CINCINNATI – Researchers at Cincinnati Children’s Hospital Medical Center have described the first integrative analysis of disparate single-cell datasets applied to a single developmental system. The team used this approach to discover a new cellular source of GDNF, a protein that is a required driver of kidney development. Their findings published online Nov. 27 in Developmental Biology.

Nathan Salomonis, PhD, of Biomedical Informatics and S. Steven Potter, PhD, of Developmental Biology led the cross-divisional research team.

The use of single-cell genomics has sharply increased in concert with new technological advances for measuring the transcriptome of individual cells, allowing for the detection and characterization of distinct cell states (the transcriptome is the sum total of all the messenger RNA molecules expressed from the genes of an organism).

Single-cell data workflow image
The integrative workflow used for multi-platform single-cell analysis in Potter, 2017.

While many technologies have been developed to conduct such analyses, researchers have not yet systematically compared those technologies to each other on the same cell types or integrated them in a holistic manner. Such approaches are required to enable the integration of diverse single-cell datasets between laboratories and technological platforms.

New tool supports cross-platform single-cell analyses

The researchers in this study applied a recently developed single-cell computational approach called ICGS in the software AltAnalyze (Olsson et al. 2016, Nature) to directly process and simultaneously analyze data from three separate technologies (Drop-Seq, Chromium 1-x and Fluidigm C1).

“Our computational innovation allows researchers to compare their single-cell experimental results to others or between different single-cell technologies, which has not been previously reported,” says Salomonis. “This paper demonstrates the success of this approach by reporting the first combined use of three independent cutting-edge single-cell genomics technologies.”

After identifying and confirming the presence of 16 distinct cell populations at the active stage of kidney development, the researchers used the new computational strategy to harmonize and compare the cell profiles across all three technological platforms, providing a unified view of embryonic kidney cellular heterogeneity.

Using this approach, the team compared the sensitivity of the different technologies to detect specific genes in specific cellular compartments and also used the cross-platform results to provide independent validations.

They first confirmed their earlier findings that early in development, single cells are undergoing multi-lineage priming, with progenitors sampling multiple available lineage options before committing to a specific fate. As cells differentiate, they activate additional genes associated with the chosen lineage and further repress expression of genes associated with inappropriate lineages.

It had been thought that there was an early and complete demarcation between the cells that make nephrons and those that make the stroma, located between the nephrons. The single cell RNA-seq results, however, show a blurring of this boundary, with nephron progenitors still showing some expression of stromal specific lineage genes, and vice versa. The results show that the complete repression of lineage-inappropriate gene expression can lag the differentiation decision.

Stromal cells generate growth factor GDNF

This study also showed blurring of the nephron/stroma boundary at the functional level. Using these integrated analyses, Potter and his lab team discovered that stromal cells, in addition to nephron progenitor cells, also make the growth factor GDNF.

Images of developing kidney
Illustration showing the creation and comparison of harmonized kidney cell states across diverse single-cell RNA-sequencing technologies.

GDNF has long been known to be a key driver of kidney development. It was previously thought that only nephron progenitor cells, the cells that will eventually make the nephrons (the functional units of the kidney), make this growth factor. This new work shows that another group of cells, the stromal cells (interstitial cells, the cells between the nephrons) also make GDNF.

“There is reason to think that this discovery is functionally important,” he says. “In a prior experiment, it was quite puzzling when a transcription factor (Foxd1) only expressed in stromal cells gave a very strong effect on nephron formation. This new data suggests it might have been the result of altered stromal GDNF expression.”

Potter also says that more recently, unpublished data by another research team has shown that compartment-specific ablation of the GDNF growth factor in the nephron progenitors has little effect on kidney development. “This is consistent with our prediction that stromal expression would compensate for the loss of GDNF from progenitors,” he says.

As such, this work has potential implications in kidney regenerative medicine and organ maturation, which remain a crucial objective of 21st century medicine.

As a resource to the public, the authors have provided an interactive website from which investigators can query these harmonized data (


Funding support

This work was supported by NIH RO1 DK099995 (Potter) and the NHLBI Progenitor Cell Biology Consortium, Administrative Coordinating Center (U01HL099997) (Salomonis).


Read more about Nathan Salomonis’s work in Body’s Cellular Building Blocks Arise from Genetic Tugs of War




Novel computational approach harmonizes data across single-cell technologies, enables insights into kidney development
Tagged on: