By Doug Macron
While Sirna Therapeutics' expertise with RNAi was the major driver behind its 2007 acquisition by Merck, the big pharma has also been leveraging that know-how to explore the microRNA field.
According to Merck researcher Lee Lim, the firm's efforts have thus far been primarily focused on miRNA mimics, rather than inhibitors, since such molecules are "somewhat analogous" to siRNAs. This allows the company to "piggyback on [Sirna's] efforts to develop chemical modification patterns … and delivery vehicles" for the gene-silencing agents, he said.
"Of course, we know microRNAs and siRNAs are not the same, especially given the fact that microRNA targeting is almost always very different from siRNA targeting," Lim noted during a presentation at last week's Oligonucleotide Therapeutics Society annual meeting in Dana Point, Calif.
This suggests that "microRNA mimics might be able to tolerate [the] sorts of things that siRNAs can't," he said. "Therefore, we think there are going to be different types of microRNA mimics … some [of which] are going to be siRNA-like and some may contain microRNA-specific-type structures."
When it comes to investigating miRNA mimics with unconventional structures, one of the approaches Merck has examined involves duplexes with a gap in one of two strands.
"Several years back, Jorgen Kjems' lab [at the University of Aarhus in Denmark] presented the idea of an siRNA that was pre-engineered to contain a passenger strand [that] was cut in half," Lim explained. "Of course, this echoes when an unmodified RNA gets incorporated into Ago2 and its passenger strand is cleaved."
"However, people haven't really looked at the seed-based activity, which is so important for microRNA activity, of structures like these," he noted.
To do so, Merck investigators made mimics of miR-124 in which either the passenger strand or the guide strand had been cut in half, or wherein both strands were left intact. They then examined their ability to knock down CD164, an endogenous miR-124 target, in vivo.
Both the intact miRNA mimic and the one with the nicked passenger strand had similar inhibitory activity.
"What was interesting was that when we broke the guide strand in two, we saw the same activity, suggesting that the break in the guide strand is still allowing for the seed-based activity," Lim said.
"One theoretical possibility [for this effect] is that the guide fragments might be ligated together, but we don't think that's the case because we can also do things like put a gap in between the two halves, or we can insert an inverted abasic or even a Cy3 label between the two halves and still get microRNA activity," he added.
Examining the mimics' activity globally, "we used microarrays to look at the targeting of the intact duplex and the duplex with a broken guide strand … [and found] there is quite good correlation on the microarrays, suggesting we are seeing similar activity globally between" the two agents, Lim said.
"We also wanted to look more closely at … protein readout using luciferase reporters … [which] also allowed us to engineer different sorts of binding sites for miR-124 — we had a full-length binding site, a seed-match binding site, a seed match plus some 3' base pairing, and a control," he said. The team found that the seed-match mimic was active, but noticed an increase in activity with the 3'-pairing or full-length matching molecules.
[ pagebreak ]
"We did see a difference when we broke the guide strand in half," he said. Specifically, "we lose the enhancement in activity that is provided by 3' matching or full-length matching," with both of those compounds "behaving like the seed match reporter."
The Merck team also examined the effects of a miRNA precursor mimic, in this case one of miR-124. They created intact hairpins, as well as ones in which the guide strand had been nicked, and saw "a loss in potency for the nicked hairpin."
These data were published earlier this month in RNA.
"It's still unknown to us whether or not both pieces of the guide strand are being incorporated into Ago," Lim said at the OTS meeting. "We do know that the seed region is there because we see seed-region activity. If the 3' end is there, the break in the guide strand is obviously preventing its usage."
In the end, "these sorts of experiments simply suggest the structural flexibility of microRNA mimics," he said.
In terms of applying Sirna technology to miRNAs, Merck has been developing mimics that incorporate certain siRNA-modification patterns designed to improve half-life and decrease immune responses. These are packaged into lipid nanoparticles for delivery.
Early experiments with such molecules included miR-34 mimics that worked well in vitro, prompted in vivo testing, Lim said.
Delivering three consecutive 3 mg/kg doses of either the mimics or controls to mice, Merck researchers found that the siRNA-like mimics led to a 40 to 50 percent knockdown of target genes, while the controls had little effect.
The team then tested single 3 mg/kg doses of mimics with alternative modifications, as well as an unmodified miR-34 mimic. While the unmodified oligo had "intermediate activity," significant activity was observed with the modified mimics, he said.
The team also tested modified mimics against miR-34 and miR-124 in combination, and saw similar target inhibition as with the miR-34 mimic alone.
"Based on these results, we are confident we're seeing target engagement, both singly and in combination," Lim concluded. "So we're currently testing these modification patterns on other microRNAs, and we're testing these in animal models."