`
`Second Edition
`
`Edited by
`
`M. Ian Phillips, PhD, DSc
`Vice President for Research
`
`University ofSouth Florida, Tampa, FL
`
`Foreword by
`
`Stanley T. Crooke, MD, PhD
`Isis Pharmaceuticals Inc., Carlsbad, CA
`
`Humana Press - ~ Totowa, New Jersey
`1
`MTX1 038
`This material was {spied
`at the Nle arm may be
`Suniem US CD wright Laws
`
`1
`
`MTX1038
`
`

`

`© 2005 Humana Press Inc.
`999 Riverview Drive, Suite 208
`Totowa, New Jersey 07512
`
`www.11umanapress.com
`
`All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted in any
`form or by any means, electronic, mechanical, photocopying, mierofilming, recording, or otherwise without
`written permission from the Publisher. Methods in Molecular BiologyTM is a trademark of The Humana Press
`Inc.
`
`The content and opinions expressed in this book are the sole work of the authors and editors, who have
`warranted due diligence in the creation and issuance of their work. The publisher, editors, and authors are not
`responsible for errors or omissions or for any consequences arising from the information or opinions presentct1
`in this book and make no warranty, express or implied, with respect to its contents.
`
`This publication is printed on acid-free paper.
`ANSI 23948-1984 (American Standards Institute)
`
`Permanence of Paper for Printed Library Materials.
`
`Cover illustration: “The principle ofantisense inhibition," Figure 1 from chapter 1, Antisense Therapeutics: A
`Promise Waiting to be Fulfilled, by M. Ian Philips
`
`Cover design by Patricia F. Cleary.
`
`For additional copies, pricing for bulk purchases, and/or information about other Humana titles, contact Hu-
`mana at the above address or at any of the following numbers: Tel.: 973—256-1699; Fax: 973-256—834l; E—
`mail: humana@humanapr.com; or visit our Website: www.humanapress.com
`
`Photocopy Authorization Policy:
`
`Authorization to photocopy items for internal or personal use, or the internal or personal use of specific cli-
`ents, is granted by Humana Press Inc., provided that the base fee of US $25.00 per copy is paid directly to the
`Copyright Clearance Center at 222 Rosewood Drive, Danvers, MA 01923. For those organizations that have
`been granted a photocopy license from the CCC, a separate system of payment has been arranged and is
`acceptable to Humana Press Inc. The fee code for users of the Transactional Reporting Service is: [1—58829-
`205-3/05 $25.00I.
`
`Printed in the United States ofAmeriea. 10 9 8 7 6 5 4 3 2 1
`
`Library of Congress Cataloging in Publication Data
`Antisense therapeutics / edited by M. Ian Phillips— 2nd ed.
`p. ; cm. —— (Methods in molecular medicine; 106)
`Includes bibliographical references and index.
`ISBN 1-58829-205-3 (alk. paper); eISBN 1-59259-854-4
`l. Antisense nucleic acids—Therapeutic use.
`[DNLMz l. Oligonucleotides,Antisense—
`therapeutic use. 2. Oligonucleotides, Antisense—pharmacology. QU 57 A6332 2005] 1.
`Phillips, M. Ian. 11. Series.
`RM666.A564A585 2005
`615‘.3 l—dc22
`
`2004006680
`
`This mate rial was {ppieel
`EithE MLM and may be
`Subject U5 Ce {wright Laws
`
`2
`
`

`

`This material may be protected by Copyright law (Title 17 U.S. Code)
`
`2
`
`Antisense Inhibition
`
`Oligonucleotides, Ribozymes, and siRNAs
`
`Y. Clare Zhang, Meghan M. Taylor, Willis K. Samson,
`and M. Ian Phillips
`
`1. Introduction
`
`Over a span of more than two decades, antisense strategies for gene therapy
`have expanded from antisense oligonucleotides (AS-ODNs) solely, to the
`addition of ribozymes and, more recently, to the inclusion of small interfering
`RNAS (siRNAs). Antisense therapeutics has also experienced its phases of high
`expectation, sudden disappointment, and meticulous rediscovery, while
`maintaining its status as a viable and effective gene therapy approach. With
`the discovery of RNA interference (RNAi) and development in delivery of
`these gene drugs, more preclinical and clinical investigations are anticipated
`to take place in the near future to finally fulfill the promise of antisense thera-
`peutics in humans.
`
`2. Antisense Oligonucleotides
`
`AS-ODNs are typically 18—25 bases in length, consisting of sequences that
`are complementary to the target RNA. They can be injected directly into tis—
`sues or delivered systemically. Once delivered into cells, oligonucleotide binds
`to its RNA counterpart and suppresses expression of the proteins encoded by
`target RNA. The specificity of this approach is based on the probability that
`any sequence longer than a minimal number of nucleotides (nt)—l3 for RNA
`and 17 for DNA—occurs only once within the human genome. The idea of
`antisense therapy for inhibiting disease—associated proteins has become par—
`
`From: Methods in Molecular Medicine, Vol. 106: Antisense Therapeutics, Second Edition
`Edited by: l. Phillips © Humana Press lnc., Totowa, NJ
`
`71
`
`This material wascnpied
`atthe NLM and may be
`Subject US Copyright Laws
`
`3
`
`

`

`12
`
`Zhang et al.
`
`ticularly appealing since Zamecnik and Stephenson (1) first demonstrated in
`1978 the reduction of Rous sarcoma viral RNA translation by a specific oligo-
`nucleotide.
`
`2.1. Mechanisms of Antisense Inhibition
`
`Gene expression can be altered by oligonucleotides by means of either
`posttranscriptional inhibition or splicing shift. Posttranscriptional inhibition is
`accomplished by several mechanisms including sterical blockade of ribosomal
`access to the target mRNA, induction of RNase H cleavage of mRNA, and
`inhibition of ribosomal assembly. The net outcome of this process is the dimin-
`ished translation of target proteins. Oligonucleotides chemically modified by
`phosphorothioation are especially effective in activating RNase H, resulting in
`sequence—specific digestion of the target mRNA molecules. This destruction
`of RNA while leaving the DNA oligonucleotide intact allows the oligonucle—
`otide to be recycled, which makes AS—ODNs long lasting. A majority of
`antisense studies so far, including most clinical trials, are aimed at reducing
`undesired disease—associated proteins by virtue of translational inhibition. Alter—
`natively, oligonucleotides that are RNase H inactive and designed toward a cer-
`tain exon—intron junction can prevent the pre-mRNA splicing at the targeted
`site and redirect the splicing to a more favored site. The therapeutic potential
`of this approach has been exemplified in the correction of the expression of
`B—globin and the breast cancer gene BCL—X in related diseases. Certain forms
`of fi—thalassemia are caused by aberrant splicing of [S-globin pre-mRNA that
`leads to abrogation of the protein production (2). AS—ODNs designed to the
`untoward splice site have been proven effective at inhibiting aberrant splicing
`and at restoring B—globin expression in thalassemic patients (3). Likewise, alter-
`native splicing of BCL-X pre-mRNA gives rise to two isoforms, BCL—XL and
`BCL-XS, with opposing antiapoptotic and proapoptotic activities. Targeting
`the BCL-XL splice site with oligonucleotides favored production of the
`proapoptotic BCL-XS protein that enhances cell death in prostate and breast
`tumor cells (4).
`
`2.2. Targeting Antisense
`
`Although antisense can be designed against any region of the target RNA in
`theory, different sequences vary markedly in efficiency of gene inhibition. The
`accessibility of oligonucleotides to RNA is considered the most important fac—
`tor in choosing the optimal antisense sequences. Computational analysis of the
`secondary structure of RNA by programs such as mfold or RNAstructure has
`been used to facilitate selection of target sites for antisense action (5); how-
`ever, it does not take into account the three—dimensional structures as well as
`the instant interaction of RNA molecules with other factors. More commonly
`4
`Th is mate rial was {npied
`attha MLM and may be
`Subiett US Copyright Laws
`
`4
`
`

`

`Antisense Inhibition
`
`13
`
`taken routes involve evaluation of accessible sites by use of RNase H mapping
`(6) or scanning oligonucleotide arrays for the best hybridization signals (7).
`Nevertheless, in general, targeting the start codon AUG, where mRNA is sup-
`posedly open for ribosomal entry, has been a successful strategy, although in
`many cases other sequences turned out to be more effective. Despite these pre-
`dictive approaches, the selection of optimal antisense sequences still requires
`trial—and-error testing initially and, in the end, needs to be confirmed in vivo.
`
`2.3. Chemical Modifications
`
`Stability and efficient delivery, prerequisites for oligonucleotides to achieve
`observable therapeutic effects, have been obstacles due to their macromolecu-
`lar nature. Numerous chemical modifications and delivery approaches have
`been developed to overcome this problem (Fig. 1). The first generation of
`antisense agents contains backbone modifications such as replacement of oxy—
`gen atom of the phosphate linkage by sulfur (phosphorothioates), methyl group
`(methylphosphonates), or amines (phosphoramidates). Of these, the phosphor-
`Othioates have been the most successful and used for gene silencing because of
`their sufficient resistance to nucleases and ability to induce RNase H func—
`tions. However, their profiles of binding affinity to the target sequences, speci-
`ficity, and cellular uptake are less satisfactory. The second generation of
`antisense modifications was aimed at improving these properties, among which
`substitutions of position 2' of ribose with an alkoxyl group (e.g., methyl or
`methoxyethyl groups) were most successful. 2'—0—methyl and 2'—0-
`methoxyethyl derivatives can be further combined with phosphorothioate link-
`age (8). The third generation contains structural elements, such as zwitterionic
`oligonucleotides (possessing both positive and negative charges in the mol-
`ecule); locked nucleic acids (LNAs)/bridged nucleic acids (BNAs) (9);
`morpholino (10); peptide nucleic acids (PNAs) (with a pseudopeptide back-
`bone) (11); and, more recently, hexitol nucleic acids (HNA) (12). All of the
`modifications enhanced AA-ODNs in terms of nuclease resistance; specific
`binding; and with agents such as PNA and morpholino, cellular uptake. How—
`ever, the ability of oligonucleotides to induce RNase H cleavage was abolished
`by these alterations. Therefore, chimeric oligonucleotides with an unmodified
`RNase H—susceptible core flanked by modified nuclease-resistant nucleotides
`have recently been proposed to address this issue and applied in a number of
`investigations (13), including clinical trials.
`
`2.4. Delivery of Antisense
`
`Oligonucleotides are primarily taken up by cells via endocytosis. Only a
`portion of oligonucleotides are able to escape endosome/lysome, enter the
`nucleus, and bind to its RNA complement. Because of the hydrophilic and
`
`This material wascopied
`atthe NLM and may be
`Subject US Copyright Laws
`
`5
`
`

`

`M
`
`Zhang et al.
`
`_EoEoEEm-Q-.NADVMBmoEHoHonmmozm1338-6-
`
`
`
`Eon220::£53mAUV6::2358AB”QEEESonmmonmAm:A<ZE<Z£mEom£2
`
`.Nav.oaoEHEonmmoaaA3”mowsofloscowzoBEES?Mo35886A.wE
`
`
`
`
`
`
`
`
`
`0::comwtfiwuvfloofiAndugwomvonmmonm
`
`
`
`A<vaEon063::658:Q:5&va
`
`O/amo
`
`oz
`
`_om_w<m\/_12/r
`_Hzmw<mowro
`
`:bzz
`o/my
`and
`
`wm<mooxaoomw<moodoomm<moo/axxw
`_/oN_/O_/OooIz/\/\o022b0
`
`mw<moDmw<moOmw<mom
`
`mam
`
`ofimo/b01mfl/On__/Ofilooosz/\/\oo226o
`
`wd_OéwGMmWM.33
`_/Omum05mm
`
`mamamm.
`
`DML
`
`
`I0
`
`wm<mOf
`/_l
`\zln_TOHZ
`
`0/
`
`mw<mo0¢0/n_
`
`I\ozz
`
`_/O
`
`O5II.—mw<mmmm
`
`6
`
`

`

`Antisense Inhibition
`
`75
`
`macromolecular nature, permeation of oligonucleotides across cell membrane
`is relatively difficult. Even after two decades of research, safe and efficient
`delivery of oligonucleotides in vivo still remains a major barrier to the clinical
`success of antisense therapies. Cationic liposomes and electroporation are com-
`monly used carriers. A large variety of liposomal formulas have been devel-
`oped to facilitate antisense delivery, some of which have entered clinical trials
`(14). More recently, nanoparticles and oligonucleotide conjugates have shown
`improved cellular uptake, biodistribution, and targeted delivery, especially in
`cancer treatment (15,16). A hydrodynamic tail vein injection has proven very
`effective in delivering oligonucleotides into liver of rodents (17). Inhalable
`and topical applications of oligonucleotides in patients have shown satisfac—
`tory profiles of uptake and distribution (18,19). However, interestingly, most
`AS-ODNs that are therapeutically valuable in animal models and in patients
`have been administered in the form of naked compounds, despite the progress
`in antisense delivery.
`
`2.5. Antisense in Therapies
`
`Antisense therapeutics has seen its ups and downs since the first antisense
`trial was planned in leukemia in 1992 (20), followed by the excitement over
`the FDA approval of the first antisense drug, Fomivirsen, for the treatment of
`cytomeglovirus (CMV) retinitis in 1998 (21). In addition, more recently, a
`phase III trial reported disappointing results for Affinitak (an antisense inhibi-
`tor of protein kinase C—Ot [PKC-orj) for the treatment of non—small cell lung
`cancer (NSCLC). Cancer is the major target of ongoing clinical trials using
`antisense therapies, followed by human immunodeficiency virus (HIV) and
`other immune-related diseases (Table l). The targets of antisense for cancer
`treatment include genes involved in cell growth, apoptosis, angiogenesis, and
`metastasis. A limitation for antisense as a therapy for cancer may be the single-
`target approach. Even if the target is successfully inhibited by antisense, other
`targets may be activated and compensate for the antisense inhibition. Another
`potential problem is that for successful suppression of cancer growth, the inhi—
`bition should be 100%. However, the mechanism of antisense inhibition is al-
`
`ways in competition with constitutive copies of mRNA, making a 100%
`knockdown difficult to achieve. It is noteworthy that after extensive efforts at
`endogenous expression of antisense RNA by plasmids and viral vectors in a
`variety of disease models, viral delivery of antisense has recently advanced to
`human patients; VRX 496 (a lentivirus vector encoding antisense to HIV-1 env
`protein) started its phase I trial in 2003. Cancer vaccine, a cell therapy using
`NSCLC cell lines genetically engineered to express transforming growth fac-
`tor-[3 (TGF-B) antisense, has also been tested in patients with lung cancer. With
`the emergence of new generations of modified oligonucleotides and delivery
`
`7
`This material was copied
`it the NLM and may he
`Subject US Copyright Laws
`
`7
`
`

`

`
`
`
`
`
`
`wnHtoEmHHHHwaHU“H22£69832
`
`
`
`mAHEoEomHHE085EHom
`
`
`
`mmtqumHHH£053EH8.9594
`
`UHUmZ
`
`
`
`
`
`KN88202:3smummoma
`
`
`
`mngchm833885BoEEEHom
`
`
`
`
`
`
`
`
`
`U<tuEwHHHIHE085EHom
`
`HH"2083@2ko
`
`douooHE93
`
`
`
`
`
`wnHEuEHNHHIHE082EHom
`
`252:3
`
`
`
`
`
` SHOP32938mououfiwcumHHuoU‘HUmZHHUAUmZ
`
`
`
`93583H$0588885
`
`mEEHEuHEumcomsfiwH>HHHH>HEH>HE
`
`
`
`
`
`U<hoEomHHIH€055EHom
`
`
`
`
`
`wnHEuEHJNHHIH‘HEU
`
`xenowooco
`
`xmgoz
`
`:35onSum
`
`059332
`
`.3B5:300
`
`825$
`
`@6me
`
`
`
`HHIHmHoEBEHow
`
`Ewamooz
`
`HHIH$0ch
`
`9:523onH><
`
`HHIH832wmucEvH.5230
`
`
`«€ng5H><
`
`b52580
`
`£395Zmi
`
`m5
`
`8253;
`
`384
`
`ll/
`
`
`
`
`
`
`
`HHIH38mm85>?“mam
`
`mEEanHonHH><
`anomazvSQH><
`
`$8.55;
`
`94-5%
`
`8.2%
`
`8393
`
`$62
`
`02-93580
`
`on?H><
`
`
`
`ocmoom>HUUGNUHHoomuo
`
`HommHBO
`
`2038qu
`
`
`
`035352338<ZQ
`
`
`
`omwmquEootnU
`
`mtg/70
`
`mam>5
`
`>5>HwH
`
`H->HHH
`
`852-0
`
`852-0
`
`HHCDHmDHU
`
`$-59
`
`”2&5
`
`.ovomHBO
`
`o:
`
`a)U)
`«sH
`3
`"Ua)H
`a.)
`'D
`'5
`0<1)
`.H
`
`u5C
`
`:O
`
`.0fi
`94
`8
`Th is m aerial was {Spied
`atthe NLM and may be
`Subject US Cfipyright Laws
`
`mHEo><EEoU
`
`8303:0vaomHm0
`
`35:EmHmH
`
`CHEEEH‘QQHmmmmHmH
`
`mommmHmH
`
`wEQ
`
`
`
`mmHmmHmH
`
`NHvaHaH
`
`Hmm2mm
`
`
`
`835<Z~HE
`
`NHHUmH
`
`mm~H&HH
`
`8.de
`
`nH<~H-o
`
`UanH
`
`UHNHéQH
`
`
`
`>929:emcem$c<Lo“—mET...FEE—UmEomco_2%;
`
`8
`
`
`

`

`
`
`mwmmtomm
`
`Sumac
`
`
`
`
`
`El:638%2:80EBEofiofiozmvNommmHmH7322E>UmE98%;mHmH>Um
`
`
`
`EIH6&6me>265”DadEbwamommH><OZéoEum0%26
`
`
`
`
`
`
`
`
`
`
`
`mmwoaoOEmSomémm~2<oEmocofié‘ mascoaE“0886.580.mEEumxmHmHwmwvofimHmHRYE/E.E«Enaxw
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`.0336band@388.Q<U”E8832msocewofimfioEoEuJED588323223832:85JAUKEEP?“0322:.23?
`
`
`
`
`
`
`
`
`
`
`
`.mowdofluscomzobwmmEonuwooamém.U<”moaomfieosmmonmd?
`
`
`
`
`
`This material was {spied
`atthe NLM arm may be
`Sub-jeit US Capyright Laws
`
`9
`
`

`

`18
`
`Zhang et a/.
`
`technologies, antisense therapeutics is closer to fulfilling its promise in the
`clinic for diseases other than cancer, such as cardiovascular disease, psoriasis,
`and Crohn’s disease.
`
`3. Ribozymes
`
`3.1. What Are Ribozymes?
`
`It was discovered in the early 19805 that some naturally occurring RNA
`molecules have enzymatic activity (22,23). These enzymatic RNA molecules
`were termed ribozymes. Ribozymes recognize specific RNA sequences and
`then catalyze a site—specific phosphodiester bond cleavage within the target
`molecule. Following cleavage, the ribozyme releases itself and binds to another
`target molecule, repeating the process. The cellular consequence varies depend—
`ing on the setting. There are many naturally occurring ribozymes, including in
`plant viroids, ribosomes, self-splicing introns, and the RNA portion of RNase
`P. In plant and animal cells, as well as in viruses, ribozymes are necessary for
`some normal cellular processes such as transcription. The goal of most syn—
`thetic ribozyme usage, however, is reduction in targeted RNA and, thus, lower
`levels of the protein encoded by the target RNA.
`Ribozyme substrate recognition occurs in the same manner as antisense pair—
`ing, through strand complementarity. Therefore, any decrease in target protein
`following ribozyme treatment could in part be due to antisense inhibition of
`translation or the recruitment of cellular enzymes to the double-stranded RNA
`(dsRNA) molecules. However, the ability of each ribozyme molecule to rap—
`idly cleave multiple target molecules gives this technology an advantage over
`classic antisense that can act only on a single RNA molecule. In fact, the rate
`constants of ribozyme cleavage reactions can approach and exceed those of
`protein enzymes, including enzymes with similar functions such as RNase A
`(24,25).
`There are multiple types of ribozymes; the two most commonly used for
`research and therapeutic purposes are the hammerhead ribozyme and the hair—
`pin ribozyme (Figs. 2 and 3). One of the smallest and most well—understood
`r1bozymes, the hammerhead ribozyme, is composed of 30—40 nt and was origi—
`nally discovered as a common sequence found in plant viroids that undergo
`Site-specific, self-catalyzed cleavage as part of their replication process (26).
`All hammerhead ribozymes have a common structure consisting of three base—
`paired helices connected by two invariant single—stranded regions forming the
`catalytic core. Helices l and 3 contain the antisense arms of the ribozyme.
`Helix 3 also contains the cleavage triplet, the site that is cut by the catalytic
`core. The triplet most commonly found in naturally occurring hammerhead
`ribozymes is GUC; however, mutagenesis studies have shown that any cleav—
`
`This material was {spied
`atthe NLM and may be
`Subject US Cepyright Laws
`
`10
`
`

`

`Antisense Inhibition
`
`79
`
`Cleavage
`
`Helix 3
`
`Helix 1
`
`G
`c
`
`u
`A
`
`c
`
`I
`
`Loop 3
`
`c G
`
`Helix 2
`
`Loop 2
`
`Fig. 2. Schematic of a natural hammerhead ribozyme. Hammerhead ribozymes con—
`sist of three helices, formed by complementary base pairing, which are connected by
`single—stranded regions. Loop 3 is removed to generate a trans—cleaving ribozyme;
`helices l and 3 then form the antisense arms. The most commonly found cleavage
`triplet, GUC, is indicated, as is the cleavage site. The single—stranded domain at the
`top of helix 2 is the catalytic core. Highly conserved GC residues in helix 2 are neces-
`sary for catalytic activity.
`
`age triplet with the sequence NUH is tolerated, in which N is any nucleotide
`and H is A, U, or C (27). Hammerhead ribozymes catalyze the hydrolysis of
`the phosphodiester bond at the 3' end of the cleavage triplet. The mechanism
`requires a divalent metal ion, usually Mg“, which plays two crucial roles in
`ribozyme function: it promotes proper folding of the catalytic core and also is
`a catalytic cofactor (28).
`Native hammerhead ribozymes are cis—cleaving enzymes, meaning that their
`targets lie within the same RNA molecule. The ribozyme structure can be engi-
`neered to create an intermolecular cleaving ribozyme consisting of two single-
`stranded antisense arms surrounding the catalytic core and helix 2 so that it
`will cleave within a different RNA molecule. Because RNA often folds into
`
`11
`This mate rial was {nu-pied
`at the N LM arm may he
`Emmett US Copyrigm Laws
`
`11
`
`

`

`20
`
`Zhang et al,
`
`Helix 3
`
`A
`
`U
`
`uA
`u
`
`A
`
`G
`
`c
`
`/Cleavage
`
`U
`
`G
`
`A
`
`Helix 2
`
`cG
`
`A
`
`G
`
`A
`
`Loop A
`
`Helix 1
`
`c
`
`A
`
`A
`
`A
`
`A
`
`c
`
`Loop B
`
`u
`
`A
`
`G
`
`Helix 4
`
`Fig. 3. Schematic of a natural hairpin ribozyme. Hairpin ribozymes consist of 4
`helices, formed by complementary base pairing, which are connected by single
`stranded regions. The small loop at the base of the ribozyme is removed to generate a
`trans-cleaving hairpin ribozyme; helices 1—4 then form portions of the antisense arms.
`The cleavage site is indicated. Loops A and B comprise the catalytic core domains.
`The sequences of both loops are highly conserved, as are the GC residues in helix 2.
`
`eomplex secondary structures, the accessibility of the target site to the anneal—
`mg arms of the ribozyme must be considered when designing a ribozyme. Arm
`lengths of 7 to 8 nt are optimal to convey both specificity and access to most
`lifb0lymes (29). These shorter annealing arms also aid in turnover of the
`rlbozyme, enhancing the ability of each ribozyme molecule to cleave multiple
`target RNA molecules (30).
`_ Hairpin ribozymes, like hammerhead ribozymes, are found in some plant
`v1r01ds that undergo self-catalyzed cleavage as part of their replication pro-
`cess. Hairpin ribozymes contain four base-paired helices and two unpaired
`loops. The ribozyme cleavage site resides within loop A. The helices can vary
`12
`This material wescepied
`etthe NLM and may be
`Subject US Copyright Laws
`
`12
`
`

`

`Antisense Inhibition
`
`27
`
`in length and will tolerate any sequence that maintains complementarity with
`the exception of a requirement for a guanine residue located at the beginning of
`helix 2, which is required for cleavage site recognition (31). Nucleotides within
`the catalytic loop regions, however, must be highly conserved to ensure cata—
`lytic activity of the ribozyme (31).
`Hairpin ribozymes catalyze site—specific hydrolysis of the phosphodiester
`bond on the complementary strand of RNA that is one base upstream of the
`conserved guanine in helix 2. Hairpin ribozymes, like hammerhead ribozymes,
`require an Mg2+ ion to activate proper secondary structure. However, unlike
`the hammerhead ribozyme, Mg2+ does not play a direct role in the catalytic
`process (32). The exact catalytic mechanism used by hairpin ribozymes is not
`yet fully understood. A greater understanding of how both hammerhead and
`hairpin ribozymes work and of methods to optimize their function will enhance
`their attractiveness as potential therapeutic agents.
`
`3.2. Delivery of Ribozymes
`
`Two major issues in the use of ribozymes for research and therapy are ensur—
`ing that the ribozyme is delivered to the target tissues and ensuring that the levels
`of ribozyme delivered are adequate to produce the desired effect. There are
`two methods for delivering the ribozyme to cells: exogenous delivery of a
`presynthesized ribozyme or endogenous expression of the ribozyme. Exog-
`enous delivery is relatively easy and rapid; however, as with antisense, there
`are two main problems with this technique; cellular uptake of the ribozyme is
`often difficult to achieve, and once the ribozyme is taken up, it is quickly
`degraded. Cellular uptake of the ribozyme can be enhanced through the use
`of cationic liposomes. These cationic lipid micelles have the added benefit of
`protecting the ribozymes from RNase present in serum. To enhance further the
`lifespan of ribozymes they are frequently chemically modified. The addition of
`a 2'-0—methyl moiety on some or all of the bases is the most commonly used
`modification. Work is currently being done to engineer DNAzymes, which
`should be more stable than their ribozyme RNA counterparts (33). One benefit
`of exogenous ribozyme delivery in vivo is that the immune system is fairly
`tolerant of foreign RNA molecules (34).
`The other method for delivering ribozymes, endogenous expression of the
`ribozyme, is most often accomplished using viral vectors; however, plasmid
`vectors may also be used. Both retroviral and DNA viral vectors have been
`used. Expression cassettes can be designed to carry cell type—specific or condi—
`tional transcription initiation sites, as well as to include reporter proteins. The
`big advantage of an endogenous ribozyme is that it can be continuously pro—
`duced, allowing for the compromise of target protein production over a long
`period of time.
`
`13
`This mate rial was {nu-pied
`at the N LM arm may he
`Emmett US Copyright Laws
`
`13
`
`

`

`22
`
`Zhang et 3/.
`
`3.3. Research and Therapeutic Uses of Ribozymes
`
`There are four main uses of ribozymes in the medical field: as a research
`tool, as a chemotherapeutic agent, as an antiviral agent, and as a method to
`overcome acquired dominant genetic diseases.
`With the recent sequencing of the Drosophila, mouse, and human genomes,
`there was a surge of newly identified proteins whose role in the organism is
`currently unknown or not fully understood. The use of ribozymes t0 selec\
`tively target these new proteins offers an attractive method to rapidly screen
`their role in vivo. This method, along with other antisense techniques, offers
`several advantages over traditional methods of screening proteins. First, only a
`partial cDNA sequence is required to design a ribozyme. Second, ribozymeS
`can be generated very rapidly, whereas both traditional and conditional knock
`out animals as well as transgenic overexpression animals require a significant
`amount of time to generate. Finally, ribozymes can lead to greater effects for
`longer periods of time when compared with antibody neutralization of the tar
`get protein.
`In addition to rapid screening of new proteins, ribozyme technology can
`also be used to overcome problems with traditional protein function studies. For
`example, we use ribozymes to target a protein that when knocked out results in
`embryonic demise in mice and for which conditional knockouts have been
`
`unsuccessful (35). Ribozymes can also be used to locally target a protein that
`is made in many tissues, such as to lower targeted protein levels in brain
`without altering protein expression in the periphery.
`The specificity of ribozymes makes them very attractive as therapeutics in
`disease states in which a protein is overexpressed or is malfunctioning.
`Ribozymes have the capability to specifically recognize single nucleotide dif—
`ferences in their targets. This special feature has resulted in the development of
`ribozymes to target oncogenes that are frequently mutated in tumors. For
`instance, the oncogene H—ras is mutated at a high frequency in many cancers;
`therefore, a ribozyme that recognizes only the mutant H-ras transcript has the
`potential to be a very efficacious treatment. Several ribozymes have been devel—
`oped that can discriminate between H-ras mutants and the normal H-ras tran—
`script and initial studies have shown that stable expression of H—ras mutant
`ribozymes leads to reduced tumor formation in athymic mice (36,37).
`Alternative uses for ribozymes in cancer therapy are to block the elevation
`of normal gene products, such as c—fos, that occur in transformed cells or to
`block angiogenic pathways. One such antiangiogenic ribozyme is targeted to
`flt-l mRNA, which encodes for the vascular endothelial growth factor receptor
`(VEGF-R). The ribozyme has been shown to be well tolerated when adminis—
`tered daily by sc injection, and this dosing schedule leads to prolonged eleva—
`
`14
`This material was {spied
`atthe NLM and may be
`Subject US Cepyright Laws
`
`14
`
`

`

`Antisense Inhibition
`
`23
`
`tion in the plasma levels of ribozyme (38). This ribozyme is now in phase II
`clinical trials in which therapeutic efficacy in breast and colorectal cancers is
`being examined.
`A different set of circumstances in which a ribozyme can offer great thera-
`peutic potential is the treatment of acquired dominant genetic diseases. Retini—
`tis pigmentosa is a genetic disease that causes carriers of the dominant P23H
`rhodopsin allele to slowly lose their vision. Hauswirth and Lewin have devel-
`oped a ribozyme that recognizes only the dominant version of the gene tran—
`script, which differs by two bases from the wild-type gene. Ribozyme treatment
`has resulted in a halt of disease progression in various species includlng rat,
`dog, and now monkey (39,40). This treatment is currently being prepared to
`enter the first phase of clinical trials. Promising results from this study could
`open the door for the development of ribozymes to treat other dominant genetIC
`disorders.
`,
`,
`One final area where ribozyme therapy holds much promise is as anthIFal
`agents, particularly in the treatment of retroviral infections. Many RNA VlfUSeS
`such as HIV have very high mutation rates throughout much of their genome
`that renders the mutated viruses resistant to current treatments. However, some
`sequences, including promoters and slicing signals, are highly conserved in
`HIV and among other RNA viruses. These regions provide excellent targets
`for ribozymes. In fact, some groups have designed ribozymes agamst conserved
`areas of HIV and have shown that ribozyme treatment can provide long-term
`HIV resistance and decrease HIV replication in infected cells (41). Several
`companies now have ribozymes directed against HIV in various stages of cllnl-
`cal trials. Other viruses for which ribozyme treatments are also being deslgfled
`include hepatitis B, hepatitis C, and the herpes viruses. Table 2 summertZeS
`ongoing clinical trials using ribozymes.
`The use of ribozymes for the inhibition of gene expression hOIdS great prom—
`ise in both therapeutics and research; however, we have only begun tolunder-
`stand the potential of these molecules. Efforts to improve the .stablllty and
`delivery of ribozymes will enhance their usefulness as therapeutic agents and
`lead to a greater recognition of the role of novel proteins in selected tlssues
`and the body as a whole.
`
`4. RNA Interference with siRNA
`
`4.1. What Is siRNA?
`
`RNAi is a form of antiviral immune response mounted by many higher eu—
`karyotes—including plants, nematodes, and insects—0n exposure to dsRNA.
`dsRNA molecules are key intermediates in the genomic replication of many
`viruses but are not normally found in eukaryotic cells. In contrast to the 1nter—
`
`This mate rial was {GDlEEl
`atthe N LM and may be
`Subject US Ca-pyright Laws
`
`V
`
`15
`
`

`

`
`
`
`
`>HEE@8235“:£88onEwEonmEb:Emnoméoq2830an
`
`
`
`355%>HmE£030me358%638::mm<coE>53:88385:8wEEEEoEH8058
`
`
`
`
`
`ofimNoflm2:353:amigo4%onofizwofim2938:Non?MmmwmEUmagnum
`
`
`
`@538an3:35253»3:0st
`
`23-93
`
` 3530an
`
`
`
`mBoZomwam$28mebamEoUWED
`
`
`
`
`
`
`
`Hows“<58
`
`
`
`
`
`8502‘wa
`
`
`
`oquowEE053wzomEEGOQEofizNomwsaVMiwmm
`
`.
`.16
`_
`ThIS material was {GDIEEX
`aithe NLMand maybe
`‘3
`Subject UV
`Cn-pyright Laws
`
`
`
`
`
`tramp:2:?an.5»m3:FEE—UwEomco
`
`
`
`
`
`.N2&8.
`
`16
`
`
`
`

`

`Antisense Inhibition
`
`25
`
`feron responses of mammalian cells in the face of viral infection, RNAi is used
`by many other eukaryotes to defend against viruses through dsRNA-induced
`degradation of viral RNAs.
`The first evidence that dsRNA could suppress gene functions came from the
`work in Caenorhabditis elegans (42). In 1998, Fire et al. (43) found that sense
`RNA was as effective as antisense RNA for inhibiting genes. Subsequently,
`Zamore et al. (44) demonstrated that dsRNA was at least 10-fold more potent
`as a silencing trigger than was sense or antisense RNA alone. Since then, gene
`silencing by dsRNA has been termed RNAi, and its mechanisms have been
`elucidated vigorously. Our current mechanistic understanding of RNAi derives
`largely from work in the Drosophila system (44,45). The first step of RNAi is
`to process longer dsRNA into 21- to 23—nt fragments that bear 3' overhangs by
`an RNase III—like enzyme called Dicer (46). These approx 21 nt dsRNAs,
`which are termed as siRNA, are essential to form a large (approx 500-kDa)
`RNA—induced silencing complex (RISC) (47). Through a yet-undefined mecha-
`nism, RISC cleaves the target mRNA that is complementary to the guide
`siRNA, whether the target RNA is a viral mRNA or a cognate gene.
`
`4.2. Application in Mammalian Cells
`
`The key characteristics of RNAi are its remarkable sequence specificity,
`and it can therefore be used to target gene expression. It was found in Droso—
`phila that artificial siRNAs can be incorporated into RISC and induce degrada-
`tion of target mRNA. However, previous efforts to induce RNAi in cultured
`mammalian cells had largely failed because long dsRNAs (>30 bp) could induce
`a potent, nonspecific interferon response and activation of the protein kinase
`PKR and 2'