FALLS CHURCH, Va. — Great strides have been made with RNAi therapeutics in the short time since the gene-silencing phenomenon was first observed. Still, a number of key issues involving efficacy, safety, and delivery — the “number one problem” — must be addressed before the field can succeed, according to a senior Merck official.
“The field is very young,” Alan Sachs, vice president of RNA therapeutics at Merck Research Laboratories, said during a keynote presentation at the second Drug Information Association Oligonucleotide-based Therapeutics Conference held here this week. “At Merck, we see [RNAi] as a game-changer. We anticipate a lot of commercial success, but we also anticipate that that’s going to take time.”
Before RNAi can become a feasible therapeutic modality, Sachs said researchers developing RNAi drugs will need to understand that the presence of the molecules in a target tissue doesn’t necessarily equate with function.
For instance, in animals, siRNAs delivered systemically in lipid nanoparticles are rapidly cleared from the plasma, but remain stable in the liver — the target organ for many RNAi drugs — for long periods.
“This on the surface would seem to be a good thing,” he said. “But is the actual concentration in the liver a reflection of the functional activity of the RNA? What we have found is that the RNA levels that you measure in an organ have nothing to do with the functional activity of the RNA.”
Citing experiments in which “significant” target knockdown was observed for up to two weeks in liver tissue-culture cells exposed to an siRNA-lipid complex, Sachs noted that if a clear relationship between pharmacokinetics and pharmacodynamics for siRNAs existed, the level of siRNAs present in the cells should remain steady.
“But in fact … less than 3 percent of the RNA remains in the cells after a week, even though the IC50 remains the same,” he said. “That’s a little confusing if you think [that] total bulk measure of siRNA is an indication of pharmacodynamic activity unless you appreciate that the RNA [in lipid-based carriers] is taken up in an endosome … and it has to get out of the endosome and into the cytoplasm to be active.”
If it remains in the endosome, the siRNA eventually ends up in the lysosome where it is destroyed, he said. “We think that the vast majority … of the RNA taken up by the cells never leaves the endosome and ends up in the lysosome … [and] only a very small amount ends up in RISC.”
As a result, measuring the total RNA doesn’t provide an accurate indication of how much of the siRNA entered RISC, “which means that to really establish a PK/PD relationship, we’re going to [have to focus on] RISC-associated siRNA directly,” Sachs said, noting that Merck is currently working on tools to measure RISC association, such as transgenic animals with tagged RISC.
Delivery, “beyond a shadow of a doubt, is the number one problem. I don’t think that we’re even close to finding a solution to this problem.”
Another important consideration for RNAi drug makers is the difference between how siRNAs work in different animals, and what this means for avoiding potentially hazardous off-target effects, he said.
Gene-expression analysis has become an invaluable tool for choosing which siRNA candidates to advance though the pipeline by allowing researchers to examine the effects of the oligos on the whole genome, Sachs said.
Yet despite the robustness of this technology, it is important to look beyond rodent data for a compound that will ultimately be designed for human use.
“Does mouse profiling predict what happens in human cells?” he asked. “The answer to that, we’ve discovered, is ‘no’. If you look at genes that are shared between human and mouse cells … you see there is very little overlap between what happens in the mouse cell and what happens in the human cell.”
During preclinical development, “if you don’t use a primate, which is the closest we can get to humans for safety, you’re running the risk that you’ll have an off-target activity that isn’t discovered because you’ve worked in a species [such as a mouse] whose 3’ UTRs are too distant from humans,” Sachs said.
“Whether or not you accept that gene-expression profiling should or shouldn’t be used as an [investigational new drug application] package component, we all need to accept that you have to use a primate as a safety species because you can’t predict off-target activities with a rodent,” he added.
To demonstrate his point, Sachs described experiments conducted by researchers at Merck subsidiary Rosetta Inpharmactics in which siRNAs that seemed highly specific in mice could hit unintended targets once evaluated in human cells.
These data were first presented last year at the third annual Oligonucleotide Therapeutics Society meeting in Berlin. As reported by RNAi News, the Rosetta team was asked to select one out of four chemically modified siRNAs against an undisclosed gene for further development (see RNAi News, 10/11/2007).
Although none of the siRNAs appeared to cause unintended silencing in mice, when examined in human cells, one of the oligos was found to hit both its target gene and a gene associated with a liposomal storage disorder.
The moral of the story: “You need to get as close as you can to the human in cell lines and in animals, and pay attention to the gene signals that are coming out,” Sachs said.
When it comes to making a successful siRNA drug, however, delivery, “beyond a shadow of a doubt, is the number one problem,” Sachs said. “I don’t think that we’re even close to finding a solution to this problem.”
Here, again, making certain that an siRNA is getting into RISC, and then examining efficacy in primates, are key, he noted.
“It doesn’t help to know that your delivery system is bringing RNA to the cell because it may never leave the endosome,” Sachs said. “You really have to look to see if the RNA is getting out and onto the RISC.”
To do this, he said, Merck is using various endosomal-escape assays such as cell fractionation, but, he added, “the standard for all of us will be using transgenic animals, as well as tagged RISC … [and] monoclonal antibodies to look at the siRNA down to RISC itself.”
In the end, however, “it comes down to what’s happening in the animal,” and all too often a compound that works in the mouse fails to be effective in primates, Sachs said
“Why?” he said. “When we understand that we will have an advance. But what is true today is that it doesn’t help … to see mouse data [alone]. What you’re saying is that you may have something that works in primates. But unless it’s tested in primates, you don’t know.”