Targeting with siRNAs

A high power image of tumor cells injected with F105-P and FITC-siRNA shows fluorescent staining in the cytoplasm.

© 2005 / NATURE BIOTECHNOLOGY

How do you get short RNAs into mammalian cells? The question has puzzled scientists since Andrew Fire and Craig Mello demonstrated the concept of RNAi in 1998. Short interfering (si)RNAs can jam the signals from specific genes, thus providing promise for exquisite genetic-based therapies. The challenges include aiming for the right mechanism without dragging other pathways into the fray, bypassing immune reactions, and simply getting past the cell membrane.

In dealing with living organisms, researchers had to start at the membrane. "Our main concern was the difficulty of getting highly charged nucleic acid inside the cell to reach the protein with the targeted activity," says Martin Woodle of Intradigm based in Rockville, Md.

Amid a rapid succession of discoveries, this issue's Hot Papers demonstrated distinct ways to package and ship siRNAs to specific cell targets. Both approaches allowed discrete siRNA delivery and silencing via intravenous injection in mice. In 2004, Woodle and colleagues designed and constructed a nanoparticle "shipping container" that used Arg-Gly-Asp (RGD) peptide ligand attached at the distal end of the polyethylene glycol (PEG) to target tumor neovasculature expressing integrins. ¹ These nanoparticles delivered a payload of siRNAs that would quash angiogenesis. They used the nanoparticles to target tumor neovasculature and deliver siRNAs that would quash angiogenesis.

A year later, a research team from the CBR Institute for Biomedical Research at Harvard University conjugated a Fab fragment from an HIV-envelope targeting antibody to protamine, the proteins that package nucleic acids in sperm. ² With this protamine-antibody fusion protein (F105-P), they were able to deliver siRNA to select cells and silence specific genes. "Until our paper came out, it was still up for grabs whether you could get siRNA's into circulating cells," says Judy Lieberman, an author on the paper.

Though the papers represent just one step in deriving applicable human therapies, some signs indicate that the pace has quickened. According to Woodle, a human therapy based on his group's work appears destined for clinical trials in 2008.

Woodle's group was able to inhibit vascular endothelial growth factor receptor-2 (VEGF R2) in mice injected with tumor cells. Researchers injected 40 mg of siRNA nanoplex every three days into the tail vein a week after tumor cell inoculations. Compared to untreated controls and mice injected with free siRNA, those who received nanoparticle siRNA had visibly shrunken tumor vasculature after two weeks. In some mice, solid tumors in different body parts diminished completely. Expression analysis showed strong endothelial cell uptake and 90% gene inhibition in the tumor.

Puthupparampil Scaria, an author on the paper, says he expects a feasible treatment to be available by 2012, if everything goes through trials successfully. And, he adds, "The beauty of the nanoparticle is it's easy to create a second drug; you only need to change the siRNA ligand to target other tumor tissues." Woodle agrees that a logical extension of the nanoparticle mechanism will be "to see if we can extend our design to payloads that inactivate tumor cells."

To test in mice, which can't be infected with HIV, Lieberman's group injected mice with B16 melanoma cells transfected with an HIV envelope. They combined siRNAs that would target three different RNA pathways - c-myc, MDM2, and VEGF - and injected the mice with this fusion protein. The results showed that tumor growth was inhibited sevenfold, and siRNAs had no significant effect on non-HIV envelope-expressing tumor cells. "We've been able to generalize this method," says Lieberman. They had also tested anti-Erb2 fusion proteins against Erb2+ breast cancer cells.

Just getting into the cells is only the first challenge. Ratna Ray, a professor in the pathology department of St. Louis University, says she worries about the toxicity of siRNA at high levels. Brian Silverman, of Lerner Research Institute in Ohio, says, "You might get interferon response in vivo," which would lead to upregulation of interferon-stimulated genes. Delivering siRNAs can potentially activate nonspecific inflammatory responses, but Lieberman says her group's repeated experiments show that the fusion protein-delivery system targets only intended cells. "If any drug can target a specific cell, that reduces the toxicity of drug, and the overall dose would be lower," says Lieberman. Ray says she hopes the nanoparticle will provide a slow release system of siRNA, reducing chances of toxicity.

Both teams still need to determine whether these siRNA therapies will have off-target effects. Aimee Jackson of Rosetta Inpharmatics published an important warning in 2003. ³ Jackson's group designed 16 different siRNAs to target the IGF1R coding region and eight siRNAs to target MAPK14. Most of the expression patterns they looked at showed silencing in untargeted transcripts having similar nucleotide sequences, even if only six to eight nucleotides matched. The most potent proof came in the form of siRNA designed to target luciferase. The human genome lacks a luciferase target, but the siRNA still regulated several gene pathways.

Woodle says the problem with studies showing off-target effects is that all experiments to this point have been in vitro and therefore inadequate to predict siRNA behavior in animal models, and especially in people. Scaria and Woodle say the mouse model showed minimal siRNA interaction in other VEGF pathways and no physiologic side effects. Woodle is confident that precise targeting of cells will make the approach viable. "The objective of the nanoparticle system is to limit biodistribution to target tissue," says Woodle. "It's the right place to be exploring a solution."