There are many more cell types than you might think. In your liver, for example, the cells in one section may be different than those in another. But researchers have only a general idea of how stem cells differentiate into broad categories (like “liver”) and divide to make new cells.

Understanding each different cell type – and how it interacts with its neighbors over time – is one of the greatest challenges in biomedical research.

A new study led by HSCI researcher Vijay Sankaran of Boston Children’s Hospital and Aviv Regev of the Broad Institute has risen to the challenge. Published in Cell, the new method allows scientists to trace the steps of a cell’s developmental history and create accurate “cell family trees”. By pinning down a cell’s specific origin, it is now possible to gain a clear view on some very complex biology.

The problem

Single-cell genomics allows scientists to study individual cells and learn how they develop into specialized cells. But the origins of specific cell types is still largely unknown.

“It’s like looking at the statistics for a college — you can determine what the average student is like, but you have no idea what any one individual student is doing.” -Vijay Sankaran

“It’s like looking at the statistics for a college — you can determine what the average student is like, but you have no idea what any one individual student is doing,” said Sankaran, who is a hematologist at Boston Children’s Hospital. “Learning about cellular relationships is critical — it can help us understand how many stem cells give rise to any tissue in our body, what cell types cancers emerge from, or how some cells can be dysfunctional in particular diseases.”

What the researchers did

Mitochondria are tiny, energy-producing organelles that are passed from mother to child. There are hundreds of them in every cell. Their DNA is naturally prone to acquiring new mutations, which can act as barcodes to identify each individual cell.

In this study, the group used single-cell gene sequencing on mutations in mitochondrial DNA. The scientists observed as the ‘barcode’ mutations were passed from mother cells to daughter cells, and established cell family trees.

The group also used the barcodes to track individual cells over time, assessing how different cells in a tissue are related to one another.

What they found

The researchers discovered that mitochondrial DNA sequencing can identify a cell’s specific lineage correctly about 95 percent of the time. Furthermore, it can be used on any cell or tissue sample from a living person, since every cell is teeming with mitochondria.

Mitochondria have fewer genes than other cell organelles, so their DNA is relatively easy (and cheap) to sequence. The new approach can be used on a scale 1,000 times greater than standard genome sequencing.

Why it matters

Wayward cell division and differentiation play a large part in many illnesses, so understanding cell relationships has implications for a range of diseases, including cancer.

“One of the problems in studying humans is that we’re limited in our understanding of what happens in them – we can’t say where one cell comes from and what it is giving rise to,” said Sankaran. “Cancer patients often have cancer cells that survive treatment – if we could survey those cells, we could learn which ones survive and what’s different about them.”