell, China) for 24 h. After 24 h, the cell culture supernatant was collected. This were then centrifuged at 1000×g for 30 min, The resulting supernatant was transferred to a 3 kDa ultrafiltration tube (Millipore, USA) and further centrifuged at 5000×g for 30 min at 4 °C. TSCC-derived exosomes were extracted from the concentrated supernatant using the exoEasy Maxi Kit (QIAGEN, Germany), according to the manufacturer’s instructions. The morphology of the exosomes was observed using a transmission electron microscope (JEM-1400Flash, Japan). The number and size of the particles in the exosomal samples were measured using a Zetaview PMX110 (Particle Metrix, Germany) which was equipped with a 405 nm laser. Photographs were taken at a rate of 30 frames per second, lasting for 1 min. The movement of the particles was analyzed using the NTA software (ZetaView8.02.28, Germany).

Exosome tracing

PKH26-labelled exosomes were administered to the hOMFs once they had reached 50–60% confluence. The interaction between the exosomes and the fibroblasts was then monitored using a laser scanning confocal microscope (Nikon, Japan).

Western blotting

CAL-27 cells, CAL-27-exosomes, hOMFs, and treated hOMFs were lysed using RIPA buffer (Solarbio, China). The protein concentrations were quantified using a BCA Protein Assay Kit (Beyotime, China). Antibodies against CD9 (#ab263019), CD81 (#ab79559), TSG101 (#ab125011), FAP (#ab207178), and IL-6 (#ab9324) were obtained from Abcam (Wales, UK). Antibodies against HSP90 (#TA500494) were sourced from OriGene Technologies (Rockville, USA). The antibodies against GAPDH (380646) were purchased from ZEN-BIOSCIENCE (Chengdu, China). The antibodies against α-SMA (bsm-33178 M) were obtained from Bioss Antibodies (Beijing, China), and the antibodies against β-tubulin (M20005) were obtained from Abmart (Shanghai, China). The antibodies against GNPDA1 (12312-1-AP) were obtained from Proteintech (Wuhan, China). All antibodies were diluted with primary antibody diluent (Beyotime, P0273, China) according to the manufacturer’s instructions. The secondary antibody was purchased from ZSGB-BIO (Beijing, China) and diluted as specified in the manufacturer’s protocol. The intensity of the protein bands was quantified through optical density analysis using ImageJ.

Quantitative reverse transcription polymerase chain reaction (qRT-PCR)

Total RNA was isolated using the TaKaRa MiniBEST Universal RNA Extraction Kit (Takara, Japan) according to the manufacturer’s protocol. The isolated RNA was then reverse transcribed into cDNA using the PrimeScript™ RT Master Mix (Perfect Real Time) (Takara, Japan). qRT-PCR was then conducted using a QuantStudio™ Real-Time PCR System (Thermo Fisher Scientific, USA) with TB Green^® Premix Ex Taq™ II (Tli RNaseH Plus) (Takara, Japan) as the PCR master mix. Melting curve analysis was conducted to confirm amplification specificity, and primer amplification efficiencies were validated to be between 90% and 110%, with an R² value of ≥ 0.98 ensured prior to data analysis. Relative expression levels were normalized to the expression level of glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and calculated using the 2^(-ΔΔ)Ct method. For microRNA (miRNA) analysis, the miRNA was isolated using the miRcute miRNA Isolation Kit (TIANGEN Biotech, China) and reverse transcribed into complementary DNA (cDNA) using the Mir-X miRNA First-Strand Synthesis Kit (Takara, Japan). The corresponding qRT-PCR was then performed using the same system and the miRcute Plus miRNA qPCR Kit (TIANGEN Biotech, China) under the same conditions, with U6 small nuclear RNA (U6) as the endogenous control. Primers were synthesized by Sangon Biotech (Table 1).

Table 1.

List of primer sequences

Gene	Forward primer (5′–3′)	Reverse primer (5′–3′)
FAP	TTATGCTGGTCGCCTGTTGG	AGGAGACCACCAGAGAGCATA
α-SMA	AGCGTGGCTATTCCTTCGTT	TGAAGGATGGCTGGAACAGG
IL6	CACTGGTCTTTTGGAGTTTGAG	GGACTTTTGTACTCATCTGCAC
CCDC97	CAACTTGCTGCTCCAGTCCTACG	CTGTCCTCCTCCTCTTCCTCTTCC
FKBP5	CAGGCGGTGATTCAGTATGGGAAG	CCAGGTTCAGAAAGGCAGCAAGG
GNPDA1	CAAGGCATTTGCTCTGTACAAG	TTTGATACTGTACAAGGGGTCC
SESN2	ACCCAGAGAAGACCACCCGAAG	CCTCCAGGAGCAGCAAGTTCAC
SLC25A34	ATGGCGTTCGTTTCTACTGCTACAG	GTCTGGTGATTGTGCTGGTGTCC
SMARCD1	CACGAGCGGGAGTTTGTCATCTG	ATAGGTTCTGGTGGCATAAGCAAGG
SP2	GCTCACTCTGCCCGTCAACAAC	TCTCGATCAGCACCGTCTCCAC
GAPDH	GAAGGTCGGAGTCAACGGATTT	GCCATGGGTGGAATCATATTGG
MiR-21-5p	CCGCTCGTAGCTTATCAGACTG	*
MiR-122-5p	ACCGAGGTTGGAGTGTGACAA	*
MiR-148a-3p	AAGTTGCATCAGTGCACTACAGA	*
MiR-143-3p	CTCGACCGTGAGATGAAGCAC	*
let-7i-5p	TCGCGCATGAGGTAGTAGTTTG	*
U6	GGAACGATACAGAGAAGATTAGC	TGGAACGCTTCACGAATTTGCG

Proliferation assay

hOMFs were seeded at a density of 5 × 10³ cells per well in a 96-well plate. After an overnight starvation period, the cells were treated with 50 µg/mL of exosomes, 10 ng/mL of TGF-β1 (Peprotech, USA), or a negative control. Cell viability was assessed using the Cell Counting Kit-8 (GLPbio, USA) by measuring the absorbance at 450 nm using a microplate reader (Thermo, USA) at 24, 48, and 72 h after treatment. The same procedure was conducted for the transfected miRNA mimics (RiboBio, China) at the same time points.

Migration assay

hOMFs were seeded at a density of 1 × 10⁴ cells per well in a 12-well plate, with a Culture-Insert (Ibidi, Germany) placed in the centre of each well. The cells were treated under the same conditions as in the proliferation assay. After 24 h, scratches of approximately 500 μm wide were created by removing the insert. Images of the same positions were captured using an inverted microscope (Olympus CKX53, Japan) at 0, 12, and 24 h post-treatment. The width of the scratches was measured manually using ImageJ software. Two investigators conducted and analysed the experiment in a blinded manner, capturing images at consistent time intervals.

Exosomal miRNA sequencing

Total RNA was extracted from CAL-27 exosomes for use in preparing a microRNA (miRNA) library for sequencing. This was performed at Novagene (Beijing, China). The samples were sequenced on the Illumina NovaSeq 6000 platform.

RNA interference

MiRNA mimics (mimics-NC, miR-21-5p mimics, miR-122-5p mimics, miR-148a-3p mimics, miR-143-3p mimics and let-7i-5p mimics) were purchased from RiboBio (Guangzhou, China), while small interfering RNAs siRNAs (si-NC, siRNAs targeting GNPDA1) were obtained from Sangon Biotech Co. (Shanghai, China). The sequences of the miRNA mimics and siRNAs are listed in Tables 2 and 3, respectively. However, RiboBio did not provided the sequence of the mimics-NC. Transfection of the mimics and siRNAs were carried out using RiboFECT™ CP (RiboBio, China). Gene interference efficiency was evaluated using qRT-PCR.

Table 2.

List of miRNA mimics sequences

Gene	Sequences
MiR-21-5p	UAGCUUAUCAGACUGAUGUUGA
MiR-122-5p	UGGAGUGUGACAAUGGUGUUUG
MiR-148a-3p	UCAGUGCACUACAGAACUUUGU
MiR-143-3p	UGAGAUGAAGCACUGUAGCUC
let-7i-5p	UGAGGUAGUAGUUUGUGCUGUU

Table 3.

List of siRNAs sequences

Gene	Sense (5′–3′)	Antisense (5′–3′)
siR-GNPDA1-1	GAACAACUUCUUCAAGCACAUTT	AUGUGCUUGAAGAAGUUGUUCTT
siR-GNPDA1-2	GCUAGGUUCUUCGAUGGAGAATT	UUCUCCAUCGAAGAACCUAGCTT
siR-GNPDA1-3	GCUGAAAGUGAAGACUGUCAATT	UUGACAGUCUUCACUUUCAGCTT
siR-NC	UUCUCCGAACGUGUCACGUTT	ACGUGACACGUUCGGAGAATT

Luciferase miRNA target reporter assay

To identify the binding site between miR-122-5p and GNPDA1, cells were transfected with a luciferase construct containing either the wild-type or mutant GNPDA1 binding site, cotransfecting with miR-122-5p mimics or an empty vector. The luciferase vectors were constructed by Sangon Biotech Co. (Shanghai, China). The 3’ UTR of GNPDA1 mRNA containing either wild-type or mutant miR-122-5p binding sites was cloned using PsiCheck2. Following that, the cells were then co-transfected with psiCheck2-GNPDA1-wt/mut plasmids and either miR-122-5p mimics or mimics-NC using Lipofectamine 3000™ (Invitrogen, USA). Approximately 48 h later, the cells were harvested, and luciferase activity was measured using the Promega Dual-Luciferase system.

Immunofluorescence

This experiment used antibodies against FAP (Abcam, #AB207178), α-SMA (ServiceBio, GB13044), IL-6 (Abcam, #AB9324), and SCCA (ServiceBio, GB111989), along with the Opal Kit (AKOYA Biosciences, USA). FAP was visualized in red at a wavelength of 620 nm, IL-6 in green at 520 nm, SCCA in yellow at 690 nm, and α-SMA in yellow at 570 nm, in the specified order of operation. The experimental steps were performed according to the manufacturer’s protocol. Immunofluorescence (IF) was visualized under a fluorescence microscope (Nikon, Japan).

Statistical analysis

All experimental data were statistically analyzed using GraphPad Prism 8.0.2. The results are presented as the mean ± SEM. All experiments included three independent biological replicates, each comprising three technical repeats. Statistical significance between groups was determined using a Student’s t-test (for two groups) or one-way ANOVA (for multiple comparisons), with p < 0.05 being considered significant.

Results

Identification of exosomes secreted by TSCC

Exosomes were isolated from the supernatant of the CAL-27 cell line culture. Their morphology was observed using a transmission electron microscope (Fig. 1A). Exosomal particle size was analysed using nanoparticle tracking analysis (NTA) (Fig. 1B). The presence of the exosomal markers HSP90, TSG101, CD9, and CD81 was confirmed by WB(Fig. 1C). These experiments demonstrated that the extracellular vesicles exhibited lipid bilayer membrane structures with diameters ranging from 50 to 150 nm and confirmed the presence of typical exosomal markers. This indicates that they are indeed exosomes.

Fig. 1.

The CAL-27-exosomes promote the transformation of hOMFs to iCAFs. A–C Under transmission electron microscopy (TEM), CAL-27-derived exosomes exhibited the characteristic cup-shaped morphology with a double-layered membrane structure. Furthermore, these exosomes displayed a lipid bilayer membrane structure with diameters ranging from 50 to 150 nm, and positive expression of exosomal markers CD9, CD81, HSP90, and TSG101. D Immunofluorescence analysis shows that CAL-27 exosomes are taken up by hOMFs. E–H CAL-27 exosomes promote the expression of the iCAF markers FAP, α-SMA, and IL-6 in hOMFs, as demonstrated using by qRT-PCR, Western blotting, and ELISA. I CAL-27 exosomes enhance cell proliferation. J, K CAL-27 exosomes facilitate cell migration. (n = 3, ^nsp ≥ 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, one-way ANOVA)

Promotional impact of CAL-27-exosomes on the transformation of iCAFs

The exosomes were labelled with PKH26, and fluorescence staining revealed their primary localisation to be within the cytoplasm (Fig. 1D). qRT-PCR and Western blot analyses showed increased expression of the typical iCAF markers, including FAP, α-SMA, and IL-6, in hOMFs treated with CAL-27 exosomes (Fig. 1E, F, H). Concurrently, IL-6 expression levels in the cell supernatant, as determined by ELISA, correlated with the Western blot results (Fig. 1G). An interesting phenomenon was observed: IL-6 expression of at the mRNA level was inversely related to the protein level in the TGF-β1 positive control group. This suggests an underlying mechanism that warrants further investigation. Following the co-culture of CAL-27 exosomes with hOMFs, cell proliferation and migration were assessed using the CCK-8 and scratch assay. The results indicated that CAL-27 exosomes enhanced both the proliferation and migratory capacity of hOMFs (Fig. 1I, J, K). These results suggest that CAL-27 exosomes can activate the conversion of fibroblasts into iCAFs and promote their proliferation and migration.

Exosomal miR-122-5p promotes the transformation of iCAFs

Exosomes contain a variety of biomolecules, including DNA, RNA, proteins, and lipids. We hypothesized that exosomal miRNAs play a role in the activation of normal fibroblasts (Fig. 2A). Three CAL-27 exosome samples were selected for miRNA sequencing, revealing that the five most abundant miRNAs were miR-21-5p, miR-122-5p, miR-148a-3p, miR-143-3p, and let-7i-5p (Fig. 2B). Transfection efficiency was confirmed using fluorescence microscopy (Fig. 2C), followed by the detection of miRNA overexpression through qRT-PCR (Fig. 2D). Western blot analysis revealed that miR-122-5p was the most effective in upregulating the expression of FAP, α-SMA, and IL-6, playing a key role in the activation of iCAFs (Fig. 2E, F).

Fig. 2.

The high abundance of miR-122-5p in exosomes may promote the expression of FAP, α-SMA, and IL-6. A Schematic representation of TSCC-derived exosomes acting on fibroblasts. B MiRNA profile indicates the high abundance of miR-122-5p in CAL-27 exosomes. C Fluorescence microscopy results following transfection with miRNA mimics. D Evaluation of gene overexpression efficiency after transfection with mimics using qRT-PCR. E and F Western blot analysis demonstrating that miR-122-5p may promote the expression of FAP, α-SMA, and IL-6. (n = 3, nsp ≥ 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, a Student’s t-test or one-way ANOVA )

Furthermore, the expression of miR-122-5p was found to be increase after hOMFs internalized CAL-27 exosomes (Fig. 3A). IL-6 levels in cell culture media increased following transfection with mimics, which correlated with the Western blot results (Fig. 3B). Additionally, qRT-PCR revealed increased expression levels of FAP, α-SMA, and IL-6 (Fig. 3C). Interestingly, the IL-6 expression in the positive control group supplemented with TGF-β1 showed an inverse relationship between mRNA and protein levels, consistent with previous observations regarding CAL-27 exosomes. Furthermore, miR-122-5p was found to promote the proliferation and migration of iCAFs, as demonstrated by the CCK-8 and scratch assays (Fig. 3D, E, F).

Fig. 3.

MiR-122-5p may promote the transformation of hOMFs into iCAFs. A CAL-27 exosomes facilitate the transfer of miR-122-5p to hOMFs. B MiR-122-5p enhances IL-6 protein expression, as measured by ELISA. C MiR-122-5p promotes the gene expression of FAP, α-SMA, and IL-6, as determined by qRT-PCR. D MiR-122-5p promotes cell proliferation. E, F MiR-122-5p enhances cell migration. (n = 3, nsp ≥ 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, a Student’s t-test or one-way ANOVA )

GNPDA1 is a target gene of exosomal miR-122-5p

To identify the target genes of miR-122-5p, we used two bioinformatics databases (miRDB and TargetScan 8.0) to predict potential targets. We then intersected the resulting target genes were then intersected with transcriptome data from iCAFs derived from hOMFs treated with CAL-27 exosomes. This analysis revealed seven potential target genes: CCDC97, FKBP5, GNPDA1, SESN2, SLC25A34, SMARCD1, and SP2 (Fig. 4A).

Fig. 4.

GNPDA1 is a target gene of miR-122-5p. A Bioinformatics analysis was used to identify target genes for miR-122-5p. B qRT-PCR results demonstrating that miR-122-5p downregulates GNPDA1 gene expression. C Western blot analysis indicating that miR-122-5p downregulates GNPDA1 protein expression. D Schematic representation of the binding sites between miR-122-5p and GNPDA1. E Luciferase reporter assay results showing that the GNPDA1-WT group significantly downregulates luciferase activity, whereas the GNPDA1-MUT group has no effect on luciferase activity. (n = 3, nsp ≥ 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, a Student’s t-test or one-way ANOVA )

Subsequent gene expression analysis by qRT-PCR was conducted following transfection with miR-122-5p mimics. This analysis revealed that SLC25A34 and SP2 expression was increased, while CCDC97, FKBP5, GNPDA1, SESN2, and SMARCD1 expression was decreased. Notably, GNPDA1 exhibited the most significant downregulation (Fig. 4B). Furthermore, Western blot analysis confirmed the downregulation of the GNPDA1 protein, which was consistent with the qRT-PCR results (Fig. 4C).

Based on these findings, we speculate that GNPDA1 may be a potential target gene for miR-122-5p. We identified the potential binding site of miR-122-5p on GNPDA1 was identified by comparing the miR-122-5p sequence with the full-length GNPDA1 sequence (Fig. 4D). Furthermore, we confirmed that GNPDA1 is a direct target of miR-122-5p by conducting a luciferase miRNA target reporter assay (Fig. 4E).

The exosomes of TSCC may upregulate α-SMA and activate CAFs through the miR-122-5p/GNPDA1 signaling axis

The silencing effect of GNPDA1 was assessed after three different siRNAs were transfected. Both qRT-PCR and Western blot analyses confirmed the successful downregulation of GNPDA1 expression (Fig. 5A, B). The expression levels of FAP, α-SMA, and IL-6 were then measured using qRT-PCR and Western blot analysis. The results indicated that only α-SMA expression was upregulated, while the expressions of both FAP and IL-6 were downregulated (Fig. 5C, D, E). Therefore, we propose that miR-122-5p may upregulate the expression of α-SMA by targeting GNPDA1. Additionally, we hypothesize that miR-122-5p regulates FAP and IL-6 expression via other target genes, thereby promoting the conversion of hOMFs into iCAFs (Fig. 5F).

Fig. 5.

GNPDA1 can specifically upregulate the expression of α-SMA. A, B Successful downregulation of the GNPDA1 gene and protein expression following treatment with siRNA was confirmed by qRT-PCR and Western blot analysis, respectively. C qRT-PCR results showing that GNPDA1 specifically upregulates α-SMA gene expression. D, E Western blot analysis indicating that GNPDA1 specifically enhances the protein expression of α-SMA. F Schematic illustration depicting the mechanism by which exosomes from TSCC promote the transformation of hOMFs to iCAFs. (n = 3, nsp ≥ 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, one-way ANOVA )

iCAFs may exist in well-differentiated TSCC

Our experiments indicate that CAL-27 exosomes may facilitate the transformation from hOMFs to iCAFs, with the miR-122-5p/GNPDA1 signaling axis potentially being involved in this process. As the CAL-27 cell line is a well-established and well-differentiated model of tongue squamous cell carcinoma (TSCC), our next aim was to determine the clinical relevance of iCAFs in human tissue samples. To confirm the presence of iCAFs, we performed IF analysis on tumor tissues from 15 patients with well-differentiated TSCC. The results revealed the presence of iCAFs in the tumor tissues of 9 patients, while the tissues of 6 patients did not show this presence, resulting in a detection probability of 60% (Table 5). The schematic diagram of the IF results is presented in Fig. 6. These findings suggest that iCAFs may indeed exist in well-differentiated TSCC.

Table 5.

Statistics table of iCAFs in tumor tissues of patients with well-differentiated TSCC

Group	Number	Ratio
Patients with iCAFs	9	60%
Patients without iCAFs	6	40%
Total patients	15	*

Fig. 6.

iCAFs may exist in well-differentiated TSCC. IF analysis revealed the presence of iCAFs in tumor tissues from 9 out of 15 patients with well-differentiated TSCC, resulting in a detection rate of 60%

Discussion

In the tumor microenvironment of tongue squamous cell carcinoma, exosomal miR-122-5p, which is derived from cancer cells, is delivered to normal fibroblasts via exosomes. This promotes their transformation into cancer-associated fibroblasts (CAFs), particularly the inflammatory CAF (iCAF) subtype. This process is mediated by the targeting of GNPDA1 by miR-122-5p. The transformed fibroblasts exhibit significantly enhanced proliferative and migratory capacities. Overall, our study demonstrates that inhibiting the generation of iCAFs by targeting either miR-122-5p or GNPDA1 represents a promising therapeutic strategy. Our findings provide new insights into cell-to-cell communication within the tumor microenvironment between cancer cells and CAFs.

Several studies have demonstrated that exosomes derived from OSCC carry various substances, such as TGF-β1, miRNAs, and long non-coding RNAs (lncRNAs), which can facilitate the conversion of normal fibroblasts (NFs) to CAFs through multiple mechanisms. For instance, Huang et al. found that exosomes carrying TGF-β1 from HNSCC promote the conversion of NFs to CAFs [26]. Zhu et al. reported that downregulation of Caveolin-1 (CAV1) in fibroblasts, induced by the uptake of hypoxic HNSCC exosome-derived miR-192 and miR-215, activates TGF-β signaling, facilitating the conversion to CAFs and remodeling the hypoxic tumor microenvironment in HNSCC [27, 28]. Chen et al. identified that OSCC-derived tumor stem cells secrete exosomes containing mTOR, PI3K, STAT3, β-catenin, and miR-21-5p, which promote the conversion of NFs to CAFs and accelerate OSCC progression [28]. Ding et al. demonstrated that lncRNA signaling from OSCC stromal cells reprograms NFs to CAFs through lncRNA FLJ22447/IL33, thereby enhancing OSCC progression [29]. Collectively, these findings suggest that OSCC-derived exosomes can promote the conversion of NFs to CAFs through various pathways, thereby reshaping the tumor microenvironment and facilitating OSCC progression. This study focused specifically on the iCAF subtype. The selection of miR-122-5p as the key candidate was based on its high abundance in cancer-derived exosomes and its potent ability to increase the expression of iCAF-specific markers in normal fibroblasts. Our findings indicate that TSCC-derived exosomal miR-122-5p can upregulate the expression of FAP, α-SMA, and IL-6 in normal fibroblasts, promoting their conversion to iCAFs. This suggests that tumor-derived exosomal miR-122-5p are a contributing factor to the activation of iCAFs.

MicroRNAs (miRNAs) exert their functions by binding to target sites in a sequence-specific manner, leading to the suppression of target protein expression or its facilitation of degradation. There have been extensive reports on miRNAs promoting the transformation of normal fibroblasts into cancer-associated fibroblasts (CAFs) across various systemic cancers through interactions with their respective targets. In lung cancer research, Fan et al. reported that exosomal miR-210 from lung cancer cells may act as a pro-angiogenic factor in CAFs by regulating the JAK2/STAT3 pathway [30]. Wang et al. demonstrated that miR-122-5p promotes the proliferation, migration, and glycolysis of kidney cancer cells by targeting PKM2 [31]. Additionally, a study indicated that differential expression of miR-122-5p in oral cancer distinguishes between TNM stages 2 and 3, suggesting its potential as a biomarker for early diagnosis through saliva analysis in a highland mestizo Ecuadorian population [32]. We discovered that exosomal miR-122-5p plays a significant role in the conversion to iCAFs, as revealed by miRNA sequencing of three Cal27-exosome samples. Through bioinformatics analysis and exosome transcriptome sequencing, We identified the following genes as potential targets: CCDC97, FKBP5, GNPDA1, SESN2, SLC25A34, SMARCD1 and SP2. Subsequent quantitative reverse transcription PCR (qRT-PCR) analysis revealed that GNPDA1 exhibited the most significant downregulation, consistent with the mechanism of microRNA (miRNA)-mediated gene silencing.

GNPDA1 is involved in regulating the metabolic fluctuations of cytoplasmic UDP-GlcNAc, which influences hyaluronan synthesis during tissue remodeling [33]. Li et al. found that GNPDA1 promotes proliferation, migration, and invasion of hepatocellular carcinoma (HCC) cells while inhibiting apoptosis, suggesting its potential as a prognostic biomarker and therapeutic target for HCC [34]. GNPDA1 is known to influence cellular metabolism, primarily within metabolic signaling pathways. Our studies showed that knocking down GNPDA1 resulted in the upregulation of α-SMA in fibroblasts, while expression levels of FAP and IL-6 were downregulated. This indicates that GNPDA1 is a target gene of miR-122-5p; miR-122-5p inhibits GNPDA1 expression, thereby promoting α-SMA expression and activating NFs. However, our results also imply that there are additional target genes of miR-122-5p involved in regulating FAP and IL-6 expression, which in turn facilitates the transformation of NFs to iCAFs. Further investigation into these mechanisms is warranted.

Our study found that the detection rate of iCAFs in tumor tissues from 15 patients with well-differentiated TSCC was 60%. This suggests that iCAFs may indeed be present in well-differentiated TSCC, supporting the validity of our earlier experiments. However, due to the limited research on CAF subtypes in IF results demonstrated that iCAFs positively expressed FAP, α-SMA, and IL-6. Our findings also revealed an intriguing phenomenon: TGF-β1 stimulates hOMFs, leading to downregulation of IL-6 mRNA as confirmed by qRT-PCR, while IL-6 protein levels were upregulated based on Western blot and ELISA analyses. This discrepancy has been observed in our experiments, and the underlying mechanism warrants further investigation.

Conclusion

Our study suggests that exosomal miR-122-5p derived from TSCC could facilitate the transformation of hOMFs into iCAFs, with GNPDA1 suggested as a downstream target gene. We propose that targeting the miR-122-5p/GNPDA1 signaling axis may suppress TSCC progression by modulating the transformation of CAFs, representing a potential therapeutic strategy.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Bin Xia and Jiamin Wu: contributed to manuscript writing, conducted literature searches, and participated in data cleaning, processing, and data visualization. Hongrong Zhang: contributed to data processing and assisted in interpreting the results of data analysis, participated in manuscript revisions and proofreading. Hefeng Yang and Biao Xu: provided major contributions, including project design, determination of research direction, supervised data analysis, and oversaw the entire research process.

Funding

These studies were supported by Yunnan Provincial Department of Science and Technology, Kunming Medical University, Joint Special Project for Basic Research (202401AY070001-202).

Data availability

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.