'! • . ·: ' )
`
`SEPTEMBER 15, 1993
`VOLUME 90
`NUMBER 18
`
`@) PROPERTY Of THE
`
`NATIONAL
`LIBRARY OF
`MEDICINE
`
`OF THE
`
`Proceedings
`National Acadeiny
`of Sciences
`
`OF THE UNITED STATES OF AMERICA
`
`

`

`Proceedings
`OF THE
`National Academy
`of Sciences
`OF THE UNITED STATES OF AMERICA
`
`Officers
`of the
`Academy
`
`Editorial Board
`of the
`Proceedings
`
`BRUCE ALDERTS, President
`JACK HALPERN, Vice Presidellt
`PETER H. RAVEN, Home Secretary
`JAMES B. WYNGAARDEN, Foreign Secretary
`MILDRED s. DRESSELHAUS, Treasurer
`
`LAWRENCE BOGORAD, Chairman
`JAMES E. DARNELL, JR.
`ERIC R. KANDEL
`IGOR B. DAWID
`PHILIP W. MAJERUS
`RONALD L. GRAHAM
`DAVID D. SADATINI
`GORDON G. HAMMES
`PAUL R. SCHIMMEL
`
`HERDERT A. SIMON
`MAXINE F. SINGER
`IRVING L. WEISSMAN
`SHERMAN M. WEISSMAN
`
`Senior Contributing
`Editors:
`
`Contributing Editors:
`
`Consulting Editors:
`
`MICHAEL T. CLEGG, JAMES E. DARNELL, JR., RAYMOND L. ERIKSON, RONALD M. EVANS,
`ERIC R. KANDEL, RICHARD A. LERNER, DAVID J. L. LUCK, PHIL W. MAJERUS,
`ARNO G. MOTULSKY, DAVID D. SADATINI, PAUL R. SCHIMMEL, ALLAN C. SPRADLING,
`JAMES A. SPUDICH, CHARLES F. STEVENS, IRVING L. WEISSMAN, SHERMAN M. WEISSMAN
`MAX D. COOPER, HERMANN. EISEN, WILLIAM E. PAUL
`JOSEPH L. GOLDSTEIN, RONALD GRAHAM, JACK HALPERN, JOANNE STUBDE
`
`Managing Editor:
`Senior Associate Editor:
`Associate Editors:
`
`FRANCES R. ZWANZIG
`GARY T. COCKS
`CAY BUTLER, JOHN M. MALLOY, MARILYN J. MASON, JANET L. MORGAN, T. PEARSON,
`DOROTHY P. SMITH, COLENE RUCH WALDEN
`Assistant Managing Editor: JOANNE D'AMICO
`Senior Production Editor: BARBARA A. BACON
`JAMIE M. FEAR, BILL FOGLE, ELLEN GoLDDERG, ScoTT C. HERMAN, AIJA OZOLINS,
`Production Editors:
`KATHLEEN RUDY, ANNE M. SUNDERMANN, DON C. TIPPMAN
`MARY E. McLAUGHLIN
`DELORES BANKS, BRENDA L. McCoY
`JACQUELINE PERRY
`JULIA LITTLE, CYNTHIA MATHEWS
`
`Proofreader:
`Administrative Assistants:
`Manuscript Coordinator:
`Circulation :
`
`Editorial Correspondence: PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES, 2101 Constitution Avenue, Wash(cid:173)
`ington, DC 20418.
`Business Correspondence: Circulation Office of the PROCEEDINGS, National Academy of Sciences, 2101 Constitution
`Avenue, Washington, DC 20418.
`Information for Contriblllors: See p. i (of this issue) and pp. i-viii of Issue Number 1, January l, 1993.
`Copyright: Volume 90, copyright © 1993 by the National Academy of Sciences; Volumes 1-89, copyright as a collective
`work only with copyright to individual articles retained by the author(s). Requests for permission to reproduce parts of
`individual articles published in Volumes 1-89 or for reprints of those articles should be addressed to the authors. Microforms
`of complete volumes are available to regular subscribers only and may be obtained from University Microfilms, Xerox
`Corporation, Ann Arbor, MI 48103. This journal is printed on acid-free paper effective with Volume 84, Issue 1.
`Subscriptions: All correspondence concerning subscriptions should be addressed to the Circulation Office of the PROCEED(cid:173)
`INGS. Subscriptions are entered on a calendar year basis only. For 1993, subscription rates are as follows-in the United States:
`personal, $230; institutional, $420; elsewhere by surface mail: personal, $325; institutional, $515; elsewhere by Air Cargo at
`a surcharge of $192. Information regarding other air mail postage rates is available from the Circulation Office. Subscribers
`are requested to notify the Circulation Office of the PROCEEDINGS 6 weeks in advance of any change of address ; also the local
`postmaster. The Academy is not responsible for nonreceipt of issues because of an improper address unless a change of address
`is on file. The notice of address change should list both the old and new addresses. Claims for replacement copies will not
`be honored more than 60 days after the issue date for domestic subscribers and not more than 90 days after the issue date
`for foreign subscribers.
`Single Copies: The price of a single issue is $25.00.
`PRINTED IN THE USA
`S econd class postage paid at Washington, DC, and at additional mailing offices.
`PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA (lSSN-0027-8424) is
`published semimonthly by THE NATIONAL ACADEMY OF SCIENCES, 2101 Constitution Avenue, Washington, DC 204/8.
`«'.J 1993 by THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA.
`POSTMASTER: Send address changes to: PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED
`STATES OF AMERICA, 2101 Constitution Ave., Washington, DC 20418.
`
`

`

`This material may be protected by Copyright law (Title 17 U.S. Code)
`
`Proc. Natl." Acad. Sci. USA
`Vol. 90, pp. 8673-8677, September 1993
`Biochemistry
`
`Restoration of correct splicing in thalassemic pre-mRNA by
`antisense oligonucleotides
`
`2BIGNIEW DOMINSKI AND RYSZARD KOLE*
`Department of Pharmacology and Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill, NC 27599
`
`Communicated by Sidney Altman, June JO, 1993 (received for review April 22, 1993)
`
`Antisense 2'~0-methylribooligonucleotides
`ABSTRACT
`were targeted against specific sequence elemen~ !n mutated
`human P-globin pre-mRNAs to restore correct sphc1~g of these
`RN As in vitro. The following mutations of the /J-globm gene, A
`- G at nt 110 of the first intron ({1110), T - G at nt 705 and C
`- T at nt 654 of the second intron (IVS2705 and IVS2654,
`respectively), which led to aberrant splicing ofthecorr~sponding
`pre-mRNAs, were previously identified as the underlymg causes
`of P•thalassemia. Aberrant splicing of {1110 pre-mRNA was
`efficiently reversed by an oligonucleotide targeted against the
`branch point sequence in the first intron of the pre-mRNA but
`not by an oligonucleotide targeted against the aberrant 3' splice
`site, In both 1vs210s and IVS2654 pre-mRNAs, correct splicing
`was restored by oligonucleotides targeted against the aberrant 5'
`splice sites created by the mutations in the second intron or
`against a cryptic 3' splice site located upstream and activated in
`the mutated background. These experiments represent an ap(cid:173)
`proach in which antisense oligonucleotides are used to restore
`the function of a defective gene and not, as usual, to down(cid:173)
`regulate the expre~ion of an undesirable gene.
`
`The potential of oligonucleotides as modulators of gene ex(cid:173)
`pression and as chemotherapeutic agents is currently under
`intense investigation. Rapidly accumulating literature has
`been surveyed in a number of recent reviews (1-4). The
`activity of antisense oligonucleotides generally relies on their
`hybridization with targeted RNA, leading either to its degra(cid:173)
`dation by cellular RNase Hor to a block in its translation (1-4).
`In another approach, so-called antigene oligonucleotides are
`directed against specific regions in DNA, where they form
`triplex structures and inhibit transcription by RNA polymer(cid:173)
`ase II (5, 6). Both antisense and antigene oligonucleotides lead
`to down-regulation of targeted genes and are applied to reduce
`the intracellular level of undesirable gene products coded by
`viruses (7, 8), other pathogens (9), or oncogenes (2). This is
`difficult to accomplish, especially if the targeted gene codes for
`a protein that is stable and has a low turnover rate.
`In the approach reported here, antisense oligonucleotides
`are designed to generate a correct gene product that had been
`rendered defective by mutations that changed the splicing
`pattern of the corresponding pre-mRNA. We have used oli(cid:173)
`gonucleotides complementary to specific sequence elements
`in pre-mRNAs coded by mutated {3-globin alleles identified in
`various cases of /3-thalassemia. In this model system of
`potential clinical significance (10), the oligonucleotides sup(cid:173)
`pressed aberrant splicing pattern of /3-globin pre-mRNA re(cid:173)
`sulting from gene mutations and restored correct splicing.
`
`MATERIALS AND METHODS
`All pre-mRNAs were transcribed by SP6 RNA polymerase
`(11) from appropriate fragments of the human /3-globin gene
`subcloned into the SP64 vector. HB.M (12) contains the
`
`The publication costs of this article were defrayed in part by page charge
`payment. This article must therefore be hereby marked "advertisement"
`in accordance with 18 U.S.C. §1734 solely to indicate this fact.
`
`whole human /3-globin gene, consisting of three exons and
`two introns. The 13no construct, carrying an A- G mutation
`at position 110 of the first intron, was obtained by subcloning
`a fragment from the original thalassemic clone (ref. 13; P. J.
`Furdon and R.K., unpublished data) into HBM plasmid. To
`construct clones containing thalassemic mutations in the
`second intron, a fragment of human {3-globin gene containing
`virtually the entire second exon, the entire second intron, and
`a major portion of the third exon was subcloned into the SP64
`vector. The resulting IVS2 clone was subsequently subjected
`to site-specific mutagenesis (14). A mutation of T- G at nt
`705 of the intron was introduced to construct the IVS2705
`clone, and a c- T mutation was introduced at nucleotide 654
`to construct the IVS2654 clone. Further details of the con(cid:173)
`s~ruction are available upon request. Transcription was ear(cid:173)
`ned out on a plasmid linearized at the BamHI site for the 13no
`clone or at the Pvu II site for the IVS2705 and IVS2654 clones.
`The oligonucleotides were synthesized at the Lineberger
`Comprehensive Cancer Center using reagents from Glen
`Res~arch (Sterling, VA) and purified using SurePure kit
`(Umted States Biochemical). The concentration of the oli(cid:173)
`gonucleotides was measured spectrophotometrically at 260
`nm. The sequence of the oligonucleotides is shown in Fig. lB.
`Nuclear extract from HeLa cells was prepared as described
`(12) except that the pellet of the nuclei was initially suspended
`in 0.5 vol of buffer C [20 mM Hepes , pH 7.9/25% (wt/vol)
`glycerol, 1.5 mM MgCh/0.2 mM EDTA/0.5 mM dithiothrei(cid:173)
`tol/20 mM KCl] and subsequently supplemented with an(cid:173)
`other 0.5 vol of buffer C containing 1.2 M KCl. Splicing of
`32P-labeled pre-mRN As ( = 105 cpm per reaction, 25 fmol) was
`carried out in the nuclear extract for 2 hr at 30°C in a reaction
`volume of 25 µ.I (12, 15). Antisense oligonucleotides were
`added together with the other components of the splicing
`reaction. Reaction products were analyzed on an 8% poly(cid:173)
`acrylamide sequencing gel and visualized by autoradiogra(cid:173)
`phy.
`All autoradiograms were captured by a Dage-MTI CCD-72
`video camera (Michigan City, IN), and the images were
`processed using National Institutes of Health IMAGE 1.43 and
`MACDRA w PRo 1.0 software. The final figures were printed
`out on Sony dye sublimation printer.
`
`RESULTS AND DISCUSSION
`Restoration of Correct Splicing in Intron 1 of Human
`P•Globin Pre-mRNA. In {3110-thalassemia, a form of the
`disease predominant in thalassemic individuals of Greek and
`Cypriot origin (10, 16), an A- G mutation at nt 110 of the first
`intron of the human /3-globin gene creates an additional,
`aberrant 3' splice site (13). In spite of the presence of the
`normal 3' splice site, the aberrant site is preferentially used
`by the splicing machinery, resulting in an incorrectly spliced
`mRNA that contains 19 nt of the intron sequence (Fig. l) . .In
`cells transfected with the {3110-globin allele, correctly spliced
`mRNA constitutes only about 10% of the spliced product (17,
`
`*To whom reprint requests should be addressed.
`
`Th is mate ri,a I wai611l ed
`
`

`

`8674
`
`Biochemistry: Dominski and Kole
`
`18), in agreement with the markedly reduced levels of normal
`hemoglobin observed in patients with this form of thalas(cid:173)
`semia (10). It was found that in 13110 pre-mRNA the aberrant
`3' splice site utilizes the normal branch point at nt 93 of the
`intron, competing with the correct 3' splice site and thereby
`reducing correct splicing (19). Significantly for this work,
`mutations inactivating this element activate a cryptic branch
`point at nt 107 and result in splicing at the correct 3' splice site
`(20). Aberrant splicing cannot proceed in the mutated back(cid:173)
`ground due to the proximity of the 3' splice site at position 110
`to the cryptic branch point. We therefore reasoned that
`antisense oligonucleotides complementary to the normal
`branch point sequence will, like mutations, block its function
`in splicing and force the splicing machinery to select the
`cryptic branch point, generating correctly spliced mRNA.
`To test this idea, a 14-mer 2'-O-methyloligonucleotide
`(oligonucleotide 1, Fig. 1) was targeted against the branch
`point sequence in intron 1 of (3-globin pre-mRN A. The branch
`point nucleotide is located at position 94 of the intron, and the
`oligonucleotide spans positions 82-95. The 2'-O-methyloli(cid:173)
`gonucleotides were selected for this and subsequent exper(cid:173)
`iments because they are resistant to nucleases and form
`A
`HBLl6
`
`110
`
`654 705
`
`130 222
`
`850
`
`241
`
`BP 110
`t
`t
`___ ....... _.""""'-...,. BamHI
`1 2 .. 2
`
`---
`
`p110
`
`rvs2705
`
`er. 3' 705
`t
`t
`
`_____ , ; ............
`--· ...
`
`rvs2654
`
`3 4
`
`--
`
`er. 3' 654
`___ , , ......
`--• ...
`t t
`....
`,
`3 5
`
`Pvull
`3
`
`Pvull.
`3
`
`B
`Ollgonueleotldes:
`1.GUCAGUGCCUAUCA
`2.AUAGACUAAUAGGC
`3.CAUUAUUGCCCUGAAAG
`4.CCUCUUACCUCAGUUAC
`5.GCUAUUACCUUAACCCAG
`
`Proc. Natl. Acad. Sci. USA 90 (1993)
`
`stable hybrids with RNA that are not degraded by RN_as~~
`(2~-23). Degradation by RNase H, seen _when a~tl:~ves
`ohgodeoxynucleotides or their phosphoroth1oate d~A and
`are used (1-4), would destroy the substrate pre-m
`in
`prevent splicing. The oligonucleotides were added to ~A
`vitro splicing reaction containing 32P-labeled pre-m cells,
`(=25 fmol, 0.5 nM) and a nuclear extract from HeLa_d
`el
`Analysis of splicing of 13110 pre-mRNA by polyacr~lamt e g rt
`electrophoresis shows that, consistent with a prevto~s re~?
`
`(19), in the control reaction without the oligonucleottde ~- :a
`
`2, lane 2) the ratio of the incorrectly to correctly sp/c M
`products is =9:1. Addition of oligonucleotide 1 at 0.0S- µ t
`.
`. h'b' • of aberran
`t
`concen rations causes dose-dependent m 1 1t1on
`te
`splicing and induction of correct splicing of the subStral
`•
`r nuc e-
`(F!g. 2, lanes 3-6). At a concentration of 5 µ,M, o igo
`tly
`ott~e 1 reverses the ratio of the incorrectly to _correch _
`sp!•~ed products to 1:5. Apparently, oligonucleoti~e .1 ~f
`bndtzes to the normal branch point and prevents bmdmg
`the splicing factors to this sequence element, forcing them t~
`oli~onucleotide 1 must be sequence specific, ~ince it t
`select the cryptic branch point downstream. The effe~t ?
`
`unhkely that its hybridization with any other site on t e
`pre-mRNA would promote the observed results (see below).
`Sequence specificity of oligonucleotide 1 is further demon-
`
`Oligo 1
`
`~ e
`
`.,,
`
`al.- OLO
`:C:C(!,,OON11>
`
`339*-r-:==~=, - .-om
`
`. '
`;
`: o:J-r{II
`493 - . . . . . . . . -a-'
`, ~ , w· ~ m-m
`\_ ' •--- • Z.,.: .;:_1 aberrant
`1
`- - - - - - - ,
`1 I 2]
`' ~ ,
`·
`, correct
`•
`
`382
`
`363
`
`~240
`
`130* -
`
`.
`
`j 1
`
`'
`
`. ' ·1
`I
`
`2
`
`3
`
`4
`
`5
`
`Fm. 1.
`(A) Structure of pre-mRNAs. Boxes indicate exons;
`heavy lines indicate introns. Positions of the mutations (110 or 654
`and 705) relative to nt 1 of intron 1 and intron 2, respectively, are
`6
`7
`shown above the HB~6 clone. Numbers below indicate the length,
`in nucleotides, of exons and introns. Antisense oligonucleotides are
`Fm. 2. Re~ersal of aberrant splicing by oligonucleotide 1 ~i-
`indicated by the numbered short bars below 13110, IVS2654 , and
`rected against the normal branch point in intron 1 of f:3-globm
`IVS2705 constructs, and splicing pathways are indicated by the
`pre-mRNA. In this and subsequent experiments, in vitro splicing was
`dashed lines. (B) Sequence ofantisense 2'-O-methyloligoribonucleo•
`carried out as described in Materials and Methods. Reaction prod-
`tides. Oligonucleotide l, complementary to nt 82-95 of intron 1, is
`ucts were analyzed on an 8% polyacrylamide sequencing gel and
`targeted against the normal branch point (BP); oligonucleotide 2 (nt
`visualized by autoradiography. The structure of the produc~s an_d
`103-116) is targeted against the aberrant 3' splice site created by 13110
`intermediates is depicted on the right; their size in nucleol!des 1s
`mutation in intron 1 of the {:3-globin gene. Oligonucleotide 3, com-
`shown on the left. An asterisk denotes aberrant mobility of lariat-
`plementary to nt 573-589 of intron 2, is targeted against the cryptic
`containing intermediates. The same designations are used in subse-
`3' (er. 3') splice site at nt 579 of the second intron; oligonucleotide
`quent figures. Lane 1, splicing of control HB~6 pre-mRNA. Lane!•
`4 (nt 697-713) is targeted against the aberrant 5' splice site created by
`splicing of 13110 pre-mRNA. Lanes 3-6, splicing of 13110 pre-mRNA m
`the mutation at nt 705 in the second intron of IVS2705 pre-mRNA;
`the presence of increasing concentrations (µ.M; indicated at the top
`oligonucleotide 5 (nt 643-660) is targeted against the aberrant 5'
`of the figure) ofoligonucleotide 1. Lane 7, splicing of {3110 pre-mRN A
`splice site created by the mutation at nt 654 in the second intron of
`in the presence of oligonucleotide 3, targeted to a sequence in intron
`IVS2654 pre-mRNA.
`Tl1ismaterial wasJ:opiMlobin pre-mRNA.
`at. the Nl JM .and m ay b e
`:.uibject US Co;pyright Laws
`
`

`

`Biochemistry: Dominski and Kole
`
`Proc. Natl. Acad. Sci. USA 90 (1993)
`
`8675
`
`Oligo2
`
`,.,:,-1.'
`
`., ,
`
`strated by the fact that at 5 µ,M an oligonucleotide targeted
`against the cryptic 3' splice site in the second intron of the
`/3-globin gene (oligonucleotide 3, Fig. 1; see also below) does
`not affect the original ratio of the spliced products (Fig. 2,
`lane 7). At higher concentrations (10-20 µ,M), oligonucleotide
`1 inhibits splicing in an unspecific manner and generates
`greater amounts of an =240-nt RNA species, which is de(cid:173)
`tectable in small quantities at a 5 JLM concentration of
`oligonucleotide 1 (Fig. 2, lane 6). This product accumulates
`only in the presence of ATP and other components of the
`splicing mixture, and its length is consistent with cleavage of
`the RNA at the site of hybridization with the oligonucleotide.
`These results suggest the presence of an ATP-dependent
`nuclease that either recognizes RNA-2'-O-methyloligonucle(cid:173)
`otide duplexes and cleaves the RNA in an endonucleolytic
`manner or degrades RNA exonucleolytically and is blocked
`at the site of duplex formation.
`The aberrant 3' splice site generated by the 13110 mutation
`at position 111 appears to be an obvious target for reversal of
`aberrant splicing by an antisense oligonucleotide. Blocking
`this sequence element should be the simplest way of forcing
`the splicing machinery to use the original 3' splice site at the
`end of the intron. However, a 14-mer (oligonucleotide 2, Fig.
`1) directed against the aberrant splice site, spanning nt
`103-116 in the intron, was not effective; at increasing con(cid:173)
`centrations of oligonucleotide 2, accumulation of both spliced
`products was inhibited; the correct one was inhibited some(cid:173)
`what more efficiently (Fig. 3, lanes 2-5). Interestingly, the
`first step of the splicing reaction, cleavage at the 5' splice site
`and formation of the lariat-exon intermediate, seems to be
`less affected by oligonucleotide 2 than the formation of the
`final spliced product. This is shown by the presence of the
`intermediates generated during this step even at 5 and 10 µ,M
`concentrations of oligonucleotide 2 (Fig. 3, lanes 5 and 6). At
`higher concentrations of the oligonucleotide, cleavage at the
`5' splice site was also inhibited (data not shown).
`The control transcript (IVS2) containing the second intron
`The lack of success with oligonucleotide 2 may result from
`of_ normal {3-globin pre-mRNA is spliced efficiently, gener-
`the proximity of the aberrant and correct 3' splice sites and,
`ati?g the expected intermediates (the 5' exon and the large
`in consequence, the concomitant interference of this oligo-
`lanats) and a 451-nt correctly spliced product (Fig. 4, lane 1).
`nucleotide with both splicing pathways. Note that in 13110
`Note that the splicing of long transcripts and the removal of
`pre-mRNA the sequence elements important for splicing
`long introns have been difficult to achieve in vitro. The high
`(i.e., the two 3' splice sites, two branch point elements, and
`efficiency of the reaction seen here and in the following
`the polypyrimidine tract) are located within the stretch of 37
`experiment is probably due to the improvement in the prep-
`nt. Conceivably, hybridization of oligonucleotide 2 in the
`aration of the nuclear splicing extracts. Splicing of 1vs2105
`middle of this region interferes with binding of a large number
`pre-mRN A is also efficient and yields an additional spliced
`of splicing factors assembling there and prevents any splicing
`product 577 nt long as well as the expected 722- and 348-mer
`by steric hindrance. Consistent with this idea, additional
`intermediates, resulting from the aberrant splicing pathway
`experiments showed that any oligonucleotide targeted down-
`caused by the mutation (Fig. 4, Jane 2). The 1:2 ratio of
`stream from the normal branch point and upstream from the
`correctly to incorrectly spliced RNAs is similar to that
`correct 3' splice site inhibited splicing without restoring the
`observed in vivo (24). Oligonucleotide 3, spanning positions
`correct pathway (results not shown). Note also that oligo-
`573-589 of intron 2 (Fig. 1) and complementary to the
`nucleotide 2, as well as some other oligonucleotides targeted
`activated cryptic 3' splice site at nt 579, is very effective; it
`to this region, block a significant portion of the polypyrimi-
`•
`dine tract, which is essential for splicing.
`mduces a dose-dependent reversal of splicing to the correct
`Restoration of Correct Splicing in lntron 2 of the Human
`splicing pathway (Fig. 4, lanes 3-5). There is a significant
`P-Globin Gene. Whether an aberrant 3' splice site can nev-
`accumulation of the correctly spliced product at a 0.12 JLM
`ertheless be used as a target for reversal of incorrect splicing
`concentration of the oligonucleotide (Fig. 4, lane 3); at 0.5
`was further tested on pre-mRNA carrying a T- G mutation
`and 2 JLM the restoration of correct splicing is virtually
`at position 705 of the second intron of human 13-globin gene.
`complete. Correct splicing is also completely restored at
`This rare mutation (IVS2705), found in Mediterranean
`similar concentrations of oligonucleotide 4 (Fig. 1), spanning
`thalassemia patients (10, 16), creates an additional, aberrant
`positions 697-713 of the intron and targeted against the
`5' splice site and activates a cryptic 3' splice site at position
`aberrant 5' splice site created by the mutation at nt 705 (Fig.
`579 of the intron (24). The incorrect splicing pathway results
`4, lanes 6-8). At higher concentrations of this oligonucleotide
`in the removal of nt 1-578 and 706-850 as separate introns
`(5 and 10 µ,M) , correct splicing remains very efficient;
`and incorporation of the remaining portion of the intron into
`however, at 20 JLM, unspecificinhibition of splicing occurred.
`the spliced product (Fig. 1). In IVS2705 pre-mRNA the
`A control experiment, in which splicing ofIVS705 pre-mRNA
`distances between each of the sequence elements involved in
`is carried out in the presence of 5 JLM oligonucleotide 1,
`splicing exceed 100 nt, and no steric hindrance effects caused
`Th is. m ateria I w as. stmwil no effect on the ratio of correctly to incorrectly spliced
`by the oligonucleotide would be expected.
`3t t:he• N LM .and may Ire
`:.ubject US Copyright Laws
`
`2
`
`3
`
`4
`
`5
`
`6
`
`Fm. 3. Effects ofoligonucleotide 2, directed against the aberrant
`3' splice site in intron 1 of 13110 pre-mRNA. Lane 1, splicing of 13110
`pr~-mRN~. Lanes 2-6, splicing of 13110 pre-mRNA in the presence
`of m~reasmg concentrations (µM; indicated at the top of the figure)
`of ohgonucleotide 2.
`
`

`

`8676
`
`Biochemistry: Dominski and Kole
`
`U)
`
`0 ...
`
`N
`N
`Cl)
`Cl)
`~ ~
`
`1019•~, .. , .. . , ·
`
`850*
`
`Ollgo3
`
`Ollgo4
`
`N
`~ U)
`C) ci N
`
`N
`~ U)
`C) C) N
`
`JJM
`
`• ~om
`
`Q_
`
`1301 - - ; ........ ..,, - -.. ,-II}-rrill
`-
`577 -1 -
`, .... ~ ---- ..... ----[]]]]
`
`~ .......
`
`m-,--rn
`-
`~
`- a
`t
`
`correct
`
`'W,
`
`722-
`
`451 -,
`
`I
`348 - 1
`
`Proc. Natl. Acad. Sci. USA 90 (1993)
`
`0
`
`.,,
`
`.,.
`.,,
`'°
`...
`"' "'
`2 2
`1019•F•Cll!li
`
`(/)
`
`(/)
`
`Oligo 3
`
`Oligo 5
`
`U) " '
`
`:::; ci ci "' µM
`
`--•-10QJ
`
`Q_
`
`- . -....... ~ [D,-r-GJ
`
`...................
`
`-
`
`•
`
`"'i'[TirU]
`'J mm
`·•·--~
`
`J aberrant
`
`correct
`
`850'
`
`1301 -
`
`.
`
`526 -
`451 -
`
`222---, - - - ----w
`
`2 3 45 6 7 8
`
`222 -...,., ........ _
`2
`
`.._ .......... - - - - ~ ........ ------ru
`
`3 4
`
`5 6
`
`7
`
`8
`
`9 10 11
`
`Fm. 4. Reversal of aberrant splicing of IVS2705 pre-mRNA by
`oligonucleotide 3, directed against the aberran~ 3' _spli~e ~it~, and
`oligonucleotide 4 directed against the aberrant 5 sphce site m mtron
`2 of the 1vs210s pre-mRNA. Lanes 1 and 2, splicing of control
`transcripts (indicated at the top of the figure). Lanes 3-5 and 6-8,
`splicing of 1vs21os pre-mRNA in the presence of oligonucleotide 3
`and oligonucleotide 4, respectively. The concentrations of the oli(cid:173)
`gonucleotides in the reaction (JLM) are indicated at the top of the
`figure. The band at the top of the gel contains unresolved interme(cid:173)
`diates and products containing large lariats. The numbers 1079 and
`850 represent the number of nucleotides in the lariat and lariat-exon
`generated by the correct splicing pathway. The RNA band at 722 nt
`contains an intermediate of the aberrant splicing pathway with the
`upstream, aberrant intron already removed. After cleavage at the
`aberrant 5' splice site, its size is reduced to 348 nt.
`
`FIG. 5. Reversal of aberrant splicing of IVS2654 pre-mRNA by
`oligonucleotide 3, directed ag~inst the aberrant 3' splice site, and
`oligonucleotide 5, directed agamst the aberrant 5' splice site in intron
`2 of the IVS2654 pre-mRNA. Lanes 1 and 2, splicing of control
`transcripts (indicated at the top_ of the figure). Lanes 3-7 and 8-11,
`splicing of IVS2654 pre-mRNA_ m the presence of oligonucleotide 3
`and oligonucleotide 5, respectively. The concentrations of the oli(cid:173)
`gonucleotides in the reaction (JLM) are indicated at the top of the
`figure. Other symbols are as described in the legends to the previous
`figures.
`
`three oligonucleotides are several times more effective than
`oligonucleotide 1 used in the experiments shown in Fig. 2.
`The higher efficacy may be due to the fact that oligonucle(cid:173)
`otides 3-5 are 3 nt longer than oligonucleotide 1 and may form
`products (data not shown). This is further evidence of the
`more stable hybrids with RNA. They also block aberrant
`high specificity of oligonucleotides used in these studies.
`splice sites, allowing the splicing machinery to use the correct
`The IVS2654 mutation (25), frequently identified in
`splice sites and, presumably, the correct branch point. In
`contrast, in /3110 pre-mRNA, oligonucleotide 1 forces the
`thalassemic individuals of Chinese origin (16), affects splicing
`via a mechanism analogous to that operating in the IVS705
`splicing machinery to use a suboptimal cryptic branch point
`sequence, which may result in relatively inefficient genera-
`mutant-i.e., by creating an additional 5' splice site at nt 653
`and activating the common cryptic 3' splice site at nt 579 of
`tion of correctly spliced mRNA.
`intron 2. Splicing patterns of both mutated pre-mRNAs are
`In the experiments presented above, the oligonucleotides
`presented side by side in Fig. 5, lanes 1 and 2. Different
`were added simultaneously with the other components of the
`lengths of incorrectly spliced products and of some of the
`splicing reaction. Prehybridization of the oligonucleotides
`intermediates reflect the different positions of aberrant 5'
`with the pre-mRNA did not increase their efficiency, and
`splice sites in IVS26s4 and IVS705 mutants. The efficiency of
`oligonucleotides added 15 min after the start of the reaction
`aberrant splicing of IVS2654 pre-mRNA is higher than that for
`[i.e., after splicing complexes had a chance to form (26)] were
`IVS2705 pre-mRNA, and only small amounts of correctly
`almost as effective (data not shown). These results indicate
`that oligonucleotides containing the 2'-O-methyl modifica-
`spliced product, relative to the aberrant one, are detectable
`during splicing in vitro (Fig. 5, lane 2). In spite of the high
`tion are able to compete effectively for their target sequences
`efficiency of aberrant splicing, oligonucleotide 3, targeted
`with the splicing factors. The high efficacy of these com-
`pounds is most likely due to their strong hybridization with
`against the cryptic 3' splice site, as well as oligonucleotide 5,
`RNA (21-23). In other experiments they also affected splice
`spanning nt 643-660 and targeted against the aberrant 5'
`site selection and/or efficiently inhibited pre-mRNA splicing
`splice site (Fig. 1), restored correct splicing efficiently at
`(27), whereas methylphosphonate oligonucleotides, which
`concentrations similar to those used in the preceding exper-
`iment (Fig. 4). At a 2 µ,M concentration of either oligonucle-
`form less stable hybrids, were much less effective (ref. 28;
`P. J. Furdon and R.K., unpublished results).
`otide, the correctly spliced product accumulates, and the
`The possibilities of alteration of splicing pathways by
`aberrant product is virtually undetectable (Fig. 5, lanes 7 and
`antisense oligonucleotides are not limited to the mutations
`11, respectively).
`The results presented in Figs. 4 and 5 demonstrate that
`tested above. Since it is estimated that up to 15% of all point
`aberrant 3' and 5' splice sites provide suitable targets for
`mutations in genetic diseases results in defective splicing
`specific reversal of incorrect splicing. Similar effects of
`(29), the approach described here may have potentially wide
`oligonucleotides 3, 4, and 5 suggest that there are no major
`application. Antisense oligonucleotides may be useful in
`differences in their accessibilities to the target splice siwi..4ilkrial w.r~l:i~~ion of correct splicing in other thalassemic mutations
`at the NLM and m ay be
`5'ubject US Co pyright Laws
`
`

`

`Biochemistry: Dominski and Kole
`
`(10) and in genetic disorders, such as Tay-Sachs ~yndro~e
`(30), phenylketonuria (31), and certain forms of cystic_ fi_bros1s
`(32), in which mutations leading to aberrant sphcmg of
`pre-mRNA have been identified. They may also p~omote
`reactivation of cancer suppressor genes that are defective due
`. ..
`to mutations affecting pre-mRNA splicing (33, 34).
`It is of interest that oligonucleotide 3 was able to mh1b1t
`aberrant splicing caused by two different mutations._ Th~re
`are a number of thalassemic mutations that lead to activation
`of a single cryptic splice site in exon 1 of the human {3-gl_o~in
`gene (10); therefore, it seems likely that a?errant sphcmg
`could be corrected with a single oligonucleottde compler_nen(cid:173)
`tary to this site. This property would be ver~ useful tf an
`antisense approach were applied for therapeutic purposes.
`The next logical step in this approa~h is t~ d_emonstrate th~t
`oligonucleotides can efficiently modify sphcmg pathways m
`vivo. A number of studies showed that transmembrane trans(cid:173)
`port of oligonucleotides is possible and can be significantly
`improved by various modifications of thesi: co~pounds,
`affecting their chemical properties (1-4). Spltce sites have
`been shown to provide effective targets for inhibition of gene
`expression by antisense oligonucleotides (35-37). A recent
`report indicates that the use of a lipofection procedure
`dramatically increases delivery of the oligonucleotides to the
`cells and leads to their increased accumulation in the nucleus,
`the site of splicing (38). These results suggest that the
`approach reported here will also be successful in vivo.
`In this work the antisense oligonucleotides were used to
`generate correctly spliced mRNA from an aberrant ~re(cid:173)
`mRNA substrate in vitro. It is important to note that ohgo(cid:173)
`nucleotides 1 and 3 were able to elicit a desired effect by
`antisense interaction at sequences distant from the actual
`mutations. These results point to the usefulness of the in vitro
`splicing system for rapid screening of suitable target se(cid:173)
`quences, thus narrowing down the number of oligonucleo(cid:173)