Preneoplastic liver colonization by 11p15.5 altered mosaic cells in young children with hepatoblastoma

Abstract

Pediatric liver tumors are very rare tumors with the most common diagnosis being hepatoblastoma. While hepatoblastomas are predominantly sporadic, around 15% of cases develop as part of predisposition syndromes such as Beckwith-Wiedemann (11p15.5 locus altered). Here, we identify mosaic genetic alterations of 11p15.5 locus in the liver of hepatoblastoma patients without a clinical diagnosis of Beckwith-Wiedemann syndrome. We do not retrieve these alterations in children with other types of pediatric liver tumors. We show that mosaic 11p15.5 alterations in liver FFPE sections of hepatoblastoma patients display IGF2 overexpression and H19 downregulation together with an alteration of the liver zonation. Moreover, mosaic livers' microenvironment is enriched in extracellular matrix and angiogenesis. Spatial transcriptomics and single-nucleus RNAseq analyses identify a 60-gene signature in 11p15.5 altered hepatocytes. These data provide insights for 11p15.5 mosaicism detection and its functional consequences during the early steps of carcinogenesis.

Fig. 1. Molecular features of 11p15.5 locus pre-malignant expansions.

a Bottom: age at necropsy (fetal livers) and surgery in 131 patients with pediatric liver tumors. Patients with mosaicism are colored by type of 11p15.5 alteration (cn-LOH: red, paternal duplication: orange, epimutation IC2: light red). Top: median age comparison between 13 mosaic HB (red) and 64 non-mosaic HB (pink). Two-sided Student’s t-test was performed. b IC1/IC2 methylation ratio in 124 non-tumor tissues from patients with pediatric liver tumors. c Spectra of 11p15.5 locus alteration in 13 mosaic livers and patients with Beckwith-Wiedemann syndrome. Two-sided χ² goodness of fit test was performed. d Heatmap representation of key gene expression in 7 fetal livers, 10 mosaic, and 23 non-mosaic livers. All IGF2 isoforms were taken into account including P0 to P4 promoters. Limma differential expression analysis was performed adjusted for age at surgery. e Gene-set enrichment analysis using fgsea R package based on the adaptive multilevel splitting Monte Carlo approach was used to assess the significance of gene-set enrichment in 10 mosaic vs. 23 non-mosaic livers. ns: non-significant. NT: non-tumor, HB: hepatoblastoma, pHCC: pediatric hepatocellular carcinoma, pFLC: pediatric fibrolamellar carcinoma, pHCA: pediatric hepatocellular adenoma, FL: fetal liver, ML: Mosaic liver, NML: non-mosaic liver, LOH: loss of heterozygosity, cn-LOH: copy-neutral LOH, pat.dup: 11p15.5 paternal duplication. Source data are provided as a Source Data file.

Fig. 2. Spatial and longitudinal heterogeneity of 11p15.5 locus mosaic hepatocytes.

a Representative non-mosaic, mosaic 11p15.5 cn-LOH, and tumor areas of FFPE slides from mosaic patient #3377. Six stainings were performed: singleplex and duplex RNAscope IGF2/H19, Hematoxylin and Eosin, β-catenin, and glutamine synthetase (GS) immunostainings. All IGF2 isoforms (P0 to P4 promoters) were considered in IGF2 singleplex and duplex RNAscope assay. In duplex RNAscope assay, 11p15.5 mosaic cells display IGF2 overexpression (in blue) and H19 expression loss (in red). Two slides from patient #3377 were analyzed. Scale bars, 50 µM and 200 µM. b Fraction of mosaic cells (%) in multiple samplings from the same patient using RNAscope IGF2/H19, MS-MLPA and WGS/WES detection methods. All samplings are synchronous samples collected at surgery except one liver biopsy labeled with an asterisk (*). c WGS-derived B-allele frequency (BAF) of heterozygous single-nucleotide polymorphisms (SNP) at chromosome 11 in mosaic patient #3370. We identified 33% of mosaic 11p15.5 cells in the liver biopsy but not in the blood or liver resection. IGF2/H19 RNAscope assay allowed for the detection of 14% of mosaic cells in the FFPE section from liver resection. Parts of Fig. 2c were drawn by using a picture from Servier Medical Art. The body color was changed. Servier Medical Art by Servier is licensed under a Creative Commons Attribution 3.0 Unported License (https://creativecommons.org/licenses/by/3.0/). Source data are provided as a Source Data file.

Fig. 3. 11p15.5 mosaic hepatocytes disrupt liver zonation architecture.

a Frequence of IGF2 positive cells (none to many) and IGF2 intensity in metabolic zones (1 to 3) where i3 stands for highest intensity in 8 non-mosaic and 9 mosaic livers (11 and 10 analyzed slides respectively). Two-sided χ² for trend were performed. CV: central vein, PV: portal vein. b Number of central veins per µm² analyzed and glutamine synthetase staining intensity around CV in 8 non-mosaic and 9 mosaic livers (11 and 10 analyzed slides respectively). “s” indicates the number of slides analyzed and “p” the number of corresponding patients. Two-sided Wilcoxon and χ² for trend statistical tests were performed. c Representative RNAscope of IGF2 and immunostainings of glutamine synthetase in paraffin-embedded liver sections from a non-mosaic patient (#3163), a non-mosaic and a mosaic area from a patient with 11p15.5 locus mosaic cn-LOH (#3377). All IGF2 isoforms (P0 to P4 promoters) were considered in IGF2 singleplex and duplex RNAscope assay. Two slides from patient #3377 and one slide from patient #3163 were analyzed. Scale bars, 50 µM and 200 µM. Source data are provided as a Source Data file.

Fig. 4. Spatial transcriptomics in patients #4001 #3559 and #3115 with mosaic cn-LOH at locus 11p15.5.

a IGF2 RNAscope in situ hybridization assay in 3 patients with mosaic cn-LOH. Three slides from patient #4001, 2 from patient #3559, and 3 from patient #3115 were analyzed. Scale bars, 1 mm. b Hematoxylin and eosin staining in FFPE sections from mosaic cn-LOH patients #4001, #3559, and #3115 realised during the spatial transcriptomics experiment. c UMAP representation of visium spatial transcriptomics (10X genomics) merged expression data from #4001, #3559, and #3115 samples and d spatial clusters visualization in FFPE slides. e Heatmap representation of differentially expressed genes between clusters. A two-sided Wilcoxon–Mann–Whitney test was performed. All IGF2 isoforms were taken into consideration in IGF2 spatial transcriptomics quantification. f Gene-set enrichment analysis between mosaic and non-mosaic hepatocytes from all 3 merged patients. Source data are provided as a Source Data file.

Fig. 5. Single-nucleus RNAseq (snRNAseq) in 3 patients with mosaic cn-LOH at locus 11p15.5 and 1 patient without 11p15.5 mosaicism.

a UMAP representation of snRNAseq (10X genomics) merged expression data from #4001, #3559, #3115 and #2996 samples. Left: annotation of the different cell types recovered. HSC: Hepatic stellate cells, VSMCs: Vascular smooth muscle cells. Right: patient of origin annotation. In total, 9760 cells were analyzed. b Proportion (0–1) of each cell population in all 4 snRNAseq samples. c In the 3 mosaic samples (#3115, #3559, #4001), B-allele frequency was assessed in 11p15.5 altered region specific of each patient in all cell types. The results indicate a cn-LOH in hepatocytes, cholangiocytes, and tumor cells. d UMAP representation of cn-LOH status per cell in the 4 patients. e Venn diagram representing the commonly upregulated genes in mosaic cells/livers using snRNAseq, spatial transcriptomics, and bulk RNAseq. The top 15 genes upregulated both in spatial transcriptomics and snRNAseq are highlighted. f Visualization of the 60-gene mosaic signature. g Functional enrichment (ToppGene suite) in mosaic hepatocytes based on 60 genes upregulated in snRNAseq and spatial transcriptomics. Source data are provided as a Source Data file.

Fig. 6. β-catenin and 11p15.5 locus oncogenic cooperation in mosaic and non-mosaic HB tumors.

a Evolution of β-catenin and 11p15.5 locus alteration spectra with age at surgery in 74 and 72 patients, respectively. One sample per patient was represented. A two-sided χ² test was performed. b Heatmap representation of CTNNB1 and 11p15.5 locus alterations in 74 HB tumors. c Proportions of fetal and adult IGF2 promoter usage in 110 pediatric liver tumors. HB RM: hepatoblastoma relapse or metastasis. d First Kaplan–Meier curve indicates progression-free survival in mosaic (n = 12) and non-mosaic patients under 3.3 years of age (n = 39). The second Kaplan–Meier curve displays progression-free survival in patients with HB with cn-LOH (n = 24) or without (n = 25) under 3.3 years of age. Log-rank statistical test was performed. e Scheme representing different routes for HB tumorigenesis. HB can develop from 11p15.5 mosaic cells or directly in isolation through β-catenin alteration. Mos.: mosaic, non-mos.: non-mosaic, pat.dup: locus 11p15.5 paternal duplication, yo: years old. alt: alteration, mut. mutation, cn-LOH: copy-neutral LOH, exp.: expression. Source data are provided as a Source Data file.