YARS as an oncogenic protein that promotes gastric cancer progression through activating PI3K-Akt signaling

Gastric cancer (GC) is one of the most prevalent malignancies around the world. With the rapid development of targeted therapy and immunotherapy, diagnosis and treatment of multiple types of cancer have been greatly improved, yet due to limited drug targets, as well as high heterogeneity, gastric cancer remains to be a major health problem (Siegel et al. 2017). Currently, HER2 (ERBB2)-targeted antibody Trastuzumab is the only first-line option for GC’s targeted treatment (Van Cutsem et al. 2016); for immunotherapy, the response rate for PD-1/PD-L1-based regimens is limited in GC population, while the choice of immune checkpoint inhibitors’ combination with chemotherapy or targeted agents were controversial (Lordick and Janjigian 2016). Therefore, uncovering novel druggable targets and mechanisms of gastric cancer progression is of urgent need for the development of precision medicine.

As housekeeping proteins prevalently found in all living organisms, members of the aaRS (aminoacyl-tRNA synthetase) family catalyze the binding of amino acids to tRNAs and translate the coding information from nucleic acids into amino acids, exerting fundamental roles in protein synthesis (Carter 1993). During evolution, tRNA synthetases incorporated multiple domains that expanded their functions beyond aminoacylation, allowing them to participate in diverse biological events, such as nuclear tRNA export, mitochondria RNA splicing, transcriptional and translational control, RNA maturation and retrovirus packaging (Cen et al. 2001; Ko et al. 2000, 2001; Liu et al. 2002; Sarkar et al. 1999). Among all aaRSs, YARS (tyrosyl-tRNA synthetase, aka TyrRS or YRS), the member that promotes the combination of tyrosine to its cognate tRNA, was recently potentiated to regulate intracellular signaling (Sun et al. 2017). Mutations of YARS have been prevalently reported to be associated with genetic diseases, including an autosomal dominant form of Charcot-Marie-Tooth (CMT) neuropathy (Fuchs et al. 2019), as well as a recently reported autosomal recessive multi-organ disease (characterized by failure to thrive, hypertriglyceridemia, developmental delay, and abnormalities/disorders in multiple organs) (Nowaczyk et al. 2017; Tracewska-Siemiatkowska et al. 2017).

On the other hand, the upregulation of aaRSs, including YARS, was observed in several types of cancer (Guo et al. 2010; Hsu et al. 2019), while polymorphisms in aaRS genes were reported to be associated with breast cancer risk (He et al. 2015). Emerging studies also implicated the involvement of aaRSs with multiple pathway networks through constituting the multi-tRNA synthetase complex (MSC) (Hyeon et al. 2019) and raised the possibility of aminoacyl-tRNA synthetases as therapeutic targets against autoimmune diseases, rare diseases, and even cancer (Kwon et al. 2019). Nevertheless, although abundantly exists and functions in all organisms, YARS’s actual linkage with cancer remains unspecified. In this study, by analyzing the omics datasets and performing experimental study with patient-derived specimens, cell lines, as well as animal models, we explored the functions and relevant mechanisms of YARS in gastric cancer.

As a housekeeping gene that facilitates tyrosyl aminoacylation, YARS is traditionally considered playing a fundamental role in maintaining basic biological activities. To date, although mutations of YARS were reported correlated with neuropathy or development disorders, YARS’s association with cancer has not yet been characterized. In this study, we explored YARS’s functions in gastric cancer by applying bioinformatics analysis and wet lab experiments. We discovered that both YARS transcript and protein were highly expressed in GC specimens, which was correlated with poor prognosis. Through analyzing GC datasets, we revealed YARS’s co-expression with EGFR amplification, specific mutations, tumor mutation burden, EBV/MSI phenotypes, as well as multiple gene sets potentially those were enriched by YARS high expression. By referring to cell line-based RNA-sequencing, we focused on PI3K-Akt signaling and further validated its association with YARS. YARS elicited multiple malignant phenotypes through activating PI3K-Akt signaling, while the YARS-induced homologous recombination and insensitivity to PARP inhibitors also depended on PI3K-Akt. To our knowledge, it is the first report to unveil the malignant roles and potential applications of aminoacyl-tRNA synthetase in cancer.

As discovered in patients carrying CMT neuropathy, missense mutations and deletions of YARS (such as Gly41Arg, Asp81Ile, Glu196Lys, Glu196Gln and 153_156del) lead to a loss of its aminoacylation activity and subsequently reduced cell growth (Gonzaga-Jauregui et al. 2015; Jordanova et al. 2006; Schabhuttl et al. 2014). Apart from mutations of YARS, loss-of-function variants of other types of aaRS such as AARS (alanyl-tRNA synthetase), HARS (histidyl-tRNA synthetase), KARS (lysyl-tRNA synthetase), and MARS (methionyl-tRNA synthetase) were also reported to be related with Charcot–Marie–Tooth disease and other types of neuropathy (Abbott et al. 2018; Gonzalez et al. 2013; McLaughlin et al. 2010, 2012; Vester et al. 2013). Interestingly, despite the fact that aaRSs were basic participants of aminoacylation and protein synthesis as suggested by limited functional studies in yeast/worm/Drosophila models, aaRS mutation-related neuropathogenesis was more likely to be caused by inducing neuronal toxicity and synaptic degeneration, rather than by impaired canonical aminoacylation activity. Nevertheless, the mutation (10/440, 2.27%) and amplification (6/440, 1.36%) of YARS were relatively rare in gastric cancer, hence our study majorly focuses on the malignant function of YARS’s expressional changes. Because our limited cases of IHC indicated that YARS was positively expressed in both gastric normal and tumor tissues, the threshold of positiveness should be specified for potential clinical applications.

As a crucial signaling node stimulating multiple pathways and transcriptional factors (such as PI3K-Akt, RAS-RAF-MEK-ERK, PLC-γ, Src, and JAK-STAT), EGFR (epidermal growth factor receptor) was prevalently mutated in lung cancer, while its genetic amplification and overexpression were also major causative events for gastrointestinal cancers (Du and Lovly 2018; Yarden and Sliwkowski 2001). EGFR-targeted regimens (tyrosine kinase inhibitors and antibodies) have been proved as effective therapeutic options against lung cancer and colorectal cancer, while also displaying prospective applications in gastric cancer (Liu et al. 2018). We discovered that YARS was highly expressed in EGFR-amplified patients. Overexpressing EGFR upregulated, while repressing STAT1 downregulated the protein level of cellular YARS, giving hint that EGFR might induce YARS expression through stimulating its downstream transcription factors (such as STAT1). Yet, the specific regulatory mechanisms demand further exploration and whether EGFR-mediated YARS upregulation influences the sensitivity or resistance to EGFR-targeted therapy remains to be answered.

Our work exhibited that PI3K-Akt was an intermediate pathway for YARS-induced phenotypes. In addition to PI3K-Akt, we demonstrated in GC cells that YARS enhanced homologous recombination (HR), a major pathway that repairs DNA double strand breaks (DSBs) in mammalian cells (Scully et al. 2019). Genomic instability and DNA breaks are representative traits in carcinogenesis and malignant progression. Since these highly genotoxic events commit cells to apoptosis, multiple mechanisms are vigorously activated in cancer to repair DNA lesions and to maintain cancer survival (O'Kane et al. 2017). Consequently, inhibitors targeting PARP (poly (ADP-ribose) polymerase) are applied as anticancer agents by disrupting PARP-mediated DNA single strand breaks (SSBs) repair. Due to that DNA SSBs can be converted to DSBs, PARP inhibitors induce a synthetic lethality and achieve a hypersensitivity in patients harboring HR deficiency (such as deletions, loss-of-function mutations or low expressions of BRCA1/2, ATM, RAD51 and other components of HR machinery) or simultaneously accepting HR inhibition (Hoppe et al. 2018; Noordermeer and van Attikum 2019). We demonstrated that depleting YARS enhanced GC cells’ sensitivity to PARP inhibitors, while YARS-directed HR and impaired PARP-i sensitivity relies on the activation of PI3K-Akt signaling. Thus, it is reasonable to believe that besides combing PI3K-Akt inhibitors (Konstantinopoulos et al. 2019), PARP-targeted drugs combining YARS inhibitors might also be a potential therapeutic option against GC.

Several studies have also addressed aaRS’s association with regulative pathways and processes. A predictive work raised the potential interaction between aaRSs (including LRS, DRS, RRS, IRS, KRS, QRS, EPRS) and MAPK/PI3K-Akt pathways (Hyeon et al. 2019). YARS was reported directly interacting with TRIM28/HDAC1 and sequestering HDAC1-mediated deacetylation on E2F1, thus enhances E2F1-mediated upregulation of homologous recombination (HR) factors and provides protection against DNA damage (Wei et al. 2014). On the other hand, aaRSs were suggested as essential mediators of Myc-directed growth control (Zirin et al. 2019), which was in concert with our findings. Besides PI3K-Akt related (PI3K-Akt/mTOR), E2F1, HR/DNA repair and Myc-targets, G2/M checkpoints associated signaling was also predicted to be positively enriched in YARS high-expression groups. Hence, YARS’s association with cell cycle control and other additional carcinogenic pathways also deserves further investigation. Besides YARS, the upregulation of several aminoacyl-tRNA synthetases has also been observed in GC and multiple other types of malignancy (Hu et al. 2013), which might be due to the increased carcinogenic metabolism and requirements for protein translation. Although we have demonstrated that YARS induced multiple signaling activation and progression of GC, the malignant roles of other aaRSs were unexplored. Also, it remains unclear whether YARS exerted its malignant functions through canonically enhancing expressions of relevant pathway members or through non-canonical routes beyond aminoacylation. Notably, our data in HGC-27 and AGS cells showed that YARS positively regulated S6, Akt, Erk, and mTOR phosphorylation instead of protein expression, while directly lifted protein levels of HR-related molecules. Considering no evidences supported the notion that aaRSs harboring kinase activity, we inferred that YARS’s impact on phosphorylation might be done through modulating the upstream kinase/phosphatases of PI3K-Akt pathway, yet the mechanistic details remain to be elucidated.

In contrast to bacterial or low-eukaryotic homologs, high-eukaryotic YARS (including human) specifically contains a C-terminal domain and can be split by proteolysis into two distinct cytokine mimics: an IL-8 (interleukin 8)-like N-domain derivative and an EMAP II (endothelial-monocyte-activating polypeptide II)-like C-domain derivative (Wakasugi and Schimmel 1999). IL-8 causes the recruitment of inflammatory cells and is also implicated in mediating cancerous progression (Chao et al. 2019). By acting on the CXCR1/2 receptor, the IL-8-like fragment (aka “mini-YARS”) induces cell migration as a monomer or inhibits migration as a dimer (Vo et al. 2011); mini-YARS was also reported potentially leading to angiogenesis by activating VEGFR2 or VEGF (Tzima et al. 2005; Zeng et al. 2014). IL-8 has been characterized to activate PI3K-Akt signaling through modulating phosphorylation/activation of Akt and S6 (MacManus et al. 2007), while PI3K-Akt-mTOR signaling was also reported to enhance IL-8 production as feedback (Lin et al. 2019), which comply with our findings in gastric cancer. On the contrary to IL-8, EMAP II is a pro-inflammatory cytokine that exerts anti-endothelial and anti-angiogenic activities through binding to VEGF receptors and disrupting fibronectin and VEGF signaling (Awasthi et al. 2013). As a consequence, YARS’s role in controlling malignant progression might be related with its cytokine-releasing capabilities. Nonetheless, we observed in GC datasets that IL-8-related signaling was insignificantly correlated with YARS, while angiogenesis-related signaling was negatively enriched in patients harboring high-YARS level (Supplementary Fig. S5). These might be due to that the actual levels of YARS-derived mimics could not be appropriately represented by value of total-YARS generated by RNA-sequencing or proteomics study or due to the functional diversities between YARS-derived cytokine mimics and authentic IL-8 or EMAP II. Since no antibodies targeting mini-YARS or EMAP II-like domains were commercially applicable, the production, release, functional details, as well as the respective proportions of YARS-derived IL-8 and EMAP II mimics in GC remained unexplored in this study. Additionally, since the production of pro-inflammatory cytokines (i.e., IL-8, IL-6 and TNF-α) is known to be mediated by NF-κB signaling (Tak and Firestein 2001), whether NF-κB promotes YARS cleavage and mini-YARS release also deserved future investigation.

Furthermore, YARS displayed a trend of high expression in high-TMB, MSI-H, and EBV subtypes of GC. Since high-TMB, MSI, and EBV patients have been proved more likely to benefit from immunotherapy (Le et al. 2015), the linkage between YARS, microenvironment, and cancer immune responses merits future investigation. Considering that alterations of DNA damage repair pathway result in increased TMB and neoantigen loads (Park et al. 2019; Wang et al. 2018), we hypothesized that YARS-mediated HR upregulation might serve as a feedback event of mutations or viruses infections.

In conclusion, our work demonstrated that YARS functions in gastric cancer beyond a fundamental gene. Through activating PI3K-Akt signaling, YARS promotes malignant progression and insensitizes PARP inhibitors. Our work shed light upon the novel functions of housekeeping proteins, and also proposed the carcinogenic involvement and druggable applications of the aminoacyl-tRNA synthetase family.