Discoveries made during the first large-scale analysis
of interactions between proteins in our cells hold promise
for identifying new genes involved in genetic diseases,
according to researchers at Johns Hopkins and the Institute
of Bioinformatics in Bangalore.
The findings, reported in the March issue of Nature
Genetics, were made using a database of more than
25,000 protein-protein interactions compiled by the Johns
Hopkins-IoB team. The result is believed to be the most
detailed human "interactome" yet describing the interplay
of proteins that occur in cells during health and
disease.
"Genes are important because they are the blueprints
for proteins, but proteins are where the action is in human
life and health," said Akhilesh Pandey, an assistant
professor at the
Institute of
Genetic Medicine and the departments of
Biological
Chemistry,
Oncology and
Pathology at the
Johns Hopkins School of Medicine. "This ability to find
links between sets of proteins involved in different
genetic disorders offers a novel approach for more rapidly
identifying new candidate genes involved in human
diseases," he said.
The analysis included interactions among 1,077 genes
coding for proteins linked to 3,133 diseases, the
researchers reported. Significantly, it showed that
proteins encoded by genes that are mutated in inherited
disorders were likely to interact with proteins already
known to cause similar disorders. In addition, the
researchers disproved the long-held belief among scientists
that the relative importance of a specific protein is
always reflected by the number of other proteins with which
it interacts in the cell.
According to Pandey, the team's comparison of almost
25,000 human, 16,000 yeast, 5,500 worm and 25,000 fly
protein-protein interactions showed that, among these more
than 70,000 links, only 16 were common to all four
species.
Researchers say this low level of interactome overlap
among species was surprising. It showed that current
rapid-testing methods for identifying protein interactions
are likely to miss true interactions.
Much of the Johns Hopkins-Bangalore work was based on
information compiled in the Human Protein Reference
Database, a repository of information on protein-protein
interactions collected from the published literature and
stored in a format suitable for rapid study and comparison
with other animal cells. HPRD was developed by the IoB and
the Pandey laboratory.
"Using HPRD and several other databases, we have been
able to develop a gold mine of new information for
researchers seeking new ways of finding candidate genes
involved in genetic diseases," Pandey said. "And our
demonstration that a protein's importance is not based on
the number of interactions it has with other proteins is an
important conceptual breakthrough. It eliminates a blind
alley that could mislead researchers investigating the
roles of specific proteins in the cell."
Pandey is the chief scientific adviser to the IoB and
senior author of the Nature Genetics article. The team's
conceptual advance was made by comparing human data with
6,014 genes in yeast and 2,284 genes in mice whose effect
on survival was known, he said. "Our much larger database
on genes and proteins gave us the information to set the
record straight on how to measure a protein's
importance."
Using this kind of comprehensive comparison of
information about human and other organisms allowed
Pandey's group to identify 36 previously unknown
protein-protein interactions, nine of which were tested in
the laboratory to verify what the analysis suggested. "We
proved they were valid," Pandey said. "By linking
computerized sleuthing to laboratory experiments to confirm
those findings, we expect to be able to eventually fill in
many blanks in human protein-protein interactions."
All the analyses were primarily carried out at the
IoB, a nonprofit research institute founded by Pandey in
May 2002. The Human Protein Reference Database was
developed with funding from the National Institutes of
Health and the Institute of Bioinformatics. Pandey serves
as chief scientific adviser to the Institute of
Bioinformatics. He is entitled to a share of licensing fees
paid to Johns Hopkins by commercial entities for use of the
database. The terms of these arrangements are being managed
by the university in accordance with its
conflict-of-interest policies.
The URL of the Human Protein Reference Database is www.hprd.org.
Other investigators on this study were Giovanni
Parmigiani, Joel Bader and Jef Boeke, all of Johns Hopkins;
and Stefan Pinkert and Joerg Schultz, from the University
of Wuerzburg.
GO TO