Chemical Modifications of Nucleic Acids Workshop

May 13, 2005
Natcher Conference Center
Bethesda, Maryland

Background and Introduction

In October 2003, NIH Director Dr. Elias A. Zerhouni expressed interest in tracking progress and obstacles in the quickly evolving arena of RNA interference (RNAi). In response, NIH held a small workshop to assess the state of the field. As follow-up action, NIH staff ascertained current NIH funding in this area, including resources and grant support to individual investigators working on both basic research and clinical applications of RNAi.

In May 2004, NIH convened a second workshop, chaired by Dr. Andrew Z. Fire and Dr. Phillip A. Sharp, that more fully explored the state of the field and what potential actions NIH could take to speed progress. At the 2004 workshop, participants suggested several ways NIH might take action toward streamlining the development and maturation of the RNAi field. The 2004 group recommended that NIH:

  • Establish RNAi screening centers/core facilities with a tight coupling between methods development and biological characterization;
  • Seek innovative and/or cost-cutting measures with the private sector to make high-throughput sequencing efforts more affordable for individual investigators;
  • Issue a solicitation aimed across a broad spectrum of diseases in which individual NIH institutes and centers could fund proposals appropriately targeted toward siRNA delivery to individual organs and/or disease states;
  • Issue a solicitation for proposals in oligonucleotide chemistry;
  • Continue to support RNAi research using model systems, including nontraditional organisms, and consider funding studies in large animals, especially primates, as a translational step; and
  • Take steps to encourage that data remains in the public domain.

Dr. Judith H. Greenberg welcomed the 2005 workshop attendees and invited the group to discuss needs and potential roadblocks in current knowledge of nucleic acid chemistry as it relates to RNAi and related biological processes. Specific questions were distributed in advance, to focus the day's discussion:

  • What lessons have we learned from gene therapy about viral vs. non-viral delivery?
  • What is the role of bioconjugate chemistry for delivery/stability?
  • What is the potential of modified backbones and sugars for RNAi?
  • How can modified nucleic acids address off-target effects?
  • How will modified nucleic acids play a role in the stability of nucleic acid therapeutics?
  • What, if anything, should NIH do to facilitate progress in chemical modifications research?

Scientific Presentations

The workshop began with five presentations that spanned current research in RNAi, chemical modifications of nucleic acids, and drug delivery. Brief summaries of these talks appear below, followed by a digest of the ensuing general discussion.

Function and Mechanism of Mammalian RNA Silencing Processes (Dr. Thomas Tuschl)

In recent years it has become clear that double-stranded RNA (dsRNA) has an important influence on gene expression through gene silencing. A key step is the processing of dsRNAs into short RNA duplexes of defined size and structure. The two predominant natural forms of small, duplex RNA include short interfering RNAs (siRNAs) and microRNAs (miRNAs). These molecules affect gene expression in a variety of ways, including mRNA degradation, chromatin modification, and/or translational repression. These dsRNA-mediated processes are evolutionarily conserved and are thought to be involved in transposon silencing in various species, although evidence of a natural role in the majority of human cells is still scant. Exogenous introduction of dsRNA, which yields the production of siRNA, has been adopted as a means to purposefully effect gene expression.

Whatever the source, maturation of dsRNA occurs through a stepwise process catalyzed by RNase III-like endonucleases such as Dicer and Drosha, the latter of which is required for the processing of miRNA precursors but not long dsRNA. Sequence dependence defines structure and efficiency of cleavage by Dicer. Small RNAs and proteins of the Argonaut family (all of which contain the PAZ and PIWI conserved domains) assemble in the cytoplasm into an RNA-induced silencing complex (RISC). Some but not all of the mechanistic requirements of this process have been identified, and various points in the process may be amenable to alterations through the chemical modification of nucleic acids.

Current issues that still need to be addressed include sequence selection, specificity, and delivery. Many aspects of delivery require further exploration and optimization, including determination of different requirements for local vs. systemic administration, clarification of varied tissue uptake properties, identification of novel expression methods, and consideration of potential, yet to be identified membrane-embedded nucleic acid transporters. Chemical modifications may impact potency, nucleic acid resistance, and host immunogenicity. In addition, it may be possible to mitigate off-target effects through "dilution" with excess on-target sequence. Structural and functional characterization of the RNA silencing machinery should provide a rational framework for guiding the positioning of chemical modification and/or conjugates. Robust biochemical, cell-based, and animal models are necessary for adequate testing and further development of promising molecules and approaches.

Progress and Obstacles in Delivery of Nucleic Acids as Therapeutics (Dr. Francis Szoka)

Previous and current studies have illuminated some of the characteristics of nucleic acid backbone modifications, various carrier-mediated systems, intracellular nucleic acid trafficking, biological effectors of off-target effects, and sequence determinants of potency. Multiple clinical trials are under way to test second-generation nucleic acid-based therapies, particularly antisense, and thus there is a relatively good understanding of the human pharmacokinetic properties of antisense molecules. The principal determinant of half-life and stability of small RNA molecules appears to be protein binding. For example, while free siRNA has an in vivo half-life of only 6 minutes, modification with a cholesterol moiety extends this time to 90 minutes, due to consequent increases in protein binding and presumed increased stability. The toxicities of antisense oligodeoxyribonucleotides (ODNs) have been relatively well established in humans. These include untoward effects on clotting, complement activation, and immune stimulation via CpG motifs (mediated through Toll-like receptors).

Current obstacles to the effective use of siRNA oligotherapies center on delivery issues (including intracellular entry), metabolism, and immunostimulation. Recent reports that have begun to define immunomodulatory sequence motifs may inform sequence selection in deriving oligotherapies. Various nucleic acid carrier systems are currently under investigation. These include nanosized nucleic acid carriers, polyethyleneglycol (PEG)-oligoconjugates, PEG liposomes, and polymeric nanospheres. It is unlikely that any of these relatively large entities (each bigger than a typical protein), would be subject to renal excretion. However, the large size of these carriers enables packaging of multiple copies of nucleic acid. Current research is investigating the utility of pH-sensitive hydrolysis and disulfide exchange, the latter of which may reduce immunogenicity thought to be caused by excess cationic charge.

Demonstrated in vivo activity of siRNAs, along with continued improvements in selectivity and potency, are encouraging. On the downside, significant current obstacles preclude parenteral administration of nucleic acids. However, carrier-mediated approaches may help to ameliorate this issue. Increased cell entry may be achieved through novel photochemical strategies, and increased understanding of sequence-specific immunological effects may guide the development of therapies with minimized toxicity.

Backbone Modifications of Nucleic Acids (Dr. David Corey)

Because oligonucleotides can be rapidly synthesized to be complementary to any genetic sequence, they offer promise as research tools and potential gene-based therapies. Many types of oligomers have been developed, each with chemical and biological properties that affect in vitro and in vivo use. Ideally, therapeutic oligomers should have the ability to enter cells, be relatively nuclease-resistant, have sufficient thermostability, and be target-specific. In addition, nucleic acid-based therapeutics should have low toxicity, be orally bioavailable with desirable pharmacokinetic properties, and be inexpensive to manufacture. However, due to a wide range of potential uses and biological targets, it is likely that no universal design will circumscribe this overall class of molecules.

Backbone modifications of nucleic acids offer flexibility in the design and utility of these molecules. The advantages and disadvantages of several types of oligomers were presented. Peptide nucleic acids (PNAs), with a peptide backbone, are easy to modify and efficiently invade duplex DNA. However, PNAs are expensive to make, and few definitive animal studies have validated their use in vivo. Phosphorothioate DNAs (PTs), which by contrast are simple and inexpensive to make, have been extensively tested in the clinic. However, PTs are notorious for cell culture artifacts, and several have failed as single treatment modalities in phase III trials. 2-O-alkyl-RNAs (2-O-RNAs) benefit from a solid track record in animal testing, and they have low toxicity as demonstrated in phase I and II studies. Although 2-O-RNAs are cheap to make, most researchers do not recognize them as a revolutionary advance. Locked nucleic acids (LNAs), which contain a methylene bridge joining the 2' oxygen of ribose with the 4' carbon, reduce the conformational flexibility of ribose and thus can increase melting temperature values up to 10 degrees C per substitution. As such, LNAs can be "plugged into" a variety of experimental systems; however, these molecules are costly to make and have not been validated decisively in cell culture or animal systems. Phosphoroamidate-linked DNAs (PAs), which contain backbone nitrogen, have shown potent inhibition of certain molecules such as telomerase. However, it is not yet clear whether PAs offer significant advantages over other existing chemical modifications, due to delays that have occurred in the development and clinical testing of these molecules. Double-stranded RNAs (dsRNAs) have been very successful in cell culture studies and show solid results in animal studies, potentially due to the fact that dsRNA-mediated gene silencing is a natural biological process. However, potential disadvantages of dsRNAs include the fact that there are two strands to consider. In addition, a clear advantage of dsRNAs over traditional antisense strategies has not yet been demonstrated.

While many backbone modifications may be applied to the development of oligotherapies, there is a striking paucity of animal data to support or refute their individual or combined use in vivo. Structure-activity relationships may also further inform progress, as would increased interactions between the chemistry and biology communities.

Intracellular Delivery of Antisense Oligonucleotide Conjugates and Complexes (Dr. Rudolph Juliano)

Antisense oligonucleotides offer promise as powerful tools to selectively control cellular and viral gene expression; however, target specificity is a crucial determinant. The multidrug resistance-1 (MDR-1) gene represents a challenging model target for gene regulation, due an abundance of message and protein, gene amplification, and high stability of the encoded protein. Different approaches have met with variable success in targeting MDR-1 with oligomers. Among these are antisense, siRNA, and "designer" transcription factors. DNA array analysis has been used to assess the specificity of antisense oligotherapies, measuring cellular response to peptide-oligonucleotide conjugates. These studies revealed a complicated picture of pharmacological effects, including not only off-target effects, but also upregulation of certain genes. Peptide-oligonucleotide conjugates have also been evaluated in vivo, to monitor biological effects such as splicing. Although these conjugates achieve effective delivery and are relatively selective in their action, results show significant variability in tissue specificity and effect.

Delivery of antisense and siRNA oligonucleotides is also a challenging problem, especially since existing data obtained in vitro, with either free oligonucleotides or carrier-bound oligonucleotides, does not always correlate with in vivo findings. One emerging technology that may show promise is a class of molecules called cell-penetrating peptides. Although some controversy exists regarding their mechanism of action and effectiveness, cell-penetrating peptides offer the potential advantage of being bioreversible, through sulfide conjugation to antisense oligonucleotides or siRNAs. Additional delivery approaches still under investigation include cationic dendrimer-oligomer complexes. While these molecules can carry multiple nucleotide copies, their main disadvantage is toxicity.

Nucleic Acid Modifications that Could be Triggered Inside the Cell (Dr. Steven Rokita)

In vivo nucleic acid modification may provide an untapped potential for oligotherapy, to facilitate transport, increase stability, reduce off-target effects, and promote long-term gene expression. It may be necessary to relax design criteria for inducing RNAi, perhaps through combining multiple entities and/or using pro-drug strategies. Potential dissociative approaches include using photochemical deprotection to induce secondary structure changes that influence RISC assembly. Another potential dissociative method is hydrolysis, in which an endogenous enzyme removes protecting groups intracellularly. Alternatively, template-directed techniques could be used to enhance the self-assembly of small molecules on the cell surface or inside the cell. Innovative chemical strategies, such as "click chemistry" or chemical remodeling of cell surfaces, may be brought to bear on this problem. Finally, conformation-dependent chemistry affords the opportunity to selectively, and potentially reversibly, cleave RNA or DNA in vivo.

Adding covalent reactivity to oligotherapy technologies may expand the range of target sequences and/or help stabilize target-probe associations. It may also be possible to develop ways to initiate covalent processes in vivo, since cells contain macromolecular surfaces for templating reactions as well as enzymes for deprotecting latent reactivity.

Discussion and Thematic Summary

NIGMS Director Dr. Jeremy M. Berg opened the discussion by confirming NIGMS' support for research on nucleic acid technologies as potential therapies. Dr. Greenberg stated that the intended outcome of the current workshop and related NIH-sponsored RNAi-funding activities is to facilitate the use of small RNAs in therapeutic applications. The group's discussion visited several themes, which are presented below in the context of comments made that are relevant to these topical areas.

(1) There is a need for greater collaboration among scientists with different expertise.

Participants agreed that a significant gulf continues to separate chemists and biologists, and an even greater gulf separates biologists and chemists from pharmacologists and physiologists. Cultural issues that cloud the success of collaborations include differing scientific interests and reward systems among academic fields. The dilemma intensifies as points of juncture are beginning to unfold within the RNAi area: What are the roadblocks? How should siRNA therapy be applied? Is systemic delivery a necessary goal? There is a pressing need to facilitate interactions so that these and other relevant questions can be answered with the best tools and approaches.

Several participants noted that the problem of interdisciplinary interactions is more general in nature and that conferences that bring together scientists with complementary expertise can be a positive force for forging connections. Venues that may be considered include gene therapy meetings, in which chemists can "learn the language" and interact with biologists working with nucleic acids in vivo. Of particular interest may be the first annual meeting of the Oligonucleotide Therapeutics Society, to be held September 15-18, 2005 at Rockefeller University, New York (for more information, see Link to external Web site).

Other existing resources may help to foster necessary interactions between the chemistry and biology communities working on RNAi, such as certain facets of the Molecular Libraries and Molecular Imaging initiative of the NIH Roadmap (for more information, see and the availability of NIH-funded conference grants. NIGMS "glue grants" may also be a viable mechanism in which multiple investigators can work synergistically toward a goal with a common pool of resources (more information can be found here). Another possibility is the formation of virtual teams, since scientific interest does not always coalesce within the boundaries of an individual institution.

(2) Scale-up of promising molecules in preparation for in vivo testing is not practical in an academic setting.

Iterative screening studies that require scale-up and animal testing are inherently expensive, labor- and time-intensive, and not typically fundable by individual research grants. The group suggested that since the necessary work required to test candidate siRNAs and other nucleic acids in vivo requires substantial quantities of starting material (e.g., more than 150 mg per animal for mouse studies), NIH could consider establishing core facilities for siRNA scale-up, analogous to existing screening centers. The suggestion arose that such arrangements could be implemented through contract mechanisms awarded to individual labs or to commercial entities. The point was made, however, that large quantities of nucleic acid are not needed for all investigations, and that in some cases candidate molecules can be "weeded out" using only a few mice. In addition, relatively small quantities are needed for sequence determination and optimization, chemical modification, and physicochemical characterization.

(3) The development of oligonucleotide therapies requires rigorous, standardized testing of molecules in animal models.

There was widespread agreement within the group that well-controlled animal studies are beyond the scope of most individual laboratories but nonetheless desperately needed to test the feasibility, specificity, and toxicity of siRNA and similar oligotherapies. Reliable animal-based research is needed to validate concepts and prepare for translational steps, and the field would benefit from the definition of a few "gold standard" assays for comparison purposes. It may also be desirable to focus on several target organs (e.g., liver, brain, skin, and others). Along these lines, transgenic models with gene targets in many organs may provide an effective strategy. To date, high cost has been a major limitation of performing whole-animal pharmacological/toxicological studies. Ideally, such approaches should be rapid and highly accurate, and have the following characteristics: positive/unambiguous physiological readout, broad tissue expression, potential for serial sampling, and limited confounding immunological effects.

Several participants expressed interest in having NIH play a leadership role by: (i) choosing model systems as centerpieces for further study, and/or (ii) establishing scale-up facilities amenable to animal testing. It was emphasized that in order to succeed, these activities must be highly standardized and performed by full-time, expert staff dedicated to the effort.

(4) Fundamental gaps in knowledge persist regarding immune responses to nucleic acids and the cell biology of nucleic acids.

Several participants questioned whether enough fundamental work had yet been done to explore the biological and chemical properties of nucleic acids. Consulting existing nucleotide analog databases may inform this process, although it was noted that a portion of this collection is proprietary and thus is not publicly accessible. The group pointed to the importance of access to both positive and negative assay results.

Further research is needed to answer specific biological questions about oligonucleotides, including their immunomodulatory effects. Investigating natural mechanisms of nucleic acid trafficking within and into cells may suggest novel strategies for the development of new delivery vehicles and/or chemical modification approaches. For example, PEG-oligonucleotide complexes, which can bypass hepatic metabolism, may increase bioavailability, if toxicity does not become prohibitive. Better tools are also needed to understand mechanisms of oligonucleotide endosomal capture and release. Finally, small-molecule inhibitors may help clarify the role and behavior of small RNAs in cellular processes and networks.

Participants noted that much of the necessary work required to clarify these basic questions can likely be accomplished via traditional research grants to individual investigators.

* * *

To conclude the workshop, Dr. Greenberg thanked participants for their thoughtful contributions and invited the group to send any additional comments to NIGMS after the meeting. She also noted that a trans-NIH Request for Applications on disease-focused RNAi delivery would appear soon in the NIH Guide for Grants and Contracts.


Mechanisms of gene silencing by double-stranded RNA Link to external Web site
Meister G, Tuschl T.

Antigene, ribozyme and aptamer nucleic acid drugs: progress and prospects
Stull RA, Szoka FC Jr.

Role of Toll-like receptors in antisense and siRNA Link to external Web site
Agrawal S, Kandimalla ER.

Challenges for RNAi in vivo
Paroo Z, Corey DR.

Novel antisense and peptide nucleic acid strategies for controlling gene expression. Link to external Web site
Braasch DA, Corey DR.

Peptide-oligonucleotide conjugates for the delivery of antisense and siRNA
Juliano RL.

Conjugates of antisense oligonucleotides with the Tat and antennapedia cell-penetrating peptides: effects on cellular uptake, binding to target sequences, and biologic actions
Astriab-Fisher A, Sergueev D, Fisher M, Shaw BR, Juliano RL.

Evaluating the specificity of antisense oligonucleotide conjugates. A DNA array analysis. Link to external Web site
Fisher AA, Ye D, Sergueev DS, Fisher MH, Shaw BR, Juliano RL.

Triggering of RNA secondary structures by a functionalized nucleobase
Hobartner C, Mittendorfer H, Breuker K, Micura R.

Light-activate​d RNA interference
Shah S, Rangarajan S, Friedman SH.

Nucleic acid-triggered catalytic drug release Link to external Web site
Zhaochun Ma and John-Stephen Taylor

A general strategy for target-promoted alkylation in biological systems Link to external Web site
Zhou Q, Rokita SE.

The growing impact of click chemistry on drug discovery Link to external Web site
Hartmuth C. Kolb and K. Barry Sharpless

Metabolic oligosaccharide engineering as a tool for glycobiology
Dube DH, Bertozzi CR.


Invited Participants

Cynthia J. Burrows, Ph.D., Chair
University of Utah
Department of Chemistry
Peter A. Beal, Ph.D.
University of Utah
Department of Chemistry

Donald E. Bergstrom, Ph.D.
Purdue University
Department of Medicinal Chemistry

Christine Chow, Ph.D.
Department of Chemistry
Wayne State University

David Corey, Ph.D.
University of Texas
Southwestern Medical Center
Department of Pharmacology

Masad Damha, Ph.D.
McGill University
Department of Chemistry

Rudolph Juliano, Ph.D.
University of North Carolina at Chapel Hill
Department of Pharmacology

Muthiah Manoharan, Ph.D.
Alnylam Pharmaceuticals, Inc.

Larry McLaughlin, Ph.D.
Boston College
Department of Chemistry

C. Russ Middaugh, Ph.D.
University of Kansas
Pharmaceutical Chemistry Department

Tariq Rana, Ph.D.
University of Massachusetts
Department of Biochemistry and Molecular Pharmacology

Kevin Rice, Ph.D.
University of Iowa
Department of Medicinal Chemistry

Steven E. Rokita, Ph.D.
University of Maryland
Department of Chemistry and Biochemistry

John Rossi, Ph.D.
Division of Chemical Biology
Beckman Research Institute of
The City of Hope

Francis Szoka, Ph.D.
University of California, San Francisco
Department of Pharmacology

Thomas Tuschl, Ph.D.
Rockefeller University


Judith H. Greenberg, Ph.D.
Director, Division of Genetics and
Developmental Biology

Philip LoGrasso, Ph.D.
Program Director, Division of
Pharmacology, Physiology, and
Biological Chemistry

Marcus Rhoades, Ph.D.
Chief, Genetic Mechanisms Branch
Division of Genetics and
Developmental Biology

John Schwab, Ph.D.
Program Director, Division of
Pharmacology, Physiology, and
Biological Chemistry