Integrated electrophysiological and genomic profiles of single cells reveal spiking tumor cells in human glioma

Abstract

Prior studies have shown that glioma cells form synapses with neurons to receive synaptic inputs. To discern if glioma cells can send outgoing electrochemical signals in the form of action potentials (APs), we employed Patch-sequencing on surgically-resected human glioma slices. Results showed that half of patched cells in IDH1 mutant (IDH1mut) tumors demonstrate select properties of both neurons and glia and fire single, short APs. To define the transcriptional profiles of these hybrid cells (HCs) and discern if they are tumor derived, we developed a computational tool, Single Cell Rule Association Mining (SCRAM), to annotate features in each cell individually. SCRAM revealed that HCs represent a heterogenous group of tumor and non-tumor cells that are uniformly defined by both GABAergic neuron and oligodendrocyte precursor cell (GABA-OPC) transcriptional signatures. We found that GABA-OPC tumor cells express requisite voltage-gated ion channels needed to fire APs. We validated our findings in human single cell and bulk RNA-seq datasets, confirming that GABA-OPCs represent 40% of IDH1mut tumor cells, correlate with survival outcomes in IDH1mut human patients and are also found in select molecular subtypes of IDH1 wild-type tumors. These studies describe a new cell type in IDH1mut glioma with unique electrophysiological and transcriptomic properties.