,
`
`lo..
`
`W.l M0194U
`v.14
`NO. 1 l,
`1994
`c. 02---:-~----irno: M43290000
`Tl: MOLECULAR AND CELLULAR
`BIOLOGY
`10/2 0/94
`
`,,
`
`~
`
`\..
`
`.
`
`@) NATIONAL
`
`PROPERTY Of THE
`LIBRARY OF
`MEDICINE
`
`"
`
`,
`
`1~£:5lx:/iTIJViiB)H\~ 'utf9;/t\
`\J©!i[!_~~) fo'iift 1/\:r
`[~Q!JAfu[I:'II[~: 111,
`
`

`

`MOLECULAR AND CELLULAR BIOLOGY
`
`Alan M. Weiner, Editor in Chief (1995)
`Yale U11ive1;1·ity
`James C. Alwine, Editor (1998)
`University o f Penmylvania School o f
`Medicine
`James R. Broach, Editor (1995)
`Princeton University
`Nouria Hernandez, Editor (1998)
`Cold Spring Harbor la/Joratmy
`Anita K. Ho1>per, Editor (1999)
`77-ic Milton S. Hershey Medical Center
`J>enn.1ylvania State University
`
`VOLUME 14 • NOVEMBER 1994 • NUMBER 11
`Tom Kadesch , Editor ( 1996)
`University o f Pen11.1ylvania School
`o f Medicine
`Robert E. Kingston, Editor (1997)
`Massachu setts General Hospital
`Adrian R. Krainer, Editor (1998)
`Cold Spring Harbor laboratory
`Douglas Lowy, Editor (1998)
`National Institutes o f Health
`Charles J. Sherr, Editor (1995)
`St. Jude Children's Research Hospital
`Nahum Sonenberg, Editor (1998)
`McGill University
`
`George F. Sprague, Jr., Editor (1998)
`University o f Oregon
`Jonatha!l R._Warner, Editor (1998)
`Albert Emstem College o f Medicine
`Mi~hae~ J. We~e~, _Editor (1997)
`Umvers1ty 1 Virgmza Medical
`Schoo
`Elt?n 1:· Young, Editor (1998)
`Umvers1ty o f Washington
`Kenneth S. Zaret, Editor (1998)
`Brown University
`
`Naomi E. Rosenberg (1994)
`Nadia Rosenthal (1995)
`Matthew P. Scott (1995)
`Bart Sefton (1995)
`Harincler Singh (1995)
`Stephen Smale (1994)
`Frank Solomon (1994)
`Allan C. Spradling (1994)
`Karen U. Sprague (1995)
`Nat L. Sternberg (1995)
`Tom H. Stevens (1995)
`Bruce Stillman (1994)
`Kevin Struhl (1995)
`nm Sugden (1994)
`Kelly G. Tatchell (1995)
`Lawrence H. Thompson
`(1994)
`Lex H. T. van cler Ploeg (1994)
`Geoffrey Wahl (1995)
`Randolph Wall (1995)
`Peter Walter (1996)
`Jean Y. J. Wang (1996)
`Lewis T. Williams (1994)
`Fred Winston (1994)
`Richard Young (1995)
`Virginia A. Zakian (1995)
`Richard S. Zitomer (1994)
`
`EDITORIAL BOARD
`Monty Krieger (1995)
`Scott Emr (1994)
`Elisabeth Lacy (1996)
`l <'rcderick W. Alt ( J 996)
`Douglas Engel (1995)
`Alan Lambowitz (1996)
`Albert Baldwin (1996)
`Ileverlv Errede (1996)
`Lester F. Lau (1995)
`Claudio nasilico (J 994)
`Stanley Fields (1994)
`Michael Lenardo (1994)
`Susan Herget (1996)
`Thomas U. Fox (1996)
`Arthur D. Levinson (1996)
`Eli:r..abeth H. lllackbum (1994)
`Bruce Futchcr (1995)
`David Levy (1995)
`Helen M. Blau (1994)
`Michael Gilman (1995)
`John T. Lis (1995)
`Stephen P. Goff (1995)
`Jef D. Boeke (1994)
`James Manley (1995)
`Paula J. Grabowski (1994)
`Joseph Holen (1996)
`Kunihiro Matsumoto (1994)
`Jack F. Greenhlatt (1994)
`Allan Bradley (1996)
`Frank McCormick (1995)
`Marjorie C. Hrandriss (1994)
`Rudolf Grosschecll (1995)
`N. Ronald Morris (1995)
`Rodrigo Bravo (1995)
`Mark Groucline (1996)
`Paul Neiman (1995)
`John n. E. Burch (1994)
`Joyce L. Hamlin (1996)
`Joseph R. Nevins (1996)
`Hidesaburo Hanafusa (1995)
`Kathryn Calame (1994)
`Carol Newlon (1994)
`Judith Campbell (1995)
`Nathaniel Heintz (1994)
`Eric N. Olson (1995)
`John A. Carbon (1996)
`Steven Henikoff (1994)
`Harvey L. Ozer (1994)
`Marian Carlson (1995)
`Alan Hinnehusch (1994)
`Mary Lou Pardue (1994)
`Nam-Hai Chua (1994)
`Peter M. Howley (1994)
`David Patterson (1994)
`Charles Cole (1996)
`Stephen H. Hughes (1995)
`Marvin R. Paule (1996)
`Jonathan A. Cooper (1995)
`James N. Ihle (1995)
`Tony Pawson (1994)
`Gerald R. Crabtree (1996)
`Alexander Johnson (1996)
`Scott Powers (1994)
`Elizaheth A. Craig (1994)
`Mark Johnston (1994)
`John R. Pringle (1994)
`Thomas Curran (1994)
`Katherine A. Jones (1995)
`Randall R. Reed (1995)
`Melvin L. DePamphilis
`James T. Kaclonaga (1995)
`Robin Reed (1996)
`Michael Karin (1994)
`(1995)
`Steven I. Reed (1994)
`Thomas Donahue (1995)
`Randall Kaufman (1994)
`Danny Reinberg (1994)
`Daniel J. Donoghue (1996)
`Jack D. Keene (1996)
`Jasper D. Rine (1994)
`Robert N. Eisenman (1994)
`Daniel Klessig (1995)
`Sarah C. R. Elgin (1994)
`Linda M. lllig, Director, Journals
`Barbara H. Iglcwski, Chainnan, Publications Board
`Thomas R. May, Assistant Production Editor
`Kenneth April, Production Editor
`Molecular and Ce!!ular Biology, a publica tion o! the. American Soc'.ety. fo_r ~icrobiology (AS~), 1325 Massachusetts Ave., N.W., Washington, DC
`20005-4171, is devoted to the advancement and d1ssem111at1on _of fund,imental kno_wledge concerning ~he molecular biology of eukaryo!ic cells, of both
`microbial and higher organisms. In struct io ns to a uthors are published m the January is~ue each ~ea_r; rep~mts are avai lab le from !he editors and th e Journals
`Divis ion . The journal is published_ monthly! one volume per year. The n~ nmei~ber pnnt subscn p'.1 on pncc~ a:e $37? (U.S.) (Canadians add 7% GST) a nd
`$485 (other countries) per year; srngle copies are $43 (Canadians acid 7 % GS1), The me'!1ber prmt subscnpt1on pnces are $80 (U.S.) (Canadians add
`GST) and $105 (other countries); sing le copies a re $ ] l (Canadians add 7% GST). For pnces of CD-ROM versions, contact the Subscriptions Unit ASM
`Correspondence relating to subscriptions, defective cop_ies, missing issues, ,~nd avai lahi lit~ of back issues should be directed to the Subscriptions Unit' ASM :
`correspondence relating to reprint ord~rs shou ld be directed _to the Repnnt Order Umt, ASM; and co!-respondence relating to disposition of subm itt ed
`manuscripts, proof5, and genera l editorial ma tte rs should be directed to the Journals Department, American Society for Microbiology, 1325 Mass ac husetts
`<?laims for missing issues !rom res ide nts_ of tl~e Un_ite_d States, Canada, and ~exico mu~t be subn~ittcd wi_thin 3 m_onths after publication of th e iss ues;
`Ave ., N.W., Washington, DC 20005-4171. Phone: (202) 737-3600.
`residents of a ll other co un tries must submit claims wJthm 6 months of publication of the issues. Claims for issues mi ss mg because of failure to rep
`d
`11
`.
`b
`not e a owe
`Second-class postage paid at Washington, DC 20005, and at add ition a l mailing offices.
`.
`POSTMASTER: Send address changes to Molecular an~ Ce!!ular Biology, ASM, 1325 Massachusetts Ave., N.W., Washington, DC 20005-4171.
`address change or for issues
`Made in the United States of America . Printed on acid-free paper.
`The code at the top of the first page of an articl e Ill this Journa l rnd,cates the copyright owner's consent that copies of the article may be made ,
`Copyright ID !994, Americ?n Society for Mi~ro~iology: All Rights Reserved.
`us consent 1s given on 1e con 1t1 on, owever, that the copier pay th e stated per-copy fee through the Co
`Cleara nce Center Inc., 222 Rosewoo d ~rive, Danw_rs, MA 01923, f~r copying beyon~ tl~at ~ermitted by S~~tions 107 and 108 of the U.S. Copyrigli(~;:'.
`·
`·
`This c~n sent does not extend to other kmd s of copymg, such asmlID'slj~tlH-1iW\\J:i~Ja/4\~jput1on, for advertismg or promotional purposes, for creating new
`use or for personal use of spec ific clients.
`
`.
`.
`'
`' n11ss111g
`
`f
`rom
`
`·11
`"
`t·1
`I es w1
`
`'fl ·
`
`·
`
`·
`
`ti
`
`.
`
`d' ·
`
`h
`
`·
`
`collective works, or for resale.
`
`%
`
`7
`
`t
`or an
`
`ISSN 0270-7306
`.
`
`CODEN: MCEBD
`4
`.
`,or persona
`1
`.
`1
`
`

`

`This material may be protected by Copyright law (Title 17 U.S. Code)
`
`MOLECULAR ANO CELLULAR OIOLOGY, Nov. 1994, p. 7445-7454
`0210. 7306/94/$04.oo +o
`Copyright r!) 1994, American Society for Microbiology
`
`Vol. 14, No. 11
`
`Identification and Characterization by Antisense Oligonucleotides of
`Exon and Intron Sequences Required for Splicing
`ZBIGNIEW DOMINSKI AND RYSZARD KOLE*
`Department of Plumnacology and Lineberger Comprehensive Cancer Cente,;
`Unive1:1·ity of North Carolina, Chapel Hill, North Carolina 27599
`
`Received 8 June 1994/Returned for modification 15 July 1994/Accepted 17 August 1994
`
`Certain thalassemic human J3-globin pre-mRNAs carry mutations that generate aberrant splice sites and/or
`activate cryptic splice sites, providing a convenient and clinically relevant system to study splice site selection.
`Antisense 2'-0-methyl oligoribonucleotides were used to block a number of sequences in these 11re-mRNAs and
`we~e tested for their a~ility to inhib_it splicing in vitro. or to affect the ra_tio between aberrantly and correctly
`spliced products. By tins approach, 1t was found that (1) up to 19 nucleotides upstream from the branch point
`adenosine are involved in proper recognition and functioning of the branch point sequence; (ii) whereas at least
`25 nucleotides of exon sequences at both 3' and 5' ends are required for splicing, this requirement does not
`extend past the 5' splice site sequence of the intron; and (iii) improving the 5' splice site of the internal exon
`to match the consensus sequence strongly decreases the accessibility of the upstream 3' splice site to antisense
`2' -0-methyl oligoribonuclcotides. This result most likely reflects changes in the strength of interactions near
`the 3' splice site in response to improvement of the 5' splice site and further supports the existence of
`communication between these sites across the cxon.
`
`Pre-mRNA splicing takes place within a large ribonucleo(cid:173)
`protein complex termed the spliceosome. The specificity and
`accuracy of splicing are determined by the interactions of small
`nuclear ribonucleoprotein particles and protein components of
`the spliceosome with a number of pre-mRNA sequence ele(cid:173)
`ments in pre-mRNA, such as the branch point sequence, the
`polypyrimidine tract, and the 3' and 5' splice sites (reviewed in
`references 15, 21, and 29). In addition, exon sequences seem to
`contribute to the specificity of splicing (references 35, 41, and
`49 and references therein). However, besides identification of
`a regulatory element in the female-specific exon of the dou(cid:173)
`blesex pre-mRNA (17, 19, 30) and characterization of purine(cid:173)
`rich motifs in exons from some other spliced transcripts (7, 12,
`27, 45, 47, 49, 51), the involvement of exon sequences in
`splicing remains unclear.
`In this work, we have used antisense 2'-O-methyl oligoribo(cid:173)
`nucleotides (see reference 43 for a review) to study the
`function of several intron and exon sequences in pre-mRNA
`splicing. This approach stems from our recent report which
`showed that the binding of 2' -O-methyl oligoribonucleotides to
`the branch point or aberrant splice sites leads to the restora(cid:173)
`tion of correct in vitro splicing of mutated [3-globin pre(cid:173)
`mRNAs
`identified
`in
`individuals with various forms of
`[3-thalassemia (11). These oligonucleotides form strong du(cid:173)
`plexes with RNA which are resistant to RNase H and RNA
`unwinding activities. In consequence, they remain stably asso(cid:173)
`ciated with the complementary regions in RNA, efficiently
`inhibiting the function of the targeted sequences. Antisense
`2' -O-methyl oligoribonucleotides were originally used as se(cid:173)
`quence-specific probes to study the structure of small nuclear
`ribonucleoproteins and their interactions with the pre-mRNA
`substrate (43). They were also used to search for novel
`cis-acting sequence elements and possible scanning mecha(cid:173)
`nisms during the splicing reaction (28).
`
`* Corresponding author. Mailing address: University of North Caro(cid:173)
`lina, Lineberger Comprehensive Cancer Center, CB#7295, Chapel
`Hill, NC 27599. Phone: (919) 966-1143. Fax: (919) 966-3015.
`
`We have targeted these oligonucleotides against human
`[3-globin pre-mRNAs carrying mutations responsible for a
`number of [3-globin thalassemia variants. The mutations gen(cid:173)
`erate new splice sites and/or activate cryptic ones, leading to
`aberrant splicing pathways (reviewed in reference 40). In the
`experiments described below, we took advantage of the fact
`that changes in the accumulation of the correctly spliced
`products relative to that of the aberrant ones provide a
`sensitive and internally controlled assay for monitoring inter(cid:173)
`ference with the activity of the sequence targeted by the
`antisense oligonucleotide. Using this approach, we have found
`that sequences upstream from the conserved branch point
`sequence as well as those within the exons are required for
`splicing and play a role in splice site selection. The experiments
`also showed that improving the 5' splice site of the internal
`exon affects interactions at the upstream 3' splice site.
`
`MATERIALS AND METHODS
`
`Plasmid construction. Previously described pSP64H[3~6
`(22) and its thalassemic derivatives cloned in the pSP64 vector
`were used in all studies. The plVSl-[3 110 clone, carrying an
`A-to-G mutation in position 110 of the first intron, was
`constructed by subcloning an appropriate fragment from the
`original thalassemic clone (14a). The pIVS2-[3705 clone was
`obtained by introducing a T-to-G mutation at position 705 of
`the large [3-globin intron in the pIVS2 clone, as previously
`described (11, 25). A C-to-T mutation at position 654 and a
`TA-to-GT mutation at positions 657 and 658 of the large
`[3-globin intron in the pIVS2 clone were introduced to create
`pIVS2-f3654 and pIVS2-[3 654co n clones, respectively. The struc(cid:173)
`tures of all mutated constructs were confirmed by sequencing.
`In vitro transcription and splicing. 32P-labeled transcripts
`were synthesized as previously described (20) with SP6 RNA
`polymerase (Promega) and DNA templates cleaved within the
`second exon at either the BamHI site ( clones Hf3~6 and f3 110),
`thePvull site (clone [37 05 ), or the Asel site (clones pIVS2, [3 654 ,
`and f3 654c 0 "). The preparation of nuclear extract from HeLa
`
`7445
`
`

`

`7446
`
`DOMINSKI AND KOLE
`
`TABLE I. Oligonucleotides used in this study
`Sequence (5'-3')
`
`Oligonucleotide
`
`% GC
`
`8Pa
`BPb
`BPc
`BPc/18
`BPd
`BPe
`BPf
`Exa
`Exb
`Exe
`Exd
`3'ss 579/14
`3'ss 579/18
`ExUl
`ExU2
`ExU3
`ExU4
`ExU5
`ExDl
`ExD2
`5'ss 705
`Intl
`Int2
`5'ss 652/18
`5'ss 652con/18
`5'ss 652con/14
`
`GUCAGUGCCUAUCA
`GUGCCUAUCAGAAA
`CCUAUCAGAAACCC
`CCUAUCAGAAACCCAAGA
`AUCAGAAACCCAAG
`AAACCCAAGAGUCU
`CAAGAGUCUUCUCU
`ACCAGCAGCCUAAG
`AGGGUAGACCACCA
`UCUGGGUCCAAGGG
`UCAAAGAACCUCUG
`UAUUGCCCUGAAAG
`CAUUAUUGCCCUGAAAGA
`GUAUCAUUAUUGCC
`UACAUUGUAUCAUU
`AGAGGCAUGAUACA
`AAUGGUGCAAAGAG
`UAUUCUUUAGAAUG
`CAGAAAUAUUUAUA
`AAUUUAUAUGCAGA
`CCUCUUACCUCAGUUAC
`UAUUAGCAAUAUGA
`UUGUAGCUGCUAUU
`GCUAUUACCUUAACCCAG
`GCACUUACCUUAACCCAG
`CACUUACCUUAACC
`
`50
`43
`50
`44
`43
`43
`43
`57
`57
`64
`43
`43
`39
`36
`21
`43
`43
`21
`14
`21
`47
`21
`36
`44
`50
`43
`
`cells and in vitro splicing were performed as previously de(cid:173)
`scribed (11 , 22).
`Synthesis and purification of oligonucleotides. 2' -O-methyl(cid:173)
`ribonucleosicle phosphoramiclates (Glen Research, Sterling,
`Va.) were used for oligonucleotide synthesis in an Applied
`Biosystems synthesizer at the Lineberger Comprehensive Can(cid:173)
`cer Center. Oligonucleoticles were purified by thin-layer chro(cid:173)
`matography (SurePure kit; U.S . Biochemicals), a~d their con(cid:173)
`centrations were determined spectrophotometncally at 260
`nm. To ascertain the quality of oligonucleotides, the 5'-32P(cid:173)
`labeled compounds were analyzed by polyacrylamide gel elec(cid:173)
`trophoresis.
`Oligonucleotide treatment. In all experiments, 2' -O-methyl
`o ligoribonucleotides were added to the reaction mixture to(cid:173)
`gether with the other components of the splicing reaction. The
`prcannealing of oligonucleoticles with transcript in the absence
`of nuclear extract has been shown not to increase their overall
`effect on splicing ( 11, 28). The extent of unspecific effects for
`each experiment was controlled by using oligonucleotides with
`no complementarity to the RNA substrate. O ligonucleotide
`3'ss 579/14 was used to control unspecific effects during the
`splicing of pre-mRNAs containing the first [3-globin intron,
`whereas oligonucleotide 13Pa was used during the splicing of
`pre-mRNAs containing the second intron. In some experi(cid:173)
`ments, a mixture of randomly synthesized oligonucleotides was
`also used.
`Sequences of oligonuclcotides. The oligonuclcotides used in
`this study are listed in 5'-to-3' orientation in Table 1.
`Data processing and analysis. All autoradiograms were
`captured by a DAGE MTI CCD72 video camera (DAGE,
`Michigan City, Ind.), and images were processed with NIH
`Image 1.47 and MacDraw Pro 1.0 software. The final figures
`were printed out on a Sony dye sublimation printer. Results
`were quantitakd with NIH Image 1.47 software and expressed
`as percentages of correct product relative to the sums of
`correct and aberrant products. Values were adjusted to ac-
`
`MOL CELL- 8101..
`
`I b I , 1 C nucleotides in
`count for the higher number of -r- a e ec
`l t'iined from
`.
`the longer, aberrantly splic~d r rod uct. Aven~gt ~: 'osures arc
`several independent expenments and multip e P
`presented.
`
`3~
`
`RESULTS
`f nctional domain
`f. I
`Mapping the upstream boundary o t 1~ 10u
`RNA sub-
`near the branch point sequence with ~ . P[fci"1 f the first
`· · 0
`strate. A G-to-A point mutation at pos1tion
`.
`.
`( vs ·1 A' ih. F' I 13) creates a new
`.
`nc located
`ig:
`intron in the [3-glob1? _gene I .
`. . '·
`-IJ
`3, splice site at pos1t1on l 09 m add1t1on to the O
`'bl f
`ns1 e or
`·. , .
`•
`.
`·
`·
`. ..
`at position 130 of this intron. This mutatH)n is_ respo
`, scs 111 southe<1stern
`•,
`h 1
`. .
`.
`· .. IVSl-[3' 10
`a s1gmhcant number of [3-t a assemia ed. : .
`Europe, Cyprus, and Lebanon ( 40) .. The spllcmg _of el 90% )
`. 1
`mutant pre-mRNA leads to predominant (approximatddY. .
`1t1ona
`t· · 1•11g •m a
`d
`.
`.
`. )
`'
`accumulat10n of a sphced pro uct con <1m
`19 nucleotides of the intron sequence (Fig. 2A, lane 1. ·
`RNA occurs via
`110
`• •
`pre-m
`Remarkably aberrant sphcmg of [3
`94 of the
`.· ·
`•
`'
`selection of the regular branch pomt at positwn
`tely 10% of the
`·
`(
`. .
`.
`.
`mtron, whereas correct sphcmg approx1ma
`f tl1c· cryptic branch
`t·
`I
`.
`,
`. , .
`.
`.
`resultant mRNA) occurs via se ec 1011 o
`point at position 107 (54). As a result, mutatwns mactiv,1~mg
`the regular branch point (38) or antisense ol_igonucleotides
`blocking it (11) stimulate the cryptic branch pomt and restore
`.
`correct splicing in the [3 uo background.
`In these studies we used 2' -O-methyl oligoribonucleotides
`to determine the importance of sequences located _up5t~e~m
`from the branch point region in the function of this sphcmg
`element. We have designed a series of 14-mers complementary
`to the region extending up to 32 nucleotides upstrea~ from the
`branch point sequence (Fig. lB). The abili~y of_ ohgonucleo(cid:173)
`tides to inhibit splicing at the aberrant 3' sphce sit~ anc~ at the
`same time to promote splicing at the correct 3' sphce site was
`taken as the measure of the function of the upstream se(cid:173)
`quences during the splicing of [3 110 pre-mRNA (11). Note that
`a concomitant switch in selection between the regular and
`cryptic branch points can be directly determined on autorad_io(cid:173)
`grams because of the variable mobility of the correspondmg
`.
`lariats (54).
`The results of in vitro splicing of [3 110 pre-mR~A earned
`out in the presence of 0.5, 2.0, and 10.0 µM ohgonucl~o(cid:173)
`tide BPa, blocking all seven nucleotides of the branch pomt
`sequence, are shown in Fig. 2A. Consistent with the r~sult_s of
`a previous report (11 ), the antisense 2 ' -O-methyl ohgonb?(cid:173)
`nucleotide targeted to this site leads to a change in the ratio
`between correct and aberrant products. A splicing react(on
`carried out at a 2 µ,M concentration of this oligonucleot1de
`results in the accumulation of approximately 55% of the
`correct product (Fig. 2A, lane 3; see Fig. 2C for quantitation).
`Although 10 µ,M oligonucleotide BPa results in a marked
`decrease of the overall efficiency of the splicing reaction, it
`does not significantly modify the ratio between correct and
`aberrant products achieved at a 2 µM concentration (Fig. 2A,
`lane 4).
`Figure 2B demonstrates the results of an in vitro splicing
`reaction of [3 110 pre-mRNA carried out in the presence of a 2
`µ,M concentration of oligonucleotides directed upstream from
`the branch point adenosine. Similar to BPa, oligonucleotides
`BPb, -c, and -d restore correct splicing to 50 to 65% (Fig. 2B,
`lanes 2 to 4, respectively, and C). Note that oligonucleotides
`BPc and BPd hybridize immediately outside the conserved
`branch point sequence and four nucleotides upstream from
`this element, respectively. While oligonucleotide BPe, hybrid(cid:173)
`ized 13 nucleotides upstream from the branch point adenosine,
`still has some effect on the splicing pathway, oligonucleotide
`
`

`

`VOL. 14, ]994
`
`SPLICING DOMAINS IN PRE-mRNA
`
`7447
`
`A
`
`Q. Q.~
`ID
`.0 ~
`
`~
`M
`♦ ♦ ♦ ♦.----,
`__ E_x_o_N_1 _ __,1---62----94-101-110--1Jo ... l __ E_x_o_N_2 _ ___,
`..
`..
`..
`..
`
`Ollgos BPa-f
`
`Ollgos Exa-<I
`
`B
`
`0110
`_g.
`(A)
`~
`♦ ♦
`♦
`(62)AGAGAAGAcucuuGGGuuucuGAUAGG~;J;~~~cuc~J\1uGGlu:u (114)
`0110 exon 2
`
`BPf
`
`BPe
`
`81-'d Pc
`
`C
`
`Ill ·"' M
`♦ +1
`+40
`(115)AUUUUCCCACCCUUAGl:GCUGCUGGUGGUCUACCCUUGGACCCAGAGGUUCUUUGAG
`Exa
`xb
`-----:::::::: .. E_xc _____ Exd
`0-globln exon 2
`FIG. 1. (A) Structure of human f3-globin pre-mRNA (not drawn to scale) containing the first intron (the line) and two flanking exons (boxes).
`The positions of the regular (BP) and cryptic (bp) branch point adenosines, (3 110 mutation, and the 3' splice site (3'ss) are shown. The regions of
`pre-mRNA used as targets for 2' -0-methyl oligoribonucleotides are indicated by double-headed arrows. Exon 1 contains 154 nucleotides, and exon
`2 contains 210 nucleotides. The exact sequence positions of oligonucleotides BPa to BPf (B) and Exa to Exd (C) are shown. The regular and cryptic
`) and correct 3' splice sites are indicated.
`branch point sequences (shaded boxes) and the aberrant ((3 110
`
`BPf, targeted 5 nucleotides further upstream, seems to be
`ineffective. Consistently, oligonucleotide BPf, in contrast to the
`other probes tested, does not stimulate the formation of the
`lariat intermediate at the branch point adenosine at position
`107, as indicated by the lack of a band migrating to the top of
`the gel (Fig. 2B, lane 6).
`The involvement of exon sequences in the splicing of 13 110
`pre-mRNA. To survey other regions of f,-globin pre-mRNA for
`the existence of functional domains involved in splicing, we
`designed 14-mers complementary to the correct junction be(cid:173)
`tween the first intron and the second exon ( oligonucleotide
`Exa) or to sequences extending downstream into the second
`exon (oligonucleotides Exb to -d) (Fig. IC). For I-lf3M pre(cid:173)
`mRNA, oligonucleotide Exa partially overlaps with the intron
`sequence; in the case of (3 110 substrate, the same oligonucleo(cid:173)
`tide has its target entirely within the aberrantly spliced second
`exon. Although the upstream end of the duplex formed
`between this oligonucleotide and its target in the r, 110 tran(cid:173)
`script is separated from the predominantly used aberrant 3'
`splice site by 14 nucleotides, Exa oligonucleotide at a 2 µM
`concentration fully blocks splicing (Fig. 3, lane 3). This is in
`agreement with the observation that mutations disrupting the
`AG dinucleotide of the normal 3' splice site inhibit the use of
`the aberrant splice site in (3 110 pre-mRNA (23, 54). An
`inhibitory effect on splicing is also observed in the presence of
`oligonucleotides Exb and Exe, which hybridize further down(cid:173)
`stream into the second exon, and, albeit to a lesser extent, in
`the presence of the most distal oligonucleotide, Exd, which
`hybridizes 44 nucleotides downstream from the r, 110 3' splice
`
`site (Fig. 3, lanes 4 to 6). The control noncomplementary
`oligonucleotide does not significantly affect the efficiency of the
`splicing reaction (Fig. 3, lane 7). A pattern of inhibition similar
`to that described above was obtained when each of the
`oligonucleotides Exa, -b, -c and -d was added to the splicing
`reaction with normal r,-globin pre-mRNA (data not shown).
`The involvement of exon sequences in the splicing of the
`13105-globin pre-mRNA. The results of Zhuang and Weiner
`(54) suggested that the splicing of r, 110 pre-mRNA may
`represent
`a special case in which the AG dinucleotide located inside
`the exon is required for splicing at the upstream 3' splice site.
`To confirm that exon involvement in splicing is not limited
`to only one type of substrate, we analyzed the effects of a
`series of antisense oligonucleotides (Fig. 4B and C) on the
`splicing of thalassemic 13105 pre-mRNA. A T-to-G mutation
`at position 705 of the large (3-globin intron creates an addi(cid:173)
`tional 5' splice site and activates a cryptic 3' splice site at
`position 579 of this intron (8). The incorrect splicing path(cid:173)
`way resulting from the utilization of both splice sites leads to
`the incorporation of nucleotides 580 to 705 of the intron into
`the accumulation of significant
`the spliced product and
`amounts of the 577-nucleotide aberrant RNA in addition to
`the correct product (Fig. 4A). Changes in the ratio between
`correct and aberrant products provide a sensitive method for
`measuring the sequence-specific effects of the oligonucleotides
`used.
`Figure 5A shows the complete pattern of in vitro splicing of
`the 13105 substrate. Quantitative analysis indicates that the ratio
`
`

`

`7448
`
`DOMINSKI AND KOLE
`
`A
`
`B
`
`0 -
`
`-~
`[]]-ill
`J aberrant
`\ [II]]
`correct
`
`_ .. ... • - -3'Exo
`
`245-
`2421
`
`_Q_130
`
`130•-
`
`111•-
`
`154 -
`
`C
`
`I-
`0 w
`a:
`a:
`0
`0
`u.
`0
`'l'..
`
`80
`
`70
`
`60
`
`50
`
`40
`
`30
`
`20
`
`10
`
`0
`
`J 3'exo
`
`-c::::L130
`107
`
`_Q_130
`
`M OL. C ELL. BIOL.
`
`.0
`a.
`Ill
`+
`
`(J
`
`a.
`Ill
`+
`
`.,,
`a.
`Ill
`+
`
`.,
`a.
`Ill
`+
`
`it
`Ill
`+
`
`2 µM
`
`0
`
`- a.
`c;,
`1ii :::
`
`"'
`Ill
`I +
`
`2
`
`3
`
`4
`
`6
`5
`4
`3
`2
`1
`FIG. 2. (A) Change in the ratio between correct and aberrant products during the splicing of 13 1 w pre-mRNA in the presence of oli~~nucleotide
`BPa, targeted to the branch point sequence. The concentration of oligonucleotide is shown at the top of each lane. Tl~e control sphcmg patter_n
`for 13 110 prc-mRNA is shown in lane 1. Bands at the top of the autoradiogram correspond to the lariat-exon 2 in~e rmediate, formed _a,t the. cryptic
`branch point at nucleotide 107 (upper) and at the regular adenosine at nucleotide 94 (lower). The 249-nucleotide RNA product vis,_ble m lanes
`3 and 4 is generated by endogenous 3'-to-5' exonuclease (3' Exo) blocked at the site of duplex formation (reference 11 a~d unpubh~he d d ata).
`The structures of splicing products and intermediates are shown on the right, and the ir sizes (in nucleotides) are on the left: An asterisk den otes
`the unusual mobility of the corresponding lariat-containing RNA species. The same designations are used in subsequent f1gur~s. ~B) Effec~s of
`oligonucleotides BPb to BPf, targeted upstream from the branch point adenosine, on the splicing of 13 1 w pre-mRNA. (C) Quant1tat1ve a.nalys1s of
`the reactions carried out in the presence of 2 1.1.M oligonucleotides. Data were calculated as the percentages of the correct product m overall
`splicing (sec Materials and Methods).
`
`of correct to aberrant products is approximately I :1 (Fig. 5B,
`lane 1 ). Aberrant splicing is almost completely suppressed
`in the presence of 2 µM 14-mer oligonucleotides targeted
`either to the 3' splice site ( oligonucleotide 3'ss 579; Fig. SB,
`lane 2) or to the exon sequence immediately downstream
`from this site (oligonucleotide ExUl; Fig. 5B, lane 3). Oligo(cid:173)
`nucleotides targeted to exon sequences between nucleo(cid:173)
`tides 7 and 20 (ExU2), 16 and 30 (ExU3), and 27 and 40
`(ExU4) also block aberrant splicing, although not as eJTicient(cid:173)
`ly as the first two probes mentioned immediately above (Fig.
`5Il, lanes 4 to 6, respectively). The most distant probe (oligo(cid:173)
`nucleotide ExUS), complementary to nucleotides 37 to 50 of
`the cxon, has only a limited effect on the aberrant splicing
`pathway of [3705 pre-mRNA and delimits the downstream
`boundary of the exon region that plays a role in splicing (Fig.
`5Il, lane 7).
`The involvement of exon sequences in splice site selection
`was further confirmed by using oligonuclcotidcs ExDl and
`ExD2, which hybridize to a region near the downstream end of
`the exon and lead to virtually full inhibition of the aberrant
`splicing pathway (Fig. SC, lanes 2 and 3). As expected (11), a
`strong inhibitory effect is also mediated by oligonucleotide S'ss
`705, targeted to the cxon-intron junction (Fig. SC, lane 4).
`Interestingly, oligonucleotidcs Intl and lnt2, targeted to re(cid:173)
`gions located entirely within the intron and almost immediately
`
`downstream from the 5' splice site, have little, if any, effect on
`aberrant splicing (Fig. SC, lanes 5 and 6).
`Testing the accessibility of 3' and 5' splice sites to 2'-0-
`methyl oligonucleotides in [3654 and J3654
`con pre-mRNAs. It has
`been shown for a variety of pre-mRNA substrates that muta(cid:173)
`tions within the 5' splice site of an internal exon affect the
`splicing of the upstream intron (1, 10, 26, 31, 46). This effect
`suggests the existence of a certain type of communication
`between splice sites across the exon and argues in favor of the
`cxon definition model (36). In order to determine whether an
`improvement of the 5' splice site to the consensus sequence
`increases interactions near the upstream 3' splice site flanking
`the internal exon, we tested the accessibility of these sites to
`antisense oligonucleotides in [3654-globin pre-mRNA and its
`modified version, f3654c
`• [3654 represents another thalassemic
`mutation which occurs in the human J3-globin gene and results
`in the generation of aberrantly spliced product, containing part
`of the second intron (5). The C-to-T mutation at nucleotide
`654 of the intron creates a new 5' splice site at position 652
`(AG/GUAAUA), activates a cryptic 3' splice at position 579,
`and results in the generation of a 73-nucleotide aberrant exon
`(Fig. 6). We improved the aberrant 5' splice site at position 652
`by two point mutations, including a highly conserved G at
`position +5, thereby creating the consensus element AG/
`GUAAGJl (substituted nucleotides are underlined). These
`
`0 11
`
`

`

`VoL. 14, 1994
`
`SPLICING DOMAINS IN PRE-mRNA
`
`7449
`
`2 fJM
`
`0
`
`~ cu .Cu"CC
`
`: : z :2 :+++++
`
`21
`!.cn~~~~s
`:
`.
`r· I:
`--. - .! -o..m
`339"--- ;l--, r ...... - ..
`- !7: . ._. ._..,... .,_ - 1 - pre-mRNA
`i [D-,ill
`. \,_
`
`493
`
`\
`
`;
`
`.
`
`'
`
`'
`
`382
`
`·.
`I
`\_ •,·~ ~ -
`,
`363 '
`I
`i
`
`·
`
`-
`
`,,
`
`-
`
`i
`
`' · I
`1.
`
`I '
`
`I
`
`· I [!]-ill
`; -1 aberrant
`~
`f""fT2l
`I ,
`·: ~
`i
`correct
`
`j
`
`I
`
`i
`I
`i
`I
`j-o._130
`
`'
`I
`i'" ,, :
`.
`r
`,·
`' ; I
`';
`'
`l .
`, I _0..111
`111•- i.' . ~
`'
`,,. ,I
`,
`f,
`· •..• ~ ...,:j _OJ
`__ .: ~---- :
`154 - k
`L..J L · - -·---- . , .... __ I
`
`. '
`
`7
`6
`5
`4
`3
`2
`FIG. 3. Inhibition of splicing by oligonucleotides targeted to aber(cid:173)
`rant cxon 2. The splicing of 13110-giobin pre-mRNA in the presence of
`2 µM oligonucleotides Exa to Exd is shown in lanes 3 to 6, respectively.
`The splicing reactions of Hl3M and 13 110 pre-mRNAs with no oligo(cid:173)
`nucleotides are shown in lanes 1 and 2, respectively. Lane 7, the
`splicing of 13 110 pre-mRNA in the presence of noncomplementary
`oligonuclcotide (control) (sec Materials and Methods).
`
`changes increase the base pairing of the 5' splice site to the 5'
`end of Ul small nuclear RNA and result in more eflicient
`selection of this site (32, 52, 53). The full splicing patterns for
`654
`) and improved (f36 54con) pre-mRNAs
`both the original (f3
`are shown in Fig. 7 (lanes 2 and 3, respectively). In addition,
`the pattern for wild-type pre-mRNA, containing no mutation
`in the intron, is presented in lane 1 of Fig. 7. As expected,
`creation of the perfect 5' splice site at position 652 leads to
`complete inclusion of the aberrant exon.
`Consistent with the results of a previous report (11), 18-mer
`oligonucleotides targeted to either 3' or 5' splice sites in (3654
`pre-mRNA inhibit the aberrant splicing pathway with similar
`efliciency. At 0.25 µM concentrations, both oligonucleotides
`lead to a substantial reduction in the amount of the aberrant
`product (Fig. BA, lanes 2 and 5); at 0.75 µM concentrations,
`they lead to full restoration of the correct pathway (Fig. SA,
`lanes 3 and 6). On the basis of these results, we conclude that
`in f36 5 4 pre-mRNA, 3' and 5' splice sites are equally accessible
`to antisense oligonucleotides. The results of a similar analysis
`654
`co n pre-mRNA are markedly different. An 18-mer
`with f3
`targeted to the 5' splice site restore