At the time a patient is first diagnosed with a tumor, the tumor may be composed of millions to billions of tumor cells.
These cell populations have already diversified likely due to various reasons such as genetics, epigenetics and environmental influences, producing a tumor that is highly heterogeneous. This diversity, known as intratumoral heterogeneity, is commonly thought to be one of the key mechanisms for therapeutic failures. An additional layer of tumor complexity is provided by the normal cells. Their pathological potential is linked to a wide variety of tumor cell properties, many of which are driven by intercellular communication mechanisms. Our laboratory studies how information from cells is conveyed to other cells in the context of glioblastoma, a representative model of tumor heterogeneity, and how this process is manifested in phenotypic heterogeneity. We also use our new knowledge to better experimental model systems faithful to human tumors and to rationally design combination therapy. In sum, we confront the tumor complexity, with a focus on cell-cell communications, that needs to be integrated into future medicine, and the testimony is below.
Mechanisms of cell-cell interactions
Communication between cancer cells and immune cells is a key determinant of the glioblastoma ecosystem and its response to therapies, but remains poorly understood.
In this work, we leveraged single-cell RNA-sequencing (scRNA-seq) of human samples and mouse models, deconvolution analysis of bulk specimens from The Cancer Genome Atlas (TCGA) and functional approaches to dissect cellular states and cross-talk in glioblastoma.
With molecular underpinnings, we demonstrated that macrophages induce a transition of glioblastoma cells into mesenchymal-like (MES-like) states. We also showed that MES-like glioblastoma states are associated with increased T cells cytotoxicity and potentially with better clinical response to immunotherapies.
Overall, our work dissects the cellular interactions within the glioblastoma microenvironment and provides opportunities to modulate intratumoral heterogeneity for glioblastoma immunotherapy.
Phenotypic heterogeneity and plasticity
Our investigation using single-cell RNA-sequencing (scRNA-seq) on clinical specimens from IDH-wild-type glioblastoma patients led to the identification of four transcriptional programs (cellular states) in malignant cells as major drivers of glioblastoma heterogeneity.
This study was the most comprehensive scRNA-seq analysis of glioblastoma at the moment of publication in 2019, extensively analyzing 28 tumors from both adults and children. Importantly, integrating multiple sources of molecular heterogeneity (e.g., genetics, epigenetics, microenvironment), the observed convergence of diverse malignant cells to a limited set of recurrent signatures opens up new therapeutic windows for glioblastoma. In addition, by coupling scRNA-seq to uniquely barcoded single cells in vivo, we demonstrated plasticity between states and the potential for a single cell to generate all four states.
This work provides a roadmap for glioblastoma by elucidating the cellular programs of malignant cells and their plasticity.
We developed a next-generation FLPo-FRT-based glioma mouse model that recapitulates malignant features seen in human patients.
This model, derived using lentiviral vectors and two recombination systems, increases the scope of glioma research by enabling (i) sequential induction of oncogenic events rather than delivering those to cell of origin at the same time, (ii) temporal depletion of genes in established tumors to perform a proof-of-concept study for therapeutic feasibility, (iii) specific depletion of genes in the stroma, and (iv) lineage tracing of tumor cell/stroma subpopulations.
The transgenic animal from this study is now available from Jackson laboratory (RRID: IMSR_JAX:033116; B6.Cg-Tg(Gfap-flpo)62Thara/J).