Skip Navigation

Pedro Lowenstein

Personal Information


  New Search

Primary Appointment: Neurosurgery
Primary PIBS Dept.: Cell and Developmental Biology
PubMed Name: Lowenstein PR
Department Website
Lab Website

  Whether cancer growing in vivo is self‐organized remains poorly understood. Our data demonstrate that brain tumors are self‐organized. We do so by illustrating the existence of various multicellular structures in brain tumors in vivo, which are counterparts of collective motion of cells in vitro. The large scale organization of brain tumors was studied in several in vivo models, including transplantable and genetically induced gliomas using the Sleeping Beauty Transposase system expressing the following genes which determine brain tumor development and progression in human patients: i.e., N‐Ras, Large T antigen from SV40 (LgT), down regulation of p53 expression (sh‐p53), overexpression of PDGF, downregulation of ATRX expression (sh‐ATRX), and expression of mutant IDH1 (IDH1‐R132H). Human GBM were also studied. Oncostreams, similar to collective motion of cells in vitro, support the existence of glioma self-organization in vivo. Oncostreams are 10‐20 glioma cells wide, exhibit variable lengths, and are evenly distributed throughout the tumors. Within oncostreams, cells are fusiform, and oncostreams that reach the tumor borders participate in glioma invasion. Oncostreams also join other streams, and are seen near pseudopalisades and at the tumor interface with normal brain tissue. Using advanced network bioinformatics we identified fyn protein kinase as a potential regulation of oncostream formation and tumor malignity. Mathematically, oncostreams can be explored using agent based modeling. Our agent based model demonstrates that cells need to be elliptical in order to allow the formation of streams and swirls. Details of the cellular interactions, including the dynamics between tumor cells and surrounding normal brain, the role of intercellular adhesion, the repulsion rate between tumor cells, and the replication rate, were carefully studied using our agent-based model. Our model mimics the formation of similar multicellular structures as seen in vivo, and will allow predictions to be tested experimentally. We believe streams enhance migration of slower moving glioma cells, block entry of immune cells, and accelerate tumor growth. We also detected the presence of vortices or swirls in which cells appear to rotate around a central axis. Vortices were found in all models, independently of causative genetic mutations. We hypothesize that vortices are cores of glioma stem cells that nucleate brain tumor growth around them. We also detected the presence of tumor cells organized as spheres, some being discharged into the ventricular fluid. Dynamics and interactions between swirls, neurospheres and streams, with the ventricles and normal surrounding brain, as well as their contribution to cancer growth and progression is an exciting novel research area. We conclude that brain tumors display large scale non‐random self‐organization, which affects tumor progression and represents a novel therapeutic target.
There are several projects offered to students who aim to rotate in my laboratory:
1. Self-organization in brain tumors: molecular and cellular basis, bioinformatics and sequencing.
2. Viral-induced brain tumors: testing new therapies in gliomas containing human mutations
3. Immune therapeutic approaches to treat brain tumors
4. Mathematical modeling of experimental brain tumors (high level math skills required for this project)