Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2020 Sep 11;11(1):4551.
doi: 10.1038/s41467-020-18355-2.

Spatial expression analyses of the putative oncogene ciRS-7 in cancer reshape the microRNA sponge theory

Affiliations

Spatial expression analyses of the putative oncogene ciRS-7 in cancer reshape the microRNA sponge theory

Lasse S Kristensen et al. Nat Commun. .

Abstract

Circular RNAs (circRNAs) have recently gained substantial attention in the cancer research field where most, including the putative oncogene ciRS-7 (CDR1as), have been proposed to function as competitive endogenous RNAs (ceRNAs) by sponging specific microRNAs. Here, we report the first spatially resolved cellular expression patterns of ciRS-7 in colon cancer and show that ciRS-7 is completely absent in the cancer cells, but highly expressed in stromal cells within the tumor microenvironment. Additionally, our data suggest that this generally apply to classical oncogene-driven adenocarcinomas, but not to other cancers, including malignant melanoma. Moreover, we find that correlations between circRNA and mRNA expression, which are commonly interpreted as evidence of a ceRNA function, can be explained by different cancer-to-stromal cell ratios among the studied tumor specimens. Together, these results have wide implications for future circRNA studies and highlight the importance of spatially resolving expression patterns of circRNAs proposed to function as ceRNAs.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1
Fig. 1. The putative oncogene, ciRS-7, is not expressed in colon cancer cells.
a, b RNA chromogenic in situ hybridization (CISH) analyses for ciRS-7 in a representative poorly differentiated (a) and a representative well differentiated (b) colon adenocarcinoma. Samples from 16 representative colon cancer patients, described in Table 1, were tested and showed similar results. Overviews (left) and higher magnifications (right), indicated in the overview with a square, are shown. The ciRS-7 signal (red dots) is observed in the tumor stroma whereas the cancer cells are negative. Scale bars, corresponding to 500 µm (overviews) and 100 µm (higher magnifications), are indicated in the lower-left corners. c, d NanoString nCounter expression analyses of ciRS-7 (two individual experiments each with n = 1) (c) and of the transcripts T1, T2 and T3 of LINC00632 (n = 1) (d) in laser capture microdissected cancer- and stromal cells pooled from four individual representative patient samples. e Violin plots showing expression levels of T1, T2, and T3 in 32 samples from colon cancer patients measured by NanoString nCounter analyses. Quartiles and medians are indicated. Statistical analysis were performed by two-tailed Mann–Whitney test. f, h Scatterplots showing the expression levels in 32 colon cancer patients samples of ciRS-7 (x-axis) and T1 (f), T2 (g), and T3 (h) (y-axis) with corresponding linear regression statistics, employing an F test to investigate if the slope was significantly non-zero, and Pearson correlation coefficients (r). Correction for multiple testing was not performed.
Fig. 2
Fig. 2. Correlations between ciRS-7 and miR-7 target genes.
ae Scatterplot showing expression levels in 32 colon cancer patients samples of ciRS-7 (x-axis) and FOS (a), NR4A3 (b), PIK3CD (c), CDK1 (d), and PAK1 (e) (y-axis) with corresponding linear regression statistics, employing an F test to investigate if the slope was significantly non-zero, and Pearson correlation coefficients (r). Correction for multiple testing was not performed. f NanoString nCounter expression analyses of 20 miR-7 target genes in fractions of stromal and cancer cells isolated by laser capture microdissection of colon cancer tissues, pooled from four individual representative patient samples (n = 1). g Schematic of approach used to knockout ciRS-7 in HEK293T cells. The expected deletion (1677 bp), the location of PCR primers used to validate deletion of the ciRS-7 exons and the expected amplicon size without the expected deletion (2719) is given. h PCR analysis of genomic DNA from HEK293T cells populations transfected with control sgRNAs or sgRNAs for knockout of ciRS-7. i, j RT-qPCR analysis of ciRS-7 (i) and FOS (j) expression in the HEK293T cells populations assessed in h. The data represent three technical replicates from n = 1 biologically independent samples. k Schematic of luciferase vector constructs. // indicates that the sequence of the plasmid on the left hand of the schematic is identical to the construct listed at the top. l Dual luciferase assay in HEK293T cells. Relative luminescence (Renilla/Firefly)) of target reporter (3x miR-7 or FOS) relative to control reporter (MCS) per replicate is shown (n = 2 biologically independent samples). Luciferase reporters were co-transfected with miR-769 (negative control) or miR-7 in a 1:1 ratio. All values are relative to that of the control setup (psicheck 3xmiR-7 + miR-769). m NanoString nCounter expression analyses of 798 miRNAs in stromal cells and cancer cells. Each dot represent a miRNA. Gray dots represents miRNAs that were expressed at levels below background. The arrows indicate the expression levels of miR-7.
Fig. 3
Fig. 3. A simple model explains correlations between ciRS-7 and miR-7 target genes.
a When ciRS-7 is co-expressed with gene X in the stromal cells, positive correlations will be observed as samples with relatively higher amounts of stromal cells relative to cancer cells will have a relatively higher expression of ciRS-7 and the gene X. b When ciRS-7 and gene Y are mutually exclusively expressed in the stromal and cancer cells, negative correlations will be observed as samples with relatively higher amounts of stromal cells relative to cancer cells will have a relatively higher expression of ciRS-7, whereas as samples with relatively higher amounts of cancer cells relative to stromal cells will have a relatively higher expression of the gene Y. Additional cell types with other expression patterns are likely to confound the analyses when analyzing tumor specimens, leading to poorer correlations than illustrated here.
Fig. 4
Fig. 4. ciRS-7 is positively correlated with stromal cell-enriched genes and negatively correlated with cancer cell-enriched genes.
a Correlation matrix showing Pearson correlation coefficients between ciRS-7 and the 20 miR-7 target genes. The genes are listed according to relative expression levels in cancer and stromal cells, displayed in the top. These data were derived from fractions of stromal and cancer cells isolated by laser capture microdissection of colon cancer tissues, pooled from four individual representative patient samples (n = 1). b Scatter plot showing log2 fold changes in stromal-to-cancer cell ratios (y-axis) and correlation with ciRS-7 (x-axis) for each of the 20 miR-7 target genes, with corresponding linear regression statistics, employing an F test to investigate if the slope was significantly non-zero, and Pearson correlation coefficient (r).
Fig. 5
Fig. 5. Other circRNAs without miR-7 binding sites also correlate with miR-7 target genes.
a NanoString nCounter expression analyses of 11 circRNAs in fractions of stromal and cancer cells isolated by laser capture microdissection of colon cancer tissues, pooled from four individual representative patient samples (n = 1). b, c Scatter plots showing log2 fold changes in stromal-to-cancer cell ratios (y-axis) and correlation with the stromal cell-enriched circRNAs, circFBXW7 (b) and circCCDC66 (c) (x-axis), for each of the 20 miR-7 target genes with corresponding linear regression statistics and Pearson correlation coefficient (r). d, e Scatter plots showing log2 fold changes in stromal-to-cancer cell ratios (y-axis) and correlation with the cancer cell-enriched circRNAs, circZKSCAN1 (d) and circZNF91 (e) (x-axis), for each of the 20 miR-7 target genes, with corresponding linear regression statistics, employing an F test to investigate if the slope was significantly non-zero, and Pearson correlation coefficient (r). Correction for multiple testing was not performed.
Fig. 6
Fig. 6. RNA chromogenic in situ hybridization (CISH) for ciRS-7 in various adenocarcinomas.
af The adenocarcinomas are derived from a tissue microarray and are originating from cervix (a), lung (b), breast (c), ovarium (d), pancreas (e), and stomach (f). For each cancer type at least two cores containing both cancer cells and tumor stroma were analyzed and showed similar results. The ciRS-7 signal (red dots) is observed in the tumor stroma whereas the cancer cells are negative. Scale bars, corresponding to 100 µm, are indicated in the lower-left corners.

References

    1. Jeck WR, et al. Circular RNAs are abundant, conserved, and associated with ALU repeats. RNA. 2013;19:141–157. - PMC - PubMed
    1. Memczak S, et al. Circular RNAs are a large class of animal RNAs with regulatory potency. Nature. 2013;495:333–338. - PubMed
    1. Salzman J, Gawad C, Wang PL, Lacayo N, Brown PO. Circular RNAs are the predominant transcript isoform from hundreds of human genes in diverse cell types. PloS ONE. 2012;7:e30733. - PMC - PubMed
    1. Kristensen LS, et al. The biogenesis, biology and characterization of circular RNAs. Nat. Rev. Genet. 2019;20:675–691. - PubMed
    1. Hansen TB, et al. Natural RNA circles function as efficient microRNA sponges. Nature. 2013;495:384–388. - PubMed

Publication types