Genomics
& Proteomics
Reed Business
Information
Morris Plains, NJ, 07950
 
 |
 |
Making Sense of RNA
Interference Methods
Four
years after its discovery, RNAi has changed how drug
discovery scientists view genes and proteins
click
the image to enlarge
siRNA
may be created synthetically and inserted into the
cell (arrow 3), or obtained by processing of dsRNA
(arrows 1, 2). One of the strands of the siRNA
duplex targets a homologous (complementary) region
in a specific gene mRNA (arrow 4). The mRNA is
then cleaved (arrow 5), and the resulting pieces
are degraded (arrow 6), effectively silencing the
gene. (Source: Compugen)
| Since the early days of
molecular biology, ribonucleic acid (RNA) has taken a
back seat to deoxyribonucleic acid (DNA). DNA, after
all, holds the “instructions,” whereas RNA merely
carries them out. RNA’s status was upgraded in the early
1990s when professor Victor Ambros at Darmouth College
found RNA strands in Caenorhabditis elegans that
interfered with the worm’s messenger RNA (mRNA). In
1999, when professor Andrew Fire at the Carnegie
Institution of Washington and professor Craig Mello of
the University of Massachusetts discovered that
double-stranded RNA (dsRNA) administered to nematodes
could effectively turn off genes, the era of RNA
interference (RNAi) was born.
Inhibiting or
“silencing” genes through chemical means is not a new
idea. Antisense methods use artificial DNA that binds to
and inactivates complementary target DNA sequences,
thereby blocking transcription. Antisense technology has
intrigued researchers since the mid-1980s because it was
the first rapid, generalized technique for blocking
protein expression at the molecular level.
Antisense molecule design is relatively
straightforward, because scientists need only identify a
critical region of the gene and create, through standard
oligonucleotide chemistry, an antisense DNA strand that
binds to it. Scientists know that an antisense oligo
with the sequence TAACCG will always bind to
complementary “sense” strands with the sequence ATTGGC.
What is RNAi?
RNAi and antisense are
similar in that both use oligonucleotide sequences
complementary to a molecular gene target. However,
instead of binding to DNA in the nucleus, RNAi
inactivates mRNA, the immediate precursor to protein, in
the cytosol. When mRNA is treated with the appropriate
RNAi agent, protein synthesis stops, producing an effect
identical to that if the gene itself had been silenced.
Originally, RNAi was thought to arise only from
long, double-stranded RNA (dsRNA). As recently as the
late 1990s, RNAi was merely an interesting biological
mechanism limited to such lowly creatures as C.
elegans and zebrafish. In addition, the effect was
short-lived and suffered interference from a competing
dsRNA pathway known as the interferon response.
RNAi’s demonstration in mammalian cells was not
realized until 2001, when scientists learned that the
interferon effect did not interfere with dsRNA sequences
shorter than about 30 base pairs. In addition to not
suffering from the interferon effect, the shorter
strands, dubbed “small interfering RNA” (siRNA), showed
consistently longer activity.
“RNA interference
is more durable than antisense,” says Joseph Fratantoni,
MD, vice president of medical affairs at MaxCyte Inc.,
Rockville, Md. “It gives you more time to do your
experiment.”
Not your grandpa’s antisense
All gene-based techniques suffer from delivery
problems. Some, like antisense, are stoichiometric. That
is, one molecule of drug is required for each DNA
target. Since so many cells contain antisense targets,
doses tend to be large, creating toxicity problems.
siRNA, by contrast, targets mRNAs which are present at
very low levels in cells. Although siRNA activity
technically not catalytic, a little bit of siRNA goes a
long way.
Whether used as a gene-knockout method
for research or as therapeutics, antisense molecules are
short-acting and require repeated large doses. siRNA, by
contrast, is longer-acting. In some systems
siRNA-induced RNAi effects persist for as long as two
weeks, which would place them among the longest-acting
non-depot therapeutics if they reach that stage.
RNAi is also reversible, which holds tremendous
advantages in research settings where cells can serve as
their own controls before and after silencing.
The combined advantages of longer duration,
reversibility, and near-catalytic activity make RNAi a
much more attractive strategy than antisense for
treating intractable diseases. “You could say that after
just a few years siRNA has already outrun antisense, and
is beginning to leave antisense behind even more
rapidly,” says Klaus Lun, project manager at Amaxa
Biosystems GmbH, Köln, Germany. The learning curve for
siRNA work has been so short because of scientists’
experience with antisense. “When RNAi came along
everyone was already at the starting line, ready to go.
Mammalian RNAi techniques could not have come at a more
opportune time, with so much progress occurring in
genomics.”
The challenge of delivery
The biggest challenge by far for siRNA/RNAi is
getting siRNA into cells. Several companies are working
on this problem, many borrowing technology from
antisense.
About 10 companies offer reagents for
delivering siRNA. Most methods are lipid-based. One such
technique, borrowed from antisense work, is
liposomes—microscopic soap bubbles—that encapsulate
siRNA and transport it through cells’ fatty membranes.
As researchers discovered with antisense, liposomal
siRNA delivery doesn’t always work as promised.
“Cell-loading techniques based on liposomes create a lot
of background noise,” says Fratantoni.
click
the image to
enlarge
Principle of
siRNA-mediated RNA interference. (Source:
Eurogentec) | MaxCyte’s GT
cell loading system uses flow electroporation—electrical
pulses—that make cell walls transiently permeable.
Contrary to popular belief electroporation does not
“poke holes” in cell membranes. Instead, it causes
lipids in cell membranes to move apart “like opening a
stage curtain,” says Fratantoni.
The other
drawback with lipid-based delivery is it only works
satisfactorily with standard cell lines. “If you want to
do experiments in primary cells where you can analyze
the biological function of genes in their original
environment, it’s difficult to do with existing
reagents,” says Amaxa’s Lun. Amaxa sells another
electroporation-based transfection system, the
Nucleofector.
“With conventional transfection
methods it can take up to 48 hours before the
transfected gene can be analyzed, but that’s too long
for busy discovery laboratories to wait,” says Lun, who
claims that Nucleofactor does the job in primary cells
and cell lines in less than four hours.
Electroporation-based delivery, being nonviral,
does not depend on cell division, which is a big plus
with nondividing cells such as resting blood cells,
neurons, and lymphocytes.
Discovery: Still
number one
As a technique that knocks out gene
function, siRNA was immediately exploited by
pharmaceutical companies as a drug discovery tool.
“Traditional drug discovery was done by trial
and error—applying materials to cells until the desired
phenotype is obtained,” says Sharon Engel, director of
genomic data at Compugen Ltd., Tel Aviv, Israel. “Drugs
discovered in this method usually are nonspecific and
have extensive side effects. With siRNA, it’s possible
to identify the reason a phenotype develops, validate
your identification, and attack the exact location in
the cell that’s responsible for disease with minimal
side effects. RNAi’s sequence specificity allows the
kind of specific treatment you would demand from a
next-generation drug.”
Compugen’s platform
technology puts genes into clusters and then assembles
them to obtain the putative mRNAs which can arise from
these genes. These putative mRNAs are called transcripts
of genes and the collection of all these transcripts is
called the transcriptome. The firm’s collaboration with
Novartis, now in its second year, involves constructing
transcriptomes based on public, proprietary and
third-party information. The partnership was recently
extended to include the design of specific siRNA ligos
for target validation.
“Having the full
structure of a gene, including exon/intron construction
and all alternative splice variants, enables the design
of siRNA specific to the target gene and that influence
all splice variants or a specific one,” says Engel.
“Having the full transcriptome allows analysis of the
homology of the siRNA to other genes, and prevents
unintentional silencing of additional pathways. The
trick is not only to know what gene sequence you wish to
target, but also to know what other genes might be
targeted and avoid it. That’s important because siRNA
needs to be very specific. One or two mismatches may
prevent it from working.”
click
the image to enlarge
RNAi
addresses critical points in drug development.
(Souce: Sequitur) | Like
other chemical gene-knockout techniques—e.g. antisense,
protein nucleic acids, and Morpholinos (see
sidebar)—RNAi may be used in research or as a
therapeutic. The pharmaceutical industry’s first
commercial application of RNAi was target validation.
Through target validation, scientists block a gene
suspected to cause disease and note if the disease
markers disappear. Using conventional target validation
methods is difficult, time-consuming, and only
occasionally occurs early enough in a development
program to be of any significant consequence.
Today’s siRNA technology works anywhere from 60%
to 80% of the time, resulting in 90% or greater
reduction in target RNA and protein levels, says
Dominique Poncelet, PhD, product manager for
oligonucleotides at Eurogentec SA, Liège, Belgium.
“There do seem to be some resistant genes which are not
silenceable, but that may be related to the design of
the siRNA rather than any inherent limit to the
technology. Some genes may, for whatever reason, be
inaccessible due to their location or secondary
structure, although theoretically folding is not
supposed to be important.”
Most pharmaceutical
companies use siRNA/RNAi for drug target validation to
support large screening efforts, for example the
Novartis-Compugen agreement. “As a validation tool, some
biotechnology companies specializing in RNAi technology
have already generated libraries of thousands of siRNA
fragments,” says Poncelet, “and screened them in model
cells for whole organisms such as C. elegans or
Drosophila . Based on these platforms, some
biotechnology companies—for example Sequitur Inc.,
Natic, Mass., and Cenix Bioscience GmbH, Dresden,
Germany—are proposing complete functional genomic
subscription programs to traditional pharmaceutical
companies to speed up drug discovery.”
Since
siRNAs exhibit all of the hallmarks for the ideal
sequence-specific therapeutics, more and more biotech
companies will become interested in drugs that work
through RNAi, says Poncelet. He mentioned Alnylam
Pharmaceuticals Inc., Cambridge, Mass., and Ribopharma
AG, Kulmbach, Germany, among others, as early
therapeutics players and Mirus Corp., Madison, Wis., and
Ribozyme Pharmaceuticals Inc., Boulder, Colo., as
delivery specialists.
”Morphing” Improves Oligonucleotide
Stability
Like first-generation
DNA-targeting antisense oligos, siRNA is degraded
by nuclease enzymes, a fact that serious siRNA
drug development must eventually address.
Antisense developers reduced their compounds’
susceptibility to degradation by creating
chemically modified antisense oligos, for example
so-called second-generation phosphorothioates.
Another approach to medicines based on small gene
fragments is to alter the backbone of the
molecule, as Gene Tools LLC, Philomath, Ore., has
done with its Morpholino synthetic
oligonucleotides. In Morpholino oligos, the sugar
components of the DNA/RNA backbone has been
replaced by morpholine, and the charged
phosphodiester linkage has given way to an
uncharged phosphodiamidate. Like antisense and
siRNA, Morpholino oligos inhibit targeted mRNAs
through antisense binding but do so by a mechanism
that involves blocking either ribosome assembly or
RNA splicing within the nucleus.
Shannon
Knuth, PhD, of Gene Tools says Morpholinos are
more active than their “natural” siRNA
counterparts and are much less toxic than
phosphorothioates. “Antisense molecules are
unstable, nonspecific, and toxic,” she says.
“There is no known degradation pathway for
Morpholinos, and they appear to be more specific
for target genes than antisense or siRNA.”
Morpholinos were developed by James
Summerton, PhD, an antisense pioneer, current
president of Gene Tools, and founder of AVI
Biopharma, which is developing Morpholinos as
therapeutics. Gene Tools’ primary customers for
Morpholinos are developmental biology research
laboratories.
| Therapeutics
siRNA therapeutics promise the specificity and
low toxicity that antisense never delivered, and of
which small-molecule are probably incapable. RNAi’s
differs fundamentally from antisense and “chemical”
drugs in that it is a natural, endogenous process that
cells already recognize and exploit. Cells were not
designed to receive small organic chemicals or to have
their genes silenced by phosphorothioate antisense
compounds. RNAi has been with cells through evolution.
Specificity is the key difference between siRNA
and other drugs, says John Maraganore, PhD, CEO of
startup siRNA therapeutic company Alnylam. “siRNA is
highly selective with respect to the mRNA it degrades as
long as you make sure that the target sequence is
unique.”
To date most siRNA work has focused on
target validation as opposed to therapeutics. But as
some have pointed out, due to the very nature of RNAi if
you have a tool for target validation, you may also have
a drug.
Because it’s so easy to create siRNA
oligos, synthesis is no longer the bottleneck in
discovery as can be is with small-molecule therapeutics.
For example, seven genes are are suspected as possible
culprits in a disease, scientists would ideally like to
test each one. Making effective inhibitors to all seven
proteins encoded by those genes would take years, so
researchers prefer to knock out the genes, one by one,
and see which knockout is effective. Because siRNA is so
easy to make, all one needs to test all seven genes is
knowledge of their (or their mRNA’s) structure, and an
oligonucleotide synthesizer. In fact, one could
synthesize many siRNA sequences for each gene and test
all of them in a fraction of the time it would take to
inhibit the corresponding proteins using small
molecules. The exciting part is that after the siRNA set
is optimized one has not just a convenient way to knock
out genes and validate targets, but a potential
therapeutic as well.
The relative ease with
which siRNA can be produced, relative to small “organic”
molecules, should not be underestimated. What makes
siRNA powerful and fast in target validation makes it
even more attractive for therapeutics. siRNA combines
drug and target validation in one molecule, one
platform, while obviating the need for traditional
medicinal chemistry. With siRNA knowledge of the gene is
sufficient, but not even necessary. When the gene is
known so is its mRNA sequence, and therefore the exact
structure of a proposed interference sequence. Narrowing
that sequence down to an optimized, 20-odd nucleotide
takes work, to be sure, but not as much work as
optimizing a small-molecule enzyme inhibitor.
Toxicity is the main reason siRNA is more
attractive than antisense (and why, way back when,
antisense appeared to hold more promise than
small-molecule medicines), says Tod Wolf, PhD, president
of Sequitur Inc., Natick, Mass. Toxicity in
small-molecule drugs and antisense arises from lack of
specificity to protein and gene targets, respectively,
as well as from poorly-understood mechanism-related
effects.
So far we have made siRNA drug
development sound easy, which of course it is not. Woolf
says that researchers must solve delivery, stability,
and tissue specificity issues before anything can come
of siRNA therapeutics. “Naked siRNA is not serum-stable,
and free oligos tend to hone in on certain tissues. So
if you want to target disease sites specifically you
need a way to protect the siRNA, for example by
encapsulating it with lipids. In cell culture, it’s
possible to use transfection agents to make RNAi more
efficient, but lipids that work in cell culture do not
always work in animals.”
As Woolf says, many
potential problems with siRNA have been resolved through
experience with antisense molecules, for example
manufacturing, toxicity, and immunogenicity.
End of history for drug discovery?
Heralding the end of history for traditional drug
discovery, or proclaiming RNAi as the pharmaceutical
industry’s microprocessor is undoubtedly premature. But
it is awfully tempting to do so. Four years after its
discovery, RNAi has changed how drug discovery
scientists view genes and proteins, transformed how we
view RNA and DNA, and greatly expanded methods for
discovering what causes disease.
Angelo DePalma
DePalma is a freelance writer based in Newton, N.J.
|
