Skip to main content
Premium Trial:

Request an Annual Quote

Merck’s New Genome-Wide RNAi Screen T’fection May Help Uncover Reliable Hits

A team of researchers at Merck has developed a genome-wide RNAi screening technique that may increase the dynamic range and decrease the variability in cell-based assays used for RNAi screening, which could facilitate reliable hit selection, according to their findings.
The investigators showed that the method, called T’fection, yielded a mean absolute deviation-based hit selection that identified physiologically relevant false negatives that standard-deviation-based methods missed.
In a separate paper, the Merck team described the development of a fully automated, high-throughput transfection method for genome-scale siRNA screens.
“The whole idea was to do replicate screens to get reliable data,” said Namjin Chung, the lead author of the two papers, which appear in the March issue of the Journal of Biomolecular Screening here and here.
While screens of several hundred thousand wells were previously not uncommon, the throughput was very slow. “To do more replicates, we needed to be able to do the screens faster than we had been,” said Chung, who was a researcher in the Department of Automated Biotechnology at Merck Research Laboratories in North Wales, Penn., when he conducted the research. He is currently a scientist at Bristol-Myers Squibb.
The researchers developed a lipid-based transfection method for siRNA that can process in eight hours 80 384-well plates in triplicate, for a total of 30,720 unique transfections. As a proof-of-principle, the scientists ran a genome-scale screen of a library of 22,108 siRNAs to identify the genes that sensitize cells to mitomycin C at concentrations of 0 nM, 20 nM, and 60 nM.
In their early siRNA screens, transfection reagents, sample siRNA, and control siRNA were sequentially transferred to an intermediate microplate, in which the lipids and siRNA are mixed together and incubated for 10 to 30 minutes, Chung said. The siRNA:lipid complex was usually coupled together, and a round of transfections for three microplates per intermediate plate usually took 25 to 30 minutes on a typical robotic platform.
The authors found, however, that if they prepared as many intermediate plates as possible in a batch before starting to transfer siRNA:lipid complexes, intermediate plate preparation could be virtually eliminated as a rate-limiting step.        
Using this “uncoupled” protocol, the researchers found that it took them about six minutes to transfect three microplates per intermediate plate, which represented a four- to five-fold increase in throughput.
In this manner, the whole transfection process was done in days, Chung told CBA News this week. “In eight hours of transfection, you could complete an entire genome-scale library screen compared to a five- to six-day transfection over a three-week timeframe.”

“This protocol allows you to do replicate screens in the same amount of time as it would take to do one screen using the earlier technique.”

He said the protocol “allows you to do replicate screens in the same amount of time as it would take to do one screen using the earlier technique.”
Drug-Discovery Applications
Genome-scale RNAi screens represent a promising area of research for novel target identification, said Chung. He said that in the past, Merck’s drug-discovery programs have relied on microarray studies or literature findings.
However, one drawback of microarray studies is that they do not yield a causative relationship between gene expression and what would happen when such gene function is obstructed, he said.
With an RNAi screen, because researchers knock down each individual gene to uncover what happens, “it depicts a causative relationship,” Chung said. “By doing this for the 20,000 genes in the human genome, you could have a lot better understanding of gene function that leads you to a better drug discovery program.”
Another area where RNAi screens are used is to identify biomarkers. “From my personal experience, and from reading about other people’s work, significant progress is being made in using RNAi screens to identify pharmacodynamic markers, as well as predictive biomarkers,” he said.  
The next step in this work would be to develop a 1,536-well siRNA transfection protocol, Chung said. “I moved to BMS about six months ago, but before I left Merck, I worked on developing 1,536-well siRNA transfection,” he said.
He said that all of the work in the paper was done in 384-well plates. The Merck researchers have conducted similar research using 1,536-well plates and “using the technologies and skills that we discussed in the paper.”     
“In a 1,536-well plate, we were able to do a whole genome-scale RNAi screen in “a matter of hours in triplicate,” said Chung. The RNAi screen is not very reliable when it is done in a non-replicate manner, he explained.
“By doing it in triplicate, you can do the statistical analysis to select hits based not only on signal size, but also on signal significance,” he added. “I am sure that there is a push in many functional genomics organizations to do that.”
He said his former colleagues at Merck plan to submit that research for publication. He did not elaborate.
Researchers are currently running genome-wide RNAi screens more frequently, Chung said. “These days we have identified so many targets, so much low-hanging fruit, that most of the targets that may be interesting may still be high up in the tree. We need to pick them.”
The only real way to get there is to do an unbiased genetic screen to try to identify genes involved in certain biological processes, and a genome-scale RNAi screen allows investigators to do this, he said.