"We thought this was really interesting and neat," said Johns Hopkins neurologist Jeffrey Rothstein, who was not part of the team. Until now, "there have been no validated risk-factor genes in ALS."
The finding links several clues, including a 2006 discovery by Penn researchers that people with ALS build up abnormal deposits of a protein called TDP-43 in their brains.
The hope is that these puzzle pieces will lead to new treatments. Today, doctors can offer little more to patients than they could to Gehrig when the New York Yankees first baseman was diagnosed in 1939.
"One needs to have all the culprits lined up if you're going to go after them with drug discovery," said John Trojanowski, a Penn brain pathologist who, with Virginia Lee, led the team making the TDP-43 finding.
Trojanowski's team found the abnormal protein deposits by analyzing hundreds of autopsies on the brains of people who died from ALS and other neurological disorders.
He said clumps of TDP-43 also appeared in the brains of people who died from Alzheimer's disease and was sometimes seen in head-trauma cases.
The trauma link is intriguing in light of a paper published last week suggesting that Gehrig and other athletes and soldiers may have died from brain injuries rather than ALS. Gehrig had several serious head injuries and was known for playing despite them.
Some suspect that brain injuries might trigger ALS through a pathway involving TDP-43.
Penn geneticist Aaron Gitler approached the problem by exposing yeast cells to TDP-43 and seeing what happened when he genetically modified the yeast.
"Yeast cells might seem like an unlikely model, because they don't have a brain," Gitler said. But, due to the relatedness of all life, they share a surprising number of genes with humans, including the neurons that make up human brains.
In their natural state, yeast cells died when exposed to TDP-43, just as human brain cells do. After modifying them in 5,500 ways, Gitler found a few changes that made them die faster and a few that helped them survive.
Colleagues called that a fishing expedition, as Gitler had no idea which changes, if any, would do anything. Eventually, though, he caught something promising.
Gitler found that when he disabled one particular gene - PBP1 - the cells were far better at surviving the assault from the abnormal protein. That gene has a human equivalent, ataxin-2.
Teaming with Penn colleague Nancy Bonini, Gitler found that disabling the ataxin-2 gene in fruit flies' nervous systems helped the insects survive TDP-43's toxic effects.
In humans, ataxin-2 is involved in a neurological disease called spinocerebellar ataxia type 2, (SCA2), which causes difficulty in walking and other coordination problems.
The disease is caused by a genetic stutter in which a three-letter stretch of the genetic code - CAG - is repeated too often. In most cases, this stretch of DNA repeats the CAG combination 22 or 23 times. In those with SCA2, it repeats more than 34 times.
That led Gitler and Bonini to wonder: "What if it's longer than normal but not long enough to give you SCA2?" Could that be the ALS link?
To test his theory in humans, Gitler went to Trojanowski, who agreed to collaborate. "It was so fascinating," said Trojanowski, who found that the protein linked to the ataxin-2 gene also builds up in the brains of ALS patients.
The team then looked at DNA from 915 ALS patients and compared it with a group of healthy controls.
They found that ALS patients were much more likely to have a genetic stutter in their ataxin-2 genes. And, just as Gitler suspected, the stutter was in that middle range - longer than the normal 22 or 23 repeats but shorter than the SCA2-causing 35.
In a different gene, the same kind of stutter causes Huntington's disease, a fatal inherited disorder.
This intermediate-length genetic stutter showed up in about 5 percent of the ALS patients, Gitler said, but it crops up in just 1 percent of healthy people.
He and Trojanowski agree that the ataxin-2 stutter may not cause ALS, but probably is a risk factor.
Rothstein, of Johns Hopkins, said the study was too small to be sure the genetic connection was real, and he would like to see the findings replicated. "Periodically, these gene studies come up with a fun, neat observation and later it turns out to be a bunch of crap," he said.
But, if validated, the finding might lead to a drug. Gitler doesn't do drug discovery - he's a basic scientist in genetics - but said, "I hope other people will read this paper and start going after this."
Contact staff writer Faye Flam at 215-854-4977 or email@example.com.