Page 13 - Read Online
P. 13

Page 4 of 15                   Chen et al. Cancer Drug Resist 2024;7:9  https://dx.doi.org/10.20517/cdr.2023.151


               RESULTS
               circRNAs that are correlated with responsiveness to PD-1 antibody therapy of CRC
               During the period from July 2018 to May 2022, 32 patients were diagnosed with CRC (20 with MSS and 12
               with MSI-H) and were provided with PD-1 antibody-based treatment. All patients received at least two
               cycles of PD-1 antibody treatment and were evaluated for efficacy by computed tomography (CT) and
               tumor-related markers. Among these patients, 2 patients obtained complete response (CR), 11 patients
               obtained partial response (PR), 12 patients remained stable disease (SD), and 7 patients exhibited
               progressive disease (PD). The ORR was 40.6% and the disease control rate (DCR) was 78.1%. To identify the
               circRNAs that might be associated with CRC immune evasion and anti-PD-1 efficacy, we conducted
               circRNA microarray analysis using 32 CRC samples (13 responsive and 19 resistant to PD-1 antibody
               treatment). A total of 2,118 circRNAs were detected, the majority with a length of 50-500 nt. Most of the
               reads mapped to exons. The circRNAs differentially expressed in CRC samples resistant or responsive to
               anti-PD-1 therapy are presented in the heat map  [Figure 1A]. Among these, 203 circRNAs were
               differentially expressed, out of which 132 were up-regulated and 71 were down-regulated in tissue resistant
               to PD-1 antibody therapy [Figure 1B]. In order to identify the circRNA that is critical for CRC immune
               evasion and anti-PD-1 efficacy and tumor progression, the circRNAs up-regulated in both anti-PD-1
               resistant and tumor tissues were validated by RT-PCR analysis [Figure 1C]. Only has_circ_0060627
               (circNCOA3) was confirmed to be up-regulated in tissues and serum of patients resistant to PD-1 antibody
               treatment in a separate cohort [Figure 1D and E]. A high circNCOA3 level was correlated with tumor size
               and liver metastasis, but was not associated with age, gender, tumor cell differentiation, and tumor location
               [Supplementary Table 3]. In addition, a high circNCOA3 level was correlated with poor OS and PFS of CRC
               patients who received PD-1 antibody treatment [Figure 1F and G]. However, there was no correlation
               between the expression of circNCOA3 and PD-L1 [Figure 1H and I].


               Characterization of circNCOA3 in CRC
               According to the UCSC and circBase database, has_circ_0060627 arises from exons 4, 5, 6, 7, and 8 of
               NCOA3 (circRNA ID: has_circ_0060627, chr20: 46252654-46256767), back-spliced with exon 4 and exon 8
               (740 bp) [Figure 2A], and thus we designated it as circNCOA3. circNCOA3 was amplified by PCR using
               divergent primers and confirmed using Sanger sequencing [Figure 2A]. PCR analysis was performed using
               genomic DNA (gDNA) and reverse-transcribed RNA (cDNA), and the results indicated that divergent
               primers amplified circRNA in cDNA but not in gDNA [Figure 2B]. For validation of the circular status of
               the circNCOA3, RNA was reversely transcribed from CRC cells using random hexamers or oligo (dT)
                                                                                                         18
               primers. The relative expression of circNCOA3, but not NCOA3 mRNA (mNCOA3), was markedly
               decreased when the primers were replaced by oligo (dT)  [Figure 2C], indicating that circNCOA3 has no
                                                                18
               poly-A tail. circNCOA3 was resistant to digestion by RNAase R, a highly processive 3’ to 5’ exoribonuclease
               catalyzing the degradation of linear RNAs but not circular RNAs, whereas NCOA3 mRNA was not
               [Figure 2D]. Moreover, when cells were cultured with actinomycin D to inhibit transcription, the expression
               of circNCOA3 and mNCOA3 was detected at different time points by RT-qPCR, demonstrating that
               circNCOA3 has a significantly longer half-life than mNCOA3 [Figure 2E]. Besides, RT-qPCR and FISH
               results revealed that circNCOA3 is predominantly localized in the cytoplasm [Figure 2F and G]. Taken
               together, the data indicated that circNCOA3 is a stable and abundant circular RNA in cells.


               CircNCOA3 stimulates CRC development via regulation of the tumor immune environment
               To evaluate the biological role of circNCOA3 in CRC, the level of circNCOA3 was determined in CRC cells
               and normal colon cells NCM460. The results showed that circNCOA3 was up-regulated in all tumor cell
               lines compared to normal cells [Figure 3A]. When HCT116 and SW620 cells were transfected with specific
   8   9   10   11   12   13   14   15   16   17   18