Multimodal profiling of single-cell morphology, electrophysiology, and gene expression using Patch-seq

Abstract

Neurons exhibit a rich diversity of morphological phenotypes, electrophysiological properties, and gene-expression patterns. Understanding how these different characteristics are interrelated at the single-cell level has been difficult because of the lack of techniques for multimodal profiling of individual cells. We recently developed Patch-seq, a technique that combines whole-cell patch-clamp recording, immunohistochemistry, and single-cell RNA-sequencing (scRNA-seq) to comprehensively profile single neurons from mouse brain slices. Here, we present a detailed step-by-step protocol, including modifications to the patching mechanics and recording procedure, reagents and recipes, procedures for immunohistochemistry, and other tips to assist researchers in obtaining high-quality morphological, electrophysiological, and transcriptomic data from single neurons. Successful implementation of Patch-seq allows researchers to explore the multidimensional phenotypic variability among neurons and to correlate gene expression with phenotype at the level of single cells. The entire procedure can be completed in ∼2 weeks through the combined efforts of a skilled electrophysiologist, molecular biologist, and biostatistician.

Phenotypic variation within and across transcriptomic cell types in mouse motor cortex

“…deep L5 Martinotti cells with thin long axons reaching all the way to L1. This was partially due to the fact that primary motor cortex in mice is thicker than primary visual and somatosensory cortices where most studies providing anatomical descriptions of neural types have been performed (Jiang et al, 2015;Gouwens et al, 2019), and partially due to the RNA extraction process requiring the aspiration of the cell contents and potentially interfering with the biocytin diffusion (Cadwell et al, 2017). As a result, our study might have missed some morphological variability in this group of cells.…”

“…We used Patch-seq (Cadwell et al, 2016(Cadwell et al, , 2017 to profile neurons transriptomically, electrophysiologically and anatomically ( Figure S1) across all layers of adult mouse MOp (mostly post-natal day P50+, median age P75, Figure S2a) using various Cre driver lines ( Figure S3) to sample as diverse a population of neurons as possible. Neurons in acute slices were patch-clamped and stimulated with brief current impulses to record their electrophysiological activity, filled with biocytin for subsequent morphological recovery and reconstruction, and their RNA was extracted and sequenced using the Smart-Seq2 protocol (Picelli et al, 2013).…”

“…In order to simultaneously obtain electrophysiological, morphological and transcriptomic data from the same neurons, we applied our recently developed Patch-seq protocol (Cadwell et al, 2017), with some modifications. In particular, changes were made to the internal solution to optimize its osmolarity in order to improve staining quality.…”

“…Before each experiment, we combined 494 µL of internal solution with 6 µL of recombinant RNase inhibitor (1 U/µL, Takara) in order to increase RNA yield. The addition of the inhibitor resulted in the increase of osmolarity to the desired value of ∼315-320 mOSM without a further dilution that was described in Cadwell et al (2017). The osmolarity of the ACSF was monitored before each experiment and adjusted to be ∼18-20 mOSM lower than the internal solution by adding sucrose when needed.…”

“…In parallel work, Yao et al perform an integrated transcriptomic analysis of the MOp, identifying 90 distinct neural types. To provide a comprehensive account of the anatomy and physiology of these transcriptomically defined cell types, we used the recently developed Patch-seq technique (Cadwell et al, 2016(Cadwell et al, , 2017Fuzik et al, 2016;Földy et al, 2016) and sampled over 1300 neurons from all cortical layers, combining single-cell RNA-sequencing, patch-clamp recordings and biocytin stainings in the same neurons.…”