First-Ever Protein Maps Will Help Scientists Understand What Makes Cells Healthy

There's a vast world inside us we've barely explored. Thousands of different types of proteins live and work in our cells, each with a unique location and function. Just as ecologists can't describe a forest without mapping its trees, we'll never understand our own biology if we can't visualize this cellular ecosystem.

That's why Brenda Andrews, a cell biologist at the University of Toronto, and her colleagues have just constructed the very first map depicting the location and abundance of roughly 3,000 different types of proteins inside the cell. Armed with robotic assistants and sophisticated artificial intelligence programs, the team charted out the location of proteins in a whopping 20 million cells and explored how this molecular ecosystem changes in response to drugs and genetic mutations.

"We first need to know how proteins move inside the cell if we are to understand what makes cells healthy and what makes them diseased," said Andrews in a statement.

These maps and the pipeline behind them, described today in the journal CellCell, offer biologists a powerful new tool for visualizing what happens when disease strikes. Ultimately, Andrews and her colleagues hope this will help researchers design better treatments.

"The major impact here will be a resource for eukaryotic cell biologists to be able to go to our website and our database, and understand where the proteins they're interested in are localized, and how they change in response to perturbations," Andrews said in a video interview.

Genes are life's blueprints, but it's the thousands of proteins those genes encode that actually make our biology work. Proteins build new structures, relay messages, transport molecules and drive nearly all metabolic reactions in the cell. They're not working in isolation—rather, each individual protein contributes to an enormously complex system we've barely begun to grasp.

To understand how an ecosystem functions, ecologists will chart out where its species live, how large their populations are, and how they migrate about. A similar approach could offer deep insights into the cell's inner workings, but only very recently have scientists developed the molecular and computational tools to seriously consider the possibility.

To build the very first cellular protein map, the researchers turned to yeast, a single-celled critter that contains roughly a quarter the proteins of most human cells, including many overlapping ones. This wasn't any garden-variety microbe, but rather, a highly engineered strain in which each protein-encoding gene has instructions for a fluorescent tag tacked onto the end. The result is that any proteins made by these genetically-augmented cells come with a tiny green lightbulb attached, allowing scientists to count and track them under a microscope.

In an effort that makes the rest of humanity look downright lazy, Andrews and her colleagues collected data on 20 million individual cells over the course of a decade, visualizing and counting roughly 3,000 proteins. Naturally, they had robotic aids at their side handling and imaging the cells, in addition to sophisticated machine-learning algorithms that chugged through the prodigious amount of data the study produced.

"The reason we need to do it on a large scale is because there simply are so many proteins," said Andrews.

This wasn't just a tremendous bean-counting exercise, but rather, a first attempt to build a holistic picture of how the cellular ecosystem responds to disturbance. Since environmental and genetic changes can both cause widespread shifts in protein expression, the researchers charted protein movement and abundance in cells exposed to several types of drugs, in addition to cells bearing a genetic mutation that affects protein expression. As expected, the protein maps of these cells differed markedly from their healthy counterparts.

"A lot of the regulation that happens within cells, which is critical for the basic functioning of the human body, influences where individual proteins are localized and how they move around. It is very important to understand how this regulation happens if we are going to be able to understand why cells are healthy and why they are sometimes diseased," said Andrews.

In the future, researchers could use the automated pipeline Andrews and her colleagues have developed to visualize cell-wide responses to any number of drugs or genetic factors.

"What this allows someone to do is follow protein localization or abundance under numerous different genetic or drug perturbations," study co-author Charles Boone told me. "This study opens the door to doing this analysis many times over so that we can say, for a given protein, what are all the genetic conditions under which its localizations is changed."

As a next step, Andrews and her team are working on the first protein maps of human cancer cells, in an effort to pinpoint the cellular origins of the disease and hunt for new treatments.

"We want to understand how all proteins are moving, at a systems level, in cancer cells upon, say, a treatment with a drug or genetic perturbation, so that we can identify vulnerabilities in cancer cells, in terms of protein localization and abundance, and start thinking about how to best target those changes," said study co-author Jason Moffat in a statement.

Given the revolutionary advances in medicine over the past generation, it's a bit strange to think we've just barely begun to understand how cells really work. Strip away all the high tech equipment and computational power, and molecular biologists today are not so unlike 19th century Darwinian explorers. We're taking our very first steps into the cellular forest, and we're going to be astounded by what we discover.