Aberrant splicing events that are induced by proviral ... - Europe PMC

Proc. Natl. Acad. Sci. USA

Vol. 84, pp. 3171-3175, May 1987

Biochemistry

Aberrant splicing events that are induced by proviral integration: Implications for myb oncogene activation (oncogenes/plasmacytoid lymphosarcomas/RNA splicing)

DAN RoSSON*, DEBBIE DUGANt, AND E. PREMKUMAR REDDY** Department of Molecular Oncology, Roche Institute of Molecular Biology, Roche Research Center, Nutley, NJ 07110

Communicated by Allan H. Conney, January 27, 1987 (received for review October 1, 1985)

ABSTRACT Activation of the mouse c-myb oncogene in Abelson virus-induced plasmacytoid lymphosarcomas was studied using cDNA cloning and nucleotide sequence analysis. The results presented here show that viral integration in the myb locus generates splicing errors at the 5' and 3' regions. Viral integration results in transcriptional initiation within the viral long terminal repeat and generation of a chimeric mRNA that lacks the rst three coding exons. The alterations at the 3' end are caused by an aberrant splicing event in which additional splice-donor and -acceptor sequences within intronic sequences are used to splice an additional 363 nucleotides into the myb transcripts. The resulting insertion of 121 amino acids is in a region of the protein where other activated forms of the myb gene product have deletions. These results suggest that alterations in the 3' end of the myb gene play a crucial role in the activation of this gene. A comparative study of retroviral oncogenes and their cellular homologues has given us invaluable information regarding the structural changes that protooncogenes often undergo to acquire transforming potential. A comparison of the c-myb cDNA sequence (1) with the v-myb sequence of avian myeloblastosis virus (AMV) (2, 3) and E26 leukemia virus (4) revealed that large stretches of coding sequences from the 5' and 3' ends have been deleted during the transduction of this gene into the viral genome. The AMV myb gene is derived from seven of the internal coding exons of the c-myb gene (3, 5) and has undergone deletions of 142 amino acids at its N-terminal end and 192 amino acids at its C-terminal end (1). Similar deletions have occurred in the E26 virus, in which myb is coded as part of a fusion polypeptide of three distinct elements, the viral gag sequences, the viral myb, and a third element derived from the protooncogene ets (4). Deletions of similar stretches of sequences in two independent viral isolates suggest that these deletions are important for the oncogenic activation of the myb gene. Another model system that has allowed us to study the activation of the myb gene is the Abelson leukemia virusinduced plasmacytoid lymphosarcomas (ABPL tumors) of the mouse (6). Detailed analysis of these tumors revealed that they do not contain the integrated Abelson leukemia virus genome, but instead have undergone rearrangements in their myb locus as a result of Moloney murine leukemia virus (Mo-MuLV) integration (5, 7, 8). To study the effect of the viral integration on the transcription of the rearranged myb locus in ABPL tumors, we have undertaken the cDNA cloning of myb mRNA from one of the ABPL tumors, ABPL-2, and determined its sequence. Our results show that the viral integration results in aberrant splicing events at both the 5' and 3' ends of the mRNA 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.

3171

transcripts producing alterations in similar positions to those seen in AMV and E26 virus.

MATERIALS AND METHODS Cell Lines. Cell lines from ABPL-1, ABPL-2, and ABPL-4 were derived from tumors induced by the injection of pristane and a mixture of Abelson leukemia virus and Mo-MuLV (6, 9, 10). Cells were grown in suspension culture in Dulbecco's modified Eagle's medium supplemented with 10% (vol/vol) fetal calf serum. cDNA Library Construction and Screening. The cDNA library was constructed from poly(A) RNA isolated from freshly excised mouse thymuses or ABPL tumor cells as described (11, 12). cDNA synthesis was as described (1). cDNA (50 ng) was ligated to 1 jig of Xgt1O arms to obtain a phage library of 5 x 105 recombinants, which was screened by standard techniques (24). RNA gel blot analysis was as described by Lehrach et al. (13). DNA Sequence Analysis. The nucleotide sequence was determined by the procedure of Maxam and Gilbert (14).

RESULTS Transcription of myb Locus in ABPL Tumors. Earlier studies indicated that all ABPL tumors have undergone rearrangements in the myb locus as a result of integration of the Mo-MuLV proviral genome with the myb gene at a point immediately preceding the first exon that exhibits homology to the v-myb gene (5, 8). RNA isolated from the tumor tissues showed the presence of normal [3.8-4.0 kilobases (kb)] and abnormal (5.0 kb) transcripts that cross-hybridized to radiolabeled v-myb probes (7). Since the tumor cells grown in vivo are often contaminated with other lymphoid cells as well as granulocytes and macrophages, it was not clear whether the ABPL tumors synthesized normal c-myb transcripts or that these RNAs are derived from contaminating normal cells. To distinguish between these possibilities, we established three of the tumor lines in tissue culture and examined their RNAs by RNA gel blot analysis. These results (data not shown) indicated that the RNA derived from mouse thymus contains a single diffuse band in the range of 3.8-4.0 kb, whereas the three tumor RNAs contain transcripts >4.0 kb. There were no detectable 3.8- to 4.0-kb RNAs, although these cells retained one normal nonrearranged c-myb allele (7, 8). Abbreviations: ABPL, Abelson virus-induced plasmacytoid lymphosarcomas; AMV, avian myeloblastosis virus; Mo-MuLV, Moloney murine leukemia virus; nt, nucleotide(s). *Present address: The Wistar Institute of Anatomy and Biology, 36th Street at Spruce, Philadelphia, PA 19104. tPresent address: Center for Neurobiology, Columbia University, 722 West 168th Street, New York, NY 10032. 1To whom reprint requests should be addressed at: The Wistar Institute of Anatomy and Biology, 36th Street at Spruce, Philadelphia, PA 19104.

3172

Biochemistry: Rosson et al. X

Proc. Natl. Acad. Sci. USA 84 (1987)

PENPA

EP

..

S N Bg )I AW''ii --

SC

E

--

I

ATG

dABPL-2 c-myb

TGA

- XPL 2-1

F-

XPL 2-6 E P

P

ATG

i

I

Sc

N Bg

Ss

E

TGA

1Kb I

3Kb

2 Kb

I

I

I

Structure and Sequence of the Aberrant c-myb Transcripts in the ABPL-2 Tumor Cell Line. To determine the structure of the aberrant transcripts, we constructed a cDNA library from the poly(A) RNA of ABPL-2. Screening of the library with a radiolabeled c-myb probe showed the presence of a high percentage of positive clones; 24 of which were grown and further characterized by restriction enzyme cleavage mapping. A restriction map of the two clones used for detailed analysis is shown in Fig. 1. A comparison to the restriction map of the normal myb gene (15, 16) shows that the ABPL-2 cDNA differs from the normal c-myb cDNA at the 5' end as well as in the middle portion where insertion of %-400 base pairs (bp) is seen. To delineate further the nature of the changes that have bccurred in the formation of the aberrant c-myb mRNAs, we carried out complete nucleotide sequence analysis of two overlapping clones derived from this aberrant mRNA. The composite nucleotide sequence is shown in Fig. 2. The sequence analysis revealed that the 5' end of these cDNA clones contain sequences derived from the Mo-MuLV proviral genome (17). Thus, the first 1596 bases of the transcript appear to be derived from the helper viral genome that is followed by the myb coding sequences. The junction between viral and myb sequences occurs at a point within the viral genome that contains the gag coding sequence CAGGCA-GGT-AGG-ACC in which the underlined sequence is homologous to the consensus splice donor sequence MAG 4 GTRAGT (20), where the arrow indicates the splice point, M is either cytidine or adenosine, and R is either adenosine or guanosine. The splice-acceptor sequence is contained within the c-myb locus and is, in fact, the natural splice-acceptor sequence present at the junction of the first viral-related exon and its upstream intron. The splicing event between the viral and myb sequences occurs such that the gag sequences are fused to the myb sequences in a reading frame that is different from that utilized for the synthesis of gag protein. This reading frame contains two in-frame ATG codons beginning at positions 1509 and 1512 either of which could be used for the ipitiation of protein synthesis by the ABPL-2 mRNA. T l homology between the normal c-myb and the ABPL2-derived c-myb sequences starts from position 1597 and continues for a stretch of 990 nucleotides (nt). At nt-2587 the myb sequence is interrupted by a stretch of 363 nt that bears no homology with the normal c-myb cDNA sequences. The interruption occurs at a point that constitutes the junction of the sixth and seventh v-rnyb-related exon. However, from

NORMAL c-myb

FIG. 1. Restriction map of ABPL-2 cmyb cDNA as derived from the two overlapping clones XPL2-1 and XPL2-6. Restriction sites are A, Aat II; Bg, Bgl II; E, EcoRI; N, Nco I; P, Pst I; Sc, Sac I; X, Xho I; Ss, Ssp I. The restriction map for normal c-myb was derived from Gonda et al. (15). The darkened portion of the bars indicates normal c-myb sequence. The open bar represents non-myb sequence. The position of the potential initiation (ATG) and terminator (TGA) codons are indicated.

nt-2950, the homology with normal c-myb sequences resumes and is continued until the end of the cDNA clone. Since the insertion of this stretch of 363 nt does not change the reading frame of the 3' end of the transcript, the protein derived from such an aberrant mRNA would contain 715 amino acids and code for a protein that contains an N-terminal region derived from the Mo-MuLV genome and a C-terminal region derived from the myb gene that is interrupted by a stretch of 121 amino acids from position 360 through 481. Origin of the Inserted Sequences Present in ABPL-2-Derived c-myb cDNA Clones. To determine the origin of the additional sequences present in the cDNA clones derived from the ABPL-2 tumor line, we initially prepared a DNA fragment specific for these sequences by cleavage of the cDNA clone with EcoRI and Aat II. A radioactive probe derived from this fragment was used to probe mouse genomic DNA. These hybridization experiments suggested that these sequences were present within the myb gene itself, presumably within the intronic sequences between viral-related exons 6 and 7. This was further confirmed by examination of mouse genomic clones (provided by J. Ihle, National Cancer Institute, Frederick Cancer Research Facility) and nucleotide sequence analysis of the intronic sequences present between viral-related exons 6 and 7. Sequence analysis of this intron revealed the presence of an identical stretch of sequence beginning :850 bases from the 3' end of the sixth exon (data not shown). The sequence of the mouse genome at this region (which normally constitutes an intron) was found to be GTTGTTTCAG 4 GATTCT=351=TCGCAG 4 GTAGGGCAG, where the arrows indicate splice points. The sequence on the 5' side of the insert corresponds well to the consensus splice-acceptor sequence (Y)6NCAG 4 GK (20), where Y is thymidine or cytidine, N is any nucleotide, and K is guanosine or thymidine. Similarly, the sequence immediately following the insertion sequences corresponds well to the consensus splice donor sequence MAG 4 GTRAGT (20). Thus, in the generation of the ABPL-2 myb mRNA, an additional set of splice-donor and -acceptor sequences within intronic sequences at the 3' end of the gene are used. The splicing event that produces such an aberrant mRNA is illustrated in Fig. 3. To determine whether the aberrant splicing observed constitutes the predominant transcriptional product or whether it represents a minor species that is produced by these tumors, we cloned and sequenced six additional cDNA

FIG. 2 (on opposite page). Nucleotide sequence of the ABPL-2 cDNA deduced from cDNA clones XPL2-1 and XPL2-6. The nucleotide sequence proceeds in the 5' to 3' direction. The deduced amino acid sequence from the open reading fiame is shown below the nucleotide sequence. Numbering of the sequence is based on the sequence of the Mo-MuLV genome (17). Vertical arrows indicate EcoRI sites, boundaries for sequences homologous to the v-myb sequences of AMV and E26 virus, and the termination point where c-myb is disrupted in the NFS-60 cell line by integration of the Cas-Br virus (18, 19). The positions of those sequences coded for by the viral-related exons 1-7 and the insertion of intron sequences are also indicated. Termination codons inframe with the reading frame beginning at positions 1482 and 3655 are indicated by ***.


Biochemistry: Rosson et al.

3173

950

1000

1100

Ci~l'lrr'li' GcACaOiw

TrrXT1CAOC CCTxA r lt'ItC'rkxCCM ACCTCA

m GI=rra"I

CTPIG AA

1200 ~X

TAG@

AGA

C

GG

W Gm

-Aa;.aa G-4GA w

A

AATO

GC

T

CG a; GCG

1300

GCXO~A ~M z CAIG =G G GIAAAITCA AOCICI

C

1400 TAC

A

ACITCAA TM1AQXLIx;GT TC1IvC1 lGX~CGTCA~r G G

1T

AAAA ATAAT~ TICTTICIGTGAc XAG

A

CAGAAk MACG= CDVAAAGC= Mo-MuLV Sequ enIesStart AMV IHomology vEl 1500 GGIGCXX G AMG AMG( IT TI C TM MG C C AMT GO MT AA OCZ COC AMG CG AAC COG AA GAT (CCCA CIIC AMC T0C CG G AM TOG M O MFT MET Gly Ala Pro Leu Asn Cys Pro NET Lys Ser MET Pro lIeu Phe Pro Ser Ser Ala Arg Ile Gly Ile Tr Pro Pro Arg Gln Asn Arg Thr Asp Start E26 Homology + vE2 1700 MG MA CM CM GAT CG AA GIC A GGM CIT GIC CG AAA TAT CIT GT CAM CCC AMC AA MrT CC CPCM COG C MAA GIG CGG AMC CCTGAA Val Gln Cys Gln His Arg Tip Gln Lys Val TMiAsn Pro Glu Leu Ile Lys Gly Pro Tip Thr Lays Glu Glu Asp Gln Arg Val Ile Glu Leu Val Gln Lys Iyr Gly TCAG

G

r

1800 MAA MT C CMC 1TA AAA A A AGCAG TGT CO GAG AM G ATT GO AAG G T; CAC AAC AM TCC1 TMG AAT a GAA GT AAG A TA Pro Lys Arg Tzp Ser Val Ile Ala Lys His ieu Lys Gly Arg Ile Gly Lys Gln Cys Arg Glu Arg Tzp His Asn His Ieu Asn Pro Gju Val Lys Lys Thr Ser Trp 1900 vE3 G (Xk MC AAG COT CIG GOG AAC AG TG CA GMAG AC QCA AAG CTGG G CCC G (XG ACT GUT AAT IT C AGM AAC AC GAM GAG GAG GAC AG AMC AT TM U Thr Glu Glu Glu Asp Arg Ile Ile Tyr Gln Ala His Lys Arg Ieu Gly Mn Arg Tip Ala Glu Ile Ala Lys Leu Ieu Pro Gly Arg Thr Asp Asn Ala Ile Lys Asn + EcoRI 2000 CM¶TGG AAT TCCCGATG CGT CGC AA GGCMGAA CAG GA X TM CTG G GMG CCT ICC MA QCC AGC CAG ACG CCA GIG GCC ACMG AC TC CG AAG AAC AAT CA His Ttp Asn Ser Thr MET Arg Arg Lyrs Val Glu Gln Glu Gly Tyr Ieu Gln Glu Pro Ser Lys Ala Ser Gln Thr Pro Val Ala Thr Ser Phe Gln Lys Asnn Hi_ 2100 VEX TTG M GGG TTT GG CAT GOC TCA C!T CCA TT G CIC ITC CCA AC QC GCTC IC GTC AAC AC GAA TAT CCC TAT TAC CAC AC GCC GA OCMCA AAC AMC Leu MET Gly Phe Gly His Ala Ser Pro Pro Ser Gln Leu Ser Pro Ser Gly Gln Ser Ser Val Asn Ser Glu Tyr Pro Tyr Tyr His Ile Ala Glu Ala Gln Asn Ile 2200 * I 1 vE5 TCC AM CAC GIT COC TAT CCT GIC QCA TTG CAT GIT AAT ATA GTC AAC GMC CT CGM GCT GM GCA G(C AC CAG AG CAM TAT AMAC GAA GAC CT GAG AAG Ser Ser His Val Pro Tyr Pro Val Ala Leu His Val Asn Ile Val Asn Val Pro Gln Pro Ala Ala Ala Ala Ile Gln Arg His Tyr Asn Asp Glu Asp Pro Glu Lys 0 "vE6 2300 * MA AAG GIA GAM A CTG GMG TG G IC CIC TG TTCA MTA GAM AMC G CIG AMWC GM GM COA ITA (CA CAGM AMC GM r TX AG TA COCGM TGG Glu Lys Arg Ile Lys Glu Leu Glu Leu Iu leu uET Ser Thr Glu Asn Glu Leu Lys Gly Gln Gln Ala lau Pro Thr Gln Asn His Thr Cys Sex Tyr Pro Gly Trp 2400 End E26 Homology4 CK: AGC ACC TICC AT GIG GAG CAGACC AGA CT CT CGGAT AMGT CIT GTT TC TIG CGA C CAT GCC ACC CIA TCT CTG CCT GCA T (CC QQC TCC G MA His Sex Thr Ser Ile Val Asp Gln Thr Arg Pro His Gly Asp Ser Ala Pro Val Ser Cys leu Gly Glu His His Ala Thr Pro Ser Leu Pro Ala Asp Pro Gly Ser 2500 Cas Br Virus Insertion + I AMCAMTCMTGAC C CIA C'CT GAA GAMr CA ATG(CA MCG It AMG TC C GM CC MG CT GT AMT CM G TA AA TIT GCA Cm MA CIC CAG TIT ATA Leu Pro Glu Glu Ser Ala Ser Pro Ala Arg Cys MET Ile Val His Gln Gly Thr Ile Leu Asp Asn Val Lys Asn Leu Leu Glu Phe Ala Glu Thr Leu Gln Phe Ile * I * Start Intron sequences ; EcoRI ATI GAT TCT GCAT ITCT TM TGTGAT CTC AMC AGT mr TTC TC G CA G (AGCG GCA GCT TIT TCA AC MCG CG GCC MA GM A GMC TIC CG CIT CG C1T Asp Ser Asp Ser Ser Trp Cys Asp Leu Ser Ser Phe Glu Phe Sex Glu Glu Ala Ala Ala Phe Ser Pro Ser Gln Gln Pro Thr Gly Lys Ala Phe Gln Leu Gln leu 2700 CG CM AGA GMG GGC CAT GXCT AG ITT GCA GOA GMG CT AMC CTG AMG GIG C AMG CA GIG CIG AGC GG GCA T GOCGC CA C TM CCC C GCG AMG Gln Gln Arg Glu Gly His Gly Thr Arg Ser Ala Gly Glu Pro Ser Lau Arg Val Thr Arg Arg Val lau Ser Glu Ala Ser Leu Gly Pro His Ser Pro Gln Ala Arg 2800 2900 MG CC CTA GCC GCT(GOM CT AMC CCC TCC cC TT GCT GA GG AMC AGC TCA ACT A AGC AMG GTT 0G G GTC GTC CIAC AAA AGMG C G CAG AGC His Ser Lys Val Arg Ieu Val Val leu Arg Iys Arg Arg Gly Gln Ala Sex Pro Leu Ala Ala Gly Glu Pro Ser Pro Ser Leu Phe Ala Asp Val Ile Ser Ser Thr End kItron sequences 4 I *vE7 3000 CM AAG CAT TCC CIT GTC AA AGC CIA CCC TIC TCT CCC TXC CG TC TC AAC ACT TCC AGC AAC CAT GCA AAC TIC GGC TTA GAT GA CCT AM TIA CC CC ACT Leu Lys His Ser Pro Vai Lys Ser Leu Pro Phe Ser Pro SerCln Ph~iLeu Asn Thr Ser Ser Asn His Glu Asn Ser Gly Leu Asp Ala Pro Thr Leu Pro Ser Thr End AMV 3100 I - 3 exons CCT CIC ATT GGT CAC AAA CIG ACA CCITT CA GM CAG ACT GTG AMA ACC CAG AAG GMA AM TCC ATC m AM ACT CA GC ATC AAA AMG ICA ATC CC GA AGC Pro Leu Ile Gly His Lys Leu Thr Pro Cys Arg Asp Gln Thr Val Lys Thr Gln Lys Glu Mn Ser Ile Phe Arg Thr Pro Ala Ile Lys Arg Ser Ile Ieu Glu Ser 3200 TCT CIT CMAMrCC AMA CCA TIC AAA CAT GCC CIT (CA A CA ATT AA TMCGGT CCC CG G AM G CIA CCT CGM MC CC TCC CAT CA GTG G GM CIA Ser Pro Arg Thr Pro Thi Pro Phe Lys His Ala Leu Ala Ala Gln Glu Ile Lys Tyr Gly Pro Leu Lys MET Ltu Pro Gln Thr Pro Ser His Ala Val Glu Asp Leu 3300 CAA GAT GTG AT AG COG GAA TCG GCAT GA TCT (G ATMT GIT CT GAG T CM GMG AMr G cA CM TTA CGG MA MA AMC AMG CAGGG GMGG TCG CA MT Gln Asp Val Ile Lys Arg Glu Ser Asp Glu Ser Gly Ile Val Ala Glu Phe Gln Glu Ser Gly Pro Pro Leu Leu Lys Lys Ile Lys Gln Ala Val Glu Ser Pro Thr 3400 GAG AAA TCG GGA AM TIC TIC TIC CA AM C TIC O GM MC AGC CTG AGC ACC CAA CTG TTC TCG CA GCG TCT CCT GIG GCA GCT GMX CA A ATT CTT MA Glu Lys Ser Gly Asn Phe Phe Cys Ser Asn His Tip Ala Glu Asn Ser Leu Ser Thr Gln Leu Phe Ser Gln Ala Ser Pro Val Ala Asp Ala Pro Asn Ile Leu Thr 3500 AMC ITCT GTT TTI AMG AA CI GTA TCA GA CAT GA GM: MT GTC CC AAA (XEC m ACC GTA CCI AAM A MG CC C GGIGT CCC TIG CAG CA TIC AMT GGT Ser Ser Val Leu El Thir Pro Val Sex Glu Asp Glu Asp Asn Val Leu Lys Ala Phe Thr Val Pro Lys Asn Arg Pro Leu Val Gly Pro Leu Gln Pro Cys Ser Gly 3600 GC TIGG GAM CIA OM TC tGT G(G AAG AA G G CG AMG AMG GX TCC G CCG GCT COG AAA TMC GIG AAC GCG TIC ICA CT CGA ACT CG GIC AMG TA Ala Trp Glu Pro Ala Ser Cys Gly Lys Thr Glu Asp Gln MET Thr Ala Ser Gly Pro Ala Arg Lys Tyr Val Asn Ala Phe Ser Ala Arg Thr Leu Val MET * 3700 TCIIMAA CGMTIAGM AGCIA ASAGCIT TIGCMT AG(CTCrIG CITC(XX TX IX AGAPAGAC G GACACTI= MAAAAGT 3900 3800 TAAAAATAAT AATA2UNAT AAICA TA CILAATTAT TMGLAAIAA7I7MGCA ACACCAGA TaTT''T GFrIurIMG MTCICACm MCT CA ACTAAAAAGM

CCG AAG COT TG TCTG

-r

Hmoogy;

4000 GIGAATITAA AAAAAAAT TACAC CIATI'ITM GXAMAMAACAICMIAIT MTMCC 4100 TATICEATCTIM CIAAMA AGAT TMTAAM avlMGA ATIATiTrGA MGIA= AATIauAP AATmI`LAT CIAMI ACTCAT TIIGrXcITAA 4300 4200 TAATGi'IA AAT TITA TATAAA XA AAGITMTM ITTMAAT AGrlTAT G AAAAAAA AITTAC MrAATITAA 1 1ATITMG A AATIT T'IICM 4400 G C T T'IA T GGTAG ICCTMGTA GIGCXXC GCAATMAXC TATA AAA=AAA GCIX GrrAITMATAM TIIGTAAAT XEAAATICA GAAATG 4500 A MG TITIKCCTIT G7CAATTGAA AAmATIT GI AATIT AACG C I CITITrITMGG ATITImITFIT MACTATXC cCTMI AGATIT TATATG= 4600 T COSTFi IGI CICATAMTI T IGM ACCXGA AGM= CITGOX mairI tTI= TrIrC TIGriG'T TAAACTE GCI

mI

ITATA 7aTAATGM ATITITIMA AAAAATAMAA

TTA

4700 X GXAMIT GCIM CTAII GoACXE InoA TIvi

4800 ITT

MTCMG GGAAAAT MACAAGIA

CAGIT

GXTIGAAT C

FIG. 2. (Legend appears at the bottom

of the opposite page.)

X

AGMCTITC AMAATCTG


3174


INTEGRATION SITE E 122EE

E B BB B

"'_-"

R\V"it'

"V'

VI

'I

N

SEQUE NCES

BE "

E

B

"-''"/ -' /vS v/

XV

v

"v'

/oi %

%

1

-

2tE ABERRANT SPLICE

5

10

15

20

25

30

Kbp FIG. 3. Comparison of the transcription events that generate the normal myb mRNA and the ABPL-2 mRNA. The seven viral-related myb exons are numbered and are represented by solid boxes (5). The open boxes represent nonviral-related exons. The integration site of Mo-MuLV is immediately before the first viral exon. In generation of the ABPL-2 mRNA, transcription is initiated within the Mo-MuLV genome. Splicing occurs normally throughout the processing of the mRNA except just after the sixth viral-related exon where an additional 363 nt are inserted into the message via an aberrant splicing event. The positions of the first three 5' exons as well as the seventh viral-related exon have been determined (5) and are as shown. The exons following the last viral-related exon have not been mapped and are illustrated for diagrammatical purposes. Hatched boxes, viral long terminal repeats.

clones containing the 3' end of the gene from the same library. All contained the identical sequence across the abnormal splicing site. Still other clones with this sequence have been identified through restriction enzyme analysis. We found no clones containing a normal 3' end. Therefore, we conclude that the aberrantly spliced myb mRNA is the vastly predominant myb mRNA in ABPL-2 cells. These conclusions are further confirmed by RNA gel blot analyses using the intron-specific probe as described below. Aberrant Splicing of Intronic Sequences Occurs in Other ABPL Tumors. Most of the ABPL tumors showing aberrations in the myb locus have undergone proviral insertion within the same 1.5-kb segment of the intronic sequences preceding the v-myb-related sequences (5, 8). To verify whether splicing aberrations such as those seen in the ABPL tumor also occur in the other ABPL tumor RNAs, we prepared radioactive probes using the DNA fragment containing the intronic sequences (EcoRI-Aat II fragment) exclusively. As shown in Fig. 4, hybridization of this probe with the RNAs derived from ABPL-1, ABPL-2, and ABPL-4 tumors showed that the aberrant myb RNAs from the three tumors hybridize to this probe whereas little or no hybrid1

28S-

18S-

2 3 4

1

FIG. 4. RNA gel blot analysis of ABPL-1, ABPL-2, and ABPL-4 mRNAs using an intron-specific probe. Lanes: 1, normal mouse thymus; 2, ABPL-2; 3, ABPL-1; 4, ABPL-4. The probe was a 300-bp EcoRI-Aat II fragment derived from the intron that appears in the cDNA sequence. Positions of 28S (5 kbp) and 18S (2 kbp) rRNA are as shown.

ization was observed with RNA from mouse thymus that contains high levels of normal myb mRNA.

DISCUSSION To understand the mechanism of activation of the myb gene, we have compared the structure of the cellular and viral myb-coding sequences (1). These studies showed that during the generation of viral transforming genes, the c-myb sequences have undergone extensive deletions at both the Nand C-terminal ends. The deletion of similar stretches of Nand C-terminal sequences in two different viral isolates raised the possibility that such deletions in the coding regions could be the mechanism for conversion of the normal c-myb gene into its transforming counterpart. In addition to the avian systems, which suggest that deletions are important in the activation of the myb gene, one other mouse model system provides additional support for this hypothesis. Cas-Br Mo-MuLV induces myeloid leukemias in NFS mice and one of the tumor lines, NFS-60, was found to contain a rearranged myb locus. This was found to be due to the integration of the proviral genome toward the 3' end of the myb gene resulting in the premature termination of myb gene transcription (18, 19). The myb-coding sequences in this aberrant mRNA terminate at a point near the end of the sixth viral-related exon that lies close to the site of deletion observed in AMV and E26 virus. In the present communication, we have examined in detail the structure of the c-myb-encoded mRNAs present in the ABPL-2 tumor line. Earlier studies (7) had indicated that these tumors had undergone rearrangements in the myb locus, resulting in the synthesis of abnormal mRNAs. Molecular cloning and structural comparison of the normal and rearranged c-myb DNA sequences revealed that the rearrangements in all ABPL tumors were due to integration of the Mo-MuLV genome into 1.5 kb of cellular DNA immediately upstream from the v-myb-related sequence (5). A detailed characterization of the mouse c-myb-coding exons revealed that this integration occurred in intron sequences that precede the first viral-related exon. The results presented here demonstrate that the viral integration results in the initiation



of transcription within the Mo-MuLV viral sequences and replacement of the c-myb N-terminal sequences by gag sequences. This results in the deletion of a minimum of three coding exons from the 5' end that is accompanied by the generation of aberrantly spliced mRNAs that contain additional sequences coded for by the intron that follows the sixth v-myb-related exon. This is particularly interesting since the deletions in the two viral transforming genes and in the NFS-60 cell line occur in this same region (1, 18). This suggests that the alterations (deletions or insertions) at the C-terminal end play a crucial role in the activation of the oncogenic potential of the ABPL-2 myb gene. The significance of the deletions at the N-terminal end are at present unclear. In the NFS-60 cell line, the 5' end appears to be unaffected, whereas the entire C-terminal end is deleted. This argues that the C-terminal deletions alone might be sufficient to activate this gene. However, it is possible that the deletions at the 5' end could enhance the transforming potential of the c-myb gene that already has an aberration at the 3' end. Fig. 5 summarizes the various alterations of the myb gene in these tumor model systems. With the availability of full-length cDNA clones of the myb gene, these questions can be addressed further by the construction of retroviruses with various mutations to test directly the affects of 5' and 3' deletions on the gene. While this work was in progress, Shen-Ong et al. (19) published the nucleotide sequence of a c-myb cDNA clone derived from the ABPL-2 tumor line. While their results agree with ours regarding the sequences at the 5' end, they did not find the presence of intronic sequences following the sixth v-myb-related exon. Their results indicate normal splicing of the 3' end of the c-myb mRNA in this tumor line. For this reason, we have analyzed 12 different c-myb clones and found only the aberrantly spliced species. Furthermore, a probe derived from the intronic sequences hybridizes well with the aberrant mRNAs found in these cells. The discrepancy could be best explained by assuming that there is a small percentage of normally spliced mRNA in these cells and the clone that Shen-Ong et al. (19) used for sequence analysis is derived from such an mRNA. Regardless of the reason for the discrepancy, it is clear from the experimental data presented here that the ABPL-2 tumor cells contain abundant quantities of c-myb RNAs that have undergone alterations at both the 5' and 3' ends. 1

2

3

a

I

I

AUG

4(kb)

UAG

myb

Splice site

I

coding sequences

I~

FIG. 5. Structural comparison of the myb-coding region of normal c-myb mRNA with activated forms of myb RNA in the AMV and the E26 virus and ABPL-2 and NFS-60 tumor cells. The hatched box in ABPL-2 represents the intronic sequences that disrupt the myb mRNA. Straight lines represent non-myb sequences. Open boxes, myb sequences. Dashed line, splicing event that occurs in the generation of the viral myb mRNA.

3175

The results presented here also have important implications regarding the role of mRNA secondary structure in the processing of high molecular weight precursors to their final form. Potential splice-donor, (Y)6NC I AGGK, and -acceptor sites, MAG I GTRAGT occur frequently along the mRNA molecule. Within the consensus donor and the acceptor sequences only the two underlined nucleotides are essentially invariant (20), whereas there is considerable fluctuation among the other residues. Therefore, the choice of which potential splicing events will predominate during mRNA formation must be governed by factors other than sequence alone. Studies using the technique of in vitro mutagenesis appear to suggest that the secondary structure of precursor mRNAs plays a crucial role in their splicing (refs. 21-23). Thus, deletion of upstream sequences from some of the mammalian genes has resulted in transcripts that are aberrantly spliced in regions that are distant from the site of deletion. Our results strongly support this hypothesis and provide, to our knowledge, the first in vivo example where the deletion of upstream sequences has resulted in splicing aberrations further downstream. We are grateful to Drs. Tom Curran and Kay Huebner for a critical review of the manuscript and Dr. Jim Ihle for helpful suggestions. We also thank Jim Averback for help in computer graphics.

1. Rosson, D. & Reddy, E. P. (1986) Nature (London) 319, 604-606. 2. Rushlow, E. K., Lautenberger, J. A., Papas, T. S., Baluda, M. A., Perbal, B., Chirikjian, J. G. & Reddy, E. P. (1982) Science 216, 1421-1423. 3. Klempnauer, K. H., Gonda, T. J. & Bishop, J. M. (1982) Cell 31, 453-463. 4. Nunn, M. F., Seeburg, P. K., Moscovici, C. & Duesberg, P. H. (1983) Nature (London) 306, 391-395. 5. Lavu, S. & Reddy, E. P. (1986) Nucleic Acids Res. 14, 5309-5320. 6. Potter, M., Reddy, E. P. & Wivel, N. A. (1978) Natl. Cancer Inst. Monogr. 48, 311. 7. Mushinski, J. F., Potter, M., Bauer, S. R. & Reddy, E. P. (1983) Science 220, 795-798. 8. Shen-Ong, G., Potter, M., Mushinski, J., Lavu, S. & Reddy, E. P. (1984) Science 226, 1077-1080. 9. Potter, M., Sklar, M. D. & Rowe, W. P. (1973) Science 182, 592-594. 10. Premkumar, E., Potter, M., Singer, P. A. & Sklar, M. D. (1975) Cell 6, 149-159. 11. Auffray, C. & Rougeon, F. (1980) Eur. J. Biochem. 107, 303-308. 12. Mushinski, J. F., Bauer, S. R., Potter, M. & Reddy, E. P. (1983) Proc. Natl. Acad. Sci. USA 80, 1073-1077. 13. Lehrach, H., Diamond, D., Wozney, J. M. & Boedtker, H. (1977) Biochemistry 16, 4743-4748. 14. Maxam, A. & Gilbert, W. (1977) Proc. Natl. Acad. Sci. USA 74, 560-564. 15. Gonda, T. J., Gough, N. M., Dunn, A. R. & deBlaquire, J. (1985) EMBO J. 4, 2003-2008. 16. Bender, T. P. & Kuehl, W. M. (1986) Proc. Natl. Acad. Sci. USA 83, 3204-3208. 17. Shinnick, T. M., Lerner, R. A. & Sutcliff, J. G. (1981) Nature (London) 293, 543-548. 18. Weinstein, Y., Ihle, J. N., Lavu, S. & Reddy, E. P. (1986) Proc. Natl. Acad. Sci. USA 83, 5010-5014. 19. Shen-Ong, G. L. C., Morse, H. C., Potter, M. & Mushinski, J. F. (1986) Mol. Cell. Biol. 6, 380-392. 20. Cech, T. R. (1983) Cell 34, 713-716. 21. Kihne, T., Wierenga, B., Reiser, J. & Weisman, C. (1983) EMBO J. 2, 727-733. 22. Solnick, D. (1985) Cell 43, 667-676. 23. Archibald, A. L., Thompson, N. A. & Kvist, S. (1986) EMBO J. 5, 957-965. 24. Benton, W. D. & Davis, R. W. (1977) Science 196, 180-182.