CircRNA amyloid precursor protein by competitive adsorption of microRNA-6838-5p mediates CDV3 expression to enhance malignant behavior and Warburg effect in Gastrointestinal Stromal Tumor

Highlights • Knockdown of circAPP inhibits GIST proliferation and the Warburg effect.• miR-6838-5p induces GIST proliferation and the Warburg effect.• CircAPP activates GIST proliferation and the Warburg effect by regulating the miR-6838-5p/CDV3 axis.


Introduction
Gastrointestinal Stromal Tumor (GIST) occurs mainly in the stomach and small intestine but also in the colon and rectum, esophagus, and other organs. 1 Mutations that activate Tyrosine Kinase (KIT) or plateletderived growth factor receptor α are reported to be the main molecular mechanism of GIST. 2 Currently, KIT inhibitor (imatinib) is the first-line chemotherapeutic drug for GIST treatment, but most GIST patients develop imatinib resistance. 3Although key aspects of the pathogenesis of GIST have been elucidated, the mechanism of oncogene overexpression and gene expression regulators in GIST remains unclear.Metabolic reprogramming is the earliest known cancer hallmark, which achieves tumor cell proliferation and survival by altering bioenergetic and biosynthetic pathways, and tumor cells exhibit a high rate of glycolysis, also known as the Warburg effect. 4Therefore, exploring the regulatory mechanism of the Warburg effect is likely to develop new strategies for GIST treatment.
CircRNAs are endogenous non-coding RNAs that differ from traditional linear RNAs in that they have covalently closed structures. 5ircRNA/microRNAs (miRNAs) interaction can regulate gene expression and participate in various physiological and pathological processes. 6urthermore, due to the high stability and conservation of circRNAs and their high abundance in body fluids, circRNAs have been considered potential biomarkers for GIST. 7CircRNA Amyloid Precursor Protein (cir-cAPP) is an abnormally highly expressed circRNA in GIST. 8In the preliminary experiments, it was also determined that circAPP was abnormally highly expressed in GIST.Therefore, this study focused on exploring the function and potential downstream molecular mechanisms of circAPP in GIST.
CircRNAs expressed in the cytoplasm contain miRNA binding sites to act as miRNA sponges, preventing miRNAs from interacting with mRNAs in the 3′Untranslated Region (UTR), thereby modifying miRNA downstream mRNAs, thereby affecting a series of cellular activities. 9iRNAs are capable of modifying post-transcriptional gene expression by recognizing and binding to the 3′-UTR of target gene mRNAs. 10Dysregulated miRNAs are key to many cancers including GIST, where they act as tumor suppressors or oncogenes. 11,12For example, miR-4510 promotes GIST cell invasion and metastasis 13 whereas miR-494 suppresses GIST development. 14miR-6838-5p has been confirmed to be involved in osteosarcoma, 15 gastric cancer, and renal cell carcinoma. 16,17However, no studies have demonstrated its actions in GIST.
Targeting to unearth the mechanism of circAPP in GIST, various studies were included, ultimately discovering the mediatory impacts of circAPP on proliferation, migration, and Warburg effect through miR-6838-5p targeting CDV3 in GIST, and hopefully replenishing the options for molecule-targeted therapy in GIST.

Clinical samples
Between April 2013 and July 2015, 52 pairs of GIST tissue samples and adjacent normal tissue (≥ 5 cm from cancer tissue) were obtained from the Affiliated Hospital of Inner Mongolia Minzu University.Inclusion criteria: 1) All the included subjects met the diagnostic criteria of GIST according to the criteria of "Chinese Expert Consensus on Diagnosis and Treatment of Gastrointestinal Mesenchymal Tumor (2017 Edition)" and "NIH2008 Modified Chinese Consensus 2017 Revised"; 2) Patients with GIST observed by abdominal CT or ultrasound endoscopy with tumor located in the stomach, and those who were diagnosed by pathology; 3) Clinical data of all the study subjects must be accurate and complete; 4) Patients not undergoing any radiotherapy or chemotherapy; 5) Patients with no history of other concomitant malignant tumors.With written informed consent, tissue sampling was done with the approval of the Ethics Committee of the Affiliated Hospital of Inner Mongolia Minzu University (n°201211Y62).After diagnosis by a pathologist, tissue specimens were frozen in liquid nitrogen and then stored at -80 °C to determine gene expression.All methods in the study were carried out in accordance with the ARRIVE guidelines.

Cell culture
The human gastric mucosal epithelial cell line GES-1 and the human GIST cell lines GIST-882 and GIST-T1 (Cosmo, Tokyo, Japan) were serially passaged for no more than 6 months.Cells were cultured in DMEM (Gibco; Thermo Fisher Scientific) in a humidified incubator at 37 °C with 5 % CO 2 .10 % heat-inactivated fetal bovine serum (Gibco; Thermo Fisher Scientific), 100 U/mL penicillin, and 100 mg/mL streptomycin were supplementary to the culture medium.The cell lines were all validated by STR profiles and tested negative for mycoplasma contamination.

RNA extraction and measurement
Under the guidelines of TRIzol (Invitrogen), total RNA was extracted.For reverse transcription, the PrimeScript RT kit (Takara, Shiga, Japan) was employed.Taking cDNA as a template, RNA quantification was done using specific primers (Table 1) and TB Green qPCR Master Mix (Takara) on a 480 II instrument (Roche Diagnostics, Basel, Switzerland).Expression levels of circRNA, mRNA, and miRNA were calculated by the 2 −ΔΔCt method and standardized to GAPDH or U6.

circRNA stability tests
For RNase R treatment, total RNA (2 μg) was cultured with 3 U/μg RNase R (Geneseed Biotech, Guangzhou, China) at 37°C for 10 min.Then, RNA purification was done with the RNeasy MinElute Cleanup kit (Qiagen, Hilden, Germany).For actinomycin D treatment, RNA was treated with 2 mg/mL actinomycin D (Sigma-Aldrich).Dimethyl sulfoxide was taken as a control.Finally, RNA expression levels were quantitatively analyzed. 18bcellular isolation Subcellular localization of circAPP in GIST-882 cells was analyzed.Nuclear and cytoplasmic fractions were isolated using the PARIS™ kit (Invitrogen), with U6 and 18S rRNA as positive references, respectively.19

Determination of colony formation
Cell samples from log phase cultures were transferred to 6-well plates at 500 cells/well and cultured for 14 days.Afterward, samples were fixed in 4 % paraformaldehyde and counted after 0.1 % crystal violet staining.

Determination of proliferation
GIST-882 and GES-1 cells were incubated with 30 μM EdU (KeyGen Biotech, Jiangsu, China) for 2h and treated with DAPI (Sigma-Aldrich) to stain the nuclear.Fluorescent images were taken to assess the EdUpositive rate. 18ble 1 Primers.

Apoptosis detection
Apoptotic cells were assessed with an annexin V-FITC/PI kit (Vazyme, Nanjing, China). 20Cells were resuspended and stained with Annexin V-FITC and PI solutions (5 μL each solution) in the dark.Apoptotic cells were detected on a FACSCanto II flow cytometer (BD Biosciences) to calculate the apoptosis rate.

Glucose consumption, lactate production, and ATP level measurements
Glucose consumption, lactate level, and ATP were determined by glucose assay kits (Sigma), lactate colorimetric/fluorometric kits (BioVision, CA, USA), and ATP assay kits (Thermo Fisher Scientific), respectively. 21P/ADP and NAD + /NADH detection ATP/ADP was measured using the ApoSENSOR ADP/ATP Ratio Assay Kit (K255-200, BioVision, CA, USA) and luminescence was measured by spectroscopy (Molecular Devices, CA USA).Cells (10 4 cells) were seeded onto luminometer plates, incubated with nucleotide release buffer, and added with 1 µL of ATP monitoring enzyme.Data were recorded at 1 min (Data A) and 10 min (Data B).ADP convertase was then added, and values were read (Data C).ATP=ADP Data A=Data C Data B.
NAD + /NADH was tested using the EnzyChrom™ NAD + /NADH Ratio Assay Kit (E2ND-100, Bioassay Systems, CA, USA).A total of 10 5 cells were homogenized and cultured with 100 mL NAD extraction buffer for NAD assay or 100 mL NADH extraction buffer for NADH assay.
Next, 20 µL of assay buffer was added to and 100 mL of back-extraction buffer was then supplemented.The mixture was then centrifuged, and the supernatant was transferred to the working reagent.OD 565 nm was read at 0 and 15 min.

RNA immunoprecipitation (RIP) experiment
The RIP assay was performed using the Imprint® RIP Kit (Sigma-Aldrich).GIST-882 cells (1×10 6 ) were lysed with RIP buffers and interacted with antibody-coated magnetic beads at 4°C overnight.Anti-IgG was the control for Anti-Ago2.After RNA isolation from the magnetic beads, analysis of RNA expression was carried out by RT-Qpcr. 22mor xenografts Twenty-four 4-weeks-old BALB/c nude mice (Vital River Laboratory Animal Co., Ltd., Beijing, China) were animal experiment subjects.The experiments were approved by the Institutional Animal Care and Use Committee of the Affiliated Hospital of Inner Mongolia Minzu University (n°201306YT2).All animal experiments complied with the ARRIVE guidelines.GIST-882 cells (1×10 6 ) stably transfected with si-NC and si-circAPP were subcutaneously inoculated into BALB/c nude mice (n = 6/group), and tumor volumes were measured weekly with vernier calipers from the second week after injection.Volume length × width2=2.Five weeks later, all mice were euthanized by cervical dislocation under isoflurane (5 %) treatment.Tumors were resected for immunohistochemistry with antibodies against HK2 (22029-1-AP, Proteintech), PKM2 (15822-1-AP, Proteintech), and Ki-67 (ab15580, Abcam). 23To determine the distant metastasis of tumors, GIST-882 cells (1×10 6 ) stably transfected with si-NC and si-circAPP were injected into nude mice via tail vein, and the nude mice were euthanized after 8 weeks to assess liver metastasis by HE-staining. 24ta analysis All statistical analyses were performed using GraphPad Prism 9.0 (GraphPad Software, CA, USA).Data represent mean ± Standard Deviation (SD).Two-tailed Student's t-test and one-way ANOVA determined statistical differences.Statistical significance represented p < 0.05.

CircAPP is upregulated in GIST
CircAPP has been revealed to be abnormally expressed in GIST. 8onsistent results were also obtained in this study, and circAPP expression in GIST tissues and cell lines was higher than that in normal tissues and cells (Fig. 1A, B).The circbase database discovered that circAPP is 678 bp in length and consisted of exons 4-7 of the APP gene located on chromosome 21q21.3(Fig. 1C).Examinations of the ring structure of circAPP revealed that RNase R treatment only decreased linear APP mRNA but not circAPP (Fig. 1D).After actinomycin D treatment, cir-cAPP had a longer half-life than linear APP mRNA (Fig. 1E).Subsequently, the location of circAPP in GIST-882 cells was examined by subcellular isolation assay, confirming circAPP mainly expressed in the cytoplasm (Fig. 1F).

Knockdown of circAPP inhibits GIST proliferation and the Warburg effect
Since GIST-882 has the highest expression level of circAPP, follow-up studies were performed using GIST-882.Focusing on the functions of cir-cAPP in GIST, si-circAPP was transected into GIST-882 cells which maintained circAPP expression in a suppressed state (Fig. 2A).Corresponding to the decrease in circAPP expression, the cloning ability and Edu-positive cell number of GIST-882 cells examined by colony formation assay (Fig. 2B) and EdU assay (Fig. 2C) were reduced, and the apoptosis rate tested by flow cytometry was augmented (Fig. 2D).For examining Warburg effect in GIST-882 cells, glucose consumption, lactate production, and ATP levels, along with ATP/ADP and NAD + /NADH were the tested indices.In circAPP-silenced GIST-882 cells, reduction in glucose consumption, lactate production, ATP, and ATP/ADP, and upregulation of NAD + /NADH were detectable (Fig. 2E, F).Moreover, key Warburg effect proteins HK2 and PKM2 were examined by western blot: after silencing circAPP, HK2 and PKM2 levels were reduced (Fig. 2G), further proving the anti-Warburg effect property of silencing circAPP in GIST.

CircAPP accelerates GIST tumor growth and metastasis
This study examined the effect of circAPP on GIST tumor growth and liver metastasis in vivo.As demonstrated in Fig. 7A-C, ablating circAPP suppressed GIST tumor volume and weight.Immunohistochemistry showed that circAPP knockdown suppressed HK2, PKM2, and Ki-67 in GIST tumors (Fig. 7D).Furthermore, knockdown of circAPP suppressed CDV3 expression in tumors (Fig. 7E).Liver metastasis of GIST tumors was assessed by HE-staining (Fig. 7F), manifesting that circAPP depletion suppressed the number of metastatic nodules in the liver.circAPP mediates GIST tumor growth and metastasis.

Discussion
GIST accounts for 1 %-3 % of gastrointestinal tumors. 25CircRNAs are imperative regulators in various physiological processes 26 and the occurrence of GIST. 27This study found that circAPP expression was augmented in GIST, and knocking down circAPP hampered GIST cell proliferation and the Warburg effect, triggered cell apoptosis, and limited GIST tumor growth and liver metastasis.
CircRNAs have become a research hotspot in recent years. 28As research progresses, the functions and mechanistic network of circRNAs in cancer have been implicated in tumor cell activities, 29,30 including Warburg effect. 31,32The Warburg effect, proposed by Dr. Otto Heinrich Warburg in 1920, refers to the shift of tumor cell metabolism from oxidative phosphorylation to aerobic glycolysis induced by mitochondrial respiration damage. 33The Warburg effect is a central contributor to the mechanisms of cancer progression, contributing to the cellular progression of cancer cells and being involved in immune responses and drug resistance. 34,35Therefore, inhibiting the Warburg effect inhibits cancer progression.This study confirms that circAPP was a positive regulator of the Warburg effect.Similar to a previous report, 8 this study also presented that circAPP was overexpressed in GIST.The in vitro results reported that downregulating circAPP inhibited GIST cell proliferation and the Warburg effect.Distant metastasis has been reported to be obstructive to cancer treatment. 36Therefore, this study further verified the suppressive impacts of silencing circAPP on GIST tumor growth and liver metastasis in vivo.miRNA sponging is the most well-studied mechanism of circRNA action, and circRNA competitively binds miRNA through complementary base pairing to inhibit the binding of miRNA to its target molecule, thereby regulating the expression of target Mrna. 37his study confirmed circAPP as a gene modifier of miR-6838-5p.miR-6838-5p is involved in a series of biological processes in cancer. 38miR-6838-5p is downregulated in osteosarcoma, and induction of miR-6838-5p suppresses tumor metastasis. 15In addition, miR-6838-5p is also involved in myocardial ischemia-reperfusion injury and cerebral hemorrhage injury. 39,40Here, this study implied that miR-6838-5p was lowly expressed in both GIST, and suppression of miR-838-5p mitigated the effect of silencing circAPP.This study further showed that, mechanistically, CDV3 was a target of miR-6838-5p.CDV3 was documented as an unidentified gene in breast cancer in 1999 41 and it is overexpressed in hepatocellular carcinoma. 42This study also observed the abundance of CDV3 in GIST, and inhibition of CDV3 impaired circAPP restorationmediated influences on GIST cell proliferation and the Warburg effect.
Although this study has partially elucidated the function of circAPP in GIST, further research is required in order to develop effective GIST treatments.On the one hand, due to the insufficient sample size, the correlation of circAPP with pathological T-staging and poor prognosis of GIST patients has not been established.On the other hand, the expression of most circRNAs is significantly correlated with GIST tumor size, mitotic number, and malignant degree by constructing circDNA expression profiles, and does not correlate with tumor site, 43 but no experiments have been performed to verify this idea.The exact role of this still needs to be further explored, and the molecular expression of the cir-cAPP/miR-6838-5p/CDV3 axis and whether its mechanism is affected by differences in primary sites is a direction for our future in-depth research.Despite its important role, miR-6838-5p is not the only target regulated by circAPP.Whether other miRNAs are involved in circAPP affecting Warburg effect of GIST and the related specific mechanism still needs to be further explored.

Conclusion
This study demonstrates that a novel circRNA, circAPP, is upregulated in GIST tissues and cells and that upregulating circAPP expression promotes the Warburg effect by targeting the miR-6838-5p/CDV3 axis,

Availability of data and materials
The datasets used and/or analyzed during the present study are available from the corresponding author upon reasonable request.

Consent for publication
Written informed consent for publication was obtained from all participants.

Ethical approval and consent to participate
All procedures performed in this study involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.All subjects were approved by the Affiliated Hospital of Inner Mongolia Minzu University (n°201211Y62).Written informed consent was obtained from each subject.
The animal experiments were complied with the ARRIVE guidelines and performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.The experiments were approved by the Institutional Animal Care and Use Committee of the Affiliated Hospital of Inner Mongolia Minzu University (n°2 01306YT2).

Fig. 1 .
Fig. 1.CircAPP is up-regulated in GIST.(A) CircAPP in GIST tissues and adjacent normal tissues, (B) CircAPP in GIST cell lines and normal cell lines, (C) Gene structure of circAPP on the circbase database, (D and E) RNaseR and actinomycin D treatment to explore the ring structure of circAPP, (F) Subcellular localization of circAPP; data are expressed as mean ± SD (n = 3).*p < 0.05.

Fig. 2 .
Fig. 2. Knockdown of circAPP inhibits GIST proliferation and the Warburg effect.si-circAPP was transfected into GIST-882 and GES-1 cells.(A) Expression changes of circAPP, (B and C) Colony formation and EdU assays to assess cell proliferation, (D) Flow cytometry to analyze apoptosis, (E) Cellular glucose consumption, lactate production, and ATP levels, (F) ATP/ADP ratio and NAD + /NADH ratio, (G) Western blot to evaluate cellular HK2 and PKM2 protein expression; data are expressed as mean ± SD (n = 3).*p < 0.05.