iBet uBet web content aggregator. Adding the entire web to your favor.
iBet uBet web content aggregator. Adding the entire web to your favor.



Link to original content: http://www.ncbi.nlm.nih.gov/pubmed/31417090
Cancer-associated mutations in DICER1 RNase IIIa and IIIb domains exert similar effects on miRNA biogenesis - PubMed 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
. 2019 Aug 15;10(1):3682.
doi: 10.1038/s41467-019-11610-1.

Cancer-associated mutations in DICER1 RNase IIIa and IIIb domains exert similar effects on miRNA biogenesis

Jeffrey Vedanayagam  1 , Walid K Chatila  2   3   4 , Bülent Arman Aksoy  2   3   5 , Sonali Majumdar  1 , Anders Jacobsen Skanderup  2   6 , Emek Demir  2   7 , Nikolaus Schultz  4   8   9 , Chris Sander  10   11   12 , Eric C Lai  13   14
Affiliations

Cancer-associated mutations in DICER1 RNase IIIa and IIIb domains exert similar effects on miRNA biogenesis

Jeffrey Vedanayagam et al. Nat Commun. .

Abstract

Somatic mutations in the RNase IIIb domain of DICER1 arise in cancer and disrupt the cleavage of 5' pre-miRNA arms. Here, we characterize an unstudied, recurrent, mutation (S1344L) in the DICER1 RNase IIIa domain in tumors from The Cancer Genome Atlas (TCGA) project and MSK-IMPACT profiling. RNase IIIa/b hotspots are absent from most cancers, but are notably enriched in uterine cancers. Systematic analysis of TCGA small RNA datasets show that DICER1 RNase IIIa-S1344L tumors deplete 5p-miRNAs, analogous to RNase IIIb hotspot samples. Structural and evolutionary coupling analyses reveal constrained proximity of RNase IIIa-S1344 to the RNase IIIb catalytic site, rationalizing why mutation of this site phenocopies known hotspot alterations. Finally, examination of DICER1 hotspot endometrial tumors reveals derepression of specific miRNA target signatures. In summary, comprehensive analyses of DICER1 somatic mutations and small RNA data reveal a mechanistic aspect of pre-miRNA processing that manifests in specific cancer settings.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1
Fig. 1
Survey of DICER1 cancer mutations reveals recurrent RNase IIIb and IIIa hotspots. a Lollipop mutation diagrams of TCGA-PanCan (pointing up) and MSK-IMPACT (pointing down) datasets. The sites labeled in black designate mutations in the four catalytic residues in the RNase IIIb domain (D1709, E1705, E1813, D1810, in black) and an adjacent residue (G1809, in green) that are the major known biallelic DICER1 events in cancer. These aggregate data also reveal an uncharacterized biallelic alteration involving S1344L (in red) in the RNase IIIa domain. The background of other somatic mutations of unknown functional significance is shown for perspective; certain other apparently recurrent mutations are recovered in both TCGA and IMPACT datasets (e.g., PAZ-R944Q, in gray), but were not associated with biallelic hits in non-hypermutated samples. b Left column, examples of biallelic hits in each of the five known RNase IIIb hotspot residues in IMPACT data, with their corresponding secondary inactivating mutations and cancer types. Right column, a selection of multiple cancer types that exhibit RNase IIIa-S1344L and carry diverse secondary inactivating mutations across TCGA and IMPACT data. Patient sequencing from metastatic tumors are noted (in blue)
Fig. 2
Fig. 2
Frequency of DICER1 mutations in TCGA and MSK-IMPACT data across tumor types. We classify alleles as variants of unknown significance, RNase IIIa/b hotspot mutations, and/or biallelic mutations in a TCGA-PanCan data and b MSK-IMPACT data. Because of the diversity of MSK-IMPACT cohort, only cancer types with >50 cases are shown, except for certain rare MSI/POLE tumors that are shown to match TCGA hypermutated data. Hypermutated tumors are designated in blue. Note the MSK-IMPACT data is generally underpowered relative to TCGA to call copy number loss; thus, only TCGA data includes biallelic calls for HETLOSS. Many commonly surveyed cancer types (n for each tumor type summarized below) are devoid of DICER1 hotspot mutations (also see Supplementary Figs. 2 and 3). However, independently in TCGA and MSK-IMPACT datasets, multiple uterine cancers were enriched for hotspot mutations (green highlighted groups on x-axis). Bootstrap resampling analysis with hypermutated cases excluded shows that several uterine cancers and sarcoma subtypes are statistically enriched in both TCGA and MSK-IMPACT datasets (red asterisks). Furthermore, in a second iteration where hypermutated cases were retained in resampling analysis, additional uterine cancers were found enriched for RNase III hotspots (blue asterisks). c Examples of statistical analyses for enrichment or depletion of hotspot mutations in MSK-IMPACT cancer groups. From random samplings of 10,000 bootstrap replicates, the average numbers of hotspot mutations obtained are in blue, while observed numbers of hotspot mutations are in orange. For enrichment, endometrial cancer is shown as an example. Statistical significance for depletion can only be estimated using percentile confidence intervals (CI) when sample sizes are large, as illustrated for breast cancer (zero hotspots, falling well below 5% CI). Statistical analyses for all other cancer groups from TCGA and MSK-IMPACT data are shown in Supplementary Figs. 2 and 3
Fig. 3
Fig. 3
Cancer hotspots in DICER1 RNase IIIb and RNase IIIa domains deplete 5p miRNAs. a Existing model for how the two RNase III domains of Dicer cleave a pre-miRNA hairpin, and for strand-specific defects in RNase IIIb cancer hotspot mutants. b Volcano plots of miRNA-5p (green) and miRNA-3p (blue) expression in sets of Uterine Corpus Endometrial Cancer (UCEC) cases. Top: control comparison of randomly selected DICER1-wt (15 vs. bulk 533 cases) shows only incidental fluctuation of miRNA levels. Bottom: comparison of 15 RNase IIIb hotspot mutant to 548 DICER1-wt UCEC cases shows downregulation of miRNA-5p species and upregulation of miRNA-3p species. c Barplots of the same UCEC comparisons showing that RNase IIIb hotspot cases with overt biallelic-inactivating mutations (four datasets) distort miRNA-5p/3p profiles more severely than other RNase IIIb hotspot cases (11 datasets). d Systematic analysis of relative miRNA strand processing in TCGA small RNA-seq data. We calculated a metric (mi53) that summarizes sample-specific relative 5p abundances. The few samples with negative values are dominated by RNase IIIb hotspot mutants. e Segregation of TCGA samples by genotype. For this analysis, the ovarian serous cystadenocarcinoma cancer (OV) cohort was omitted, due to tissue-specific low mi53 score (Supplementary Fig. 6). DICER1 mutant samples stratify according to whether they have RNase IIIb hotspot mutants with biallelic-inactivating mutations (in red), or have RNase IIIb hotspots accompanied by other alterations of unknown consequence or lack secondary mutations (in orange). The remainder of DICER1 mutant samples behave similarly to DICER1-wt samples, with the exception of S1344L cases (in green), which also exhibit miRNA-5p depletion. f The mi53 metric varies according to the miRNAs expressed in a given tissue. Segregating DICER1-RNase III hotspot cases by tumor type illustrates that biallelic hotspot cases (which include all S1344L cases) typically exhibit lowest mi53 scores within their cohort, whereas other RNase III hotspot cases are more heterogeneous in their behavior. Box plot elements: center line, median; box limits, upper and lower quartiles; whiskers, 1.5x interquartile range
Fig. 4
Fig. 4
DICER1 RNase IIIa/IIIb hotspots selectively affect miRNA-5p processing and activity. a Western blot validation of accumulation of wt and mutant DICER1 proteins in Dcr-KO MEFs. Cells were transfected with the indicated Dicer and miRNA constructs blotted using human Dicer antibody and ß-tubulin as loading control. b, c Northern blot assays of small RNAs from Dcr-KO MEFs transfected with the indicated Dicer and miRNA constructs. The top two blots are for 5p and 3p probes directed against mir-151 (b) and mir-199a-1 (c), which detect different mature 21–23 nt species but co-detect their cognate pre-miRNA hairpins. The lack of mature species without Dicer transfection confirms their full Dicer dependence. Both miRNAs yield relatively similar 3p species with wt and mutant Dicer proteins, while 5p species are selectively impaired. mir-144-3p serves as another canonical miRNA control, miR-451 is a Dicer-independent miRNA, and U6 is a loading control. Note that some structured pre-miRNA hairpins run faster than their predicted linear sizes; pre-mir-151 is expected to be 58 nt, pre-mir-151 is expected to be 63 nt, pre-mir-144 is 57 nt. d, e Luciferase sensor assays. Dcr-KO MEFs were transfected with the indicated human DICER1, miRNA and sensor constructs. With reference to the lack of repression of these miRNAs on their sensors in Dcr-KO MEFs, wt and mutant DICER1 proteins were able to rescue 5p/3p activity of miRNAs with the exception of strongly diminished or lack of 5p rescue by RNase IIIa/b mutant DICER1 proteins. Standard error of triplicate luciferase sensor experiments is shown
Fig. 5
Fig. 5
Structural rationale for how RNase IIIa-S1344 affects RNase IIIb function. a (Left) The human DICER1 cryo-EM structure 5zam [https://www.rcsb.org/structure/5zam] is shown, and part of the RNase IIIa domain (in green)/RNase IIIb (in purple) interface is enlarged. Inspection of this region reveals that RNase IIIa-S1344 resides on the inter-molecular heterodimeric interface with the RNase IIIb domain, closest to F1706, which is adjacent to the active site residue E1705. E1705, D1709, and E1813 coordinate the Mg2+ ion in higher resolution structures of the RNase IIIb domain (i.e., 2eb1) and the Mg2+ ion is modeled here by overlaying 2eb1 [https://www.rcsb.org/structure/2EB1] with 5zam [https://www.rcsb.org/structure/5zam]. This places S1344 within 8–9 Å of the RNase IIIb domain active site Mg2+. Modeling the side chain of S1344 and correct placement of the Mg2+ in the catalytic state may bring the hydroxyl of the serine even closer. b Contact map summarizing highly evolutionarily coupled residues, analyzed across the equivalent of DICER1 aa1271-1829, comprising the RNase IIIa/b domains. Highly coupled residue pairs (black) were displayed on top of residues located <5 Å apart in the 5zam [https://www.rcsb.org/structure/5zam] cryo-EM structure (aqua). Note that the RNase IIIa domain contains a large flexible insertion whose structure is unknown (gray). This analysis reveals not only functionally coupled residues within each RNase III domain, as expected, but also a prominent interface of RNase IIIa/IIIb interactions that includes S1344. These observations provide an evolutionary explanation for how cancer hotspot RNase IIIa-S1344L mutations impair RNase IIIb activity
Fig. 6
Fig. 6
Gene expression signatures of DICER1 hotspot UCEC tumors. a Gene-set enrichment analysis (GSEA) on mRNA profiles of DICER1 RNase III hotspot cases in uterine corpus endometrial cancer (UCEC). Shown are all enriched sets in RNase III hotspot samples; five are miRNA target sets and two relate to Notch pathway targets. b The expression distributions of all enriched gene sets were preferentially upregulated in DICER1 mutants compared to wildtypes. HMGA2, a well-known oncogene and let-7 target, was the most-deregulated gene. c For each of the three mature-5p miRNA families, we saw consistent downregulation of 5p strand (green) and upregulation of 3p strand (red) miRNA members in DICER1 hotspot mutants. For the mature-3p miRNA families, both arms were downregulated in mutants. mut: DICER1 hotspot mutant; wt: DICER1 wildtype; Diff. Exp.: Differential expression (log2 ratio of mRNA/miRNA levels); p-value: The probability for the null hypothesis that the genes in the set are not differentially upregulated in mutants compared to wildtypes; FDR: p-value corrected for multiple hypothesis testing
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Bartel DP. Metazoan microRNAs. Cell. 2018;173:20–51. doi: 10.1016/j.cell.2018.03.006. - DOI - PMC - PubMed
    1. Yang JS, Lai EC. Alternative miRNA biogenesis pathways and the interpretation of core miRNA pathway mutants. Mol. Cell. 2011;43:892–903. doi: 10.1016/j.molcel.2011.07.024. - DOI - PMC - PubMed
    1. Czech B, Hannon GJ. Small RNA sorting: matchmaking for Argonautes. Nat. Rev. Genet. 2010;12:19–31. doi: 10.1038/nrg2916. - DOI - PMC - PubMed
    1. Lai EC. microRNAs are complementary to 3' UTR sequence motifs that mediate negative post-transcriptional regulation. Nat. Genet. 2002;30:363–364. doi: 10.1038/ng865. - DOI - PubMed
    1. Agarwal, V., Bell, G. W., Nam, J. W. & Bartel, D. P. Predicting effective microRNA target sites in mammalian mRNAs. eLife4, e05005 (2015). - PMC - PubMed

Publication types