PHILADELPHIA, May 17 – Many proteomics companies are promising more than they can deliver because they don’t yet have the technology to isolate and characterize low abundance proteins, Ian Humphery-Smith, a protein chemist at the University of Utrecht and the organizer of the Human Proteome Organization, said Thursday.
While companies, such as Large Scale Biology and Myriad, have touted the ability of their technology to pick out most of the proteins in certain cells, they haven’t identified more than five percent of the proteins in the human body, Humphery-Smith said.
“At present we’re at five percent and we’ve got 95 percent that are defying current technologies,” he told GenomeWeb, during a conference entitled "Proteomics: Delivering New Routes to Drug Discovery."
“Proteins are difficult…but for a number of reasons it’s not an easy job and [some companies are] overselling the deliverables,” he said.
The problem, said Humphery-Smith, is that most cells overwhelmingly contain the same kind of proteins and important, low-abundance proteins are hard to pick out with current 2-D gel electrophoresis and mass spectrometry technology.
Solving this problem will require designing affinity agents, or molecules that capture certain classes of proteins, to screen out the high abundance proteins found in cells before analyzing the remaining proteins with mass spectrometry, he said.
“If you have antibodies or other ligands [to act as affinity agents], you have the potential to enrich your signal to do good mass spec,” he said. “But we can’t do PCR and we can’t hybridize [proteins to make them more easily identifiable] so we must have a ligand affinity enrichment of signal.”
Humphery-Smith also stressed that biological systems are naturally variable, and current 2-D gel and mass spec techniques for finding proteins must be repeated many times for the results to be statistically significant. Many proteomics researchers have overstated the significance of their results, he said, because they haven’t replicated their experiments often enough.
“If you’re going to conclude on the causalities of cancer and human ailments, you need large numbers of replicants,” he said.
Humphery-Smith said protein microarray technology is the most likely to discover therapeutically important proteins because it will allow researchers to precisely study how the proteins in cells change with disease.
Affymetrix will soon have a DNA microarray that covers the entire human genome, he said, and “we’ve got to compete with Affymetrix and others until we’ve got a reporter matrix that can look at your proteome in six organs, before and after dinner, seven days of the week, with three drugs.”