Multiple mechanisms disrupt the let-7 microRNA family in neuroblastoma

Poor prognosis in neuroblastoma is associated with genetic amplification of MYCN. MYCN is itself a target of let-7, a tumour suppressor family of microRNAs implicated in numerous cancers.

LIN28B, an inhibitor of let-7 biogenesis, is overexpressed in neuroblastoma and has been reported to regulate MYCN. Here we show, however, that LIN28B is dispensable in MYCN-amplified

neuroblastoma cell lines, despite de-repression of let-7. We further demonstrate that MYCN messenger RNA levels in amplified disease are exceptionally high and sufficient to sponge let-7,

which reconciles the dispensability of LIN28B. We found that genetic loss of let-7 is common in neuroblastoma, inversely associated with MYCN amplification, and independently associated with

poor outcomes, providing a rationale for chromosomal loss patterns in neuroblastoma. We propose that let-7 disruption by LIN28B, MYCN sponging, or genetic loss is a unifying mechanism of

neuroblastoma development with broad implications for cancer pathogenesis.

G.Q.D. is supported by National Institutes of Health grant R01GM107536, Alex’s Lemonade Stand Foundation, and the Ellison Medical Foundation. G.Q.D. is an affiliate member of the Broad

Institute, and an investigator of the Howard Hughes Medical Institute and the Manton Center for Orphan Disease Research. J.T.P. was supported by Alex’s Lemonade Stand Foundation. K.M.T. was

supported as a Howard Hughes Medical Institute International Student Research Fellow and as a Herchel Smith Graduate Fellow. D.S.P. and R.E. were supported by award number T32GM007753 from

the National Institute of General Medical Sciences.

Division of Pediatric Hematology/Oncology, Boston Children’s Hospital, Boston, 02115, Massachusetts, USA

John T. Powers, Kaloyan M. Tsanov, Daniel S. Pearson, Richard Ebright, Marc Seligson, Yvanka de Soysa, Patrick Cahan, Ho-Chou Tu, Areum Han, Grace S. LaPier, Jihan K. Osborne, Samantha J.

Ross, Marcella Cesana & George Q. Daley

Department of Pediatric Oncology, University Hospital Köln, Köln, 50937, Germany

Wyss Institute for Biologically Inspired Engineering, Boston, 02115, Massachusetts, USA

Department of Pathology, Boston Children’s Hospital, Boston, 02215, Massachusetts, USA

Department of Biological Engineering, Massachusetts Institute of Technology, Broad Institute of MIT and Harvard, Cambridge, 02142, Massachusetts, USA

Stem Cell Transplantation Program, Dana Farber Cancer Institute & Boston Children’s Hospital, Boston, 02115, Massachusetts, USA

Department of Biological Chemistry & Molecular Pharmacology, Harvard Medical School, Boston, 02115, Massachusetts, USA

Harvard Stem Cell Institute, Cambridge, 02138, Massachusetts, USA

G.Q.D. provided support and guidance for this work; G.Q.D. and J.T.P. conceived the hypothesis, designed the study, and wrote the manuscript; J.T.P. performed and interpreted most of the

experiments and generated figures; K.M.T. helped perform the shRNA and siRNA experiments and generated most of the plasmid constructs; D.S.P. helped perform the RNA sequencing experiments;

F.R., J.T., and F.B. generated the aCGH and survival data on neuroblastoma patients; C.S.S., K.C.K., and J.J.C. acquired tissue samples and assisted with the IHC analysis; R.E. assisted with

the CRISPR experiments; M.S., Y.d.S., G.S.LP., and S.J.R. provided technical help; P.C. processed RNA sequencing data and helped with data analysis; H.C.T. helped perform in vitro

transfection experiments; A.H. assisted with RNA sequencing data analysis; J.K.O. and C.S.S. performed xenograft experiments; M.C. assisted with the RNA sequencing experiments.

G.Q.D. holds options and intellectual property relating to 28/7 Therapeutics, a company seeking to develop inhibitors of the LIN28/let-7 pathway.

a, Schematic of human MYCN ORF and 3′ UTR, indicating let-7 sites 1 and 2 and their approximate location. b, Predicted base pairing patterns of let-7a, let-7f, and let-7g with MYCN let-7

sites 1 and 2. A–G base pairs, common in RNA, are represented by an asterisk. c, Alignments of let-7 sites 1 and 2 in 100 vertebrate MYCN 3′ UTRs (ENCODE, https://genome.ucsc.edu/ENCODE/).

a, MYCN, LIN28B, and LIN28A mRNA expression levels in neuroblastoma (n = 649; see Source Data (ED Fig 2) in Supplementary Information). b, Immunoblot for indicated proteins in human

embryonic carcinoma cells (PA1), normal human fibroblasts (HF), SK-N-SH (SH), SK-N-AS (AS), SK-N-F1 (F1), BE2C (BE), SK-N-DZ (DZ), Kelly (Ke), and human chronic myeloid leukaemia cells (K5).

For gel source data, see Supplementary Figures. c, Representative LIN28B immunohistochemical staining of human neuroblastoma by stage (left), percentage LIN28B positive neuroblastoma by

disease stage (right); (n = 36). GNB, ganglioneuroblastoma. d, LIN28B expression by neuroblastoma stage (n = 64; Source Data (ED Fig 2)). e, Immunoblot for LIN28B in inducible LIN28B SH-SY5Y

cells and GFP- or LIN28B-expressing SK-N-AS cells (left) and corresponding qPCR analysis of relative let-7 family levels (right) (mean plus s.e.m. of three independent experiments shown).

f, Relative growth rate (BrdU incorporation, right) of SH-SY5Y and SK-N-AS neuroblastoma cells from d (*P