Skip to main content
Premium Trial:

Request an Annual Quote

Stanford Team Finds Robust shRNA Expression Can Be Toxic or Fatal; Sequence Selection, Dosing Key

Premium

A team of researchers from Stanford University has shown that sustained, high-level shRNA expression in murine livers can result in severe toxicity and death -- possibly as a result of the shRNAs interfering with endogenous microRNA processing machinery.

Although the findings, which appear in this week's Nature, have raised questions over the use of shRNA as a therapeutic modality, Mark Kay, a professor at Stanford's School of Medicine and co-author of the paper, called them "encouraging."

"If you take the paper as a whole, it's actually encouraging because we were able to get therapeutic effects without toxicity if the right sequences were selected," he told RNAi News this week. "In no way do I feel [the paper] should limit the enthusiasm for using shRNA clinically. I think it just points out that, like any drug, you have to monitor how much you give."

In the paper, the researchers were looking to build off of earlier experiments in which plasmid-expressed shRNAs were used to inhibit hepatitis B viral replication in cell culture and in normal and immunodeficient mice transfected with an HBV plasmid.


"If you take the paper as a whole, it's actually encouraging because we were able to get therapeutic effects without toxicity if the right sequences were selected. I think it just points out that, like any drug, you have to monitor how much you give."

"What we were working on was a more clinically relevant way to do this, and to also show in a more clinically relevant way that we could get efficacy," Kay said of the new Nature paper. "We wanted to show in a more bona fide preclinical model that we could develop a gene-transfer approach that would be efficacious" for hepatitis B.

To do so, the researchers used hepatitis B transgenic mice in which "every single hepatocyte has the … transgene … [and] the potential to make hepatitis B viral particles," Kay said. ShRNA-expression cassettes were delivered using an optimized delivery vector based on duplex-DNA-containing adeno-associated virus type 8 and a U6 promoter.

"The fact is that we tried to use the highest doses we could generate in the first studies because we didn't know what to expect," he noted.

"We found two things: In some cases, the mice had substantial … reductions in their viral titers, long term [and] without toxicity," Kay said. In other cases, "the mice died [due to what] appeared to be somewhat of a sequence-specific result."

Uncertain as to whether the observed liver toxicity was associated with a hepatitis B-specific phenomenon or with the transgenic model, the researchers tested the AAV vectors in normal mice, according to Kay. They found that the vectors that weren't toxic in the transgenic mice weren't toxic in normal mice, while the ones that killed transgenic mice also killed normal mice.

"Obviously, we were worried about off-target effects so we started making shRNAs against all different genes unrelated to hepatitis B," Kay said. "Again, some sequences worked really well and lasted the life of the mouse without toxicity and others were toxic -- the most severe toxicity [resulting in] death."

Further investigation revealed a correlation between shRNA levels and toxicity, Kay noted, adding that "for reasons that are unclear, if you use the same construct, but switch out the shRNA sequences, some accumulate to higher levels than others."

Given this correlation, the researchers wrote in the Nature paper that "a possible reason for hepatocyte death was oversaturation of the endogenous shRNA-processing machinery … [which is] shared by miRNAs. We therefore speculated that highly expressed 'toxic' shRNAs competed with miRNAs for intracellular processing, to such an extent that affected cells died."

Using two miRNAs -- miR-122, the most highly expressed miRNA in the liver, and the ubiquitous miRNA let-7a -- as guides, the researchers confirmed that the toxic shRNA vectors were reducing endogenous miRNA levels in the mice.

Follow-on in vitro and in vivo transfection studies indicated that competition for the nuclear karyopherin exportin-5, which is believed to be responsible for shRNA and miRNA nuclear export and stabilization, may be behind the liver toxicity.


"A possible reason for hepatocyte death was oversaturation of the endogenous shRNA-processing machinery … [which is] shared by miRNAs. We therefore speculated that highly expressed 'toxic' shRNAs competed with miRNAs for intracellular processing, to such an extent that affected cells died."

"We provide some evidence that if you over-express exportin-5, you may get around some of the inhibition," Kay noted. "But what we don't know is if the exportin-5 is the absolute rate-limiting step for toxicity. That's what we're trying to prove now."

Further research is also required to find out why different shRNAs accumulate at different levels in hepatocytes even when the same expression cassette is used, Kay noted.

Additionally, it remains unclear whether the toxicity observed with high levels of shRNAs would be seen with similar levels of siRNAs.

"Everybody asks [whether the effect is seen with siRNA], and the issue is that we've never been able to get the same level of siRNA into all the hepatocytes of the mouse -- there's no way to do that," he said. "If the rate-limiting step is at exportin-5, then it might be different between siRNA and shRNA. But again, we need to do studies to prove or disprove that, and we don't have data yet."

Although the shRNA-associated toxicities were surprising, Kay said their discovery is a step forward and said that the Nature paper reinforces his belief that shRNAs can be used effectively for therapeutic purposes.

"In the paper … we show that with the right sequence -- [one] with high sequence-specific activity -- and the [right] dose … we can get therapeutic long-term knockdown of hepatitis B virus replication [with] no toxicity," he said. "I think this [work points to] the fact that we have reached the limit of toxicity in some cases, and that this may be helpful in understanding the microRNA processing pathway."

The findings point "to the fact that you should spend time finding sequences that have the highest specific activity, and not to over-express them," Kay said. "One of our things now is to use promoters that are weaker. When we do that, we're not seeing the same kind of toxicity.

"Like any drug, you just need to do good pharm/tox testing," Kay added. "I think people were really worried about off-targeting effects, and maybe people should just be more concerned about dose effects, which is a major classical issue with drug development."

-- Doug Macron ([email protected])

Despite Toxicity, Benitec Cautions Against Jumping to Conclusions; 'Everything is a Poison'

Despite the toxicities detailed in the Nature paper, Benitec CEO Sara Cunningham this week cautioned against drawing conclusions about the safety of expressed RNAi as a therapeutic.

Benitec is currently developing therapeutics, including one for hepatitis C and one for HIV/AIDS, based on an RNAi technology similar to the one used by Kay and his colleagues in their experiments. Kay is currently chairman of Benitec's scientific advisory board and serves as a strategic advisor on the company's hepatitis C program. Avocel, a company Kay co-founded, was acquired by Benitec in mid-2004 (see RNAi News, 5/21/2004).

"The intent of the study was specifically to understand the mechanism of toxicity -- that's why the mice were overdosed," Cunningham told RNAi News this week. "Therefore sequences were selected at random, there was no prescreening, in many cases the sequences didn't even have targets in the cell, [and] they were hundreds of times over-expressed -- specifically to elicit toxicity to understand what a potential side effect could be if you overdosed.

"Everything is a poison, it just depends on the dose," she added, referring to a quote from Paracelsus, a Swiss physician and alchemist known as the father of toxicology. "That's precisely the case here. You can't make a broad statement that shRNAs are toxic.

"It's very good that we know what the upper limits are" of shRNA toxicity, Cunningham said, noting that "we are nowhere near those in the hep C work that we've been doing."

Kay's paper "should not be misinterpreted as a sweeping statement about shRNA," she added.

In a press release issued by Benitec this week, John Rossi, a researcher at the Beckman Research Institute of the City of Hope and principal investigator on Benitec's HIV/AIDS therapy, said that "the take-home message [of Kay's paper] … is to carefully test the expression or dosing of sh/siRNAs for any potential clinical application in a relevant cellular or animal model."

The Scan

Shape of Them All

According to BBC News, researchers have developed a protein structure database that includes much of the human proteome.

For Flu and More

The Wall Street Journal reports that several vaccine developers are working on mRNA-based vaccines for influenza.

To Boost Women

China's Ministry of Science and Technology aims to boost the number of female researchers through a new policy, reports the South China Morning Post.

Science Papers Describe Approach to Predict Chemotherapeutic Response, Role of Transcriptional Noise

In Science this week: neural network to predict chemotherapeutic response in cancer patients, and more.