Select Country
Follow Us
Personalized Medicine

Human Protein Atlas Charts the Road to Personalized Medicine

Over a decade ago, the Human Genome Project gave us the first blueprint of our genetic code, opening the door to a future where medical interventions could be personalized for each patient’s genetic composition. Today, programs like the Human Protein Atlas are zooming in even deeper, mapping out not just the DNA that defines our bodies, but also the building blocks – specifically, the proteins – that make them tick (or sick).

We’ve known for a while that human DNA holds about 20,000 human genes that code for proteins. But it was only late last year that scientists led by Mathias Uhlén of the KTH Royal Institute of Technology in Stockholm, Sweden, published the first comprehensive open-source map of 17,000 human proteins, showing where they are, and how they function in the human body.

Uhlén and his team are now drilling deeper into the data and seeking to analyze the remaining proteins. Their work, aided by super-fast protein purification technology developed by GE Healthcare Life Sciences, could have broad implications, from improving the understanding of human tissue at the molecular level to developing new diagnostic methods and treatments for cancer and other difficult-to-treat diseases.

“The protein is the work horse of the cell and the body,” Uhlén says. “We’ve provided for the first time to the scientific community a comprehensive list of all the proteins which are in the kidney, the brain and [other organs] so you get a descriptive view of the molecular constituents of each cell in the human body.”


Above: Human DNA is made from just four nucleic acids: Adenine, Cytosine, Guanine and Thymine. Top: Proteins like hemoglobin have extremely complex molecules.

In addition to creating a “periodic table” of the cellular building blocks, their work uncovered several findings related to the distribution of human proteins.

For example, the project has shown that almost half of all proteins are found in all tissues, while very few are unique to any one kind of tissue. That means, Uhlén says, that 30 percent of all drugs used today target proteins that are found in almost every cell in the human body.

“That’s, of course, very interesting and important for the pharmaceutical companies when they’re developing new drugs for new targets,” he says. “[They] can now go to the Protein Atlas and look at where these target proteins are in the human body.”

Beyond that, the Atlas could help advance personalized medicine by providing more information about biomarkers for different diseases. It could give medical researchers a better understanding of cancer since it includes samples from the 20 most common forms of the disease.

Uhlén his team started by identifying all of the 20,000 human genes that code for a protein and then cloning the genes in E.coli bacteria to produce the different proteins.


An illustration of E.coli bacteria. 

The next step involved using GE Healthcare’s ÄKTAexpress protein purification system, which allowed them to purify the individual proteins.

Once purified, Uhlén and team vaccinated animals with the proteins, causing them to make protein-specific antibodies that show up in different tissues and give away a protein’s location in the body. “Our system is a perfect fit for this type of study,” said Jill Simon, Global Product Manager for ÄKTAexpress at GE Healthcare.

Producing and purifying each protein can be a very time-intensive manual process. By using 14 GE Healthcare instruments, Simon and her team estimate GE helped save the Human Protein Atlas more than 23,000 hours – or 2.7 years – of manpower.


Antibodies (blue and yellow) attacking a cancer cell (red).


“This project would have been impossible if we didn’t have the fantastic instruments from GE Healthcare,” Uhlén says. “In order to do this, you need to automate different steps or it gets too expensive. Together with GE Healthcare, we developed an automated way of doing purification of the antibodies which is really working very well.”

Still, antibodies can be tricky to work with, Uhlén says. Sometimes, they help achieve good results, but other times they lead to false positives. Says Uhlén: “We want to clean that up in the next five or six years and, by 2020, have a complete map of where the proteins are in the human body.”

Subscribe to our GE Brief