to BioTechniques free email alert service to receive content updates.
Exosomes: a new hope in RNAi delivery

Vincent Shen

With no feasible delivery mechanism in sight, pharmaceutical companies are pulling the plug on RNAi research programs. Vincent Shen looks at how an exosome delivery system may revive RNAi therapeutics.

Bookmark and Share

Acuity Pharmaceuticals introduced the first RNA interference (RNAi)–based drug candidate, Cand5, into clinical trials in 2004, after having raised $20 million dollars. Cand5 was designed to reduce vision loss in adults with wet age-related macular degeneration (AMD). Other pharmaceutical companies also began introducing RNAi therapeutics into clinical trials and expanding RNAi research programs. The future looked bright for RNAi.

But in 2009, clinical trials of Cand5 were stopped after the guidance committee found the drug was unlikely to significantly prevent vision loss. The small interfering RNA (siRNA) was not being taken up by the targeted cells in the quantities necessary to be effective. Such siRNA delivery problems plagued other RNAi clinical trials as well, leading some pharmaceutical companies to exit RNAi research by 2010.

S. Patrick Walton is designing a modular delivery platform for the clinical development of small RNAs. Source: Michigan State University

But S. Patrick Walton, an associate professor of chemical engineering and materials science at Michigan State University, believes it’s too early to bury RNAi therapeutics just yet. “If we look at small molecule drug development, we are still not perfect at it, and we have been doing it for 50-odd years or more,” Walton says. “If we look at RNAi drug development, we have been doing it for 15 years or less. It is still a fairly short horizon from initial discovery to what it would be if it were in the market now.”

The problem underlying delivery of siRNA stems from RNA’s inherent vulnerability; the small RNAs trek through an extracellular jungle filled with ribonucleases. And even if the siRNAs reaches their target tissue intact, it’s still unknown exactly how they penetrate the cell membrane.

To help bring the first RNAi-based therapeutic to patients, some researchers are looking at how cells transfer RNA between one another using the extracellular vesicles called exosomes. Researchers hope these vessels can solve the RNAi delivery problem and open the floodgates for a multitude of RNAi-based drugs that could target diseases such as HIV, hepatitis B, acute renal failure, and AMD.

High-hanging fruit

Acuity Pharmaceuticals was not naive about RNAi delivery problems in 2004. Founded in 2002 by researchers from the Scheie Eye Institute of the University of Pennsylvania (UPenn) Medical School and Samuel Reich, a doctoral candidate in UPenn’s ophthalmology department, the company knew all about extracellular ribonucleases and their destructive effect on siRNA.

So Reich and his cofounders sought to minimize this effect by administering Cand5 through local injection into the patient’s eye. They hoped that reducing the distance the siRNA drug had to travel would improve its deliverability. Once injected, the siRNAs would target retinal cells and knock down vascular endothelial growth factor (VEGF), an extracellular molecule that stimulates the formation of new blood vessels and leads to the overgrowth of these vessels in AMD patients.

The outlook for Cand5 was promising. A preclinical study demonstrated that Cand5’s sequence actively knocked down the expression of VEGF messenger RNA (mRNA) in human cervical and kidney cells. A subsequent study in 2008 found that Cand5—later renamed bevasiranib—could reach its targeted tissue through local injection. The clinical trials found the treatment reduced patient symptoms without significant toxicological effects.

Although Phase I and Phase II trials showed positive results, the company—which was renamed Opko Health—stopped Phase III clinical trials of bevasiranib because it was found to be ineffective. In 2008 University of Kentucky researchers found evidence that the initial positive results might have been from off-target effects and questioned whether the drug actually entered the targeted cells at all (1).

Despite this setback, Opko Health continues its development of bevasiranib, working on “new dosing schedules, combining it with marketed products for AMD, and enhancing delivery with novel siRNA delivery vehicles,” according to the company’s web site. Furthermore, in February 2011, Opko Health paid $10 million to acquire Curna, Inc., a Jupiter, Florida–based company that is developing RNAi drugs to upregulate proteins to treat various diseases.

“RNAi is one of those boom-or-bust kinds of technologies at this point. If somebody finds one that does work and can be delivered accurately to the appropriate tissues and cells, then it has the potential to change the game of therapeutics entirely,” Walton says. “This first-to-market potential is very attractive, even if it’s a high-hanging fruit.” But some pharmaceutical companies have decided that the high-hanging fruit of RNAi therapeutics might not be worth the investment.

Roche, one such company, is terminating its effort to discover and develop drugs through RNAi, affecting research programs in its Kulmbach, Germany, Nutley, NJ, and Madison, WI, facilities. It also ended partnerships with RNAi-biotechnology companies Alnylam Pharmaceuticals, Inc. and Tekmira Pharmaceuticals Corp. Alnylam received $335 million from Roche in 2007 alone; Tekmira received $18.4 million from Roche between 2009 and 2010.

Bursting the RNAi delivery bubble

Electron micrographs of MC/9 and HMC-1 exosomes. Source: Nature Cell Biology

For pharmaceutical companies like Opko Health that remain vested in RNAi drug development, their efforts continue to focus on improving deliverability. Three RNAi delivery vehicles still hold the spotlight: viruses, cationic liposomes, and cationic polymers. While each can deliver siRNA, each also has the potential to cause more harm than good.

For example, nonpathogenic viral particles packed with DNA sequences can hijack the host’s cellular machinery to produce the siRNAs that knock down a targeted gene, but these transcribed siRNAs—which are produced as short hairpin RNAs—have caused liver toxicity and morbidity in preclinical models. Liposome-based methods unload their siRNA cargo directly into the cell via RNA-bearing endosomes, but these endosomes might trigger an immune response through the activation of toll-like receptors (TLRs), marking the cell for destruction by the immune system. A cationic polymer such as synthetic polyethylamine can deliver siRNA into a cell through an endosome, releasing its payload by rupturing the endosome; however, this could release the lysosomal protease cathespin B, which can then cleave procaspases and trigger apoptosis. So the need for more options has remained constant in RNAi drug development.

Hadi Valadi was the first to discover RNAs in exosomes. Source: Hadi Valadi

In 2007 Hadi Valadi was a postdoctoral researcher in Jan Lötvall’s lab at the Krefting Research Centre, part of the University of Gothenburg; he was working on translational research when he discovered a potentially new form of siRNA delivery via extracellular vesicles called exosomes (2).

Exosomes, first identified in sheep cells in 1983, were believed to be a mechanism for the exchange of proteins between cells. A wide range of cells secrete these small vesicles by endocytosing a portion of their own cell membrane, which they then expel as small vesicles from late endosomes .

But while studying these exosomes in human and mice mast cells, Valadi discovered that proteins weren’t the only molecules hitching a ride in them—mRNAs were tagging along as well. Using microarrays, Valadi and colleagues found mRNA from about 1300 different genes in the exosomes.

When the exosomal mRNAs were introduced into rabbit reticulocyte lysates with radiolabeled methionine, Valadi observed freshly synthesized proteins, indicating that cells could alter the gene expression of neighboring cells by sending out exosomes with functional RNAs. In short, exosomes can deliver functional RNA to another cell without any negative side effects.

“The exosomes have evolved to transfer genetic materials between cells in our bodies,” Valadi says.

Valadi’s discovery sparked interest in this poorly understood function of exosomes. In 2008, researchers from Massachusetts General Hospital discovered that human glioblastomas isolated from tumors in the brain secreted RNA-carrying exosomes. In 2011, Valadi, now an associate professor of rheumatology and inflammation at the University of Gothenburg, found that RNA-filled exosomes exist in human breast milk, saliva, and plasma.

Putting exosomes to work

Schematic representation of production, harvest and re-administration of targeted self-exosomes for gene delivery. Source: Nature BioTechnology

In 2011, Matthew Wood, an investigator in the Department of Physiology, Anatomy, and Genetics at the University of Oxford, and colleagues decided to put exosomes to the RNAi delivery test.

To start, Wood’s team turned dendritic cells—which present foreign antigens to other immune cells—into factories producing exosomes that target brain cells. To do this, the researchers transfected a DNA sequence that encoded a fusion protein made of rabies virus glycoprotein (RVG)—which targets brain cells—and Lamp2b, a protein expressed on the surface of exosomes. The transfected DNA tricked the dendritic cells into secreting RVG-bearing exosomes that home in on the acetylcholine receptors of the brain’s neurons.

Next the researchers loaded the exosomes with their siRNA cargo. Wood’s group electroporated the exosomes with siRNAs that targeted the metabolic enzyme glyceraldehyde 3-phosphate dehydrogenase (GAPDH) found in the majority of cells.

Then these RVG-coated, siRNA-charged exosomes were introduced into a mouse’s circulatory system and peripheral tissues by means of intravenous injection in the tail. This contrasts starkly with Opko Health’s localized injection of RNAi therapeutics close to the targeted tissue. Wood’s team harvested the mouse brain three days post-injection and analyzed the distribution of exosomes in the striatum, cortex, and mid-brain tissues.

Quantitative PCR analysis revealed that the siRNA-containing exosomes knocked down GAPDH mRNA preferentially in all three samples of those brain tissues compared to other tissue samples, including muscle, liver, spleen, and heart. Similarly, Western blotting showed that repression of GAPDH protein synthesis occurred in the cortex. The RVG-coated exosomes had concentrated in the brain and released their gene-silencing payload intracellularly in the target tissue.

“These exosomes can be practically nonimmunogenic for patients. Cells can be isolated from one person and one tissue and modified in vitro,” Valadi says. “We can then give it back to the same person.”

Furthermore Wood and colleagues also designed a disease-treating exosome. This time, they incorporated an siRNA sequence that targeted BACE1, an extracellular protease that triggers the formation of plaque in Alzheimer’s disease. Mice that received an injection of this exosome showed repressed levels of BACE1 protein in the cortex.

In principle, Wood’s exosome method could be applied to a variety of diseases. Swapping the RVG module for another tissue-specific homing peptides could fine-tune the vehicle’s tissue targeting. The exosome’s bilayer shields siRNAs from the extracellular ribonucleases during the nanoparticle’s commute to the target tissue while bypassing any TLR-triggered immune response by entering a cell through membrane fusion, using the dendritic CD9 receptor.

“The exosome field is a very new area, so we need much more time to develop the methodology for using the exosome as a vector,” Valadi says.

Currently researchers are expanding the range of targets to tissues other than the brain. In particular they are searching for surface receptors that exhibit muscle-specificity and mediate entry into skeletal muscle cells. Such a surface tag could lead to treatments for muscular dystrophy, a collection of inheritable human disorders that weaken the skeletal muscles of both children and adults.

“If someone comes up with a technology that is a game changer from the standpoint of delivery, then that will revive some interest,” Walton says.


1. Kleinman ME, et al. 2008. Sequence- and target-independent angiogenesis suppression by siRNA via TLR3. Nature 452:591–7.

2. Valadi H, et al. 2007. Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells. Nat Cell Biol. 9:654–9.

3. Alvarez-Erviti L, et al. 2011. Delivery of siRNA to the mouse brain by systemic injection of targeted exosomes. Nat Biotechnol. 29:341–5.

Keywords:  clinical RNAi