------
`
`BIOCHIMICA ET BIOPHYSICA ACTA
`
`INTERNATIONAL JOURNAL OF BIOCHEMISTRY, BIOPHYSICS
`AND MOLECULAR BIOLOGY
`
`VOL. 1489 (1) (1999)
`
`GENE STRUCTURE AND EXPRESSION
`
`Executive Editors: P.C. van der Vliet (Utrecht, The Netherlands)
`J.D. Oralla (Los Angeles, CA, USA)
`
`Section Editors:
`M. Busslinger
`N.H. Colburn
`J. Egly
`R. Haselkorn
`R. Lilhrmann
`P.E. Nielsen
`
`K.H. Nierhaus
`N. Osheroff
`C. L. Peterson
`A.C.G. Porter
`A.G. Porter
`S.T. Smale
`
`L.L. Spremulli
`H.G. Stunnenberg
`M. van Montagu
`M.J. Waring
`A.P. Wolffe
`
`ELSEVIER
`Amsterdam - Lausanne - New York - Oxford - Shannon - Tokyo
`
`

`

`© 1999 Elsevier Science B. V. All rights reserved.
`
`This journal and the individual contributions contained in it are protected under copyright by Elsevier Science B. V .. and the following terms and conditions
`apply to their use :
`
`Photocopying
`Single photocopies of single articles may be made for personal use as a llowed b} nationa l copyright law,. Permission of the Publisher and payment of a fee is
`required for a ll o ther photocopying. including multiple or systematic copying. copying for advertising or promotional purposes. resale. and all forms ~f
`document delivery. Special rates arc available for educational institutions that wish to make photocopies for non-profit educational classroom use.
`
`Permissions may be sought directly from Elsevier Science R ights & Pern1ission Department. P.O. Box 800, Oxford OX5 !DX. UK: phone: (+44) 1865
`843830. fax: ( +44) 1865 853331. e-mail: permissions(<ielsevier.co.uk. You may also contact Rights & Permissions directly through Elsevicr's home page
`(http://www.elsevier.nl). selecting first ·Customer Support'. then ·General Information·. then 'Permissions Query Form' .
`
`In the CSA. users may clear permissions and make payments through the Copyright C learall(;c Ct!nler. Inc .. 222 Rosewood D rive, Danvers. MA 0 1923.
`USA; phone: (+I) (978) 7508400. fax: (+I) (978) 7504744. and in the U K through th~ Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90
`To1tenham Court Road. London WI P OLP. UK: phone: (+44) 171 631 5555; fax: (+44) 171 63 1 5500. Other countries may have a local reprographic right s
`agency for payments.
`
`Dcril·ative Works
`Subscribers may reproduce tables of contents or prepare lists of articles including abstracts for internal circulation within their institutions. Permission of the
`Publisher is required for resale or distribution o utside the institution.
`Permission of the Publisher is req uired for all other derivative works. including compilations and translations.
`
`Eleclronic S1oragc or L:sage
`Permissio n of the Publisher is required to store or use electronically any material contained in this journal. inducting any article or part of an article.
`
`Except as outlined above, no part of this publication may be reproduced. stored in a retrieva l S}Stcm or 1ransmi1ted in any form or by any means. electronic.
`mechanical. photocopying. recording or otherwise, without prior wri1ten permission of the Publisher.
`
`Address permission requests to: Elsevier Science Rights & Permissio ns Department. at the mail. fax and e-mail addrcsse, noted above.
`
`:'lloticc
`No rcspon,ibility is assumed by the Publisher for any injury and/or damage to person, or property as a mailer of products liability. negligence or otherwise,
`or from any use or o peration of any methods. products. instructio ns or ideas contained in the materia l herein. Because of the rapid advances in the medical
`sciences. in particular. independent verification of diagnoses and drug dosages should be made.
`
`Although all advertising material is expected lo conform 10 ethical (medica l) standards, inclusio n in this publication docs not constitute a guarantee or
`endorsement of the quality or value of such product or of the claims made of it by its manufact urer.
`
`Ad,•ertising Information
`Advertising orders and enqLtiries can be sent lo: USA, Canada and South America: Mr T ino de Carlo. The Advertising Department. Elsevier Science Inc ..
`655 Avenue of the Americas, New York. NY 10010-5107, USA: phone: (+ I ) (212) 633 38 15; fax: (+I) (212) 633 3820: e-mail: l.dccarlo@elsevier.com.
`J a pa n : The Advertising Department. Elsevier Science K.K .. 9- 15 Higashi-Azabu 1-chome. Minato-ku. Tokyo 106-0044. Japan: phone: (+81) (3) 5561 5033;
`fax: ( +8 1) (3) 5561 5047. E:urope and ROW : Rachel Le;,veson-Gower. The Advertising Department. Elsevier Science Ltd .. The Boulevard . Langford Lane.
`Kidlington. Oxford. OX5 1G B. U K : phone : (+44) (1865) 843565: fax: (+44) (1865) 843976: e-mail : r.leveson-gower(<u,elscvicr.co.uk.
`
`O rders, claims, a nd product enquiries: please contact the Customer Support Department at the Regional Sales Office nearest you :
`New York: Elsevier Sciem:e. PO Box 945. New York. NY 10159-0945. USA: phone: (+I) (212) 633 3730 [toll free number for North American customers:
`1-888-4ES-INFO (437-4636)) : fax: (+I) (212) 633 3680: e-mail: us info-f(<iclscvier.com
`Amsterdam: Elsevier Science. PO Box 211. 1000 AE Amsterdam. The Netherlands: phone: (+31) 20 4853757: fax: (+31) 20 48514.~2: e-mail:
`nlinfo-f@,clsevic,r.nl
`Tokyo: Elsevier Science. 9-15 Higashi-Azabu 1-chome. Minato-ku. Tokyo 106-0044. Japan : phone : (+81) (3) 5561 5033: fax : (+81) 0) 5561 5047: e-mail:
`info({velsevicr.co.jp
`Singapore: Elsevier Science. >lo. I Temasek Avenue. #17-01 Millcnia T ower. Sing,1pore 039192: phone: (+65) 434 3727: fax : (+65) JJ7 2230: e-mail:
`asiainfo@elsevicr.com.sg
`Rio de Janeiro: Elsevier Science, Rua Setc de Setembro 111 /16 Andar. 20050-002 Centro. Rio de Janeiro - RJ. Brazil: phone: (+55) (2 1) 509 5340: fax:
`( +55) (21) 507 1991; e-mail: clsevicr:iYcampus.com.br [Note (Latin America): for orders, daims a nd help desk information. please contact the Regional
`Sales Office in New York as listed above]
`
`USA mailing notice: Bi()c/,i111im <'t Biophysim A,w I Gnw Smu·tw·c a11d Lrpressiort ) ( ISSN 0167-4781) is published monthly by Elsevier Science 8 . V. (P.O. Box
`211. 1000 i\E Amsterdam. The Netherlands).
`Annual subscription price in the USA US$ 2050.00 (valid in North. Central and South America). including air speed delivery. Periodical, postage pa id at
`Jamaica. NY 11431.
`USA POSTMAST ER: Send address changes to Bi()clti111ita et Bi()ph_rsim Aclll ( Gene S1ruc111re and l:.'xpre.uio11J. Publications Expediting. Inc .. 200 Meacham
`Avenue. Elmont, NY 11003.
`AIR FREIGHT Al'D MAILING in the USA by Publications Expediting. Inc .. 200 Meacham A,cnue. Elmont. Y 11003.
`
`Printccl in The Netherlands
`0 The paper used in thi, publication meet> the requirements of ANSIIN ISO 239.48- 1992 ( Permanence of Paper).
`
`

`

`This material may be protected by Copyright law (Title 17 U.S. Code)
`
`ELSEVIER
`
`Biochimica et Biophysica Acta 1489 (1999) 141 - 158
`
`BIOC HIMICA ET BIOPHYSIC A ACTA
`
`BB~
`
`www.elsevier.com/loca te/b ba
`
`Review
`Morpholino antisense oligomers:
`the case for an RNase H-independent structural type
`James Summerton *
`Gene Tools, P. 0. Box I 186. Corvallis, OR 97339. USA
`Received 2 December 1998 ; received in rev ised form 16 March 1999 ; accepted 18 May 1999
`
`Abstract
`RNase H-competent phosphorothioates (S-DNAs) have dominated the antisense field in large part because they offer
`reasonable resistance to nucleases, they afford good efficacy in cell-free test systems, they can be targeted against sites
`throughout the RNA transcript ofa gene, and they are widely available from commercial sources at modest prices. However,
`these merits are counterbalanced by significant limitations, including: degradation by nucleases, poor in-cell targeting
`predictability, low sequence specificity, and a variety of non-antisense activities. In cell-free and cultured-cell systems where
`one wishes to block the translation of a messenger RNA coding for a normal protein, RNase H-independent morpholino
`antisense oligos provide complete resistance to nucleases , generally good targeting predictability, generally high in-cell
`efficacy, excellent sequence specificity, and very preliminary results suggest they may exhibit little non-antisense
`activity. © 1999 Elsevier Science B.V. All rights reserved .
`
`Keywords: Antisensc; Morpholino ; RNase H-independent ; Specificity; Efficacy; Delivery
`
`Contents
`
`I.
`
`2.
`
`3.
`
`RN ase H cleavage: origins of the broad acceptance of S-DNAs . . . . . . . . . . . . . . . . . . . .
`
`Limitations of S-DNAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`2.1. Degradat ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`2.2 . Cleavage of non-targeted RNA sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`2.3. Prom iscuous binding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`Is RN ase H competency necessary? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`3. I. Efficacy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`3.2. Targeting versatility
`3.3. Availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`. . . . . . . . . . . . . . . . . .
`4. Advantages of RNase H-independent morpholino antisense oligos
`4.1. Predictable ta rgeting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`4.2. Reliable efficacy in cultured cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`142
`
`143
`143
`143
`143
`
`143
`143
`144
`145
`
`145
`145
`14 7
`
`* Fax : (541) 7536360; E-mail: mail@gene-tools.com
`
`0167-4781 /99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved.
`PII : S O I 6 7 - 4 7 8 I ( 9 9 ) 0 0 I 5 0 - 5
`
`

`

`142
`
`J. S11111mertonl Biochimica er Biophysica A cra 1489 ( /999) 14/-/58
`
`4.3. High specificity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`4.4. Little non-antisense activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`148
`152
`
`5.
`
`Positive test system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`152
`
`6. Delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`6. 1. Delivery into cultured cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`6.2. Delivery in vivo: the final challenge? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`153
`153
`154
`
`References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`156
`
`1. RNase H cleavage: origins of the broad acceptance
`of S-DNAs
`
`A key requirement for effective antisense oligos is
`that they remain intact for many hours in the extra(cid:173)
`cellular medium and within cells. The methylphos(cid:173)
`phonate-linked DNA analogs developed by Miller
`and Ts'o in the late 1970s constituted a major ad(cid:173)
`vance in the emerging antisense field by providing the
`first antisense type having good stability in biological
`systems [I]. However, concerns subsequently devel(cid:173)
`oped that methylphosphonates might be inadequate
`for many antisense applications, particularly thera(cid:173)
`peutics, because of their low efficacy -
`typically re(cid:173)
`quiring concentrations in excess of 20 µM for good
`activity in a cell-free translation system [2]. Some
`time later phosphorothioate-linked DNA analogs
`(S-DNAs) were introduced [3]. These S-DNAs were
`enthusiastically embraced because
`they achieved
`good efficacy at concentrations a 100-fold lower
`than corresponding methylphosphonates [4]. The
`S-DNAs also had good water solubility, reasonable
`resistance to nucleases, and they were readily pre(cid:173)
`pared on standard DNA synthesizers with only mod(cid:173)
`est modification of the oxidation step.
`The surprisingly high efficacy of the S-DNAs rel(cid:173)
`ative to methylphosphonates was not readily ex(cid:173)
`plained by their moderately higher target binding
`affinities. Investigations into this large discrepancy
`in efficacies led to the discovery that while methyl(cid:173)
`phosphonates and most other antisense types act
`only by a steric block mechanism, DNA and S(cid:173)
`ONA oligos instead act predominantly by an RNase
`H-cleavage mechanism wherein after the oligo pairs
`to its RNA target sequence the enzyme RNase H can
`cleave the paired RNA target sequence [5]. It was
`also discovered that those structural types which
`
`function only by a steric block mechanism (RNase
`H-independent types) are generally effective in block(cid:173)
`ing translation only when targeted against mRNA
`sequences in the region extending from the 5' cap
`to a few bases past the AUG translational start site
`(see Fig. 2). In contrast, structural types which utilize
`an RNase H-cleavage mechanism (RNase H-compe(cid:173)
`tent types) could also be effective against target se(cid:173)
`quences elsewhere in the RNA transcript of a gene.
`Representative RNase A-competent and RNase H(cid:173)
`independent antisense types are shown in Fig. I.
`The explanation for these differing targeting prop(cid:173)
`erties lies in the mechanism of protein translation. In
`eukaryotic systems an initiation complex recognizes
`and binds to the 5' cap structure and then scans
`down the 5' leader sequence until it encounters the
`AUG translational start site, at which point the full
`ribosome is assembled, followed by translation of the
`amino acid coding region of the mRNA. Antisense
`oligos apparently can physically block progression of
`the initiation complex down the mRNA leader and
`block assembly of the ribosome at the AUG transla(cid:173)
`tional start site. However, once ribosome assembly
`occurs at the translational start site that ribosome is
`capable of displacing almost any bound antisense
`oligo it encounters as it traverses the amino acid
`coding region of the mRNA. Presumably this oli(cid:173)
`go displacement is effected by the very robust ATP(cid:173)
`driven unwindase activity of translating ribosomes
`[6].
`Thus, most RNase H-independent antisense types
`can block translation only when targeted against se(cid:173)
`quences in the region from the 5' cap to about 25
`bases past the AUG translational start site of an
`mRNA. In contrast, RNase H-competent oligos
`can also be effective against sequences elsewhere in
`the RNA transcript by virtue of their effecting deg-
`
`

`

`J. Summerton/Biochimica et Biophysica Acta 1489 (1999) 141-158
`
`143
`
`radation of the paired RNA target sequence by
`RNase H.
`To summarize: Based on the higher efficacy and
`greater targeting versatility of S-DNAs relative to
`the early RNase H-independent oligos, many work(cid:173)
`ers in the antisense field have concluded that RNase
`H competency is essential for good antisense activity.
`This, combined with their ready availability and low
`cost, have established S-DNAs as the structural type
`currently most used in antisense studies.
`
`2. Limitations of S-DNAs
`
`With continued study of S-DNAs it is now widely
`recognized that their good efficacy and targeting ver(cid:173)
`satility are counterbalanced by a variety of disadvan(cid:173)
`tages.
`
`2.1. Degradation
`
`S-DNA oligos are sensitive to nucleases, being de(cid:173)
`graded in biological systems over a period of hours
`[7]. Such instability can complicate interpretation of
`experimental results and may require either shorter(cid:173)
`than-desired experiments or multiple dosing.
`
`2.2. Cleavage of non-targeted RNA sequences
`
`RNase H cleaves DNA/RNA and S-DNA/RNA
`duplexes as short as 5 or 6 base pairs in length and
`is highly active against such duplexes only 9- 10 base
`pairs in length [8]. As a consequence, essentially
`every RNase H-competent oligo has the potential
`to form transient complexes with and induce cleav(cid:173)
`age of 'non-targeted' cellular sequences having parti(cid:173)
`al homology to the intended target RNA. It seems
`reasonable to expect that this RNase H cleavage
`could compromise the sequence specificity of S(cid:173)
`DNAs. In simple cell-free translation systems with
`added RNase H poor sequence specificity is indeed
`seen with S-DNAs [9]. This same RNase H cleavage
`might also be expected to cause disruptions in more
`complex cellular systems and in patients.
`
`2.3. Promiscuous binding
`
`While both DNA and S-DNA support RNAse H
`
`cleavage, because DNA oligos undergo rapid degra(cid:173)
`dation in biological systems S-DNAs have become
`by default the choice for RNase H-competent anti(cid:173)
`sense oligos. The problem this presents is that the
`pendent sulfurs in the phosphorothioate linkages of
`S-DNAs interact with a wide variety of proteins, in(cid:173)
`cluding laminin, bFGF, protein kinase C, DNA
`polymerase, telomerase, fibrinogen, phospholipase
`A2, HIV gpl20, HIV reverse transcriptase, CD4,
`Taq polymerase, T4-polynucleotide kinase, fibronec(cid:173)
`tin, many tyrosine kinases, and proton-vacuolar
`ATPase [10]. For this and other reasons S-DNAs
`can cause multiple non-antisense effects.
`In addition, S-DNAs containing the sequence Pu(cid:173)
`Pu-C-G-Py-Py have been shown to trigger B cell ac(cid:173)
`tivation [11] and S-DNAs containing four or more
`contiguous guanines have been shown to form a tet(cid:173)
`rameric complex which can cause a variety of non(cid:173)
`antisense effects [12]. S-DNAs within cells have also
`been reported to rapidly induce Sp 1 transcription
`factor [13].
`The non-antisense effects caused by S-DNAs can
`result in control oligos exhibiting biological activities
`on a par with that of the antisense oligos [14]. Fur(cid:173)
`ther, because S-DNAs can effect multiple non-anti(cid:173)
`sense activities it is difficult to confirm that a given
`biological response is truly due to an antisense mech(cid:173)
`anism - leading to considerable uncertainty and pos(cid:173)
`sible misinterpretations in antisense experiments uti(cid:173)
`lizing S-DNAs [15].
`To summarize: Because of their sensitivity to nu(cid:173)
`cleases, limited sequence specificity, and multiple
`non-antisense effects it appears that S-DNAs are
`less than optimal antisense tools.
`
`3. Is RNase H competency necessary?
`
`3.1. Efficacy
`
`A key property of the RNase H-competent S(cid:173)
`DNAs which led to their broad adoption by the anti(cid:173)
`sense community was their greatly increased efficacy
`(likely a consequence of their RNase H competency)
`relative to methylphosphonates. However, since then
`at least two RNase H-independent types (PNAs [16]
`and morpholinos [17] shown in Fig. 1) have been
`developed which often match or exceed the efficacy
`
`

`

`144
`
`J. Summerton!Biochimica el Biophysica Acta 1489 (1999) 141- 158
`
`RNase H-Competent
`
`DNA
`
`S-DNA
`
`•••
`
`Methylphosphonate
`• ••
`
`RNase H-lndependent
`
`2'-0-Methyl RNA
`
`PNA
`
`Morpholino
`
`. . .
`
`. . .
`
`B = adenine, cytosine, guanine, thymine, uracil
`
`Fig. I. Representative RNase H-competent and RNase H-independent types.
`
`of S-DNAs m a cell-free translation system when
`said oligos are targeted to sequences in the region
`from the 5' cap to about 25 bases 3' to the AUG
`translational start site [9].
`In our cultured cell test system these advanced
`RNase H-independent antisense types (morpholinos
`and PNAs) show an even greater efficacy advantage
`over the RNase H-competent S-DNAs. To illustrate,
`both S-DNA (RNase H-competent) and morpholino
`(RNase H-independent) antisense oligos, all of which
`had been shown to be highly active in a cell-free
`translation system, were scrape-loaded into HeLa
`cells and assessed for efficacy in blocking their re(cid:173)
`spective RNA target sequences therein. In this study
`in cultured cells [ I 8] the two different morpholinos
`had IC50 values of about 60 nM and near quantita(cid:173)
`tive target inhibition at 300 nM. In contrast, neither
`of the corresponding S-DNAs achieved significant
`target inhibition within cells at concentrations up
`to 3000 nM. In similar experiments we have found
`PNAs (another advanced RNase H-independent
`type) to exhibit a similar large efficacy advantage
`over S-DNAs in scrape-loaded cells (unpublished re(cid:173)
`sults).
`In a different study involving inhibition of TNF-a,
`Kobzik and coworkers also found morpholinos to
`achieve appreciably higher efficacies
`than corre(cid:173)
`sponding S-DNAs in cultured cells [19].
`
`3.2. Targeting versatility
`
`While S-DNAs and other RNase H-competent
`
`antisense oligos can target and destroy (via RNase
`H cleavage) sites throughout the RNA transcript of a
`gene, including splice sites, only RNase H-independ(cid:173)
`ent antisense oligos, such as morpholinos, can efTect
`correction of splicing errors [20].
`Both RNase H-competent and RNase H-inde(cid:173)
`pendent types can be used to block translation of any
`specific mRNA by targeting the 5' leader/translation(cid:173)
`al start region of that mRNA.
`It is commonly assumed that only S-DNA and oth(cid:173)
`er RNase H-competent oligos are suitable for study(cid:173)
`ing point mutations and polymorphisms more than
`about 20 nucleotides 3' to the translational start site
`in mRNAs. However, this perceived limitation of
`RNase H-independent oligos can be circumvented
`by using a gene switching strategy (P. Morcos, Meth(cid:173)
`ods Enzymol., in press). Typically this entails using
`cells containing a normal gene and transf ecting in a
`plasmid containing a mutant or polymorphic form of
`that same gene. Key to this scheme is to use a leader
`sequence in the transfected gene's mRNA which dif(cid:173)
`fers by at least a few bases from the leader sequence of
`the endogenous gene's mRNA. One then uses a high(cid:173)
`specificity RNase H-independent antisense oligo, such
`as a morpholino, to selectively block translation of
`either the endogenous or the exogenous mRNA, after
`which one assesses for phenotypic changes. By this
`means one can exploit the exceptional specificity of
`the RNase H-independent morpholino antisense oli(cid:173)
`gos to carry out rigorous and well controlled studies
`of a wide variety of mutations and polymorphisms
`positioned anywhere in the mRNA.
`
`

`

`J. Summerton/Biochimica et Biophysica Acta 1489 (1999) 141- 158
`
`145
`
`3. 3. Availability
`
`Another factor which led to the widespread use of
`S-DNAs was their ready availability at moderate pri(cid:173)
`ces from commercial sources. While lack of commer(cid:173)
`cial sources, high prices and slow delivery have in the
`past been significant barriers to use of advanced
`RNase H-independent antisense types, this situation
`is now changing. Both PNAs [21] and morpholinos
`[22] are now commercially available, and prices of
`morpholinos are now competitive with prices of ad(cid:173)
`vanced mixed-backbone S-DNAs (i.e., chimeras).
`To summari::e: Relative to S-DNAs, properly tar(cid:173)
`geted morpholinos often achieve equal or better effi(cid:173)
`cacy in cell-free systems and often achieve substan(cid:173)
`tially better efficacy in cultured cells; of the two types
`only morpholinos can be used for correcting splicing
`errors; a new gene switching strategy gives morpho(cid:173)
`linos targeting versatility on a par with S-DNAs for
`selected applications; and morpholinos are now com(cid:173)
`mercially available at moderate prices with reason(cid:173)
`able delivery times.
`
`4. Advantages of RNase ff-independent morpholino
`antisense oligos
`
`4. 1. Predictable targeting
`
`A problem which has plagued antisense research
`with S-DNAs is the difficulty of predicting which
`antisense sequences will be effective in cells. As a
`consequence, multiple S-DNAs may need to be pre(cid:173)
`pared and empirically tested in order to identify an
`oligo with good in-cell activity [23,24]. Further, when
`one does find an effective oligo through such an em(cid:173)
`pirical search it is not unusual to find that it is tar(cid:173)
`geted in a region of the RNA, such as the 3' untrans(cid:173)
`lated region [25], which one would not normally
`expect to afford inhibition by an antisense mecha(cid:173)
`nism. The poor showing of those many antisense
`oligos which do not show good efficacy in such
`searches has been attributed to their having restricted
`access to their RNA target sequences within cells due
`to secondary structure of the RNA [26] and/or due to
`proteins bound to the RNA. Alternatively, this lack
`of inhibitory activity by many S-DNAs could be due
`to some activity which disrupts RNA/S-DNA du-
`
`plexes within cells [27] or due to S-DNAs being effi(cid:173)
`ciently sequestered in a partially sequence-specific
`manner by some nuclear structure [28], or due to
`activation of Sp l transcription factor [ 13] which
`acts to overshadow the translational inhibition by
`the S-DNA.
`A need to sift through multiple S-DNAs in order
`to find one that effectively inhibits its target in cul(cid:173)
`tured cells seriously limits the utility of S-DNAs as
`routine tools for the study of gene function and con(cid:173)
`trol. Further, in the absence of rational and reliable
`targeting rules for S-DNAs and in light of their par(cid:173)
`tially sequence-specific non-antisense effects, the need
`to test multiple S-DNAs in order to find an effective
`one raises the specter that one may not be selecting
`for an accessible antisense target, but instead one
`may be selecting for an oligo sequence effective to
`generate some non-antisense activity which is then
`misinterpreted as the desired antisense effect [15].
`In contrast to the difficulty in predicting effective
`targets for S-DNAs in cultured cells, we have found
`targeting of morpholinos (lacking undue self-comple(cid:173)
`mentarity) to be reasonably predictable both in cell(cid:173)
`free systems and within cultured cells. This is illus(cid:173)
`trated by a targeting study using a target mRNA
`comprising an 85-base segment of the leader se(cid:173)
`quence of hepatitis B virus (HBV) fused to the cod(cid:173)
`ing sequence of firefly luciferase [18]. This HBV lead(cid:173)
`er sequence presents a substantial targeting challenge
`because it contains a region of quite stable secondary
`structure extending from positions - 47 to +3 (where
`+ I is the A of the AUG translational start site).
`Experimental procedures and the mRNA target
`used in this study are detailed in [18].
`Fig. 2a shows the 5' leader region and 24 bases of
`the amino acid coding sequence of this mRNA, and
`indicates with bold lines the target sequences for sev(cid:173)
`en of the morpholino oligos tested in this study. Fig.
`2b shows the linear positioning of the morpholino
`antisense oligos along this HBV-luciferase mRNA,
`as well as each oligo's percent inhibition achieved
`in a cell-free translation assay, with oligo present at
`a concentration of 1 µM and target mRNA present
`at I nM.
`A representative subset of these morpholinos were
`tested in cultured cells stably transfected with a plas(cid:173)
`mid coding for the same HBV/luciferase mRNA con(cid:173)
`struct. Fig. 2c shows inhibition of luciferase produc-
`
`

`

`146
`
`J. Summerton/Biochimica et Biophysica Acta 1489 (1999) 141- 158
`
`tion in cells scrape-loaded in the presence of 3 µM
`morpholino oligos, assessed as described in [29].
`The results in Fig. 2b show that the tested mor(cid:173)
`pholinos were reasonably effective along the entire 5'
`leader and up to a few bases 3' to the AUG trans-
`
`(a)
`
`A o U
`A o U
`4 C • G
`C • G
`U o G,
`C • G- U
`
`• u
`• u ·c
`•A o U
`U
`
`_J, ,.,rnL· .
`
`A oU
`G=c=AA=-ec=u=uu=uu=c=Ac=c=uc=u"='Gc='=cuAAUCAUCUCUUGUAC • GGACCUCGAGGACGCCAAAAAC
`2
`
`85
`
`7
`
`+24
`
`(b)
`100
`
`C
`0
`;l
`;§ 50
`.c
`.!:
`~ 0
`
`0
`
`(c)
`100
`
`C
`0
`;l
`;§ 50
`.c
`.!:
`~ 0
`
`0
`
`4
`
`5 - -7
`-r-
`
`~
`
`10 - -
`- -11
`
`_1_2_
`
`20
`
`40
`
`60
`
`-80
`
`-60
`
`-40
`
`-20
`mRNA
`
`3
`
`-4 - -
`
`6
`
`_§_
`
`9-=-11
`
`_1_2_
`
`20
`
`40
`
`60
`
`-80
`
`-60
`
`-40
`
`-20
`mRNA
`
`lational start site. Oligos targeted to sites beginning
`more than 20 bases 3' to the AUG translational start
`site showed no significant activity. Of particular note,
`the results demonstrate that oligos 3, 4, and 5 appear
`to have effectively invaded the quite stable secondary
`structure within the HBV leader sequence.
`The results shown in Fig. 2c suggest that those
`morpholinos which are effective in a cell-free assay
`are generally also effective in cultured cells, while
`those morpholinos which are inactive in the cell(cid:173)
`free assay (i.e., those targeted more than a few bases
`3' to the translational start site) are also inactive in
`cultured cells.
`It should be noted that the IC50 of morpholinos in
`the cell-free translation system are typically about 3-
`7-fold lower than the IC50 in scrape-loaded cells. We
`postulate that this reflects limited entry of oligos
`through the very small [30] transient [29] holes be(cid:173)
`lieved to be generated in the cell membrane during
`the scrape-load procedure.
`A possible basis for this relatively predictable tar(cid:173)
`geting of morpholinos, even in regions of quite stable
`secondary structure, may be their high affinity for
`RNA. Fig. 3 illustrates the comparative affinities of
`S-DNA, DNA, and morpholino 20-mer oligos for
`their complementary RNA (assay conditions detailed
`in [9]).
`I postulate that the apparent ability of long mor(cid:173)
`pholino oligos to effectively invade RNA secondary
`
`+-
`Fig. 2. Translational inhibition as a function of target position
`on HBV/luciferase mRNA. (a) Leader and translational start
`region of mRNA with target sequences for oligos 1- 7 indicated
`by numbered bold lines; (b) percent inhibition of luciferase syn-
`thesis in cell-free translation system by 1 µM morpholino oligos
`1- 12; (c) percent inhibition of luciferase synthesis in cultured
`HeLa cells expressing HBV/luciferase mRNA and scrape-loaded
`in the presence of 3 µM morpholino oligos I, 3, 4, 6, 8, 9, 11 ,
`and 12. The sequence of the HBV/luciferase mRNA from -85
`to +24 is : (- 85) 5'-GCAAC-UUUUUCACCU-CUGCCUAA-
`UC-AUCUCUUGUA-CAUGUCCCAC-UGUUCAAGCC-UC-
`CAAGCUGU-GCCUUGGGUG-GCUUUGGGGC-AUGGAC-
`CUCG-AGGACGCCAA-AAAC (+24). The 14 oligos used in
`this experiment are targeted against the following sequences in
`this mRNA: oligo I (-85 to -66); 2 (-73 to -53); 3 (-68 to
`- 44); 4 (- 41 to - 18); 5 (- 5 to +18); 6 (- 2 to +24); 7 (+3 to
`+24) ; 8 (+6 to +27); 9 (+9 to +30); 10(+12 to +33); 11 (+15
`to +36); 12(+18 to+39); 13 (+192 to +216); 14 (+528 to +553).
`Oligos 12, 13, and 14 showed no significant activity in either
`the cell-free or cultured-cell tests.
`
`

`

`J. Summerton/ Biochimica et Biophysica Acta 1489 (1999) 141- 158
`
`147
`
`0
`
`•
`
`•
`
`0
`
`•
`
`C:
`
`0 = ca
`
`.. :I -ca
`
`C:
`CD
`C
`~ 0
`
`100
`
`80
`
`60
`
`40
`
`20
`
`0
`
`cell-free test system (IC50 in the 10-30 nM range). In
`sharp contrast to the case for morpholinos and
`PNAs, the scrape-loaded S-DNAs typically show lit(cid:173)
`tle in-cell efficacy in our test system, and then only at
`concentrations typically over 3000 nM [18]. In fact,
`at low to moderate concentrations both antisense
`and control S-DNAs often strongly increase produc(cid:173)
`tion of the protein product of the targeted mRNA -
`possibly via activation of Sp 1 transcription factor
`[13] which could then generate a net increase in the
`targeted mRNA.
`Initially we suspected that perhaps the poor in-cell
`activity by scrape-loaded S-DNAs might be due to
`their multiple negative backbone charges preventing
`good cell entry during the scrape-load procedure.
`However, when
`fluorescein-labeled oligos were
`scrape-loaded into cells it was seen that this proce(cid:173)
`dure achieves delivery of S-DNAs as well as or better
`than delivery of morpholinos [ I 8].
`It seems possible that the poor in-cell efficacy we
`have seen with S-DNAs and the good in-cell efficacy
`of morpholinos might be at least in part a conse(cid:173)
`quence of the S-DNAs' sensitivity to nucleases [7]
`and the morpholinos' complete resistance to nucle(cid:173)
`ases [31 ].
`Another possible explanation for the apparent
`poor activity of S-DNAs in cells relates to RNase
`H. In our cell-free translation studies we add Esche(cid:173)
`richia coli RNase H (4 units/ml) because S-DNAs are
`only poorly active in reticulocyte lysates in the ab(cid:173)
`sence of added RNase H. Conceivably, in the HeLa
`cells, which we typically use for our in-cell studies,
`mammalian RNas