Transcriptome profiling (‘NeuroSeq’ method)
For deep transcriptome profiling of neuronal populations we use the method for cell isolation described by Hempel et al. We typically isolate and pool 50-150 neurons from mouse, and around 250-500 neurons for flies and zebrafish, to obtain at least 500 pg RNA for downstream processing. Depending on how broadly labeled the cell population is, this can be done in as little as ½ day, and up to 1 week or more of cell sorting.
After adding the ERCC spike-in controls, we then use the NuGEN Ovation RNA-seq v2 kit to generate cDNA using single-primer isothermal amplification (SPIA), and we make sequence-ready libraries using the Ovation Rapid Library System. Barcoded libraries are pooled and sequenced on the NextSeq to achieve approximately 30M-50M reads per sample. This un-stranded assay provides whole gene body coverage suitable for isoform-level analysis.
For single-cell transcriptome profiling we have implemented the Drop-seq method described by Macosko et al. Cell suspensions of neuronal populations are prepared as done for the NeuroSeq method and are fed into a microfluidic device to capture single cells in a water-based droplet containing a bead coated with the primers necessary for 3’ digital gene expression (3’DGE).
We recently implemented a hybrid scRNA-seq approach based on Smart-seq2 (Picelli et al) and SCRB-seq (Soumillion et al) to produce 3’ DGE data for single mouse neurons, at a fraction of the cost of commercially available reagent kits. This method can be adapted for un-stranded, whole gene body coverage as well. Together with our colleagues in the Cell Culture shared resource, we have implemented a FACS-based version to complement our hand-sorted method, and we are also pursuing a “low cell” version to profile as few as 10 pooled fly neurons, with the ultimate goal of profiling single fly neurons.
We are working with several labs at Janelia to adapt ATAC-seq, Nascent-seq, and other assays for use with specifically labeled neuronal populations at low cell inputs. We have also generated libraries and sequence data for several ongoing genome and transcriptome assembly projects, and we are currently investigating the use of long reads from nanopore sequencing to improve genome assembly.