Identification of tRF-29-79MP9P9NH525 as a biomarker and tumor suppressor of gastric cancer via regulating KIF14/AKT pathway

Abstract

Gastric cancer (GC) is one of the most common malignancies with a poor prognosis. The development of novel biomarkers is of utmost importance to screen patients with GC. Molecular mechanism study of GC may provide a research basis for the development of targeted drugs. We identified tRF-29-79MP9P9NH525 (tRF-29) as a GC-associated tRNA-derived fragment (tRF). The specific hair-pin structure reverse primer and amplification primers were first designed and then applied for tRF-29 quantification. Receiver operator characteristic curve, Kaplan-Meier survival curve, and multivariate Cox analysis were applied to analyze the diagnostic and prognostic values of tRF-29 in GC. Ethynyl-2'-deoxyuridine, cell cloning, Transwell assay, and flow cytometry were used to detect the effects of tRF-29 on proliferation, migration, and cell cycle distribution of GC cells. Xenograft tumor formation in NOD-SCID mice was applied in determining tRF-29's effects on tumor growth. Fluorescence in situ hybridization, dual luciferase reporter assay, Western blot, immunohistochemistry, and RNA-binding protein immunoprecipitation were conducted to explore the molecular mechanism underlying tRF-29 regulating GC development. It was found that tissue tRF-29 showed effective diagnostic efficiency in GC and could discriminate different gastric mucosa. Besides, plasma tRF-29 improved GC diagnostic values of common tumor markers and had prognostic values in GC. tRF-29 was found to suppress proliferation and cell cycle progression. tRF-29 inhibited the growth of xenograft tumors. Mechanically, tRF-29 exerted Kinesin family member 14 (KIF14) mRNA destabilization by combining with argonaute 2 (Ago2) and regulated AKT/P27 pathway. In conclusion, tRF-29 inhibited GC progression by combining with Ago2 and regulated AKT/P27 pathway by silencing KIF14 expression. In normal cells, tRF-29, derived from tRNA-ValACC, targets the 3'UTR region of KIF14 mRNA by forming RNA silencing complex with Ago2. Reduced KIF14 results in less phospholation of AKT. Subsequently, the expression of P27 is increased, while the expression of MMP-2 is decreased. Finally, the cell cycle is arrested, and the cell proliferation is suppressed, as well as the metastasis is inhibited. In gastric cancer cells, due to the downregulated of tRF-29, the expression of KIF14 is increased, thus the cell proliferation and metastasis are promoted via AKT pathway.

In normal cells, tRF-29, derived from tRNA-ValACC, targets the 3’UTR region of KIF14 mRNA by forming RNA silencing complex with Ago2. Reduced KIF14 results in less phospholation of AKT. Subsequently, the expression of P27 is increased, while the expression of MMP-2 is decreased. Finally, the cell cycle is arrested, and the cell proliferation is suppressed, as well as the metastasis is inhibited. In gastric cancer cells, due to the downregulated of tRF-29, the expression of KIF14 is increased, thus the cell proliferation and metastasis are promoted via AKT pathway.

Fig. 1. Flow chart of the study design.

ROC receiver operator characteristic.

Fig. 2. The abundance and clinical values of tRF-29-79MP9P9NH525 (tRF-29) in various samples.

A tRF-29 was downregulated in advanced gastric cancer (AGC) tissues than in adjacent normal tissues. Training set, n = 53. B tRF-29 also showed a reduced level in enlarged AGC group. Validation set, n = 119. C tRF-29 level showed a gradient change from HCs group to AGC group. Healthy controls (HCs): n = 47, Benign lesion (BL): n = 47, Precancerous lesion (PL): n = 48, Early gastric cancer (EGC): n = 46, Advanced gastric cancer (AGC): n = 119. D tRF-29 abundance was observed in plasma samples from pre-operative (Pre-op), post-operative (Post-op) gastric cancer patients and healthy controls. Healthy: n = 88; Pre-op: n = 84; Post-op: n = 84. E Receiver operator characteristic (ROC) curve of tRF-29. F–I The joint ROC curves of tRF-29 and carcinoembryonic antigen (CEA), carbohydrate antigen 125 (CA125), CA19-9, and α-fetoprotein (AFP), respectively. J The joint ROC analysis of tRF-29, CEA, CA125, CA19-9, and AFP. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; ns no significance.

Fig. 3. The effects of tRF-29-79MP9P9NH525 (tRF-29) on the growth and metastasis of gastric cancer (GC) in vitro and in vivo.

A The effects of tRF-29 on cell cycle distribution of GC were determined by flow cytometry. n = 3. B tRF-29 demonstrated inhibitory effects on the migratory and invasive viability of AGS and HGC-27 cells. Conversely, the knockdown of tRF-29 resulted in the opposite effects. n = 3. C HGC-27 cells with different concentrations of tRF-29-agomir were hypodermically injected into the right axilla of mice. D Tumors obtained from mice after euthanized. E Tumor volume change curve. n = 7. F Tumor weight after being euthanized. n = 7. G The body weight of the mice. n = 7. Comparing with NC, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.01; ns no significance.

Fig. 4. Exploring potential targeted genes of tRF-29-79MP9P9NH525 (tRF-29).

A, B qRT-PCR results of nucleoplasmic separation (A) and FISH assay (B) identified the distribution of tRF-29 in gastric cancer cells. C, D The heatmap (C) and volcano plots (D) of the RNA sequence results in AGS cells treated by tRF-29 mimics. E The overlapped genes that were predicted by Miranda & TargetScan, and the downregualted genes in AGS cells treated by tRF-29 mimics. F The bubble diagram exhibited the biological process classification of tRF-29’s targeted genes. G The expression level of aquaporin 11 (AQP11), kinesin family member 14 (KIF14), sphingosine-1-phosphate receptor 3 (S1PR3), and thrombospondin 1 (THBS1) were detected in gastric cancer cells with tRF-29 upregulation or downregulation. n = 3; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns no significance.

Fig. 5. tRF-29-79MP9P9NH525 (tRF-29) regulates AKT pathway via silencing KIF14 expression.

A The expression level of KIF14 in 413 cases of gastric cancer and 35 cases of healthy controls (Analyzed data from TCGA). B The expression level of KIF14 (−ΔC_q) in paired advanced gastric cancer tissues and adjacent normal tissues was detected by qRT-PCR. n = 64. C KIF14 is upregulated in gastric cancer AGS and HGC-27 cells compared with GES-1 cells. n = 3. D The co-location of tRF-29 and KIF14 mRNA was detected by FISH. E The results of luciferase activity assay. n = 4. F tRF-29 overexpression reduced the protein level of KIF14, p-AKT(S473), and MMP-2, and increased the protein level of P27; while the protein level of AKT was not significantly changed. tRF-29 downregulation showed the reversed effects. n = 3. *P < 0.05, **P < 0.01, ***P < 0.001; ns no significance.

Fig. 6. tRF-29-79MP9P9NH525 (tRF-29) silences KIF14 via incorporating with Ago2.

A Immunohistochemistry (IHC) results of the protein levels of KIF14, Ki-67, P27, MMP-2, and Ago2 in tumor samples obtained from xenograft euthanized mice. n = 3. B tRF-29 and KIF14 mRNA were significantly enriched in Ago2-RIP group compared to IgG-RIP group. n = 3. C The binding between tRF-29 and the Ago2 protein. PIWI and Piwi-Argonaut-Zwille (PAZ) domains of Ago2 were two binding domains predicted by the online database HDOCK (http://hdock.phys.hust.edu.cn/). *P < 0.05, **P < 0.01, ***P < 0.001; ns no significance.

Fig. 7. Reduced KIF14 expression reversed the effects of tRF-29-79MP9P9NH525 (tRF-29) on the proliferation of gastric cancer.

A The downregulated effect of si-KIF14 was determined by qRT-PCR. n = 3. B EdU assay revealed the rescue effects of KIF14 knockdown in cells treated by tRF-29 inhibitor. n = 3. C Reduced KIF14 expression trapped more cells in G0/G1, which was reduced by tRF-29 inhibitor. n = 3. D The activated p-AKT(S473) pathway by tRF-29 inhibitor was restrained by KIF14 downregulation. n = 3. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; ns no significance.