|
For RNAi, the headlines
keep coming. Within the last few months, researchers
at Alnylam published the first paper demonstrating
systemic delivery of small interfering RNA (siRNA) in
primates. In another major development, two academic
groups just published on a new class of small RNAs (piRNAs)
found in mouse testes. Meanwhile, a group at Stanford
reported sobering news: In a test of their RNAi
treatment for hepatitis B, many mice died from liver
complications.
“From 2001 to 2004,we
had pure excitement in RNAi,” says Dharmacon’s
Anastasia Khvorova. During these early golden years,
scientists were delighted by the power and ease of
siRNA for silencing. But then by 2004 came setbacks
— off-target effects and triggering of the
interferon response.
Drugs need to be
specific. In the lab, “People could easily be fooled
by off-target effects,” says City of Hope’s John
Rossi. “You see your target is knocked down and you
think that [the knock down] is what is causing the
phenotype.”
Unintended
Consequences
Part of the problem is
study design. “Some people use siRNA the way they
used antisense oligonucleotides,” warns Cy Stein of
New York’s Albert Einstein College of Medicine.
“They add however much they need to downregulate.”
Stein notes that “We saw plenty of papers using high
concentrations of oligos that caused all kinds of
things to happen.” The vehicle used to get the
sequence
into cells can also have multiple effects, and these
can vary “depending on the cell type you are
studying,” says Stein.
Too much foreign RNA in
the cells could be what caused the mouse deaths in the
Stanford study, which was led by Mark Kay. “It shows
that there is a danger to being greedy,” says Rossi.
“You need the lowest possible concentration to get
the work done.” Unfortunately, the therapies
sometimes don’t work at lower doses.
For those using siRNAs
at the bench, Khvorova says specificity can be
improved through bioinformatics, upfront experimental
design, assay validation, pooling (using mixes of
siRNAs), and new
reagents. Dharmacon scientists and others recently
determined that some siRNAs go “off-target” by
hitting the microRNA pathway. “New chemical
modifications can enhance siRNA specificity,” says
Khvorova. Dharmacon’s new On-Targetplus reagents,
“can help reduce up to 90 percent of the off-target
effects,” she says.
Rossi says one of the
best controls for confirming siRNA knockdown is to use
two or more siRNAs to different sequences in the same
target, in separate experiments. To be really certain,
“A codon
modified cDNA in the sequence of the siRNA target can
be used which should be resistant to the siRNA and
restore the original phenotype,” he says. Margolis
and colleagues describe one such technique in the
Technical Brief on page 35. “This approach was also
recommended by an expert panel several years ago,”
Rossi says. His group demonstrated another gene
replacement technique targeting EGFP in mouse cells
(Kim, D.H. and Rossi, J. Antisense Nucleic Acid
Drug Dev 13, 151-5; 2003).
Another problem, says
William Earnshaw at the University of Edinburgh, is,
“When you do RNAi you never get rid of 100 percent
of the protein.” In addition, “Enzymes that were
binding that protein,
may be binding something else, and that could be
causing the phenotype,” he warns.
Whole New Worlds
Two major types of
synthetic small RNA are usually used for silencing —
siRNA and short hairpin RNA (shRNA). The Technical
Brief on page 31 discusses a new approach using
shRNAmirs.
The key to using any of
the tools properly is figuring out what’s going on
in the cell. Scientists like Greg Hannon of Cold
Spring Harbor are broadening our view of that. His
group used 454’s platform to do “deep
sequencing” of mouse testes cells. “We didn’t
have to look very hard,” says Hannon. But because
the tool provides hundreds of thousands of sequence
reads in a few hours, “We discovered a new class of
RNAs that is abundant in these cells.” Those are the
Piwi-interacting RNAs (piRNAs).
“Understanding how
the pathway works has always been really important to
apply these things as tools,” says Hannon. “A
couple of years ago the enzymatic machinery for siRNA
was discovered,
now we are trying to do that for piRNAs.”
|
