Based on previous studies (19), a list of 861 CRC-associated dysregulated lncRNAs was obtained from The Cancer LncRNome Atlas (TCLA) (http://tcla.fcgportal.org/). The information on lncRNA names were confirmed using the GENCODE data databasehttps://www.gencodegenes.org/# FPKM RNA expression values of 861 lncRNAs were accessed from the MiTranscriptome (http://mitranscriptome.org/) database. They were collated with the patients’ ID of 268 CRC cases (196 colon cancer and 72 rectal cancer) from the 7256 RNA sequencing library obtained in previous studies (20). Finally, the CRC dysregulated lncRNAs (including upregulated and downregulated lncRNAs) with prognostic information were obtained. The optimal cutoff point values of the lncRNA expressions were calculated using R (“Maxstat” package). Kaplan-Meier curves were generated to evaluate the association of these lncRNAs with the overall survival of patients with colorectal cancer using R (“survminer” and “survival” survival’ packages). The expression of candidate lncRNAs was downloaded from UCSC Xena (https://xena.ucsc.edu/) to verify the differential expression of lncRNAs in colorectal cancer and normal tissues.

Fifty pairs of colon cancer and matched para-carcinoma tissues were acquired at Changhai Hospital, Second Military Medical University, from colon cancer patients who underwent surgical resection between January 2015 and January 2020. Two pathologists independently identified the resected samples. Baseline information on the patients included age, sex, disease location, tumor family history, CEA, CA-199, mutation of B-raf and K-ras, postoperative therapy, tumor size, depth of invasion, and number of examined lymph nodes. None of the patients received local or systemic therapy before surgery. This study conformed to the standards set by the Declaration of Helsinki and was reviewed and approved by the Ethics Committee of Changhai Hospital. Written informed consent was obtained from all participants.

Human CRC cells (SW620, RKO, HT29, SW480, HCT116, and Caco2), a normal colorectal cell line (FHC), and HEK293T cells were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). The above cell lines were authenticated using short tandem repeat analysis by the Genetic Testing Biotechnology Corporation (Suzhou, China). The RKO, HCT116, HEK293T, SW620, and HT29 cells were maintained in RPMI-1640 medium (Gibco, US), and SW480 and Caco2 cells were maintained in DMEM (Gibco, US) supplemented with 10% fetal bovine serum (FBS) (Gibco, US) and 1% penicillin/streptomycin (Gibco, US). All cells were cultured at 37°C in a humidified incubator with 5% CO₂.

Total RNA was extracted from human CRC tissues or cultured cell lines using RNAiso Plus reagent (Takara, Shiga, Japan) according to the manufacturer’s instructions. RNA concentration and purity were evaluated using a Nanodrop 2000 (Thermo Scientific, Wilmington, DE, USA). An aliquot of 2 μg total RNA was reverse-transcribed into complementary DNA (cDNA) according to the manufacturer’s protocol using a PrimeScript RT reagent kit (Takara, Shiga, Japan). The relative expression levels of lncRNAs were examined by qRT-PCR using TB Green Premix Ex Taq™ II (Tli RNaseH Plus) (Takara, Shiga, Japan) in a LightCycler 480II system (Roche, Basel, Switzerland). The candidate lncRNA expression levels were calculated using the 2^−ΔΔCt method, which was normalized to β-actin mRNA. All assays were performed in triplicate. The details of the qRT-PCR primers used in this study are listed in Table S1.

Two pairs of siRNAs targeting FAM222A-AS1 were designed on an online website (http://sirna.wi.mit.edu/home.php) and synthesized at Shanghai Genepharma (Shanghai, China). The siRNA-FAM222A-AS1 and siRNA control sequences were as follows: siRNA1, 5’-CAGCUUCAGUACUACGAUGUU-3’; siRNA2, 5’-UCUGAGAAUCGAUACUGAAUU-3’; siRNA-control, 5’-UUCUCCGAACGUGUCACGUTT-3’. Transient transfection of small interfering RNAs was performed using the Lipofectamine RNAiMAX reagent (Invitrogen, Carlsbad, CA, USA) at a final concentration of 20nM, according to the manufacturer’s instructions.

To construct FAM222A-AS1 stably silenced colon cancer cells, the shRNA targeting FAM222A-AS1 was generated and synthesized as previously described (21) based on the sequence of siRNA1. The templates were subjected to PCR amplification with primers containing XhoI or EcoRI restriction sites (the details of the primers are listed in Table S1). The PCR products were purified and cloned into the pINDUCER10 vector (22). Lentiviruses were packaged into HEK293T cells using the Lenti-X HTX Packaging System (Clontech Laboratories, Inc., CA, USA) to produce lentiviral particles. SW620 and HCT116 cells were infected with FAM222A-AS1 shRNA lentivirus. Ninety-six hours after infection, the cells were treated with puromycin (2.5 μg/ml) for four weeks to select FAM222A-AS1 depleted cells. The shFAM222A-AS1 sequence used was: AGCCACAGATGTAAAGTTAACGTATGGCGAGATCTGCCTACTGCCTCGGA.

To construct colon cancer cells with stably overexpressed FAM222A-AS1, the total length cDNA sequence of FAM222A-AS1 (NCBI reference sequence NR_026661.2 for variant 1) was synthesized and commercially confirmed (Sangon Biotech, Shanghai), and was subcloned into the pENTR3C entry vector. Thereafter, sequences of FAM222A-AS1 in pENTER3C were Gateway-recombined into the pINDUCER20 vector (22). All recombinant vectors were re-sequenced by Sangon Biotech (Shanghai, China). FAM222A-AS1 overexpression plasmid pINDUCER20-FAM222A-AS1 was transfected into HT29 and Caco2 cells. Forty-eight hours after transfection, HT29 and Caco2 cells were treated with G418 (1000 μg/ml) for four weeks to select FAM222A-AS1 overexpressed cells. FAM222A-AS1 depleted cells and FAM222A-AS1 overexpressed cells were induced using 2.5 μg/mL doxycycline (Dox). The overexpression and depletion efficiencies were verified using qRT-PCR, as described above.

A Cell Counting Kit 8 (CCK8) assay (Dojindo, Kumamoto, Japan) was used to measure cell proliferation in 96-well plates. Cells were seeded in triplicate at 2500 cells per well. CCK8 reagent was added at 0, 24, 48, and 72 h after transfecting FAM222A-AS1 siRNA and control siRNA, and incubated at 37°C for 2 h. Cell numbers were determined by measuring the absorbance at 450 nm using the Multiskan FC 96-well plate reader (ThermoFisher, USA). In the transwell invasion assay, 1 × 10⁵ CRC cells were seeded into the upper chamber of an insert (24‐well plates, 8 μm pore size; Corning, NY, USA) coated with Matrigel (BD Biocoat, Corning) in serum-free media, which is unnecessary for the migration assay. Media containing 20% FBS was added to the lower chamber. After culturing for 24 h at 37°C, the cells remaining on the upper membrane were removed with cotton swabs, and migrated or invaded cells were fixed and stained with 0.1% crystal violet and counted in three random high magnification fields by microscopic observation. About 2 × 10⁵cells were seeded in 12-well plates to do the wound healing assay. Draw three parallel lines on the bottom of the 12-well plates before seeding cells, and three parallel lines perpendicular to it in order to accurately locate the cell imaging area when taking pictures later. A straight scratch was made using a 10 μl pipette tip before imaging. The scratch areas were photographed again after 24h, 48h, 72h, and analyzed. For the colony formation assay, CRC cells (500 cells/well) transfected with the indicated vector were plated in 6 cm culture dish and incubated at 37°C. The cell medium was replaced every three days. Two weeks later, the cells were fixed and stained with 0.1% crystal violet. The number of visible colonies was counted using the ImageJ software.

Male BALB/c nude mice (4-5 weeks old) were purchased from the SLAC Laboratory Animal Co. (Shanghai, China). All mice were housed in laminar flow cabinets under specific pathogen-free conditions at room temperature with a 12 h light/dark cycle, with food and water available ad libitum. To establish human tumors in nude mice, SW620 and HCT116 cells (1 × 10⁷ cells) with or without FAM222A-AS1 knock down and HT29 cells with or without FAM222A-AS1 overexpression were suspended in 200 μl PBS and subcutaneously injected into the right back flank of mice. The mice were then divided into treatment (+Dox) and control groups (-Dox) (five mice per group). The mice were fed water with 2 mg/mL Dox and 5% sucrose (+Dox) or standard water and 5% sucrose (-Dox). The tumor volume was measured every three days using external measurements. The mice were sacrificed after 27 days, and the tumor volume was calculated by a formula {1/2[length(mm) × width(mm)²]}. The efficiency of FAM222A-AS1 knockdown was examined by qRT-PCR using tissue lysates. All procedures were performed in accordance with the National Research Council’s Guide for the Care and Use of Laboratory Animals.

Cell smears were prepared for FISH using standard methods. HCT116 cells were washed twice with phosphate-buffered saline (PBS) and fixed in 4% paraformaldehyde at room temperature for 15 min. The pre-hybrid solution was added to the slides and incubated at 37°C for 1 h. The slides were incubated overnight at 60°C in a hybridization solution with the probe, washed, and dehydrated. After staining with DAPI, the slices were scanned and imaged using a fluorescence microscope (Nikon, Japan). FAM222A-AS1 emits green light, and the cell nucleus emits blue light. All experiments were repeated three times. All FISH probes were commercially synthesized by Zeheng Co., Ltd. The sequence probe used was 5’-AGTGTCATCTGGTGGCATCCCTCCTTCCCTCTTTTGGTCTCCATTTCCAT-3’.

The RNA pulldown assay was performed using an RNA pulldown kit. All subsequent experiments were performed following the manufacturer’s instructions. The sequence of the biotin-coupled FAM222A-AS1 probe was AACTTCTGAACTTGCCATCC. The sequence of the negative control probe was TTCTCCGAACGTGTCACGT. Cells were crosslinked with formaldehyde, equilibrated in glycine buffer, and scraped with lysis buffer. The cell samples were sonicated and centrifuged. The supernatant was transferred to a 2 ml tube, and separately saved 50 μl as input analysis. Cell lysates were incubated with a biotin-tagged specific probe or control probe for 3 h. The supernatant lysate was incubated with streptavidin beads for 1 h with rotation. The bead-sample mixture was washed. Subsequently, 10% of the bead sample mixture was re-purified using TRIzol. The purified mRNA was detected using qRT-PCR. The bead-sample mixture (90%) was resuspended in DNase buffer, and protein was eluted with a cocktail of RNase A, RNase H, and DNase I at 37°C for 30 sec. To explore the molecular characteristics of proteins between samples, the protein eluent was analyzed using liquid mass spectrometry (LC-MS/MS). Thereafter, GO and KEGG pathway analyses were performed to describe and analyze the basic information and functions of FAM222-AS1.

The human genes FAM222A-AS1, and MYH9 3’-UTR sequences were synthesized. Synthesized sequences were ligated into the pmiR-GLO dual luciferase miRNA target expression vector (Promega, Madison, WI, USA) to obtain the recombinant vectors which were named pmiR-GLO-222A and pmiR-GLO-MYH9, respectively. Luciferase reporter assays were performed as described previously (23). Cells plated in a 48-well plate were co-transfected with 50 nM miRNA mimics or negative control oligonucleotides and, 50 ng of firefly luciferase reporter using the JetPRIME reagent (Polyplus-transfection). Cells were collected 36 h after the last transfection and analyzed using the Dual-Luciferase Reporter Assay System (Promega). All experiments were performed in triplicate. The sequences of miR-let-7f and control mimics were as follows: miR-let-7f mimics, 5’-UGAGGUAGUAGAUUGUAUAGUU-3’ and 5’-CUAUACAAUCUACCUCAUU-3’; negative control mimics, 5’-UUCUCCGAACGUGUCACGUT-3’ and 5’-ACGUGACACGUUCGGAGAATT-3’.

The human phospho-kinase array kit (Art.No.: ARY003B; R&D Systems; MN Minnesota) was used to identify the FAM222A-AS1-relatived phospho-kinase proteins according to the instruction. Briefly, 1.0 mL of array buffer was first added to each well for 1 h at room temperature. After washing, CRC cell lysates (400μg/well) were added to each well overnight in the array membranes at 4°C. The membranes were then washed and incubated with the detection antibodies for 2 h at room temperature. Next, the membranes were exposed to streptavidin-HRP for 30 min on a rocking platform. After washing, protein bands were detected by enhanced chemiluminescence for 1 min and exposed to the film. The experiment was performed in triplicate.

All statistical analyses were performed using SPSS version 21.0 software (SPSS, IBM, Armonk, NY, USA), R (R, Version 4.0.1), and GraphPad Prism 7 software (La Jolla, USA). Data represent the mean ± standard deviation (SD). The significance of the differences between the groups was evaluated using a paired two-tailed Student’s t-test or χ2 test. The optimal cut-off values of the relative expression of candidate lncRNAs in CRC were determined by R software using “maxstat,” “survival”, and “survminer” packages. The Kaplan-Meier method was used to evaluate overall survival (OS), and the differences were compared using the log-rank test. The Cox proportional hazards model was used to determine independent factors, which were based on the variables selected by univariate analysis. Differences were considered statistically significant at p < 0.05.

To explore the CRC prognosis-associated lncRNAs, we utilized bioinformatics analysis and discovered that 18 candidate lncRNAs were differentially expressed between normal colorectal cells and CRC cells, among which 11 lncRNAs were downregulated and 7 lncRNAs were upregulated (Figures 1A, B). The details of these lncRNAs are shown in Tables S2, S3. In addition, compared to low FAM222A-AS1 expression, high FAM222A-AS1 expression in CRC patients was significantly associated with a lower overall survival rate and a lower disease-free survival rate (Figure 1C). FAM222A-AS1, FAM83H-AS1, and FEZF1-AS1 were the most upregulated lncRNAs among the 18 candidates (Figure 1D). According the previous study, FAM222A-AS1 was the only unexplored lncRNA in CRC, suggesting that FAM222A-AS1 may be a potential target of CRC.

To validate the bioinformatics results, we assessed the expression of FAM222A-AS1 in 50 CRC samples and pair-matched normal samples using qRT-PCR. FAM222A-AS1 expression in CRC tissues was higher than that in para-carcinoma tissues (Figure 1E) and higher FAM222A-AS1 expression was associated with poor overall survival of CRC patients (Figure 1F). The results of GEPIA also showed that the expression of FAM222A-AS1 in tumor tissues was higher than that in normal tissues (Figure S1). However, there were no significant differences in the characteristics of CRC patients with high and low FAM222A-AS1 levels (Table S4).

Further sequence analysis showed that FAM222A-AS1 was 868 nt long and contained four exons. The secondary structure and minimum free energy of FAM222A-AS1 according to the RNAfold database (http://rna.tbi.univie.ac.at/cgi-bin/RNAWebSuite/RNAfold.cgi) are shown in Figure S2.

Furthermore, we investigated the expression of FAM222A-AS1 in six CRC cell lines (RKO, SW480, SW620, Caco2, HCT116 and HT29) and a normal colorectal cell line (FHC). The results showed that FAM222A-AS1 expression in the RKO, SW480, SW620, Caco2, and HCT116 cell lines was higher than that in FHC and HT29 cells (Figure 2A).

To identify the biological function of FAM222A-AS1, we designed and synthesized siRNAs (siRNA1 and siRNA2) specifically targeting FAM222A-AS1 and found that siRNA stably inhibited the expression of FAM222A-AS1 in two CRC cell lines (Figure 2B). Among the siRNAs, siRNA-1 had the highest knockdown efficiency. We then investigated the effect of FAM222A-AS1 on CRC cell proliferation. A CCK-8 assay demonstrated that knockdown of FAM222A-AS1 slowed the growth of CRC cells (Figure 2C). Colony formation assays showed that colony numbers were significantly reduced after FAM222A-AS1 silencing (Figure 2D). Transwell assays were performed to detect the effects of FAM222A-AS1 on the migration and invasion of CRC cells. Our results showed that the migration and invasion abilities of CRC cells were significantly decreased by FAM222A-AS1 downregulation (Figure 2E). Taken together, these results indicate that FAM222A-AS1 knockdown repressed CRC cell proliferation, invasion, and migration.

To investigate the role of FAM222A-AS1 in the growth of CRC in vivo, stable HCT116 and SW620 cell models transfected with Dox-inducible shFAM222A-AS1 were constructed. To identify the knockdown effect and the cell function of the stable cell model, we performed qRT-PCR to assess the expression of CRC cells before and after Dox induction. Our results demonstrated that Dox-induced efficiency was more than 95% (Figure S3), and the FAM222A-AS1 expression in the stable cell model after 2.5 µg/ml Dox induction was lower than that in the non-Dox-induced group (Figure 2F). The CCK-8 assays and colony assays (Figures 2G, H) showed that FAM222A-AS1 knockdown inhibited CRC cell proliferation, while transwell assays demonstrated that the migration and invasion abilities decreased in the FAM222A-AS1 knockdown group (Figure 2I).

To confirm the effect of FAM222A-AS1 upregulation on CRC growth in vitro, we constructed a stable cell model transfected with the Dox-induced pINDUCER20 plasmid containing the FAM222A-AS1 sequence. After Dox induction, FAM222A-AS1 expression was significantly higher, approximately 200-fold (Figure 2J), in CRC cells than in the control group (-Dox). The colony assays showed that the colony numbers and size were larger for FAM222A-AS1 overexpression (+Dox) than in the control group (-Dox) (Figure 2K). The scratch assays demonstrated that FAM222A-AS1 overexpression (+Dox) promoted CRC cell migration (Figure 2L).

The stable cell models HCT116 and SW620 were injected into nude mice to establish a xenograft tumor model. In the xenograft tumor model, the Dox-induced group tumors (+Dox) derived from cells transfected with shFAM222A-AS1 were smaller than those without Dox induction (-Dox). Knockdown of FAM222A-AS1 significantly suppressed tumor volume. The body weights of nude mice in the two groups did not differ significantly (Figures 3A, B). This study demonstrated that FAM222A-AS1 may play an important role in promoting CRC growth in vivo.

Furthermore, a subcutaneous xenograft model was constructed to validate the biological function of FAM222A-AS1 overexpression in vivo. Consistent with the in vitro results, FAM222A-AS1 overexpression (+Dox) significantly increased tumor volume compared to that in the control group (-Dox). The body weights of mice in the two groups were not significantly different (Figure 3C). Our findings indicated that FAM222A-AS1 upregulation promoted CRC cell proliferation both in vitro and in vivo.

To elucidate the molecular mechanism underlying FAM222A-AS1, we performed FISH and bioinformatic analysis to detect the subcellular localization of FAM222A-AS1. As shown in Figure 4A, we found that FAM222A-AS1 was predominantly located in the cytoplasm, indicating that FAM222A-AS1 may act as a ceRNA to capture miRNAs, leading to the release of specific miRNA-targeted transcripts. To investigate this hypothesis, we carried out RNA pull-down assays and found that miR-let-7f had the highest abundance among 21 potential binding miRNAs (Figure 4B and Table 1). In addition, we constructed FAM222A-AS1 luciferase reporters to detect the role of miR-let-7f in the regulation of FAM222A-AS1 activity. After FAM222A-AS1 was knockdown by transfecting the siRNA of FAM222A-AS1, the expression of MYH9 was assessed by qRT-PCT. We found that MYH9 was downregulated when the FAM222A-AS1 was silenced (Figure 4D). After transfection, we observed that overexpression of miR-let-7f significantly inhibited FAM222A-AS1 luciferase reporter activity (Figures 4E, F). These experiments revealed that FAM222A-AS1 may function as a sponge for miR-let-7f.

Based on the interaction between FAM222A-AS1 and miR-let-7f, we explored the potential roles of miR-let-7f in CRC. A previous study indicated that miR-let-7f acts as a tumor suppressor in the progression of CRC, inhibiting the proliferation, migration, and invasion of CRC cells. Furthermore, miRNAs can regulate the expression of their target mRNAs by binding to the 3’-UTR of mRNA. We predicted the target genes of miR-let-7f via three miRNA prediction websites (TarBase, microRNAorg, and miRTarget) (Figure 4C), and found that MYH9, which was previously shown to be involved in the promotion of CRC cancer cell migration or invasion, was one of the target genes of miR-let-7f. To confirm that MYH9 is a possible target of miR-let-7f, a sequence of the 3’-UTR of MYH9 containing the miR-let-7f-binding sequence was employed to synthesize a luciferase reporter plasmid. The results of luciferase reporter analysis showed that co-transfection of a miR-let-7f mimic and an MYH9 plasmid strongly decreased luciferase activity (Figure 4G).

To explore which signaling pathway FAM222A-AS1 can regulate, we screened human protein kinase phosphorylation variation with the overexpression of FAM222A-AS1. We found that the phosphorylation of AKT1/2 (S473) and GSK3α/β (S21/S9) was upregulated, along with the overexpression of FAM222A-AS1 in CRC cells (Figure S4A), which indicates that FAM222A-AS1 may regulate CRC cell proliferation and invasion via downstream AKT- and GSK3α/β-associated pathways. The results of western blotting verified that phosphorylation of AKT1/2 (S473) was indeed altered along with the expression of FAM222A-AS1 (Figure S4B).