Among the bloodletting boxes, ether inhalers, kangaroo-tendon sutures and other artifacts stored at the Indiana Medical History Museum in Indianapolis are hundreds of scuffed-up canning jars full of dingy yellow liquid and chunks of human brains.

Until the late 1960s the museum was the pathology department of the Central Indiana Hospital for the Insane. The bits of brain in the jars were collected during patient autopsies performed between 1896 and 1938. Most of the jars sat on a shelf until the summer of 2010, when Indiana University School of Medicine pathologist George Sandusky began popping off the lids.

Frustrated by a dearth of postmortem brain donations from people with mental illness, Sandusky—who is on the board of directors at the museum—seized the chance to search this neglected collection for genes that contribute to mental disorders.

Sandusky is not alone. Several research groups are now seeking ways to mine genetic and other information hidden in old, often forgotten tissue archives—a handful of which can be found in the U.S., along with many more in Europe. Several technical hurdles stand in the way, but if these can be overcome, the archives would offer several advantages. Beyond supplying tissues that can be hard to acquire at a time when autopsies are on the decline, the vintage brains are untainted by modern psychiatric drugs and are often paired with detailed clinical notes that help researchers make more accurate post hoc diagnoses.

"There are probably a fair number of these collections around the country that grew out of state hospitals," says John Allman, professor of biology at the California Institute of Technology. "It is an untapped resource. If it were carefully planned and reasonably funded, it could become quite a valuable thing."

Celloidin solution
About a dozen facilities in the U.S. today receive postmortem brain donations from people with schizophrenia; collectively, they hold about 700 brains, according to Joel Kleinman, chief of the section on neuropathology at the National Institute of Mental Health (NIMH). His institute's collection is one of the largest, with 232 brains. But getting access to these modern samples is not easy for everyone, he says. "These institutions collected them at great expense, and are not just going to give them up."

A few researchers have instead tapped into much older brain collections, especially for anatomical studies, such as measuring the relative size of white and gray matter or counting neurons. The National Museum of Health and Medicine in Silver Spring, Md., holds one such collection of old brains: thousands of decades-old human brain specimens, many of which originally came from Saint Elizabeth's Hospital, a psychiatric facility in Washington, D.C.

Many of the samples are preserved in celloidin, a hard, rubbery and highly flammable form of cellulose. "The benefit of working with celloidin is that tissue shrinkage is very minimal, and you can see [tissue and cell] structures very clearly under the microscope," says Archie Fobbs, neuroanatomical collection manager at the museum.

But using celloidin-covered samples for genetic analysis is much trickier, as Sandusky's team found out.

Their collection contains brain, heart, liver and spleen tissues from about 1,400 autopsies, 95 of which are labeled "dementia praecox," an antiquated diagnosis similar to schizophrenia. After death most patients' bodies were transferred to built-in metal ice chests located throughout the building, then autopsied within 24 hours.

Many of the samples fixed in celloidin also float in a liquid preservative that is probably formaldehyde or alcohol. If one scooped a gelatinous chunk out of a jar and tossed it against a wall, it would probably bounce like a rubber ball, Sandusky says.

In the summer of 2010 his team tried to break down the preservative with several different chemical concoctions, but to no avail. Frustrated, they put the study on hold and placed the samples on the backburner—or, rather, chilled them in a tank of liquid nitrogen at –180 degrees Celsius.

But nine months later, in the summer of 2011, the researchers decided to take another stab at the analysis. When they retrieved the samples, the celloidin had broken down into little beads. After a few chemical washes, however, the researchers successfully extracted DNA from the tissues.

Sandusky does not know exactly why the method worked, and it has not yet been peer-reviewed. But he says the DNA is of high quality: "I was totally, totally shocked."

If verified by independent groups, Sandusky's technique would be useful because many old specimens, from brains to tumors to animals, are preserved in celloidin, according to Mary Herman, a neuropathologist at the NIMH who has been inspecting brains since 1962. By the 1970s most researchers switched to paraffin wax because celloidin's flammability makes it somewhat dangerous to work with.

The usefulness of celloidin-encased brains "depends on how well they've been cared for and preserved. Some maintenance is required," Herman says. Formaldehyde can be contaminated with bacteria, for example. "The quality of DNA in old celloidin specimens will require careful evaluation."

Diagnostic dilemma
As proof of principle, Sandusky is first screening the DNA for a handful of genetic glitches that researchers have already linked to schizophrenia in blood studies. So far, the search for the genetic culprits of mental illness has been inconclusive and disappointing, pushing researchers to gather as much data as possible, which is one reason old overlooked brain collections are so valuable. He also plans to look for RNA, which is produced when genes are turned on. Whereas DNA analysis reveals genetic mutations written into a person's genome, RNA analysis reveals which genes are actively overexpressed or underexpressed compared with the general population.

Several other experts, however, are skeptical that Sandusky will be able to extract RNA; the fragile molecule degrades more quickly than DNA does. "If DNA is like holding a rock, RNA is like holding a wine glass," says Carlo Colantuoni, an investigator at Lieber Institute for Brain Development in Baltimore.

The DNA could be interesting on its own, though. Researchers are learning that an individual's DNA is not always the same in blood cells and brain cells. For instance, sometimes large genomic deletions and duplications—called copy number variations, or CNVs—crop up in certain tissues after conception. Francine Benes, director of the Harvard Brain Tissue Resource Center, says she has found interesting CNV differences from one brain region to another in 20-year-old schizophrenia brains preserved in paraffin wax. These differences could reflect atypical brain development that might contribute to schizophrenia.

On the one hand, old clinical diagnoses do not line up with modern ones, which could make it difficult to compare data from old and new brains, according to several experts. On the other hand, century-old clinical records are often much more descriptive than modern ones. In the early 1900s patients were often confined to institutions and followed much more closely than are patients today, giving psychiatrists the chance to record intimate details. "For example, one record describes a female patient who, over the course of a couple of weeks, ate all of the fibers in a broom," Sandusky says.

"Some of the old cases are very good in terms of quality," says Manuel Graeber, chair of brain tumor research at the University of Sydney. In 1997 he found slides of 91-year-old brain tissue from the first described Alzheimer's patient stashed in a basement at the University of Munich. His team confirmed the telltale plaques and tangles of the disease and ran DNA tests on the tissue. "These were meticulous scientists. It's inspiring."

When the pathology building of the Indiana asylum opened in 1896, the Indianapolis Sentinel newspaper ran an announcement. In it, a local doctor described the purpose of the lab: "to gain a clearer insight into diseases of the mind, which must result in the cure of a much larger percent of cases than is now possible." That goal, at least, has been perfectly preserved.