First Meeting of the Oligonucleotide Therapeutics Society

Reading Room

Approaches for the sequence-specific knockdown of mRNA

Lisa J. Scherer & John J. Rossi
Division of Molecular Biology, Beckman Research Institute of the City of Hope, Duarte, California 91010, USA.

Copyright (2003) Nature Publishing Group. This article may be downloaded for personal use only. Any other use requires prior permission of the authors and the Nature Publishing Group.
This article appeared in Nature Biotechnology 21, 1457-1465 (2003).

Abstract: Over the past 25 years there have been thousands of published reports describing applications of antisense nucleic acid derivatives for targeted inhibition of gene function. The major classes of antisense agents currently used by investigators for sequence-specific mRNA knockdowns are antisense oligonucleotides (ODNs), ribozymes, DNAzymes and RNA interference (RNAi). Whatever the method, the problems for effective application are remarkably similar: efficient delivery, enhanced stability, minimization of off-target effects and identification of sensitive sites in the target RNAs. These challenges have been in existence from the first attempts to use antisense research tools, and need to be met before any antisense molecule can become widely accepted as a therapeutic agent.

Over the past two decades, the use of nucleic acid-based inhibitors of gene expression (antisense agents) has come in and out of fashion. Initial excitement in this area came in the mid-1980s as synthetic DNA chemistry hit full stride, enabling sequence-specific antisense ODNs to be rapidly synthesized and tested for target-specific knockdown of gene expression. It was quickly realized that certain backbone modifications were necessary for full activity of these compounds and that efficient delivery to target cells was a critical requirement. Chemists quickly addressed these challenges by developing a variety of backbone modifications that stabilized antisense ODNs without inhibiting their biological activities. Delivery challenges were also addressed with the introduction of anionic and cationic lipid formulations for packaging and delivering the net negatively charged ODN compounds to a variety of cells in culture. In subsequent years, however, interest declined because the predicted utility of these compounds as therapeutic agents was slow to materialize and, in fact, remains limited to a handful of compounds.

The second wave of interest in nucleic acid-based inhibitors of gene expression followed the discoveries of catalytic RNAs (ribozymes) in the early 1980s. The full potential of ribozymes for target-specific inhibition of gene expression was not completely realized until the late 1980s and early 1990s when simplified catalytic motifs were defined, making these molecules amenable to chemical synthesis. The exploitation of ribozymes as therapeutic agents also depended heavily upon stabilizing backbone modifications that did not inhibit activity and efficient delivery. Fortunately, these issues could be addressed by drawing on the extensive experience of the antisense ODN field. Ribozymes have an advantage over ODNs in that ribozyme genes can be delivered to cells with plasmid or viral vectors, and ribozyme expression can be controlled with promoter-based expression.

The most recent explosion of interest in the antisense world followed the discoveries of Mello and colleagues1 in Caenorhabditis elegans in 1998, and of others in mammalian cells in 2001, that double-stranded RNAs (dsRNAs) elicit potent targeted degradation of complementary RNA sequences, termed RNA interference (RNAi). Moreover, it was shown that the active component of the RNAi pathway, termed small interfering RNAs (siRNAs), can be chemically synthesized or expressed from vector backbones, similar to ribozymes. The interest in RNAi has been fueled—to an even greater extent than interest in antisense ODNs and ribozymes—by the completion of the human genome sequence initiative because siRNAs can elicit potent, target-specific knockdown of any mRNA, creating a useful and proven surrogate genetic tool. Although RNAi provides a powerful new tool for targeted inhibition of gene expression, there are, nevertheless, concerns and limitations in the use of this technology as well, including efficient delivery and potential side effects.

There have been important developments in the other areas of antisense technologies, giving the investigator several options, depending upon the experimental system and desired outcome. This article explores the basic mechanisms of action of only the popular antisense inhibitory agents. We then compare the advantages and disadvantages for each class of inhibitory agent. We do not intend to present a comprehensive review of the antisense world, but rather to provide a framework for thinking about which agent best matches the goals of an experimental or therapeutic application.

Copyright (2003) Nature Publishing Group. This article may be downloaded for personal use only. Any other use requires prior permission of the authors and the Nature Publishing Group.
This article appeared in Nature Biotechnology 21, 1457-1465 (2003).

Web Sites

Ambion, The RNA Interference Resource
This site, maintained by Ambion, Inc., publishes recent news about RNAi research.

Genome-Wide RNAi Global Initiative
A press release from Dharmacon, Inc. describing the formation of an international alliance of biomedical research centers to make use of recent advances in siRNA research.

Nature Focus RNAi Interference
Three animations, on gene expression, gene silencing, and RNAi amplification.

NOVA ScienceNow RNAi
Contains video of a segment on RNAi broadcast on NOVA, hosted by Robert Krulwich. The site also offers additional background information.

Oligonucleotide Therapeutics Society
A nonprofit forum created to foster academia- and industry-based research and development of oligonucleotide therapeutics.

The RNAi Consortium
A public-private consortium based at the Broad Institute in Cambridge, MA, created to develop and validate tools and methods for using RNAi to study the function of human and mouse genes.

RNAi.net
The "world of RNA interference"—including headlines, interviews, upcoming events, and information on suppliers.

The RNA World Website
Published by the Institute for Molecular Biology in Jena, Germany, this site features extensive links concerning research on RNA.

Companies conducting research in oligonucleotide therapeutics

Acuity Pharmaceuticals
Ambion
Alnylam Pharmaceuticals
Archemix (aptamers)
Coley Pharmaceuticals
Dharmacon RNA Technologies
Ester Neuroscience
Eyetech
Genta
Hybridon
Invitrogen RNAi Central
ISIS Pharmaceuticals
OncoGeneX Technologies, Inc.
Sirna Therapeutics
VasGene

Journal Articles

Abel, K., T. Wang, L. Fritts et al. 2005. Deoxycytidyl-deoxyguanosine oligonucleotide classes A, B, and C induce distinct cytokine gene expression patterns in rhesus monkey peripheral blood mononuclear cells and distinct alpha interferon responses in TLR9-expressing rhesus monkey plasmacytoid dendritic cells. Clin. Diagn. Lab Immunol. 12: 606-621.

Adams, A. 2005. RNA Therapeutics Enter Clinical Trials. The Scientist (January 17).

Amarzguioui, M., J. J. Rossi & D. Kim. 2005. Approaches for chemically synthesized siRNA and vector-mediated RNAi. FEBS Lett. 579: 5974-5981.

Arzumanov, A., D.A. Stetsenko, A.D. Malakhov et al. 2003. A structure-activity study of the inhibition of HIV-1 Tat-dependent trans-activation by mixmer 2'-O-methyl oligoribonucleotides containing locked nucleic acid (LNA), α-L-LNA or 2'-thio-LNA residues. Oligonucleotides 13, 435-453.

Barik, S. 2004. Development of gene-specific double-stranded RNA drugs. Ann. Med. 36: 540-551.

Chen, X., N. Dudgeon, L. Shen & J. H. Wang. 2005. Chemical modification of gene silencing oligonucleotides for drug discovery and development. Drug Discov. Today 10: 587-593.

Cui, Z., S. J. Han, D. P. Vangasseri & L. Huang. 2005. Immunostimulation mechanism of LPD nanoparticle as a vaccine carrier. Mol. Pharm. 2: 22-28.

DeFranco, A. L., R. M. Locksley & M. Robertson. The toll-like receptor family of innate immune receptors. In Immunity: The Immune Response to Infection (Primers in Biology). New Science Press, London. FULL TEXT

Du, T. & P. D. Zamore. 2005. microPrimer: the biogenesis and function of microRNA. Development 132: 4645-4652.

Egli, M., G. Minasov, V. Tereshko et al. 2005. Probing the influence of stereoelectronic effects on the biophysical properties of oligonucleotides: comprehensive analysis of the RNA affinity, nuclease resistance, and crystal structure of ten 2'-O-Ribonucleic Acid Modifications. Biochemistry. 44: 9045-9057.

Elbashir, S. M., W. Lendeckel & T. Tuschl. 2001. RNA interference is mediated by 21- and 22-nucleotide RNAs. Genes Dev. 15: 188-200. FULL TEXT

Gleave, M. E. & B. P. Monia. 2005. Antisense therapy for cancer. Nat. Rev. Cancer. 5: 468-479.

Hartmann, G., A. Krug, K. Waller-Fontaine & S. Endres. 1996. Oligodeoxynucleotides enhance lipopolysaccharide-stimulated synthesis of tumor necrosis factor: dependence on phosphorothioate modification and reversal by heparin. Mol. Med. 2: 429-438.

Hede, K. 2005. Blocking cancer with RNA interference moves toward the clinic. J. Natl. Cancer Inst. 97: 626-628.

Hornung, V., M. Guenthner-Biller, C. Bourquin et al. 2005. Sequence-specific potent induction of IFN-alpha by short interfering RNA in plasmacytoid dendritic cells through TLR7. Nat. Med. 11: 263-270.

Huang, Z., W. Li, J. A. MacKay & F. C. Szoka, Jr. 2005. Thiocholesterol-based lipids for ordered assembly of bioresponsive gene carriers. Mol. Ther. 11: 409-417.

Jackson, A. L. & P. S. Linsley. 2004. Noise amidst the silence: off-target effects of siRNAs? Trends Genet. 20: 521-524.

Jahrsdorfer, B., J. E. Wooldridge, S. E. Blackwell et al. 2005. Immunostimulatory oligodeoxynucleotides induce apoptosis of B cell chronic lymphocytic leukemia cells. J. Leukoc. Biol. 77: 378-387.

Jepsen, J. & J. Wengel. 2004. LNA-antisense rivals siRNA for gene silencing.Curr. Opin. Drug Discov. Devel. 7: 188-194.

Judge, A. D., V. Sood, J. R. Shaw et al. 2005. Sequence-dependent stimulation of the mammalian innate immune response by synthetic siRNA. Nat. Biotechnol. 23: 457-462.

Juliano, R. L., V. R. Dixit, H. Kang et al. 2005. Epigenetic manipulation of gene expression: a toolkit for cell biologists. J. Cell. Biol. 169: 847-857.

Juliano, R. L. 2005. Peptide-oligonucleotide conjugates for the delivery of antisense and siRNA. Curr. Opin. Mol. Ther. 7: 132-136.

Karagiannis, T. C. & A. El-Osta. 2005. RNA interference and potential therapeutic applications of short interfering RNAs. Cancer Gene Ther. (Epub, May 13).

Kim, V. N. 2005. MicroRNA biogenesis: coordinated cropping and dicing. Nat. Rev. Mol. Cell Biol. 6: 376-385.

Kole, R., M. Vacek & T. Williams. 2004. Modification of alternative splicing by antisense therapeutics. Oligonucleotides 14: 65-74.

Lai, J. C., L. Benimetskaya, A. Khvorva et al. 2005. Phosphorothioate oligodeoxynucleotides and G3139 induce apoptosis in 518A2 melanoma cells. Mol. Cancer Ther. 4: 305-315.

Langlois, M. A., C. Boniface, G. Wang et al. 2005. Cytoplasmic and nuclear retained DMPK mRNAs are targets for RNA interference in myotonic dystrophy cells. J. Biol. Chem. 280: 16949-16954.

Latz, E., A. Visintin, T. Espevik & D. T. Golenbock. 2004. Mechanisms of TLR9 activation. J. Endotoxin Res. 10: 406-412.

Leaman, D., P. Y. Chen, J. Fak et al. 2005. Antisense-mediated depletion reveals essential and specific functions of microRNAs in drosophila development. Cell 121: 1097-1108.

Meister, G. & T. Tuschl. 2004. Mechanisms of gene silencing by double-stranded RNA. Nature 431: 343-349.

Morizono, K. & I. S. Chen. 2005. Targeted gene delivery by intravenous injection of retroviral vectors. Cell Cycle 4 (Epub).

Moseman, E. A., X. Liang, A. J. Dawson et al. 2004. Human plasmacytoid dendritic cells activated by CpG oligodeoxynucleotides induce the generation of CD4+CD25+ regulatory T cells. J. Immunol. 173: 4433-4442.

Okano, F., M. Merad, K. Furumoto & E. G. Engleman. 2005. In vivo manipulation of dendritic cells overcomes tolerance to unmodified tumor-associated self antigens and induces potent antitumor immunity. J. Immunol. 174: 2645-2652.

Rafflo, A., J. C. Lai, C. A. Stein et al. 2004. Antisense RNA down-regulation of bcl-2 expression in DU145 prostate cancer cells does not diminish the cytostatic effects of G3139 (Oblimersen). Clin. Cancer Res. 10: 3195-3206. FULL TEXT

Rossi, J. J. 2005. Receptor-targeted siRNAs. Nat. Biotechnol. 23: 682-684.

Rossi, J. J. 2005. RNAi and the P-body connection. Nat. Cell Biol. 7: 643-644.

Sage, C., M. Huang, K. Karimi et al. 2005. Proliferation of functional hair cells in vivo in the absence of the retinoblastoma protein. Science 307: 1114-1118.
This paper served as basis for collaboration between Harvard Medical School and SiRNA to stimulate inner ear hair follicle growth by knocking down expression of Rb.

Scherer, L. J. & J. J. Rossi. 2003. Approaches for the sequence-specific knockdown of mRNA. Nat. Biotechnol. 21: 1457-1465. FULL TEXT (PDF, 9 pages, 982.2 KB)

Schwarz, D. S., G. Hutvagner, T. Du, et al. 2003. Asymmetry in the assembly of the RNAi enzyme complex. Cell 115: 199-208.

Smith, A. E. & A. Helenius. 2004. How viruses enter animal cells. Science 304: 237-242.

Song, E., P. Zhu P, S. K. Lee et al. 2005. Antibody mediated in vivo delivery of small interfering RNAs via cell-surface receptors. Nat. Biotechnol. 23: 709-717.

Tomai, Y. & P. Zamore. 2005. Perspective: machines for RNAi. Genes Dev. 19: 517-529.

Turner, J.J., A.A. Arzumanov & M.J. Gait. 2005. Synthesis, cellular uptake and HIV-1 Tat-dependent trans-activation inhibition activity of oligonucleotide analogues disulphide-conjugated to cell penetrating peptides. Nucl Acids Res. 33, 27-42. FULL TEXT

Wang, H., E. Rayburn & R. Zhang. 2005. Synthetic oligodeoxynucleotides containing deoxycytidyl-deoxyguanosine dinucleotides (CpG ODNs) and modified analogs as novel anticancer therapeutics. Curr. Pharm. Des. 11: 2889-2907.

Zamecnik, P. & M. L. Stephenson. 1978. Inhibition of Rous sarcoma virus replication and cell transformation by a specific oligodeoxynucleotide. Proc. Natl. Acad. Sci. USA 75: 280-284. (PDF, 834 KB) FULL TEXT

Books

Agrawal, S., Ed. 1996. Antisense Therapeutics (Methods in Molecular Medicine series). Humana Press, Totowa, NJ.

Crooke, S. T., Ed. 2001. Antisense Drug Technology: Principles, Strategies, and Applications. Marcel Dekker, New York.

Engelke, D. 2004. RNA Interference (RNAi): The Nuts & Bolts of siRNA Technology (Nuts & Bolts Series). DNA Press, Eagleville, PA.

Gibson, I., Ed. 2003. Antisense and Ribozyme Methodology: Laboratory Companion (Laboratory Companion Series). Wiley-VCH, Hoboken, NJ.

Hannon, G. J., Ed. 2003. RNAi: A Guide to Gene Silencing. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

Conference Contributors

Presented by

Silver Sponsors

Bronze Sponsors

Alnylam Pharmaceuticals

ALTANA Pharma AG

Avatar Pharmaceutical Services Inc.

Calando Pharmaceuticals

Dharmacon Inc.

Dowpharma

Dynavax Technologies

Ercole Biotech Inc.

Eurogentec

Eyetech Pharmaceuticals

GE Healthcare

Genta

Glen Research Corporation

IDT

Invitrogen

Kinovate Life Sciences, Inc.

Merck

Monomer Sciences Inc.

Pierce Nucleic Acid Technologies

PrimeSyn Lab Inc.

Promega

Sigma-Aldrich Corp.

Topigen

Exhibitors

Avatar Pharmaceutical Services Inc.

Avecia

Dowpharma

Eurogentec

GE Healthcare

Invitrogen

Kinovate Life Sciences, Inc.

PrimeSyn Lab Inc.

Media Sponsor

Cell Press

Conference Dinner Sponsor

Coley Pharmaceutical Group, Inc.

Lunch/Poster Session Sponsor – Friday September 16, 2005

Topigen

Exclusive Tote Bag Sponsor

Genta

Coffee Break Sponsor

Invitrogen

Platforms for drug design

Aspirin, birth control pills, and insulin are medicines with completely different chemical structures. Therefore, they must be manufactured by radically different processes. Imagine an entirely new class of therapeutics, made from just four chemically similar building blocks, strung into chains twenty or so building blocks long. Imagine that just by changing the arrangement of the four building blocks you could make a host of different therapeutic compounds. With one sequence, you could make a drug for knocking down production of a secreted protein involved in heart disease. Using the same blocks in a different array you could shut down a mitochondrial protein that prevents cancer cells from committing suicide. And so forth.

In oligonucleotide therapeutics, drug designers go after the messenger RNAs that encode the proteins that cause disease.

This sounds like an elegantly simple, inexpensive platform for drug development. Indeed, researchers conceived such an approach in the late 1970s, drawing on the discovery by Paul Zamecnik and Martha Stephenson that a synthetic oligonucleotide with a base sequence that was complementary to a particular viral RNA could inhibit replication of the virus when taken up by the virally infected cells. The oligonucleotide, as Zamecnik and Stephenson discovered, formed a Watson-Crick type duplex with the viral mRNA, and consequently inhibited translation of the viral protein. That finding, along with availability of the rapidly expanding genome sequence information, paved the way for the field of oligonucleotide therapeutics in which drug designers would bypass disease-associated proteins and go after the messenger RNAs that encoded the culprit proteins.

The first oligonucleotide therapeutic to gain FDA approval—Isis Pharmaceuticals's Vitravene, which treats a viral infection of the eye—went to market in 1998. The race to become the second approved drug in this class awaits a winner. Developers of oligonucleotide therapeutics are holding their breath as many of these new products move through clinical testing. Genasense, an antisense oligonucleotide (ASO) drug aimed at pushing cancer cells towards suicide, recently did well in a Phase III leukemia trial, prompting its maker, Genta Inc., to announce its intention to file for FDA approval by December 2005. Two ASOs (so-called because they carry the sequences that could form base pairs with the sense messenger RNA strands) from Isis Pharmaceuticals, built to be stronger and longer-lasting drugs than those of the Genesense generation, are showing favorable results in early clinical trials for treating Type 2 diabetes and hyperlipidemias. In addition, members of the fledgling class of oligonucleotide drugs called siRNAs (small interfering RNAs, produced by the company aptly named Sirna) are doing well in early phase trials on macular degeneration and Hepatitus C. Also, immune stimulatory oligonucleotides from Coley Pharmaceuticals are showing robust clinical responses in treating cancers and as a vaccine adjuvant against Hepatitus B viral infections. The list goes on.

At the same time, many ASO therapeutics have fallen out of the pipeline and siRNA drug developers are unexpectedly encountering some of the same challenges faced by ASO developers. Therefore, a spirit of cautious optimism mixed with a desire to create a clearinghouse for useful information about developing oligonucleotide-based medicines has spurred the formation of the Oligonucleotide Therapeutics Society. The Society brings together molecular biologists intent on developing oligonucleotides drugs with chemists who can tweak the compounds to make them more potent, pharmacologists and chemical engineers who are devising better ways of delivering them to their intended target cell or tissue, and physicians who are testing the compounds in patients. In its first meeting, which was cosponsored by the New York Academy of Sciences, more than 270 academic and industrial researchers from around the world gathered at The Rockefeller University on September 15-18, 2005.

The meeting, which covered everything from the basic mechanisms to the clinical trials, had three principal areas of focus, ASO, siRNAs, and immunostimulant oligonucleotides. Highlights in the ASO area included the following:

• Advances in chemistry are leading to ASO drugs with longer half-lives and greater affinities for their targets than earlier compounds. Some of these next-generation compounds have moved beyond the bench and into clinical testing.
• ASOs can successfully target nuclear mRNA precursors as well as cytoplasmic mRNAs. This finding opens the door for treating diseases stemming from inappropriate pre-mRNA splicing, of which there may be many.
• Satisfactory strategies for predicting which sequence within a given RNA would be the best ASO target are still lacking. However, a novel method for identifying optimal target sequences within the mRNA of interest has been developed.
• Early ASO work was often mired by unwanted off-target ASO actions, both on non-target RNAs and on substances other than RNA. Reports at the meeting showed that many ASO investigators are finding ways to determine if observed therapeutic affects in humans stem from on-target ASO action. Studies on prostate cancer patients, for example, gave encouraging results of an ASO drug that appear to be efficacious and on-target.

SiRNA-based drugs differ from ASOs in that they mimic naturally occurring RNAs that silence gene expression. Researchers are still unraveling the mechanisms involving these small, non-coding RNAs. Therefore, highlights of the meeting included new information about mechanisms as well as reports on the delivery and applications of the compounds:

• The siRNA therapeutics are double-stranded (ds) RNAs, one strand of which becomes the "antisense" guide strand that leads the target mRNA to its demise. If both siRNA strands had equal likelihood of becoming the guide strand, half of the drug material would be wasted. Now researchers have deciphered some rules to help them design drugs in which they can predetermine that only one of two strands can become the guide strand.
• Like ASOs, many siRNAs show off-target effects. Meeting presentations emphasized how extensive these off-target effects can be, and gave some clues about how to avoid them.
• Drug delivery poses a major challenge for the siRNA therapeutics field. Presenters at the meeting offered several approaches for improving delivery, including the assembly of drug complexes that carry ligands to specific cell membrane receptors and can therefore hone in on the cells of interest.
• Like ASOs, siRNAs may be more useful as therapeutic cocktail components than as monotherapies. At the meeting some researchers described efforts to develop RNAi cocktails comprised of siRNAs against three different targets.

Off-target actions of ASOs have posed serious hurdles for developers of ASO drugs. However, one off-target effect, the immunostimulatory action of oligodeoxynucleotides that contain CpG sequences, is now being harnessed for its own therapeutic purposes. (CpG refers to the dinucleotide of cytidine linked to guanidine through a phosphodiester bond.) Indeed, OTS meeting organizers gave equal time for the immunostimulatory oligonucleotide talks as they did for the ASO and the siRNA presentations. Stimulating aspects of those presentations included the following:

• The CpGs within oligodeoxynucleotides can activate two types of immune cells—dendritic cells and B lymphocytes. Dendritic cell activation has a totally different effect on the body than does B-cell activation. Therefore, researchers are finding ways to design compounds that activate just one or the other of those immune cell types.
• The CpG-dependent oligodeoxynucleotide activation occurs through a specific protein receptor (the Toll-Like Receptor 9, or TLR9) in endosomes of dendritic cells and B cells. Researchers reported at the meeting that siRNAs also have immunostimulatory actions, via a different endosomal receptor, TLR7. SiRNA drug developers took that finding as a cautionary note whereas investigators working on immunostimulatory oligonucleotides see it as another potential therapeutic tool.
• CpG oligonucleotide drugs have moved rapidly into clinical trials and currently show encouraging results as immune boosters for vaccines, both for preventing infections and for treating cancers. Early phase results of other studies show that some chemotherapeutic agents may be more effective when used along with a CpG oligonucleotide drug. Finally, other trials suggest that this drug class may be useful as monotherapy for treating a chronic viral disease such as hepatitis C as well as certain metastatic cancers.

The take home message from the meeting? John Rossi, chair of the division of molecular biology at City of Hope and president of the Oligonucleotide Therapeutics Society, stated it well: "The first annual meeting of the Oligonucleotide Therapeutics Society captured the breadth and excitement as well as problems and concerns of oligonucleotide therapeutics. We have begun an exciting new era to unify the many oligonucleotide factions in the direction of applied therapeutics."

The wealth of information shared by participants at the meeting, covering both the progress in the field as well as some approaches to avoid, give reason to believe that the efforts at mechanism-based oligonucleotide therapeutics design will offer new platforms for drug development.

Beth Schachter, PhD, writes about life science, medicine and biotechnology.

Immunostimulant oligonucleotide mechanisms of action

In the mid-1990s—the heyday of academic ASO research—many researchers started seeing surprising results in their ASO studies on cells in culture. Some oligodeoxynucleotides (ODNs) that were intended to be negative controls (for example, ODNs made with the same set of nucleotides as the test ASO, but arranged in a scrambled order) stimulated the cell systems far more than the test ASO did. This situation particularly plagued researchers studying cells of the immune system.

The solution to this conundrum came from studies done independently by Arthur Krieg, Günther Hartmann, and Sudhir Agrawal. Their work showed that certain oligonucleotides, particularly those with one or more cytidine-phosphothymidine dinucleotide (CpG) sequences, had immunostimulatory effects, such as prompting B lymphocytes to proliferate rapidly.

The immunostimulatory effects of CpG-containing ODNs turned out to occur, not through antisense actions on specific mRNAs, but by binding to an endosomal receptor in the Toll-like receptor family of signal transducers. The Toll-like receptor (TLR) family comprises a set of receptors expressed principally in cells of the immune system, and recognizes biochemical structures of infectious microbes. TLR-9, the receptor that CpGs activates, is the TLR family member that detects microbial DNA.

Immunostimulatory effects of CpG-containing ODNs occur by binding to a receptor in the Toll-like receptor family.

Microbial genomes, as Krieg would report, carry the expected prevalence of CpG sequences. In contrast, vertebrates have an unexpected dearth of this dinucleotide sequence. Furthermore, in vertebrate genomes, many of the cytidine residues are methylated, creating a further distinction between microbial and vertebrate DNAs. This set of distinctions, which underpins TLR-9's ability to distinguish between microbial and vertebrate DNA, makes the CpG-containing synthetic oligonucleotides potent B-cell mitogens.

Like turning lemons into lemonade, these investigators have applied their unexpected research findings to develop oligonucleotide immunotherapeutics, including several that are currently being tested in the clinic. The relative importance of this class of oligonucleotide therapeutics was reflected by the fact that oligoncleotide immunostimulation garnered equal time with ASOs and RNAi on the OTS meeting program.

The original findings about immunostimulatory oligonucleotides dealt with ASO molecules that mimic microbial DNA, acting through TLR-9. However, recent reports, including presentations at the OTS meeting, show that another TLR sibling, TLR-7, recognizes infectious viral RNA sequences or molecules that mimic viral RNA sequences, such as siRNAs. Not surprisingly, that finding sent up red flags to the investigators working on RNAi drug development and was therefore a topic of discussion at the meeting.

Deciphering the rules for fine-tuning immunostimulant responses

Arthur Krieg pioneered the field of oligonucleotide immunostimulation. As cofounder and CSO of Coley Pharmaceuticals, Krieg is blazing the trail for immunostimulatory ODNs to become new therapeutics. At the OTS meeting, he reported on several of Coley's ongoing clinical trials.

First, Krieg reviewed key features of immunostimulatory ODNs that provide the foundation for developing and refining this drug class. He noted, for example, that CpG-containing ODNs activate two different immune cell types, plasmacytoid dendritic cells (PDCs) and B lymphocytes. As Krieg pointed out, ODN activation prompts PDCs to secrete immunoregulatory compounds such as interferons, whereas activation of B cells causes them to release a different set of regulators, including the cytokines IL-6 and IL-10.

Reasoning that therapies would need to be tailored to stimulate PDCs or B cells selectively, but not both, Krieg and colleagues have studied ODNs with the CpG motif in different contexts. As he reported, this effort is helping researchers learn ways to develop ODNs with the desired PDC or B-cell selectivity.

Turning to clinical work, Krieg reported encouraging results from a variety of oncology trials, using immunonstimulatory ODNs as monotherapy, in combination with a cancer vaccine, or along with chemotherapy. Similarly, Coley's current infectious disease trials involve both monotherapy and use of CpG oligos as adjuvant therapy.

The antiviral adjuvant studies that Krieg described were particulary noteworthy, given the worldwide concern about viral epidemics. Discussing the hepatitis B vaccination trials, Krieg noted that CpG adjuvant treatment shortens the time it takes to mount an immune response. In addition, subjects who got vaccine with adjuvant developed antibodies that had greater avidity for antigen than the antibodies developed by individuals receiving just the vaccine. Krieg noted that these adjuvant-dependent differences occurred both in healthy volunteers and in individuals whose immune systems were compromised by HIV infection.

Krieg also had some good news for researchers who want to develop ASO-based drugs that are free of immunostimulatory effects. As he explained, some of the newer chemistries for synthesizing ASOs yield products that lack these unwanted effects.

Unwanted immunostimulatory effects of siRNA drugs

Just like some oligodeoxynucleotides mimic microbial DNA in prompting an immune response in vertebrates, so too do certain synthetic siRNAs mimic viral RNA. Günther Hartmann (University of Bonn), another pioneer in the field, alerted siRNA drug developers at the meeting to be aware of this potential problem. Whereas DNA mimicry occurs via activation of the immune cell TLR-9 receptors, immunostimulatory siRNA mimicry works through TLR-7 in these cells.

A potent, nine-nucleotide immunostimulatory motif stimulates interferon-α production in plasmacytoid dendritic cells.

Hartmann described how his group has screened an siRNA oligonucleotide library to identify immunostimulatory sequences. The group tested each RNA for the ability to stimulate interferon-α production in plasmacytoid dendritic cells (PDCs) in culture, and found a potent, nine-nucleotide immunostimulatory motif, with variations on that sequence showing lesser but still substantial activity. As Hartmann explained, the PDCs can detect the single-stranded motifs within double-stranded RNAs.

From these results, the researchers have developed an algorithm for predicting the extent of immunostimulatory activity for 19mer RNA sequences, and substantially confirmed their predictions with experimental and published data. The vast majority of 19mers show some immunostimulatory potential. Scanning across all possible 19mers that would target a test mRNA (in this case the mRNA encoding TLR-9), the algorithm predicts very few immunoquiescent oligonucleotides.

Finally, Hartmann noted that the immunostimulatory activity of siRNAs occurs when they are complexed with cationic lipids for delivery. Thus, a different delivery system may eliminate the problem, but Hartmann opts for avoiding it altogether by choosing the appropriate siRNA sequence, if possible.

The Oligonucleotide Therapeutics Society was founded to foster the development of novel oligonucleotide-based medicines. For many OTS members, achieving that aim involves following a course that starts with understanding basic biological mechanisms that oligonucleotide therapeutics will tap into—in particular, the production and function of both coding and non-coding RNA. In the keynote address, molecular biology guru Tom Maniatis (Harvard) discussed the mechanisms involved in the maturation of precursors for coding and non-coding RNAs.

The cellular machinery behind RNA maturation involves a network that researchers can now describe in some detail.

Maniatis explained that the 21st-century, "macromolecular systems" view of pre-mRNA synthesis and processing reveals highly interconnected events. The many steps in the process—transcription, 5' capping, splicing to remove introns, 3' polyadenylation, transport from the nucleus to the cytoplasm, translation, and degradation—were once viewed more or less independently. But now, for example, we know that capping a new RNA transcript at the 5' end does more than protect it from being degraded; it also provides a physical tag for use in nuclear export. Splice site selection—choosing which exons stay and which will be tossed out during mRNA maturation—begins early during transcription, as some proteins jump onto splicing junctions. Others bind within exons destined to remain in the mature mRNA, and help to usher the mature mRNA out of the nucleus. Thus, the cellular machinery involved in RNA maturation is part of a physical and functional network that researchers are now able to describe in some detail.

Turning to the noncoding RNAs, Maniatis focused on the microRNAs, one of the RNA classes involved in the silencing mechanisms on which oligonucleotide therapeutics is based. As he explained, microRNAs are single-stranded RNAs approximately 22 nucleotides in length, which, by interacting with complementary sequences in mRNAs, prevent the mRNAs from being translated into protein.

These non-coding RNAs, Maniatis explained, are variously derived from intergenic DNA regions and from introns within pre-mRNAs. This class of RNAs is incredibly complex, encompassing more than 90,000 different transcripts in eukaryotes.

As Maniatis, Tom Tuschl (The Rockefeller University), and other speakers reminded the audience, microRNA-dependent inhibition of mRNA function seems to be a very common regulatory event; a double-digit percentage of all mRNAs may be targets of microRNA regulation in vertebrates, including humans. Thus, drug developers intent on using strategies like RNA interference (RNAi), precisely because they mimic natural processes, are on an even stronger footing than they were a few years ago when the first siRNA start-up was born.

But, will tapping into naturally occurring mechanisms with the new RNAi-based drugs be a source of unwanted drug side effect? We don't know that answer but the interconnectedness of macromolecular systems suggests that it is an issue worth watching carefully.

In the corporate symposium session on the final day of the meeting, C. Frank Bennett of Isis Pharmaceuticals posed this list of questions.

Antisense Technology: What Don't We Know? Frank's top 10 list

What are the factors that contribute to good and bad antisense oligonucleotides?

What are the factors that mediate distribution of oligonucleotides to tissues?

How do oligonucleotides cross cellular membranes once they distribute to the tissue?

Is hybridization to target mRNA facilitated?

What are the rate limiting steps for antisense oligonucleotides?

What are the oligonucleotide sinks?

Can medicinal chemistry provide further significant improvements in ASOs (beyond LNA and MOEs)?

What are the non-TLR receptors through which oligonucleotides signal?

What are the long-term toxicology issues for oligonucleotide drugs?

What will be the next antisense drug to be approved?

ASO mechanisms of action

Antisense oligodeoxynucleotides can interfere with the function of mRNA or pre-mRNA. As C. Frank Bennett (Isis Pharmaceuticals) explained, they may do so by more than one possible mechanism. Most notably, RNA silencing can occur via an enzyme-mediated destruction of mRNA. In this case, the ASO binds to its complement in the mRNA, prompting an endogenous enzyme, RNAse H, to cleave the RNA within the RNA/DNA duplex region.

Liver

Adipose tissue

Spleen

 

MOE ASOs broadly inhibit gene expression in mice.

Most attempts at ASO drug development have focused on this RNAse H-mediated silencing. That focus has dictated that any chemical modification made to the ASO had to preserve the ability of RNAse to cleave the RNA within the RNA/ASO duplex.

Early in the development of ASO therapeutics, to increase the half-life of the compounds, many investigators switched from using unmodified oligodeoxynucleotides to ones with a sulfur substitution to the phosphodiester backbone, the so-called phosphorotioate compounds. A more recent chemical modification, which several speakers discussed, is 2'-O-(2-methoxyethyl)-modified oligonucleotide (2'-O-MOE). By substituting a few phosphorothioate nucleotides in ASOs with a limited number of MOE-containing nucleotides, investigators have created products with much greater metabolic stability than the phosphorothioate ASOs. Using just a few MOE-substituted nucleotides, and spacing them appropriately to create the so-called MOE-gapmers yields ASOs that retain their RNAse H-inducing ability.

An alternative antisense mechanism involves ASO binding to its complement in the mRNA (or pre-mRNA) and inhibiting its function without causing its destruction, perhaps through steric hinderance, as Bennett suggested. Since this mechanism does not depend on enzyme activity, it tolerates a greater range of ASO chemical structures than the RNAse H-dependent approach does.

Antisense-assisted cell suicide

For oligonucleotide therapeutics to become full-fledged templates for new drug development, rather than a collection of unrelated new medicines, the compounds must meet several criteria. Most importantly for the ASO and siRNA drug classes, each such drug must be shown to act by specifically reducing expression of the intended target mRNA. Fulfilling this requirement in human subjects poses challenges because tissue specimens for evaluation during the course of treatment are often not available. Therefore, the report given by prostate cancer researcher Martin Gleave (Vancouver General Hospital, OncoGeneX) was particularly encouraging.

As Gleave explained, prostate cancer typically starts as an androgen-dependent cancer. Consequently, standard drug treatment uses a hormone that shuts down the body's androgen production. However, with time, many patients develop hormone-resistant tumors that are lethal.

Gleave determined that hormone treatment increased the expression of certain stress-response proteins, which appear to protect tumor cells from the apoptotic actions of the hormone therapy. One such protein is clusterin, a chaperone protein that shields other proteins from untoward precipitation or aggregation.

Gleave and coworkers tested the notion that a clusterin ASO would delay or block the development of hormone-resistant prostate cancer. In collaboration with Brett Monia at Isis Pharmaceuticals, the group developed a so-called "Moe-gapmer" ASO compound designated as OGX-011. As described earlier at the meeting by Isis's Frank Bennett, compounds made with this chemistry tend to be relatively long-lived and nontoxic in the circulatory system, and once inside cells, have a high target affinity, making them more potent than ASOs built using earlier chemistries.

OGX-011 treatment reduced clusterin levels and delayed the onset of hormone-resistant tumor development.

Working with a mouse model of prostate cancer that becomes hormone-resistant over time, Gleave and colleagues found that OGX-011 treatment reduced the level of clusterin and substantially delayed the onset of hormone-resistant tumor development. Turning to studies in human patients with advanced cancers, they first showed that they could successfully deliver a concentration of ASO by intravenous injections that was similar to the dose needed to obtain protective results in mice.

The group then studied newly identified prostate cancer patients who were about to receive hormone therapy prior to prostatectomy. Working with this group of patients meant that the researchers could get biopsy tissue for baseline information, test for the effects of administering the ASO (or a negative control) during the period of hormone treatment, and then inspect tissue obtained from the prostatectomy.

This study was small (N=25) but, according to Gleave, gave very encouraging results. Patients who got the highest OGX-011 doses had approximately 10% the level of clusterin mRNA in the tumor tissue and many fewer clusterin-staining cells than did the control tumor tissue. Also, tumor tissue from OGX-011-patients had many more apoptotic cells than tissue from patients who did not get the clusterin ASO drug. Based on this evidence that OGX-011 knocks down clusterin expression and increases the apoptotic potential of the hormone therapy, the researchers are moving forward in their clinical trials.

Nuclear targets

Well over half of all human genes give rise to multiple products through alternative splicing. Given this high frequency of splice choice variation, researchers now suspect that a significant percentage of diseases may stem from aberrant splicing. Therefore, having a means to manipulate splicing may prove to be a useful therapy.

Ryszard Kole (University of North Carolina, Chapel Hill) reported that his group successfully used ASOs to manipulate pre-mRNA splice site selection inside cell nuclei. To develop this approach, they used transgenic mice carrying an aberrantly spliced gene for the enhanced green fluorescent protein (EGFP). Untreated animals were incapable of producing EGFP. Only those animals treated with ASOs that could bind to the EGFP pre-mRNA and force a change in splicing could produce the fluorescent protein.

The researchers found conditions in which MOE-gapmer ASOs, administered systemically, induced correct EGFP expression in cells of several different tissues whereas negative control oligonucleotides did not. The investigators are now developing a similar strategy for treating thalassemia, a common human blood disease caused by one of several different mutations that lead to aberrant splicing of the globin gene.

A sign of things to come

A positive outcome of a Phase III clinical trial marks a major milestone for a drug company. The biopharmaceutical company Genta Inc. reached that milestone with its ASO compound called Genasense shortly before the OTS meeting, reported Genta's chief medical officer, Loretta Itri.

Genasense attacks the mRNA for Bcl2. It is being tested for use alongside other anticancer agents.

Genasense, an anti-cancer drug, was designed to attack the mRNA for Bcl2, a protein that blocks apoptosis. Like many oligonucleotide drugs for treating cancer, Genasense is being tested for use alongside other anticancer agents to see if the combination proves more effective than the older drug on its own. Genasense showed significant benefit in a study of patients with chronic lymphocytic leukemia (CLL) who were treated with another chemotherapeutic combination, with or without Genasense. The results of this study of 241 randomly assigned patients showed that the Genasense group remained in remission significantly longer and showed a significantly lower risk of relapse than did the group who did not receive the ASO.

Genta plans to file the documents for regulatory approval, both in the United States and in the European Union, before the end of 2005. At the time of the OTS meeting, Genta was the only oligonucleotide therapeutics firm ready to seek approval for an unapproved drug.

Genasense (aka G3139), which was first described in 1997, is a phosphorothioate oligodeoxynucleotide. Now chemists have made further improvements in the chemical structure of ASOs, extending compound half-life and affinity for the mRNA target, for example by incorporating the MOE modification mentioned previously, the so-called locked nucleic acid (LNA) structure, or oxetane-modified nucleotides. Thus, even before a drug such as Genasense gains regulatory approval it may be less effective than a similar compound designed today. That pre-launch obsolescence may be one of the challenges inherent in the platform strategy of drug development.

RNAi-related mechanisms of action

In the last decade, RNA interference (RNAi) has frequently made headlines as a new suite of mechanisms for regulating gene expression, including fine-tuning development and influencing host-microbe interactions. Once scientists saw how widespread these interference mechanisms were among eukaryotes, the notion surfaced that some form of RNAi might be harnessed for therapeutic purposes. New companies such as Alnylum have sprung up and some existing ones morphed (for example, Ribozyme became Sirna) with the aim of applying RNAi therapeutically.

Like the ASO-mediated silencing of mRNA, RNAi involves a short oligonucleotide that interacts with its complement in mRNA, making that mRNA untranslatable. Naturally occurring RNAi processes convert long RNAs into short (approximately 21 nucleotides long) double-stranded RNAs that are part of a dynamic (i.e., changing) ribonucleoprotein complex. These short dsRNAs originate either from long dsRNAs or from single-stranded transcripts with self-complementary regions by enzymatic processing. The former dsRNAs are referred to as short interfering RNAs (siRNAs), and the latter are the microRNAs.

In either case, the complex matures, creating an "RNA-attack" machine. The attack machine interferes with specific mRNA function, reducing the level of protein that is translated by that mRNA.

SiRNA-based interference involves interaction with an mRNA target that has perfect or nearly perfect base sequence complementarity. This interaction prompts an enzyme within the complex to cleave the target mRNA.

RNAi post-transcriptional gene silencing. (Adapted from Tom Tuschl.)

MicroRNA-mediated interference uses a somewhat different mechanism than siRNA does. The original dogma about microRNA-mediated mRNA inhibition was much more complex, and included the following elements: 1) Most microRNAs formed imperfect duplexes with their target mRNAs; and 2) translation of an mRNA target into protein can be inhibited without immediately destroying the target mRNA. One idea, mentioned by Tom Maniatis, Tom Tuschl, John Rossi, and others at the meeting, is that interaction between the mRNA and the microRNA-containing complex does mean the kiss of death for the mRNA, by a previously unrecognized mechanism. New findings suggest that when the microRNA complex binds to the mRNA, the mRNA is dragged into so-called P-bodies, specialized cytoplasmic organelles, where it is degraded by an enzyme in the P-body.

Currently in therapeutic RNAi, investigators design their guide strand RNA to be fully complementary to the intended target mRNA. However, a better understanding of microRNA action in vivo may open new strategies for therapy.

Marking the guide strand

The process of siRNA maturation in the ribonucleoprotein complex involves getting rid of one of the two RNA strands for the RNAi complex. (The strand that remains in the RNAi complex is dubbed the guide strand and the one that leave is the passenger strand.) Then, the guide strand, as part of the complex, is available to find a complementary sequence within a long mRNA, leading to an enzymatic cleavage of the mRNA within the region of duplex with the siRNA. In other words, siRNAs help to kill their mRNA targets.

How do you designate one RNA strand as the guide strand and the other as junk to be discarded?

Anyone who contemplates using an RNAi-like strategy for knocking down expression of specific mRNAs quickly comes upon a head-scratching question: When making a duplex RNA for mRNA silencing, how do you designate one strand as the guide strand for attacking the target and the other as junk to be discarded? The answer comes largely from the work of Phil Zamore and his group at the University of Massachusetts.

Zamore explained that the functional asymmetry of the double-stranded siRNA intermediate comes about if one end forms a less-than-perfect duplex, allowing it to "breathe." This creates a functional asymmetry, instructing proteins in the complex to distinguish between the two RNA strands. (The strand with the 5'PO4 at the "breathable" end becomes the passenger, and is discarded during maturation of the complex.) 

RNAi drug developers might also ask whether the passenger-guide strand issue can be avoided altogether by using a synthetic single-stranded siRNA. Tom Tuschl's presentation provided the answer. Work form his lab says no; for in vivo situations, starting with double-stranded RNA works more efficiently than starting with single-stranded siRNA silencers. By contrast, Tuschl noted, studies done with the purified protein involved in helping the siRNA form a duplex with mRNA allows the single-stranded siRNA to load more efficiently.

Engineering an siRNA delivery system

Speakers described several approaches for delivering siRNAs to their intended targets in intact animals including: the use of viruses encoding the siRNA sequences, siRNAs complexed with a collagen derivative, cholesterol, and even an Apolipoprotein carrier of cholesterol. Mark Davis described a "systems" approach to delivery that was breathtaking for its complexity and ingenuity. Davis, a chemical engineer at Caltech and Calando Pharmaceuticals, has designed and assembled a multi-component siRNA-polymer intended for systemic use to treat conditions such as metastatic cancer, and which tackles several delivery problems at once.

Optimizing an siRNA-containing polymeric system must consider particle size and charge.

As a starting point, Davis wanted to devise an siRNA-containing polymeric system of optimal size. The particles had to be large enough to avoid clearance through the kidney on first passage after injection, but small enough to access micrometasteses fed by tiny new blood vessels. He suggested that the optimal particle size is approximately 50 nanometers across.

In optimizing the system Davis also had to consider its charge. Because he chose to use chemically unmodified siRNA, which carries a strong negative charge, he coupled the siRNA with a cationic compound that would neutralize it. He chose a cyclodextrin-containing polymer—a polycation built from molecular blocks (cyclodextrins) that have FDA approval for use as drug solubilizers—in part because of its excellent safety profile in humans.

Testing the siRNA/cyclodextrin-polymer complexes on live cells, he found that they were successfully internalized, but most of the siRNA failed to reach the subcellular locale of its target mRNA. Instead, it mainly went to lysosomes. By modifying the cyclodextrin polymer termini with imidazole groups, which buffer within an acidic pH range, Davis enhanced the delivery and release of the siRNAs, enabling them to reach their target mRNA. Although the cyclodextrin-polymer can activate the complement cascade on its own, it had no such adverse effect in the fully formulated particulate form.

The group performed additional modifications, including work to stabilize the particles and prevent clumping. Lastly, they introduced a cell-targeting ligand to help the siRNA home in on its intended site of action, for example onto cells that have surface receptors for the chosen ligand. Importantly, the researchers also determined that the particles they created are free of immunostimulatory or toxic effects and didn't cause long-term organ damage in mice.

Davis described recent results on a mouse model of Ewing's sarcoma, a deadly, highly metastatic cancer that typically strikes young people. The experimental siRNA, administered in the cyclodextrin-based delivery system, substantially reduced the tumor burden in experimental animals. In contrast, the siRNA alone or particles that either included a control siRNA sequence or lacked the cancer cell-targeting ligand had no such effect. Thus, Davis concluded from these data that the self-assembling siRNA system should be adaptable for testing in humans.

Act locally, for a long duration

Whereas Davis is developing product for systemic delivery, Judy Lieberman (CBR Institute, Harvard) has a different set of problems to tackle and therefore has taken a different approach. Lieberman wants to devise a siRNA-based anti-HIV vaginal microbicide. Because she's aiming to protect just the cells in vaginal tissue, Lieberman need not concern herself with whole-body biodistribution issues. What she does want is a drug that will be long-lasting, protecting women for days or weeks after they apply the medication.

Lieberman uses siRNAs mixed with a transfection lipid to induce silencing in the vaginal and cervical tissue. Her group found that siRNAs were taken up by cells throughout the vagina and cervix. Moreover, using transgenic mice that ubiquitously expressed green fluorescent protein (GFP), they found that siRNAs induced silencing throughout the tissue that lasted for at least nine days.

Anti-HSV siRNAs gave excellent protection against lethal doses of HSV infection.

Lacking an animal model for vaginal transmission of HIV, Lieberman and her group are using herpes simplex virus (HSV) as a surrogate in their studies on mice. Therefore, they developed siRNAs that silenced essential HSV genes. She reported that the anti-HSV siRNAs gave excellent protection against lethal doses of HSV infection. Mice were well protected against HSV infection if they were pretreated with one or another of the siRNA drugs. Moreover, while single agent treatment used after HSV exposure did not protect the animals, the use of siRNAs against two different HSV mRNA target sites did afford protection even when given after HSV infection.

Finally, Lieberman reported that the siRNA-lipid complexes do not induce an (unwanted) interferon response or visual evidence of inflammation. These and other findings that Lieberman discussed encourage the group to continue on their path to developing an HIV microbicide with siRNAs as the active ingredient.

The devil is in the details

Peter Linsley (Rosetta Inpharmatics, Merck) came to the RNAi field a few years ago, intending to use siRNAs as tools to assess the specificity of new drugs that his company was developing. Inherent in this plan was the widely held notion of the day, that siRNA knockdown took place only on mRNAs carrying a sequence that could form a perfect (or at least nearly perfect) duplex with the 21 nucleotide siRNA guide strand. Instead, Linsley's work has revealed that off-target action (i.e., effects involving the siRNA interacting with mRNAs other than the intended target mRNA) is a common feature of synthetic siRNAs.

To illustrate the problem, he showed data comparing the transcriptional read-out of cells tested with each of several siRNAs against different regions of the mRNA encoding the protein MAPK14. As the figure shows, all the test siRNAs substantially reduced the levels of MAPK14 mRNA and protein.

However, each MAPK14 siRNA knocked down many other transcripts as well, creating a unique and varied transcript set that differed widely from those produced by the others. The heterogeneity of MAPK14 siRNA read-out implied that there was considerable off-target activity coming from the siRNAs. Perhaps the notion that each siRNA exhibited mRNA knock-down activity only when it met its perfect match was grossly oversimplified.

Since Linsley and his colleagues started this work, a great deal has been learned about the siRNA relatives, the microRNAs. In particular, the current consensus is that most naturally occurring microRNA targets are not complete, 21 nucleotide-long stretches within mRNAs. Rather, they are much shorter stretches—perhaps just six to eight bases complementary to the 5' portion of the guide strand of the siRNA.

Extrapolating from these findings, Linsley wondered if many of the off-target effects were due to actions on transcripts that complemented just the 5' region of the test siRNA guide strands. As Linsley explained, the data they gathered supported this notion.

Mutating the base at the tip of the 5' end of the siRNA guide strand reduced off-target effects.

Now the researchers are hunting for chemical modifications to increase the specificity of synthetic siRNAs. As a start, they found that mutating the base at the very tip of the 5' end of the siRNA guide strand substantially reduced off-target effects. Linsley also showed evidence that another proprietary chemical change, made at Dharmacon, further enhances the target signal-to-noise ratio. This modification, Linsley noted, selectively reduces the siRNA interaction with weak targets.

Linsley ended his talk by showing results of in vivo studies with modified siRNAs, done in collaboration with Mirus Bio Corp, in which the target of drug action is PPARα, a nuclear receptor involved in metabolic disorders. In vivo delivery of siRNA against the PPARα transcript showed modest reduction of the target. Importantly, it gave the predicted phenotype, including reduced levels of triglycerides and increased glucose levels in the treated mice. Comparing the gene expression profiles on tissues from siRNA-treated animals with those from animals getting a PPARα antagonist or from PPARα knock-out mice showed substantial similarity. These findings give the researchers reason for optimism as they go on to more carefully scrutinize the in vivo actions of the modified siRNAs.