Seattle’s Allen Institute for Brain Science has released the first database of brain cell types, along with new analysis to improve treatment of deadly brain cancer.

Share story

Three years ago, Microsoft co-founder Paul Allen doubled down on brain research, luring some of the nation’s top experts to Seattle and pledging an additional $300 million to study the world’s most complex organ at a level of detail never before possible.

Now, the Allen Institute for Brain Science is releasing initial results from that effort: The first online database that profiles individual neurons based on shape, activity and other factors — similar to the way baseball fans size up players from their batting averages and positions.

The institute is also unveiling a more in-depth analysis of a deadly type of brain cancer to help spur development of new treatments.

“Worldwide, this is utterly unique,” the institute’s chief scientific officer, Christof Koch, said of the cell database. “But it’s only a down payment.”

The first catalog includes profiles of about 240 neurons, but the goal is to expand quickly by using industrial-scale methods. “In order to do this systematically, we have to look at tens of thousands of cells,” said Koch, who left Caltech to lead the project.

As always, the institute’s new resources are freely available to scientists around the world.

“You can be at big pharma or working in Timbuktu — anybody can download it,” Koch said.

The human brain is made up of 200 billion cells, on a par with the number of stars in our galaxy. Almost half those cells are neurons, which are often lumped together as gray matter.

But neurons come in a dazzling array of shapes and sizes, and scientists are just beginning to understand the differences between them.

“They look a little bit like trees in a forest,” Koch said. “Some are dense, some are open, some have branches close to the ground, others have branches high above the ground.”

One of the institute’s long-term goals is to characterize all those cell types and assemble the neurological equivalent of chemistry’s periodic chart of elements.

“If we want to understand brain disease, what goes wrong in schizophrenia or Alzheimer’s, it’s critical to understand how many different cell types there are and how they differ in function,” Koch said.

Because the human brain is so complex, the initial database was drawn from the portion of the mouse brain responsible for vision. But work is already under way on a human version.

Koch and his team estimates the mouse visual cortex contains at least 42 types of cells. In humans, estimates range up to 1,000 types of neurons.

The new database is both visual and quantitative, including high-resolution micrographs of each neuron, three-dimensional models that can be rotated on screen, and graphs that trace the electrical firing patterns through which brain cells communicate with each other and the rest of the body.

“When you feel pain or happiness, when you hear or see, that is all conveyed by electrical activity in the brain,” Koch said.

Future iterations of the database will add information on which genes are most active in each type of cell.

Initially, the database will be most useful to researchers working to build computer models that attempt to mimic the function of the normal brain and which may help reveal what goes haywire in disease, Kock explained.

One disease the Allen Institute had explored in great detail is glioblastoma multiforme, an aggressive and deadly type of brain cancer. The majority of people diagnosed with the disease survive little more than a year and there are few treatments.

In collaboration with Seattle’s Swedish Neuroscience Institute and the Ben and Catherine Ivy Foundation, scientists at the Allen Institute have been working for several years to understand the disease’s extraordinary malignancy and find ways to stop it.

Now, they’ve updated an online atlas with new details on the anatomy of tumors from 41 patients. They also teased out the genetic differences between, for example, cells at a tumor’s core and cells on the periphery that invade healthy tissue. And they’ve compiled genetic data on what are suspected to be cancer stem cells — cells within a tumor that may drive metastasis and regrowth after treatment or surgery.

The hope is that other researchers will be able to take the new insights and use them to develop new drugs or treatments, said Ralph Puchalski, a leader of the glioblastoma studies at the Allen Institute.

“We hope people will leverage this important information and build on it to develop … personalized tools for patients,” he said.

Several promising new therapies are being tested now, said Nameeta Shah, a research scientist at Swedish involved in the project. But most appear to work only on a subset of patients.

She and her colleagues want to develop genetic profiles to identify which of the new treatments will work best for individual patients. They’re also launching clinical trials to identify new combinations of existing drugs that might work in some people.

“I think in the next two to five years, we should see some dramatic impact of all the collective research,” Shah said.

The stem cell work is particularly intriguing, she added. Many scientists are convinced that the most effective way to stop tumors from recurring may be to wipe out cancer stem cells.

“If you can develop drugs that will target these stem cells, then maybe you can cure the disease,” she said.