Jens Kurreck is professor for nucleic acids technologies at the University of Stuttgart (Germany). He received his PhD at the Technical University Berlin in 1998 and was a postdoc at the Arizona State University in Tempe, USA. From 1999 to 2007 he was assistant professor (‘Habilitand’) at the Free University Berlin. Furthermore, he holds a master’s degree in philosophy. His group has been working in the field of gene silencing for many years with a focus on pain research and heart muscle infections. In 2005, Jens Kurreck received the award for the best teaching at the biochemistry department as well as the Young Scientist Lectureship Award of the European Society for Neurochemistry (ESN).

Therapeutic Oligonucleotides

By Jens Kurreck

The Royal Society of Chemistry

Copyright © 2008 Royal Society of Chemistry
All rights reserved.
ISBN: 978-0-85404-116-9

Contents

Chapter 1 The Role of Backbone Modifications in Oligonucleotide-Based Strategies Jens Kurreck,
Chapter 2 Genasense (G3139): An Antisense Bcl-2 Oligodeoxyribonucleotide with Substantial Clinical Activity and a Complex Mechanism of Action Cy A. Stein, Noah Kornblum, Johnathan Lai and Luba Benimetskaya,
Chapter 3 Antisense Morpholino Oligomers and Their Peptide Conjugates Hong M. Moulton and Jon D. Moulton,
Chapter 4 Peptide–Peptide Nucleic Acid Conjugates for Modulation of Gene Expression Martin M. Fabani, Gabriela D. Ivanova and Michael J. Gait,
Chapter 5 Locked Nucleic Acid: Properties and Therapeutic Aspects Troels Koch, Christoph Rosenbohm, Henrik F. Hansen, Bo Hansen, Ellen Marie Straarup and Sakari Kauppinen,
Chapter 6 Immune Stimulatory Oligonucleotides Eugen Uhlmann,
Chapter 7 Decoy Oligodeoxynucleotides to Treat Inflammatory Diseases Markus Hecker, Swen Wagner, Stefan W. Henning and Andreas H. Wagner,
Chapter 8 AS1411: Development of an Anticancer Aptamer Nigel Courtenay-Luck and Donald M. Miller,
Chapter 9 Spiegelmer NOX-E36 for Renal Diseases Dirk Eulberg, Werner Purschke, Hans-Joachim Anders, Norma Selve and Sven Klussmann,
Chapter 10 Strategies for the Delivery of Oligonucleotides in vivo Christian Reinsch, Evgenios Siepi, Andreas Dieckmann and Steffen Panzner,
Chapter 11 Lipid-Mediated in vivo Delivery of Small Interfering RNAs Ian MacLachlan,
Chapter 12 Vector-Mediated and Viral Delivery of Short Hairpin RNAs Henry Fechner and Jens Kurreck,
Chapter 13 Development of an RNAi-Based Gene Therapy against HIV-1 Olivier ter Brake and Ben Berkhout,
Chapter 14 RNA Based Therapies for Treatment of HIV Infection Lisa Scherer, Marc S. Weinberg and John J. Rossi,
Subject Index, 329,

CHAPTER 1

The Role of Backbone Modifications in Oligonucleotide-Based Strategies

JENS KURRECK

Institute for Chemistry and Biochemistry, Free University Berlin, Thielallee 63, 14195 Berlin, Germany, and Institute of Industrial Genetics, University of Stuttgart, Allmandring 31, 70569 Stuttgart, Germany

1.1 Introduction

Inside cells, long deoxyribonucleic acid (DNA) or ribonucleic acid (RNA) molecules are enzymatically generated from monomeric nucleotides by DNA and RNA polymerases, respectively. The development of a method for solid-phase synthesis in the 1970s allowed the artificial generation of oligonucleotides (ONs) of up to ~100 nucleotides in length. Moreover, this technology enabled the incorporation of modified building blocks into the growing chain composed of nucleotides. The chemical synthesis and (partial) modification of ONs opened the road for new research applications and novel therapeutic strategies.

Various classes of ONs have been developed in the meantime: antisense oligonucleotides (AS ONs), ribozymes and small interfering RNAs (siRNAs) specifically inhibit gene expression by Watson–Crick base pairing to a complementary messenger RNA (mRNA), but in contrast, decoy ONs and aptamers bind to their target by structural recognition. ON-based applications have been used widely for research purposes and some approaches have proceeded to the status of clinical investigations, which will be the focus of this review. Approximately 30 clinical trials of various phases with AS ONs are currently underway and an AS drug (Vitravene) to treat cytomegalovirus-induced retinitis was the first ON ever to be approved by the US Food and Drug Administration (FDA). Several chemically pre-synthesized ribozymes that target mRNAs of oncogenes or viral RNAs have been tested in early clinical phases. In 2004, only three-and-a-half years after the first demonstration that siRNAs can be used to specifically silence a target gene in mammalian cells, the first clinical trials based on RNA interference (RNAi) have been initiated. In the same year, the approval of Macugen, an aptamer to treat age-related macular degeneration, was another breakthrough in the field of ON therapeutics.

With only a few exceptions, the above-mentioned ONs are composed of modified building blocks. Biological fluids, like blood serum or intracellular liquids, contain highly active nucleases to destroy deleterious nucleic acids. As can be seen in Figure 1.1 (upper row), an unmodified DNA ON is completely degraded within only a few hours in a solution that contains 10% fetal calf serum. ONs composed of RNA are even more susceptible to nucleolytic degradation. The usability of unmodified ONs in animals or human patients is thus limited, since the ONs will be mainly degraded before they even reach their destination.

To overcome this problem, modified nucleotides have been developed that possess higher resistance against enzymatic degradation, since they are not recognized as substrates by nucleases. Early attempts in this direction focussed on the phosphodiester linkage that connects two nucleotides. Since then, the 2′-position of the ribose has been used widely as a site for the introduction of functional groups that enhance the stability of ONs. Figure 1.1 (rows 2 and 3) shows that the incorporation of modified building blocks into an ON increases its resistance against nucleolytic degradation. Even after two days of incubation at 37 °C in a medium that contains fetal calf serum the intact full-length ON can still be detected.

The introduction of modified building blocks, however, not only increases nuclease resistance of an ON, but also changes further pharmacokinetic parameters that are highly relevant for in vivo applications. Important features are the circulation time in the blood stream and the biodistribution (including cellular uptake). Furthermore, the use of non-natural nucleotides can lead to toxic side effects either of the complete ON or of breakdown products. It is also important to maintain the biological function that depends on the mode of action of the ON. For example, alterations of the binding site of an aptamer can result in decreased target affinity, ribozymes tend to lose their catalytic activity when modified nucleotides are introduced into the catalytic centre and siRNAs do not tolerate the addition of functional groups in certain positions while they tolerate this addition in other positions. It is therefore a great challenge to optimize ONs with respect to nuclease stability, functional activity, pharmacological properties and toxic side effects. In the following sections basic principles to fulfil this task are described for the different types of ONs.

1.2 Antisense Oligonucleotides

ONs, being 15 to 20 nucleotides in length, can be employed to inhibit gene expression specifically. This was originally discovered in the late 1970s, when Zamecnik and Stephenson demonstrated that an antisense agent can be used to inhibit virus replication in cell culture. Two major mechanisms of action have been identified to mediate post-transcriptional gene silencing by AS ONs: first, most AS ONs are designed to activate ribonuclease H (RNase H), which is primarily located in the nucleus. RNase H recognizes hybrids composed of DNA and RNA and cleaves the RNA moiety of this heteroduplex. Second, AS ONs that do not induce target RNA cleavage by RNase H can be designed to inhibit translation by a steric blockade of the ribosome. For this approach to be highly efficient it is advisable to direct the AS ONs against the 5′-end or the AUG initiating codon region of the target RNA to prevent binding and assembly of the ribosome. Furthermore, AS ONs can be used to correct aberrant splicing.

Although the initial antisense experiments were carried out with unmodified DNA it soon became clear that ONs have to be protected against nucleolytic degradation for prolonged silencing. In the meantime, several hundred analogues of naturally occurring nucleic acids have been described in the literature. Space restraints in this review mean that only those building blocks that have reached the status of testing in clinical trials will be discussed here: phosphorothioates (PS), 2′-O-methoxyethyl RNA (MOE), locked nucleic acids (LNAs), phosphorodiamidate morpholino oligomers (PMOs) and N3′ -> P5′ phosphoramidates (NPs; Figure 1.2). Peptide nucleic acids (PNAs) are another class of widely used ONs, in which the ribose phosphate backbone is replaced by polyamide linkages. Chapter 4 describes the development of peptide conjugates of PNAs for enhanced cellular uptake and intracellular activity by a steric block mechanism.

1.2.1 Phosphorothioates

One of the first and still widely used modifications to stabilize AS ONs is the introduction of phosphorothioates (for a review, see Eckstein). In this class of ONs, one of the non-bridging oxygen atoms is replaced by sulfur. Phosphorothioates were first synthesized by solid-phase chemistry in the 1960s. They are easy to synthesize, highly water soluble and resistant against nucleolytic degradation (Figure 1.1). Just like unmodified DNA ONs, phosphorothioates bind to complementary RNAs by Watson–Crick base pairing and activate target RNA cleavage by RNase H.

As a result of their favourable properties, phosphorothioates have the longest history in clinical testing and most of the advanced studies are bbased on this class of AS ONs. The only AS ON approved by the FDA to date is the 21-mer phosphorothioate Vitravene (fomivirsen) which targets the immediate early mRNA of the human cytomegalovirus (CMV). The ON is injected intravitreally and is used to treat CMV-induced retinitis in immunodeficient patients with acquired immunodeficiency syndrome (AIDS). The complicated mode of administration and the existence of efficient alternative drugs, however, hinder its broad application. Two additional phosphorothioates that have been tested in advanced stages of clinical investigations to treat cancer are Genasense (see Chapter 2) and Affinitak, which target Bcl-2 and PKC-α, respectively. The results of clinical trials with these ONs, however, did not meet the expectations. The primary mode of action of Genasense still remains somewhat ambiguous and may not even be antisense inhibition of the targeted gene. But, despite this uncertainty, it might be a general problem for AS therapeutics for cancer that the single-target approach might be too narrow. Incomplete knockdown of the target gene might be insufficient to stop tumour growth and the loss of function may be compensated for by other pathways in the cancer cell.

Several disadvantageous properties of phosphorothioates further limit their broad applicability. First of all, they display a reduced affinity towards complementary RNA molecules in comparison to their isosequential unmodified DNA counterpart. Even more important is the tendency of phosphorothioates to bind to certain proteins. This feature has some positive effects for the pharmacokinetic profile because binding to plasma proteins protects them from rapid filtration from the blood stream, but it may also cause cellular toxicity.

1.2.2 2′-O-Methyl and 2′-O-Methoxyethyl Ribonucleotides

To overcome these limitations, building blocks with modifications at the 2′-position of the ribose have been introduced into AS ONs. RNA derivatives with a methyl or a methoxyethyl group at the 2′-position of the ribose (Figure 1.2) have been used to obtain AS ONs with improved properties. These modifications confer high nuclease resistance with reduced toxicity as compared to phosphorothioate ONs. A major disadvantage of this second generation of modified nucleic acids, however, is their inability to induce RNase H cleavage of the targeted RNA. This problem can be circumvented by the use of so-called gapmers (Figure 1.3): blocks of nucleotides with a modified ribose at both termini protect the ON against dominant exonucleases and increase RNA-binding strength, while the stretch of deoxyribonucleotides in the centre of the ON is sufficient to activate RNase H. Interestingly, replacement of the phosphodiester bond by phosphorothioates throughout the gapmer allows further gain in nuclease stability without increasing the toxicity of the ON – a phenomenon that is not yet fully understood. Gapmer ONs that consist of MOE and phosphorothioate DNA monomers are currently in clinical development against a broad range of diseases, including diabetes, high cholesterol level, multiple sclerosis, psoriasis and cancer (Table 1.1).

However, even ONs that do not recruit RNase H have been shown to be potent antisense agents. In one of the first attempts in this direction, a fully modified 2′-O-methoxyethyl RNA ON that targeted the 5-end of the mRNA of intercellular adhesion molecule 1 (ICAM-1) efficiently inhibited translation, most likely through interference with the assembly of the ribosome. Furthermore, the seemingly undesirable property of 2′-O-methyl RNA not to activate RNase H is an indispensable prerequisite for the attempt to correct an mRNA rather than to destroy it. Roughly 60% of human genes are alternatively spliced, and close to 50% of genetic disorders are considered to result from mutations that cause defects in pre-mRNA splicing. Chemically modified AS ONs have successfully been employed to correct splicing by blocking aberrant splice sites.

1.2.3 Locked Nucleic Acids, Phosphorodiamidate Morpholino Oligomers and N3′ -> P5′ Phosphoramidates

In recent years, antisense strategies have received increasing attention because of the advances made by the development of new types of modifications. A large number of DNA or RNA analogues have been tested for their potential to improve antisense agents. Here, only LNAs, PMOs and NPs are discussed in more detail, since these building blocks have already made their way into clinical trials. Readers interested in further modifications are referred to previous review articles, which have exhaustively dealt with modern nucleic acids chemistry used in ONs for biological and therapeutic applications.’

LNAs were initially synthesized in the laboratories of Imanishi and Wengel in 1998. They are conformationally restricted in a 3′-endo/N-type sugar conformation by a methylene bridge that connects the 2′-oxygen atom of the ribose with the 4′-carbon atom (Figure 1.2). LNAs combine a number of desirable properties, including nuclease resistance (Figure 1.1) and an unprecedented hybridization affinity towards complementary ONs (for reviews, see Jepsen et al. and Karkare and Bhatnagar). Just like most of the nucleotides with modifications at the 2′-position of the ribose, LNAs do not activate RNase H. Gapmers that consist of five LNA monomers at both ends and a central stretch of eight DNA nucleotides in the centre, however, were shown to be potent inducers of RNase H cleavage. Furthermore, LNA gapmers were found to be significantly more efficient inhibitors of gene expression than phosphorothioates or 2′-O-methyl RNA gapmers. LNAs do not only confer high target affinity, but also enhance cellular as well as nuclear uptake of ONs after transfection with cationic lipids. This property further accounts for their good antisense potency.

The high efficiency of LNA ONs as antisense agents was also confirmed in vivo. Chimeric AS LNA/DNA ONs that target the delta opioid receptor mRNA were found to efficiently reduce the antinociceptive effect of the agonist deltorphin II. In this study, the LNA ONs did not elicit any histologically detectable toxicity when injected into rat brains. Furthermore, gapmers directed against H-ras with standard β-D-LNA or its diastereomer, α-L-LNA, at the termini inhibited tumour growth at very low dosages and did not show toxic side effects. These promising findings prompted the development of an LNA ON against Bcl-2 for the treatment of B-cell lymphoma in patients with chronic lymphocytic leukaemia (CLL), as is outlined in Chapter 5. However, signs of hepatotoxicity after intraperitoneal (i.p.) injection of LNA gapmers were reported recently.

In PMOs, the five-membered ribose ring is replaced by a six-membered morpholino moiety and a dimethylaminophosphoroamidate (phosphorodiamidate) intersubunit linkage is used instead of the phosphodiester bond (Figure 1.2). PMOs are resistant to nucleases, but, like most of the third-generation modifications, they do not activate RNase H. They are thus usually targeted against the 5′-untranslated region (UTR) or the first bases downstream of the AUG start codon to inhibit translation by preventing ribosomes from binding. PMOs are a widely used knockdown tool in developmental biology because of their efficient cytosolic delivery into embryos by microinjection (for a review, see Karkare and Bhatnagar). Furthermore, Avi BioPharma is developing PMO AS ONs in clinical trials (see Table 1.1 and Chapter 3). In addition to antisense applications, PMO ONs have been employed to correct aberrant splicing of β-globin precursor mRNA in blood cells from patients with β-thalassemia. Ex vivo treatment of erythroid progenitor cells with a PMO ON restored correct splicing and synthesis of haemoglobin A.

NPs are DNA analogs, in which the 3-hydroxyl group of the 2′-deoxyribose ring is replaced by a 3′-amino group (Figure 1.2). NPs exhibit high affinity towards a complementary RNA strand and good nuclease resistance. Since NPs do not activate RNase H, they have been employed for strategies that do not depend on this classical mode of antisense inhibition. Human telomerase is a reverse transcriptase that maintains telomers in rapidly dividing cells. The enzyme is inactive in most somatic cells, but more than 90% of all cancer cells display robust activation of telomerase, thus making it a suitable target for anticancer drugs. Telomerase consists of an RNA component and a protein-aceous catalytic subunit. A N3′ -> P5′ thiophosphoroamidate with a palmitoyl moiety conjugated to the 5′-end was directed against the template region of telomerase RNA and was found to inhibit telomerase activity in a human lung cancer cell line and to prevent lung metastases in vivo in xenograft animal models. A more recent study suggests, however, that the antimetastatic potential of this AS ON, named GRN163L, might rather be by antiadhesive effects conferred via specific structural determinants than by its inhibition of telomerase. According to the website of Geron Corporation, clinical Phase I studies with GRN163L for patients with CLL as well as for patients with solid tumours have been initiated.

(Continues…)Excerpted from Therapeutic Oligonucleotides by Jens Kurreck. Copyright © 2008 Royal Society of Chemistry. Excerpted by permission of The Royal Society of Chemistry.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.