The gene expression of individual neurons can be examined following patch clamp experiments by performing single cell PCR on the cytosol aspirated from the patched cell. We have successfully used this approach to correlate the expression of more than 30 genes coding for ion channels and neuronal markers with the electropysiological and morphological properties [1-4]. In addition, we show that there are rules of combinatorial gene expression related to the morpho-electrical properties of different neuron subtypes , which must reflect underlying regulatory principles. But since this technique is limited to a set of approximately 100 genes out of the more than 30,000 genes in the mouse genome, we cannot in this way hope to examine these fundamental regulatory principles in an exhaustive manner.
Furthermore, we have found that the microcircuit contains more than two hundred major morpho-electrophysiological classes with further molecularly defined subtypes. The full extent of this neuronal diversity remains unknown. Our results point to the fact that the gene expression pattern is one of the most reliable ways to map the cellular identity of individual neurons, since classification based on morpho-electrical properties fails to capture all the neuronal subtypes. It is our belief that the transcriptional profile of individual cells reflect the neuronal state and that it can be used to classify the neurons in the brain. Therefore, we wish to apply new techniques that enable us to screen the entire transcriptome within individual neurons and correlate this with the other neuronal characteristics.
The single-cell transcriptome
We are engaged in an extensive effort to implement a new high throughput method that allows us to obtain the full transcriptome of up to 96 individual neurons in one experiment. This method was first described by the group of Sten Linnarsson at Karolinska Institutet using immortalized cell lines (Islam et al, 2011), and we are currently collaborating to implement this on cells originating from tissue. Microfluidics is being explored as a tool to implement this protocol on cell suspension originating from dissociated tissue.
Using this method, we aim to map the neuronal diversity of the somatosensory column based on the expression patterns of individual cells, which, if successful, will provide an exhaustive map of neuronal subtypes in the rodent neocortex. This data will also help identify basic regulatory mechanisms relating to other cellular characteristics, such as electrophysiological properties and morphology.
Single-cell PCR classifier
From the expression profiles of all the neuron subtypes, we aim to identify the minimal set of genes whose cellular expression pattern uniquely defines the neuron subtype. Such a set, together with the single-cell expression pattern, will constitute a novel neuronal classification scheme that can be used to classify subtypes based on single cell PCR, e.g. obtained in combination with patch clamp experiments.