Despite of the substantial improvements in medical technology in the past decades, there is still a rapid growth in global cancer incidence and mortality. In general population worldwide, gastric cancer (GC) ranks the fifth in terms of newly diagnosed cases, while it accounts for 7.7% of cancer deaths in 2020, which brings a huge burden for individuals and social welfare system (Sung et al., 2021). The pathophysiology of GC is multifactorial and the risk factors include but not limited to long term of Helicobacter pylori infection, smoking, obesity, imbalanced diet such as processed meat or high-salt intake and insufficient intake of fruit and vegetables (Smyth et al., 2020). The main therapeutic strategies for GC are quite limited, including endoscopic resection for early GC and surgery or adjuvant chemotherapy for advanced stages (Smyth et al., 2020; Tan, 2019). There is a large difference in prognosis between early and advanced GC, 5-year survival rate can reach over 95% in patients with early GC whereas the median survival time for advanced stages is <1 year (Smyth et al., 2020; Song et al., 2017). Therefore, development of early screening methods and comprehensive understanding of the molecular mechanisms underlying GC progression are major concerns for improving the overall survival of patients.

Circular RNAs (circRNAs) are a class of single-stranded non-coding RNAs without a 5′ end cap and a 3′ end polyA tail that are covalently bonded to form a loop structure in eukaryotic cells (Memczak et al., 2013). Most of them are derived from exons and located in the cytoplasm, with s small portion intron-origin circRNAs present in the nucleus (Zhang et al., 2014). CircRNA is generated in a pattern of back-splicing of precursor mRNA transcripts, which differs from the formation of linear RNA (Li et al., 2020a). The peculiar closed loop structure of circRNAs gives them the special resistance to RNA exonuclease, ensuring a more stable expression patterns compared with linear RNAs (Memczak et al., 2013; Li et al., 2020a; Jeck and Sharpless, 2014). Consequently, circRNAs are promising biomarkers for many different human diseases, especially cancers (Patop et al., 2019; Chen and Yang, 2015). Over the years, numerous evidence have noted abnormal expression of different circRNAs in various tumor tissues or cells and potentially implicated in the pathogenesis of different kinds of cancers as evidenced by in vitro and in vivo data from both human and experimental animals (Qu et al., 2018; Wang et al., 2017), such as bladder cancer (Xu et al., 2019), lung adenocarcinoma (Wang et al., 2020), liver cancer (Huang et al., 2020), esophageal squamous cell carcinoma (Liu et al., 2019), and gastric cancer (Ma et al., 2020). Functionally, circRNAs are well known by their richness in microRNA (miRNA) binding sites, reducing miRNAs' repressive effect on its effector genes (Hansen et al., 2013; Kristensen et al., 2022). Thus, targeting circRNA/miRNA axis in human cancers has gained widespread attention in the past years.

In this study, through analysis of a microarray dataset, we identified a low expression of has_circ_0003356 which is derived from the exons of serine/threonine protein phosphatase 5 catalytic subunit (PPP5C), in gastric cancer tissues. circ_0003356 was predicted to target miR-556-5p and regulate FK506 binding protein 51 (FKBP5) expression. Then, utilizing functional assays and nude mice model, we attempted to further examine the potential involvement of circ_0003356/miR-556-5p/ FKBP5 axis and the associated mechanisms in GC progression.