PharmD, PhD
PharmD, PhD
MD resident
PhD
Our research explores the genomic and molecular basis of pediatric liver tumors to advance knowledge and improve treatment.
In the Genomics of Pediatric Liver Tumors group, we study tumor evolution using data from patient clinical features, whole-genome and exome sequencing, bulk RNAseq, single-cell, and spatial transcriptomics. We focus primarily on hepatoblastoma (HB), the most common pediatric liver cancer, as well as pediatric hepatocellular carcinoma (HCC), fibrolamellar carcinoma (FLC), and hepatocellular adenoma (HCA).
We partner with clinicians across France to gather samples and address key research questions, and work with Japanese teams to validate findings. We develop computational tools to integrate multi-scale data, exploring tumor evolution and chemotherapy resistance (Figure 1).
Context:
Tumor cells carry molecular changes (mutations, chromosome alterations) that affect gene function. We use computational methods to find recurring altered genes driving cancer, which can be somatic (tumor-specific) or germline/mosaic (predisposing), especially in early childhood cancers.
Published results:
By analyzing 122 tumor samples from 84 patients using whole-genome or whole-exome sequencing, we pinpointed key driver alterations in pediatric liver cancers (Hirsch et al, Cancer Discov 2021) (Figure 2).
HB and HCC converge on pathways like Wnt/β-catenin and IGF2 but differ in alteration types: HB often have mutations (e.g., CTNNB1), while HCC show deletions (e.g., AXIN1). Some HB patients have germline APC mutations linked to familial adenomatous polyposis, and a somatic second hit (Morcrette et al, Oncoimmunology 2019). Rare driver alterations predict poor chemotherapy response and survival (Pire et al, Eur J Cancer 2024). Alterations at the 11p15.5 locus, causing IGF2 overexpression, are a major driver. We found mosaic 11p15.5 changes in ~20% of HB patients’ non-tumor liver, marking preneoplastic cells and affecting liver function (Pilet et al, Nat Commun 2023).
Ongoing projects:
We aim to identify new driver alterations of pediatric liver cancers by expanding our cohort analyzed by whole-genome sequencing and bulk RNAseq. We have specific projects to further explore mosaic alterations at the single-cell and spatial level.
Context:
Tumor evolution is a multi-step process driven by mutations under immune and treatment pressures, leading to diverse phenotypes. We study how clonal changes and phenotypes connect to tumor evolution and resistance.
Published results:
Using RNAseq on 100 HB samples, we defined three transcriptomic groups tied to differentiation, proliferation, and immune response (Hirsch et al, Cancer Discov 2021) (Figure 3a). Multiple samples from the same tumor showed varied phenotypes despite shared drivers, indicating plasticity. Chemotherapy boosts immune infiltration in ‘Hepatocytic’ tumors but not ‘Liver Progenitor’ ones. Single-cell analysis confirmed these groups, revealing a continuum of cell states and subclonal diversity (Roehrig et al, Nat Commun 2024) (Figure 3b).
Ongoing projects:
We currently explore the intra-tumor heterogeneity of hepatoblastoma at the spatially-resolved single-cell level, by combining high-plex immunofluorescence, single-nucleus RNAseq and spatial transcriptomics.
Context:
Hepatoblastoma is treated with cisplatin-based chemotherapy, but some cases resist treatment, with few therapeutic alternatives.
Published results:
Whole-genome sequencing revealed cisplatin’s SBS35 mutational signature (Figure 4a) in a subset of primary hepatoblastomas post-chemotherapy, associated with poor treatment response (Hirsch et al, Cancer Discov 2021 ; Pire et al, Eur J Cancer 2024). In primary tumors, SBS35 mutations were subclonal, meaning they appeared in only a fraction of tumor cells, specifically within ‘Liver Progenitor’ sectors, while ‘Hepatocytic’ and ‘Mesenchymal’ areas lacked them. In contrast, relapse samples showed thousands of clonal SBS35 mutations, present in all tumor cells, indicating relapses arise from a single resistant cell that accumulated cisplatin-induced mutations during treatment (Figure 4b). Overall, the longitudinal analysis of cisplatin-induced mutations, integrated with the transcriptomic classification, pinpoints the ‘Liver Progenitor’ cells as being chemoresistant and at the origin of relapses.
Targeting PLK1, a ‘Liver Progenitor’ marker, reduced tumor growth in proof-of-concept experiments developed with Sandra Rebouissou’s group (Hirsch et al, Cancer Discov 2021).
Ongoing projects:
We’re refining detection of cisplatin mutations with machine learning and, with Sandra Rebouissou’s group, seeking drugs to reverse chemoresistance.
PhD
MD
MD - PhD
Roehrig A, Hirsch TZ, Pire A, Morcrette G, Gupta B, Marcaillou C, Imbeaud S, Chardot C, Gonzales E, Jacquemin E, Sekiguchi M, Takita J, Nagae G, Hiyama E, Guérin F, Fabre M, Aerts I, Taque S, Laithier V, Branchereau S, Guettier C, Brugières L, Fresneau B, Zucman-Rossi J, Letouzé E. Nat Commun. 2024 Apr 8;15(1):3031. doi: 10.1038/s41467-024-47280-x. PMID: 38589411
Pire A, Hirsch TZ, Morcrette G, Imbeaud S, Gupta B, Pilet J, Cornet M, Fabre M, Guettier C, Branchereau S, Brugières L, Guerin F, Laithier V, Coze C, Nagae G, Hiyama E, Laurent-Puig P, Rebouissou S, Sarnacki S, Chardot C, Capito C, Faure-Conter C, Aerts I, Taque S, Fresneau B, Zucman-Rossi J. Eur J Cancer. 2024 Mar;200:113583. doi: 10.1016/j.ejca.2024.113583. Epub 2024 Feb 1. PMID: 38330765
Pilet J, Hirsch TZ, Gupta B, Roehrig A, Morcrette G, Pire A, Letouzé E, Fresneau B, Taque S, Brugières L, Branchereau S, Chardot C, Aerts I, Sarnacki S, Fabre M, Guettier C, Rebouissou S, Zucman-Rossi J. Nat Commun. 2023 Nov 6;14(1):7122. doi: 10.1038/s41467-023-42418-9. PMID: 37932266
Hirsch TZ, Pilet J, Morcrette G, Roehrig A, Monteiro BJE, Molina L, Bayard Q, Trépo E, Meunier L, Caruso S, Renault V, Deleuze JF, Fresneau B, Chardot C, Gonzales E, Jacquemin E, Guerin F, Fabre M, Aerts I, Taque S, Laithier V, Branchereau S, Guettier C, Brugières L, Rebouissou S, Letouzé E, Zucman-Rossi J. Cancer Discov. 2021 Oct;11(10):2524-2543. doi: 10.1158/2159-8290.CD-20-1809. Epub 2021 Apr 23. PMID: 33893148
Hirsch TZ, Negulescu A, Gupta B, Caruso S, Noblet B, Couchy G, Bayard Q, Meunier L, Morcrette G, Scoazec JY, Blanc JF, Amaddeo G, Nault JC, Bioulac-Sage P, Ziol M, Beaufrère A, Paradis V, Calderaro J, Imbeaud S, Zucman-Rossi J. J Hepatol. 2020 May;72(5):924-936. doi: 10.1016/j.jhep.2019.12.006. Epub 2019 Dec 18. PMID: 31862487
APC germline hepatoblastomas demonstrate cisplatin-induced intratumor tertiary lymphoid structures.
Morcrette G, Hirsch TZ, Badour E, Pilet J, Caruso S, Calderaro J, Martin Y, Imbeaud S, Letouzé E, Rebouissou S, Branchereau S, Taque S, Chardot C, Guettier C, Scoazec JY, Fabre M, Brugières L, Zucman-Rossi J. Oncoimmunology. 2019 Mar 28;8(6):e1583547. doi: 10.1080/2162402X.2019.1583547. eCollection 2019. PMID: 31069152
| Item | Data | Details |
|---|---|---|
| Cell lines | LCCL_description | Description of liver cancer cell lines. |
| Cell lines | Somatic_alterations | List of putative somatic genomic alterations identified in 34 Liver Cancer Cell Lines by whole-exome sequencing. |
| Cell lines | Fusion_transcripts | Fusion transcripts identified in 34 liver cancer cell lines by RNA-sequencing. |
| Cell lines | HBV_fusions | HBV RNA fusions identified by RNA-sequencing among 34 liver cancer cell lines. |
| Cell lines | Drivers | List of the 252 cancer driver genes used to analyze association between drug sensitivity and genetic alterations in 34 liver cancer cell lines. |
| Cell lines | RNAseq_vsn | RNA expression in 34 liver cancer cell lines (vsn) |
| Cell lines | RNAseq_FPMK | RNA expression in 34 liver cancer cell lines (FPKM) |
| Cell lines | microRNAseq_vsn | microRNA expression in 34 liver cancer cell lines (vsn) |
| Cell lines | microRNAseq_RPM | microRNA expression in 34 liver cancer cell lines (RPM) |
| Cell lines | Protein_expression | Expression of 126 proteins analyzed by RPPA in 34 liver cancer cell lines (log2-transformed). |
| Cell lines/Drugs | AUC | Drug sensitivity in 34 liver cancer cell lines. Table of AUC values for each drug |
| Cell lines/Drugs | GI50 | Drug sensitivity in 34 liver cancer cell lines. Table of GI50 values for each drug |
| Drugs | Screened_compounds | Description of screened compounds. |
Illumina MiSeq
Manufacturer’s description: The MiSeq desktop sequencer allows you to access more focused applications such as targeted gene sequencing, metagenomics, small genome sequencing, targeted gene expression, amplicon sequencing, and HLA typing. New MiSeq reagents enable up to 15 Gb of output with 25 M sequencing reads and 2×300 bp read lengths.
Applications currently used: Targeted sequencing (paired-end 2x150bp) using multiplexed PCR for library preparation.
Manufacturer’s description: Based on one of the most widely used, widely trusted sequencing methodologies available (Sanger sequencing) the 3500 Series Genetic Analyzer is designed to deliver the accuracy you demand. The 3500 platform can run a wide variety of applications, including de novo sequencing and resequencing (mutational profiling), microsatellite analysis, MLPA, AFLP, LOH, MLST, and SNP validation or screening.
Applications currently used: Sanger sequencing (read length up to 1000bp)
Manufacturer’s description: The nCounter® Analysis System offers a simple, cost-effective way to profile hundreds of mRNAs, microRNAs, or DNA targets simultaneously with high sensitivity and precision. The digital quantification of target molecules and high levels of multiplexing eliminate the compromise between data quality and data quantity, producing excellent sensitivity and high reproducibility for studies of hundreds of targets. The system uses molecular “barcodes” and single molecule imaging to quantitate up to 800 unique transcripts in a single reaction.
Applications currently used: Gene expression and miRNA analysis on FFPE tumor samples
Manufacturer’s description: Fluidigm’s revolutionary integrated fluidic circuits (IFCs) empower life science research by automating PCR reactions in nanoliter volumes. This means using less sample and reagent, and a single microfluidic device, to achieve the high-quality, consistent results your work depends on. The Biomark HD system runs IFCs in either real-time or end-point read modes, bringing flexible, efficient and economical PCR solutions to a range of applications such as digital PCR, gene expression, genotyping, library preparation for next generation sequencing.
Application currently used: Gene expression, genotyping and library preparation
Manufacturer’s description: The Applied Biosystems 7900HT Fast Real-Time PCR System is the only real-time quantitative PCR system that combines 384-well plate compatibility with fully automated robotic loading. Acknowledged as the gold standard in real-time PCR, the 7900HT system combined with TaqMan®Assays enables you to achieve unprecedented throughput and flexibility, allowing you to pursue projects beyond the scope of previous real-time instruments
Application currently used: Gene expression analysis, allelic discrimination, library quantification
Manufacturer’s description: The Operetta CLS system combines speed and sensitivity with the powerful and intuitive data analysis. It is a combination of technologies with a powerful, stable 8x LED light source for optimal excitation of fluorophores and confidence in results. It contains also a proprietary automated water-immersion objectives with very high numerical aperture enable high resolution and fast read times with minimal photodamage. The confocal spinning disk technology provides a fast and gentle imaging process, enabling efficient background rejection, live cell experiments, and 3D imaging. Its large format sCMOS camera delivers low noise, wide dynamic range, and high resolution for sensitive and quantitative measurements at short exposure time.
Application currently used: Fixed-cell assays, Live-cell assays, Complex cellular models, FRET assays, Phenotypic fingerprinting
Manufacturer’s description: MultiFlo™ FX is an automated multi-mode reagent dispenser for 6- to 1536-well microplates. MultiFlo FX incorporates several unique technologies in its modular design, such as Parallel Dispense, RAD™ Random Access Dispense and the new, patent-pending AMX™ Automated Media Exchange modules to facilitate a variety of liquid handling applications from 2D and 3D cell culture to concentration normalization assays, ELISA, bead-based assays and more. A fully configured MultiFlo FX replaces up to five liquid handlers, saving space, time and instrumentation budgets.
Application currently used: Cell culture for automated reagent dispensing and washing (6to 384 wellplates)
Manufacturer’s description: Fast and accurate determination of a candidate compound’s IC50 provides drug discovery biologists with valuable information for the development of new pharmaceuticals, yet traditional techniques are both time consuming and laborious, with no standardization across the industry. The HP D300 Digital Dispenser offers a simple method for streamlining your workflow, offering picoliter to microliter non-contact dispensing of small molecules in DMSO directly into your assay plate. Using HP’s Direct Digital Dispensing technology, this convenient benchtop solution requires almost no set up time, and single use T8 Dispenseheads virtually eliminate the risk of crosscontamination. It allows rapid delivery of any dose to any well, saving time, minimizing waste of valuable compounds and accelerating drug discovery.
Application currently used: Cell culture for delivery of pharmacological compounds in 96 and 384 well plates
PhD
Pleural mesothelioma is a rare tumor mainly linked to asbestos exposure, characterized by a poor prognosis and an urgent need for precision medicine strategies. Therefore, predicting response to current treatments and developing new therapies that account for the molecular and cellular heterogeneity of pleural mesothelioma is crucial. To meet these challenges, we are focusing our research on 3 major axes:
(1) Deciphering intra-tumour heterogeneity: Single-cell omics and emerging spatial omics approaches will help define the tumour cell subpopulations present in a single biopsy. This will allow us to better understand their plasticity and phenotypic evolution, as well as to dissect the landscape of immune and stromal cells composing the tumour microenvironment.
(2) Developing therapeutic strategy: Functional screening using knockdown and knockout approaches, along with pharmacogenomics studies using our large patient-derived primary cell line biobank will lead to the identification of new therapeutic targets and new effective anti-tumour drugs that consider the phenotypic diversity of tumours.
(3) Identifying biomarkers of response to treatment: We will uncover signatures or biomarkers that predict response to treatment through multi-omics integrated analysis of tumor sample collection from patients enrolled in clinical trials or treated in real-life settings.
These projects are carried out in partnership with several clinical departments and associations. Close collaborations with other research laboratories allow us to explore areas such as the contribution of specific immune subpopulations to treatment resistance in preclinical models, and the identification of risk factors for pleural mesothelioma beyond asbestos exposure.
Our research works has contributed to a better understanding of the molecular alterations but more importantly of the molecular heterogeneity of pleural mesothelioma. We were the first to propose a molecular classification of pleural mesothelioma that goes beyond the histologic classification and identifies specific molecular subtypes linked to mutational status. We also proposed a novel way to describe mesothelioma heterogeneity as a continuum using histo-molecular gradients that consider the main histologic types (epithelioid/sarcomatoid). We highlighted that these histo-molecular gradients identify tumours classified epithelioid at the histologic level, which are engaged towards the sarcomatoid phenotype. They do have strong prognostic value and may guide therapeutic strategies. Recent works have contributed to a better description of pleural mesothelioma anatomic intra-tumour heterogeneity. Importantly, we revealed genetic heterogeneity involving the major tumour suppressor gene NF2, as well as“hot” and “cold” immune profile of the tumour microenvironment depending on tumour positions in the thoracic cavity. Our results support the need to analyse multiple samples from distinct anatomical sites in order to estimate the prognosis or implement precision medicine strategies.
de Reynies et al, Clinical Cancer Research, 2014; Tranchant et al, Clinical Cancer Research, 2017; Blum et al, Nat Commun, 2019 ; Meiller Genome Med. 2021.
PhD
MD, PhD
MD, PhD
PharmD, PhD