Abstract
Deciphering mesoscopic connectivity of the mammalian brain is a pivotal step in neuroscience. Most imaging-based conventional neuroanatomical tracing methods identify area-to-area or sparse single neuronal labeling information. Although recently developed barcode-based connectomics has been able to map a large number of single-neuron projections efficiently, there is a missing link in single-cell connectome and transcriptome. Here, combining single-cell RNA sequencing technology, we established a retro-AAV barcode-based multiplexed tracing method called MEGRE-seq (Multiplexed projEction neuRons retroGrade barcodE), which can resolve projectome and transcriptome of source neurons simultaneously. Using the ventromedial prefrontal cortex (vmPFC) as a proof- of-concept neocortical region, we investigated projection patterns of its excitatory neurons targeting five canonical brain regions, as well as corresponding transcriptional profiles. Dedicated, bifurcated or collateral projection patterns were inferred by digital projectome. In combination with simultaneously recovered transcriptome, we find that certain projection pattern has a preferential layer or neuron subtype bias. Further, we fitted single-neuron two-modal data into a machine learning-based model and delineated gene importance by each projection target. In summary, we anticipate that the new multiplexed digital connectome technique is potential to understand the organizing principle of the neural circuit by linking projectome and transcriptome.
Highlights
MERGE-seq recovers single-neuron transcriptome and projectome simultaneously
MERGE-seq partitions projection pattern into neuronal layer and subtype
MERGE-seq bridges projection pattern to differentially expressed genes
Machine learning models interpret relationship between digital projectome and gene expression
Competing Interest Statement
The authors have declared no competing interest.