Pancreatic stellate cell‐derived exosomal tRF‐19‐PNR8YPJZ promotes proliferation and mobility of pancreatic cancer through AXIN2

Abstract The pancreatic stellate cells (PSCs) play an important role in the development of pancreatic cancer (PC) through mechanisms that remain unclear. Exosomes secreted from PSCs act as mediators for communication in PC. This study aimed to explore the role of PSC‐derived exosomal small RNAs derived from tRNAs (tDRs) in PC cells. Exosomes from PSCs were extracted and used to detect their effects on PC cell proliferation, migration and invasion. Exosomal tDRs profiling was performed to identify PSC‐derived exosomal tDRs. ISH and qRT‐PCR were used to examine the tRF‐19‐PNR8YPJZ levels and clinical value in clinical samples. The biological function of exosomal tRF‐19‐PNR8YPJZ was determined using the CCK‐8, clone formation, wound healing and transwell assays, subcutaneous tumour formation and lung metastatic models. The relationship between the selected exosomal tRF‐19‐PNR8YPJZ and AXIN2 was determined by RNA sequencing, luciferase reporter assay. PSC‐derived exosomes promoted the proliferation, migration, and invasion of PC cells. Novel and abundant tDRs are found to be differentially expressed in PANC‐1 cells after treatment with PSC‐derived exosomes, such as tRF‐19‐PNR8YPJZ. PC tissue samples showed markedly higher levels of tRF‐19‐PNR8YPJZ than normal controls. Patients with PC exhibiting high tRF‐19‐PNR8YPJZ expression had a highly lymph node invasion, metastasis, perineural invasion, advanced clinical stage and poor overall survival. Exosomal tRF‐19‐PNR8YPJZ from PSCs targeted AXIN2 in PC cells and decreased its expression, thus activating the Wnt pathway and promoting proliferation and metastasis. Exosomal tRF‐19‐PNR8YPJZ from PSCs promoted proliferation and metastasis in PC cells via AXIN2.


| INTRODUC TI ON
Pancreatic cancer (PC) is considered the 'king' among malignancies as its 5-year survival rate is less than 5%. 1,2 Additionally, early metastasis and rapid invasion commonly occur in patients with PC. 3 Although curative resection is the most common treatment for PC, only 15%-20% of patients benefit from it in the early stages. 4 Furthermore, most patients have PC recurrence within 2 years after resection, so understanding the molecular mechanisms involved in PC development may aid in diagnosis and treatment.
Stellate cells of the pancreas (PSCs) are stromal cells that contribute to the cancer microenvironment in PC tissue. 5 PSCs have been found to be involved in PC cell invasion and metastasis in several studies. 6,7 Exosomes, small membranous vesicles, are mediators secreted by PSCs that affect PC progression. Studies have shown that PSC-derived exosomes promote PC proliferation, migration, and drug resistance by delivering proteins, mRNA, and non-coding RNA, leading to changes in gene profiles and activation of oncogenic pathways. [8][9][10] It has been reported that small ncRNAs can be derived from the precursors and mature sequences of small RNAs derived from tRNAs (tDRs). 11,12 RNA polymerase III transcribes tRNAs, which are typically 76-90 nucleotides long in eukaryotes. It has been found that pre-tRNAs and mature tRNAs are extensively modified before and after they are exported to the cytoplasm in order to create two different kinds of can generally be divided into tRNA halves (tiRNAs) and tRNA-derived fragments (tRFs). 13,14 Exosomal tDRs have been used as biomarkers in liquid biopsies to differentiate cancer patients from healthy controls. 15 However, it is still unclear how tDRs regulate the biological properties of malignant PC.
In the present study, we detected the role of PSC-derived exosomal tDRs in PC. We demonstrated that PSCs secreted exosomal tRF-19-PNR8YPJZ and delivered them to PC cells, thereby enhancing proliferation and mobility via regulating AXIN2. Our present study suggests that exosomal tRF-19-PNR8YPJZ may be a potential biomarker and an effective target for the diagnosis and clinical therapy of PC.

| Clinical specimens and ethical statement
The present study used 80 PC and corresponding adjacent non-tumour tissues. All samples were collected prior to chemotherapy and radiotherapy in patients with PC. All samples were stored at −80°C before performing research. This study was approved by the ethical committee at Guizhou Medical University, which followed the Declaration of Helsinki. Informed consent was obtained from all participants.

| CCK-8 assay
A density of 4 × 10 3 AsPc-1 and PANC-1 cells per well, with eight parallel wells per group, was applied in 96-well plates. For exosome treatment group, exosome with concentration as 1 μg/mL was added. Various times (6,24,48,72 and 96 h) of culture in the incubator followed by a 1-h incubation in fresh medium containing 10% CCK-8 solution (Dojindo, Japan) was conducted. An automatic multifunctional enzyme labeller (Varioskan LUX, Thermo Fisher Scientific, USA) was used to measure absorption at 450 nm.

| Colony formation assay
Three parallel wells were plated with PANC-1 and AsPC-1 cells (400 cells/well). For exosome treatment group, exosome with concentration as 1 μg/mL was added. Cells were incubated at 37°C for 14 days, followed by 10 min of methanol fixation and 30 min of crystal violet staining (Solarbio, China). Finally, a camera was used to record the condition of cell colonies in per plate.

| Wound healing assay
Cells were grown in three parallel wells of a 6-well plate. After the cells reached 80%-90% confluency, sterile micropipettes were used to scrape off the cells to create a wound. Fresh medium was added after washing the cells with PBS to remove cell debris. For exosome treatment group, exosome with concentration as 1 μg/mL was added. Three fields of view were selected for each group and photographed to calculate the wound edge distance at 0 and 24 h after scraping.

| Transwell assay
Total 300 μL cell suspension with the density of 1 × 10 5 cells/mL was transferred to the upper chamber (Costar, USA) which pre-coated with 8% matrigel (Thermo Fisher Scientific, USA). Total 700 μL DMEM containing 10% FBS was set in the lower chambers. For exosome treatment group, exosome with concentration as 1 μg/mL was added in lower chambers. After 48 h, upper chambers were taken out and incubated with 4% paraformaldehyde for 15 min. Then, cotton swabs were used to remove the non-invasive cells, and then crystal violet was used to stain cells. Under a microscope (magnification times, 100×), three randomly selected views of transwell were recorded and invasive cells were counted.

| Isolation of exosomes
The Invitrogen Total Exosome Isolation Kit (ThermoFisher, USA) was used to isolate exosomes from PSCs culture medium. PSCs

| Transmission electron microscope (TEM)
Exosomes isolated from culture medium were examined using TEM.
Copper electron microscopy grids (Servicebio, China) coated with Formvar carbon were dipped into 1% glutaraldehyde once exosomes had been fixed in 1% glutaraldehyde. To stain the exosomes, 2% uranyl oxalate and PBS were added in sequence, followed by air-drying overnight. Finally, the images of exosomes were photographed using a TEM (Electronics Co. LTD, Japan).

| Small RNA-seq analysis
The manufacturer's instructions were followed for preparing a sequencing library to study tDRs expression, with NEXTflex Small RNAseq Kit v3. Raw sequencing data were cleaned by removing 5′ and 3′ adapters, filtering low-quality reads, merging identical reads and counting the unique reads. Only reads with 16-40 nt insertion were retained for further analysis. To identify tDRs, reads were first aligned to mature tRNAs and downstream sequences using BLAST. Only the reads perfectly matched to the genome were counted and classified into 5′-, 3′-, I′-and 3′-U of tRNA according to the positions from which tDRs were generated. To allow quantitative comparisons, the expression level of specific tDR was normalized to the total tDRs. Differentially expressed tDRs were identified using the R package Deseq2. Statistical significance was determined by log-fold change ≥1 and p < 0.05. The differentially expressed tDRs were presented using a heatmap.

| Dual-luciferase reporter gene assay
Wild-type and site-mutation AXIN2 luciferase reporter plasmids were constructed by GeneChem (Shanghai, China). Then, these plasmids were transfected into NC and tRF-19-PNR8YPJZ overexpression cells. Dual-Luciferase Reporter Gene Assay Kit from Promega (Madison, WI, USA) was used to measure luciferase activity based on the manufacturer's instructions.

| Quantitative real-time PCR
The total RNA was extracted with TRIzol (Invitrogen, Carlsbad, CA, USA). The cDNA was synthesized using the RiboSCRIPTTM Reverse Transcription Kit (Ruibo, China). Guangzhou Ruibo Biotechnology Co., Ltd. provided the primers used in current study. For qRT-PCR, SYBR green reagent (Ruibo, China) in combination with a microplate reader (BioTek, Epoch, USA) were used based on the manufacturer's instructions. 2 −∆∆Ct method was used to calculate the relative gene expression. Table S1 includes the primer sequences used for qRT-PCR.

| Western blotting
Bicinchoninic acid method was used to measure the protein con-

| Statistical analysis
An analysis of all results was performed using GraphPad Prism 9 (GraphPad Software, USA). To identify significant differences between two groups and multigroups, unpaired t-test and one-way analysis of variance were conducted. Results were presented as mean ± standard error of mean (SEM). The difference was considered significant if p < 0.05.

| Identification of prominent exosomal tsRNAs secreted from PSCs
To identification of the tDRs from PSC-derived exosomes involved in the progression of PC cells, small RNA-seq analysis was performed in PANC-1 cells with/without PSC-derived exosome treatment.

| The tRF-19-PNR8YPJZ is upregulated in PC
Based on ISH and RT-qPCR analysis in 80 human PC tissues and adjacent non-cancerous tissues, the expression of tRF-19-PNR8YPJZ in tumour tissues was higher than in normal tissues ( Figure 3A,B).
In 58 (72.5%) of the paired clinical samples, tRF-19-PNR8YPJZ was more abundant in tumour tissues than in non-cancerous tissues ( Figure 3C). tRF-19-PNR8YPJZ expression in PC patients was correlated with clinicopathological parameters through correlation analysis. Results showed that tRF-19-PNR8YPJZ expression was higher in PC tissues with regional lymph node invasion ( Figure 3D), metastasis ( Figure 3E), perineural invasion ( Figure 3F) and an advanced clinical stage ( Figure 3G). Age, gender and tumour size did not show any significant differences (Table 1). Additionally, we used the median values of expression levels of tRF-19-PNR8YPJZ as cut-off values for classifying PC tissues with high or low expression levels. tRF-19-PNR8YPJZ expression was associated with poor outcome for PC patients in Kaplan-Meier survival analysis ( Figure 3H). According to receiver operating curve analysis, the area under the curve was 0.8388, which indicates that tRF-19-PNR8YPJZ may be able to diagnose PC ( Figure 3I).

| PSC-derived exosomal tRF-19-PNR8YPJZ exhibits stimulative effect on PC proliferation and metastasis in vivo
To demonstrate the effect of exosomal tRF-19-PNR8YPJZ on PC phenotypes in vivo, we pre-treated tRF-19-PNR8YPJZknockdown and control cells with or without PSC-derived exosomes, and injected them into mice. Subcutaneous xenograft models and lung metastasis models were constructed. Results
Next, we transfected PC cells with wild-type and mutant AXIN2 fluorescent reporter plasmids based on their binding site genotypes ( Figure 6G). We found that tRF-19-PNR8YPJZ reduced the fluorescence intensity in PC cells transfected with AXIN2 dual-fluorescent reporter plasmids containing wild-type binding sites, but not in those containing mutant binding sites ( Figure 6H). In addition, the expression levels of AXIN2 in 80 paired PC tissues and adjacent tissues were determined using qRT-PCR, and the mRNA levels of ANXIN2 were found to be elevated in PC tissues ( Figure 6I). Furthermore, we found that low AXIN2 expression positively correlated with poor prognosis ( Figure 6J). Moreover, we found that tRF-19-PNR8YPJZ expression was negatively associated with AXIN2 expression

| PSC-derived exosomal tRF-19-PNR8YPJZ/AXIN2 axis promotes tumour progression
Then, we investigated whether AXIN2 played a role in the biological functions of tRF-19-PNR8YPJZ in PC cells. According to CCK-8 and colony formation assays, AXIN2 knockdown reversed the suppressive effects of anti-tRF-19-PNR8YPJZ on cell viability and colony formation ( Figure 7A,B). Additionally, anti-tRF-19-PNR8YPJZ inhibited migration and invasion of PC cells in wound healing and transwell assays, but AXIN2 knockdown reversed this inhibition ( Figure 7C,D).

| DISCUSS ION
Despite significant advancements in PC therapy, the prognosis of patients with PC remains poor. 16,17 Furthermore, most patients lost the optimal opportunity for therapy because of early metastasis. 18 Therefore, identification of novel biomarkers may help the therapy of PC.
Evidence showed that PSC can secrete exosomes to promote the progression of PC cells. Herein, consistent with previous studies, we our great concern, while its expression was significantly increased in the PC cells after treatment with PSC-derived exosomes.
tRNA-derived small RNAs (tDRs) are fragments of precursor or mature tRNAs that are usually 14-50 nucleotides (nt) in length. Previous studies indicated that tDRs can serve as diagnostic and prognostic biomarkers for cancers. For example, while 5′-tRF-GlyGCC expression was elevated in colon cancer tissues and blood samples, it can serve as prognostic biomarker. 20 Similarly, tDR-0009 and tDR-7336 was increased in cells under hypoxia, and predicted doxorubicin resistance in triple-negative breast cancer. 21 However, key tDRs for PC were still unknown. After reviewing of clinical characteristics, we found that the expression of exosomal tRF-19-PNR8YPJZ was also increased in patients with PC compared  roles. 23 It has also been observed that dysregulation of tDRs is associated with several cancers, including PC. [24][25][26] tRFs are part of tDRs and formed by the processing of tRNA by nucleases such as DICER and angiopoietin under certain conditions. 27,28 In addition to the simple transport function, tRFs can regulate gene expression through transcriptional, post-transcriptional and epigenetic approach. 29,30 It has been found that the 5′ tiRNA-His-GTG levels are elevated in colorectal cancer, whereas the levels of 5′ tiRNA-His-GTG are correlated with the size of the tumour. 31 In addition, tsRNA-MetCAT-37 and tsRNA-ValTAC-41 levels were elevated in PC tissues, which were used to distinguish PC tissues from adjacent non-tumour tissues. 32 In the current study, we found that PSC-derived exosomes increased the proliferation and metastasis of PC cells, while knockdown of tRF-19-PNR8YPJZ in PC cells relieved the stimulative effects of PSC-derived exosomes.
Therefore, our study identified that exosomal tRF-19-PNR8YPJZ was an oncogenic mediator in PSC-derived exosomes to promote PC cell proliferation and metastasis. AXIN2 is a homologous protein of the Axin family and has a high degree of structural and functional similarity with AXIN. 33 AXIN2, a negative inhibitor of the Wnt signalling pathway, regulates the phosphorylation and degradation of β-catenin, thus promoting the overexpression of target genes, affecting cell proliferation and differentiation, and promoting the occurrence of tumours. 34

CO N FLI C T O F I NTE R E S T S TATE M E NT
The authors declare that they have no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

CO N S E NT FO R PU B LI C ATI O N
All authors have agreed to publish this manuscript.