Image-based transcriptomics allows the quantitative analysis of transcripts from thousands of genes in thousands of single cells at single-molecule resolution. The information obtained on the subcellular patterning of transcript molecules, and the cell-to-cell variability in this patterning, can be used to recognize genes with a similar biological function. Image-based transcriptomics can also be combined with other fluorescent stains to link these properties of the single-cell transcriptome with the biology of single cells.

Monitoring the activity of many genes by measuring the transcriptome has become a routine activity in many biological and medical research labs. Transcriptomics is a very powerful method to identify mechanisms that regulate genes and to identify genes, whose altered activity contributes to disease (Soon et al. 2013). However, current methods of transcriptomics, such as next-generation RNA-seq, only measure the abundance of the transcript molecules. As they destroy the tissue, they lose information regarding cellular microenvironments, single cells and the subcellular localization of single transcript molecules. This makes it difficult to combine current methods of transcriptomics with microscopic staining methods that are complementary and equally important tools for biological and medical research.

Image-based transcriptomics is designed to preserve as much biologically meaningful information as possible. Single molecule branched DNA fluorescence in situ hybridization creates a signal for each transcript molecule that is specific, robust and bright enough for automation. Reagents can be obtained by Advanced Cell Diagnostics and Affymetrix. You can find details of the experimental protocol in Battich et al. and on our protocol page.

ComparisonOnucVsBDNA_webpage

Image analysis can be highly parallelized with iBrain and combines open-source software and custom algorithms. With the automated setup described in Battich et al. it is possible to perform over 2000 experiments on a single day, where each experiment monitors a different transcript or a different experimental perturbation in at least 5000 single cells.

Pioneering studies with a low number of genes or a low number of single cells have observed that the amount of transcript molecules is different among single cells. Thus the average amount of transcript molecules across a tissue is an incomplete readout of gene activity and can mask properties of gene activity vital to living organisms (Raj et al. 2010). Image-based transcriptomics enables, for the first time, to absolutely quantify the amount of many different transcript molecules in thousands of cells. We found that 1000 cells are required to accurately describe the differences in transcript molecule abundance for most human genes. Thus image-based transcriptomics allows to identify the causes of the differences in transcript abundance between single cells.

However, it is not clear if the current focus of transcriptomics on transcript abundance only reflects biological importance or if it also reflects the bias of existing methods. As posttranscriptional regulation and function of transcripts occurs in spatially confined compartments (Wippich, Bodenmiller, Trajkovska, Wanka, Aebersold, Pelkmans, 2013), we developed computational tools to characterize the subcellular localization of individual transcript molecules. We found that transcript molecules of genes with a similar biological function also have similar spatial distributions and similar variability in such distributions among individual cells. These spatial properties in the localization of individual transcript molecules within single cells can outperform measures of transcript abundance and variability in abundance when identifying genes that have a similar biological function (Battich et al.).

OverviewPopulation_webpage

Thus image-based transcriptomics is a versatile method for observing transcripts with single molecule resolution in thousands of individual cells. Since it can be incorporated in a wide variety of high-throughput image-based approaches, we expect it to be broadly applicable in studies that quantify phenotypes of single cells and cell populations.

 

Current lab members involved:

Nico Battich

Thomas Stoeger

 

Some relevant publications from the lab:

Battich N*, Stoeger T*, Pelkmans L. Image-based transcriptomics in thousands of single human cells at single-molecule resolution. Nat Methods. 2013 Nov;10(11):1127-33. doi: 10.1038/nmeth.2657. Epub 2013 Oct 06. *Contributed equally [Journal]