Cellular diversity in the <i>Drosophila</i> midbrain revealed by single-cell transcriptomics.

Croset V, Treiber CD, Waddell S. · 2018

Read Cellular diversity in the <i>Drosophila</i> midbrain revealed by single-cell transcriptomics. by Croset V, Treiber CD, Waddell S. (2018) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc5927767.txt.

Introduction

Croset and co-authors reveal cell diversity in the fruit fly midbrain by single-cell methods. Their article appeared in eLife. They profiled thousands of single cells with Drop-Seq. Wiring maps alone cannot show how brain cells signal. Gene profiles add the missing molecular layer. Profiles were assigned to major regions and cell types.

Each neuron could be typed by its transmitter, they report. Some peptide signals clustered with certain fast transmitters. Receptor patterns hint at rich circuit modulation options. The atlas links molecular traits to brain regions. It covers hubs for memory and navigation. Researchers get a resource for circuit-level questions.

Catalog record

CREATORCroset V, Treiber CD, Waddell S.
YEAR2018
COLLECTIONPAPERS
FORMATTXT
FILE SIZE101 KB
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC5927767

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Cellular diversity in the Drosophila midbrain revealed by single-cell transcriptomics By Vincent Croset, Christoph D Treiber, Scott Waddell, Mani Ramaswami Published in: eLife DOI: 10.7554/eLife.34550 ABSTRACT To understand the brain, molecular details need to be overlaid onto neural wiring diagrams so that synaptic mode, neuromodulation and critical signaling operations can be considered. Single-cell transcriptomics provide a unique opportunity to collect this information. Here we present an initial analysis of thousands of individual cells from Drosophila midbrain, that were acquired using Drop-Seq. A number of approaches permitted the assignment of transcriptional profiles to several major brain regions and cell-types. Expression of biosynthetic enzymes and reuptake mechanisms allows all the neurons to be typed according to the neurotransmitter or neuromodulator that they produce and presumably release. Some neuropeptides are preferentially co-expressed in neurons using a particular fast-acting transmitter, or monoamine. Neuromodulatory and neurotransmitter receptor subunit expression illustrates the potential of these molecules in generating complexity in neural circuit function. This cell atlas dataset provides an important resource to link molecular operations to brain regions and complex neural processes. INTRODUCTION Neuroscience is typically studied at the systems, cellular, or molecular level. However, it will be necessary to bridge these traditional boundaries to fully understand how the brain operates. Such a momentous task is somewhat simplified if analyses are focused on an animal with a relatively small brain, but where systems-level processes are evident. In many respects, the vinegar fly Drosophila melanogaster fits the bill (Haberkern and Jayaraman, 2016). Drosophila have an estimated 150,000 neurons in the entire brain, of which the optic lobes, or visual neuropils, comprise two thirds of this neural mass. The remaining approximately 50,000 neurons, or midbrain, houses many key neural structures such as the mushroom bodies and central complex, which are, amongst other things, critical for memory-directed behavior (Cognigni et al., 2018) and navigation (Seelig and Jayaraman, 2015), respectively. Recent large-scale electron-microscopy projects have generated wiring diagrams, or connectomes, of parts of the larval and adult fly nervous system (Berck et al., 2016; Eichler et al., 2017; Ohyama et al., 2015; Takemura et al., 2013; Takemura et al., 2017a; Takemura et al., 2017b; Tobin et al., 2017; Zheng et al., 2017). While these efforts are an essential part of the quest to decipher brain function, they are not enough. Genes determine the anatomy and mode of connectivity, the biophysical properties, and the information-processing limits of individual constituent neurons. Therefore, understanding any given wiring diagram requires a systematic view of gene expression within their functionally relevant cellular context. With this knowledge in hand, investigators can begin to examine how gene products contribute to cell- and circuit-specific functions and, ultimately, organismal behavior. New developments in single-cell sequencing technology provide a unique means to generate such a brain-wide view of gene expression with cellular resolution. Massively parallel approaches, such as Drop-seq (Macosko et al., 2015), permit simultaneous analysis of the transcriptomes of 1000 s of individual cells. In brief, each cell from a dissociated tissue is first captured with an oligonucleotide bar-coded bead in a nanoliter aqueous droplet. Inside each droplet, the same cell identifier sequence becomes attached to all mRNA molecules from an individual cell. Following this critical cell-specific hybridization step, all the material from 1000 s of individual cells can be pooled and processed together for mRNA sequencing. Drop-seq therefore provides the means to access the transcriptomes of a representation of most cells in the fly midbrain. A key hurdle in generating a single-cell atlas of the brain is the ability to assign individual transcriptome profiles to the correct cell, or at least cell-type. Again, using an animal whose brain has an intermediate number of neurons and presumably neural diversity simplifies the task. Moreover, years of genetic analyses in Drosophila have provided a considerable number of established transgenic and intrinsic markers for specific brain regions and cell-types. These identifiers often allow one to extract the relevant cell profiles from the larger dataset. Here we report the application and an initial analysis of Drop-seq data to investigate the cellular diversity of the Drosophila midbrain. We demonstrate the ability to assign many single-cell profiles to identified cell-types and brain regions, and identify novel markers for these regions. Moreover, cells can be robustly classified based on their neurotransmitter profile. We find that certain neuropeptides preferentially accompany particular fast-acting transmitters, or monoamines. In addition, we detail the apparent complexity of modulatory and neurotransmitter receptor subunit expression. This single-cell dataset provides an indication of the extent of neural diversity in the fly brain, and provides essential cellular context linking molecules to neural circuits and brain function. RESULTS Drop-seq analysis of the Drosophila midbrain We first optimized the conditions required to effectively dissociate and capture individual Drosophila melanogaster cells with DNA bar-coded microparticles in aqueous droplets, using a commercially available apparatus. Drosophila neurons are about a tenth of the size of mammalian cells. We therefore first verified the efficiency of processing insect cells and of single-cell capture by generating single-cell transcriptomes attached to microparticles (STAMPs) from a cell suspension comprised of a 1:1 mixture of Drosophila S2 and Spodoptera frugiperda (fall armyworm) Sf9 cultured cells. We then sequenced these S2/Sf9 STAMPs (Figure 1—figure supplement 1A). This procedure retrieved 764 barcode-associated transcriptomes, of which 368 were identifiable as Drosophila and 384 as Spodoptera. Importantly, only 12 transcriptomes contained cDNA coming from both species (Figure 1—figure supplement 1B), indicating that only 3.2% of all sequenced transcriptomes resulted from capturing two cells together. This analysis suggested that the Drop-seq system and our chosen parameters are suitable for barcoding single insect cells and are optimized to minimize capture of cell doublets. We next used these same parameters and concentrations to collect STAMPs from thousands of cells from the Drosophila midbrain, in eight independent biological replicates, over eight different days. Each day we isolated single-cells from 80 to 100 dissected brains taken from an equal number of male and female flies. Brains were removed from the head capsule, optic lobes were manually dissected away and a single-cell suspension was prepared from the remaining fly midbrains. Larger tissue fragments were removed by filtration, and the eluant was inspected under a microscope to confirm the presence of single cells and the absence of clumps. Cells were individually paired with DNA barcoded beads and cDNA libraries were generated from bead-bound single-cell transcriptomes, and sequenced (Figure 1A) (Macosko et al., 2015). Pooling the data from the eight independent experiments resulted in a dataset of 19,260 cells, with each containing between 200 and 10,000 unique molecular identifiers (UMIs) and therefore, single mRNA molecules. We performed a Principal Component Analysis (PCA) on these transcriptomes and reduced the top 50 PCs into two dimensions using t-SNE (Van Der Maaten, 2014) (Figure 1—figure supplement 2A). We selected the cut-off for the optimal number of UMIs per cell to be included in our analyses by generating t-SNE plots from data with a variety of quality thresholds. These analyses revealed that discarding cells with less than 800 UMIs, resulted in a data set of 10,286 high quality cells, segregated with k-means clustering into 29 cell clusters, with several corresponding to most of the known iterative, or large populations of, cell types in the Drosophila brain (Figure 1—figure supplement 2B). More stringent

Read the complete text