the 3' splice site of exon 4 does not change the processing choice. These sequence elements are functionally equivalent. CT-specific splicing requires the ...
MOLECULAR AND CELLULAR BIOLOGY, Feb. 1994, p. 951-960
Vol. 14, No. 2
0270-7306/94/$04.00+0 Copyright © 1994, American Society for Microbiology
Two Different Sequence Elements within Exon 4 Are Necessary for Calcitonin-Specific Splicing of the Human Calcitonin/Calcitonin Gene-Related Peptide I Pre-mRNA CONNIE C. M. vAN OERS, GOSSE J. ADEMA,t HANNIE ZANDBERG, TESSA C. MOEN, AND PIETER D. BAAS* Institute of Molecular Biology and Medical Biotechnology and Laboratory for Physiological Chemistry, Utrecht University, 3584 CH Utrecht, The Netherlands Received 20 July 1993/Returned for modification 3 September 1993/Accepted 5 November 1993
The calcitonin (CT)/calcitonin gene-related peptide I (CGRP-I) gene (CALC-I gene) is subject to alternative tissue-specific processing of its primary transcript. CT mRNA is the predominant mRNA produced in thyroid C cells, whereas CT gene-related peptide I mRNA is the main product in neurons of the central and peripheral nervous systems. The CT-specific exon 4 is surrounded by weak processing sites. In this study we have investigated whether exon 4 sequences are involved in the tissue-specific selection of the exon 4 splice acceptor site. The results indicate that two separate elements, termed A and B, in the 5' part of exon 4 are required for production of CT-specific RNA. These sequences are located between nucleotides 67 and 88 (A) and nucleotides 117 and 146 (B) relative to the 5' end of exon 4. Variation of the distance between these sequence elements and the 3' splice site of exon 4 does not change the processing choice. These sequence elements are functionally equivalent. CT-specific splicing requires the presence of both sequence A and B or duplicates of either sequence element in exon 4. The effect of these sequences on the RNA processing choice is overruled by mutation of the CT-specific uridine branch acceptor nucleotide into a commonly preferred adenosine residue.
splicing efficiency in nonneural 293 cells (4) and in the predominant production of CT mRNA in otherwise CGRPI-producing F9 cells (1). Replacement of the weak exon 4 poly(A) site with stronger poly(A) sites also results in an increase in CT-specific processing (59). The presence of weak processing sites at both ends of the CT-specific exon 4 may be a prerequisite for the proper regulation of alternative processing of the CALC-I pre-mRNA. cis-acting elements involved in alternative processing are often located in the processing signals. However, sequences located within exons can also influence the processing choice of alternatively processed pre-mRNAs both negatively or positively. In vivo and in vitro experiments performed on a large number of genes have indicated a role for exon sequences in splice site selection (15, 16, 21, 27, 31, 32, 36, 38, 43, 46, 55, 58, 60, 62). With the exception of Drosophila doublesex pre-mRNA and the Drosophila P element, the role of exon sequences in splice site selection remains unclear. Two trans-acting factors, transformer and transformer-2 bind to regulatory elements in the femalespecific exon of the doublesex pre-mRNA, thereby promoting spliceosome formation (33, 50, 57). The somatic inhibition of Drosophila P-element third intron splicing involves RNA-protein complexes bound to an exon pseudo-5' splice site (11, 51). This complex interferes with binding of Ul small nuclear ribonucleoprotein (RNP) to the authentic 5' splice site. In several other cases, mutations in exon sequences may result in a change in the secondary structure of the pre-mRNA, which either sequesters or exposes the splice sites (21, 36). In this study we have investigated the involvement of exon 4 sequences in the alternative RNA processing of the human CALC-I gene. Transfection studies with several deletion and substitution mutants show that specific exon 4 sequences are required for CT-specific processing.
Alternative splicing of mRNA precursors is an important mechanism in regulation of gene expression (for reviews, see references 40 and 41). The differential incorporation of exons into mature mRNA is often under tissue-specific or developmental control. Progress has been made towards understanding the biochemistry of pre-mRNA splicing (for a review, see reference 30). However, the mechanism by which the splicing machinery selects the correct pairs of splice sites as well as identification of cis- and trans-acting factors involved in tissue-specific or developmentally regulated RNA processing is still a major issue in splicing research. The alternative processing of the human calcitonin (CT)/ calcitonin gene-related peptide I (CGRP-I) pre-mRNA (CALC-I pre-mRNA) results in cell-type-specific expression of two different proteins. The CT mRNA is generated in thyroid C cells by splicing of the three most 5' exons to the CT-encoding exon 4. The poly(A) site at the end of exon 4 is used in this RNA. CGRP-I mRNA results from splicing of the first three common exons to exon 5 and the polyadenylated exon 6 in particular neural cells (5, 22, 49, 54). We have previously studied the alternative processing of a minigene, containing the exon 3 to exon 5 region of the human CALC-I gene both in vivo (4) and in vitro (7, 8). Analysis of the branchpoint used in the splicing of exon 3 to exon 4 showed that a uridine residue is used as the major branch acceptor nucleotide (3). Mutation of this unusual uridine branch acceptor nucleotide into a commonly preferred adenosine residue results in an increase in CT-specific * Corresponding author. Mailing address: Institute of Molecular Biology and Medical Biotechnology and Laboratory for Physiological Chemistry, Utrecht University, Padualaan 8, 3584 CH Utrecht, The Netherlands. Phone: (0)30-533430. Fax: (0)30-537797. t Present address: Division of Immunology, Cancer Institute, 1066 CX Amsterdam, The Netherlands.
951
952
MOL. CELL. BIOL.
vAN OERS ET AL.
MATERUILS AND METHODS
A
pRSV dI3 86NcoI
Nhel Ncol
Nsil 498 +
Plasmids. Construction of the CALC-I minigene fragments and their introduction into the expression vector pRSV328A have been described previously (4). The exon 4 deletion mutants pRSV dI3E4(50-251) and pRSV dI3E4(152-251) were constructed by using fragments made by PCR on pRSV dI3I4P as a template with the following oligonucleotide primers: pRSV dI3E4(50), 5'TAATCGATGCATGCATGCA AGTACTCAGATTA3', and pRSV dI3E4(152), 5'TAATC GATGXCAICCAAGTCGCTGGACATATCC3'. These primers were used in combination with an oligonucleotide complementary to part of the multiple cloning (MC) site (HindIII primer, 5'CACCTCCAAfL(lCGACTC3'). The underlined HindIII and NsiI restriction sites were used for cloning the PCR fragments into the HindIII and NsiI sites of pRSV dI3 (Fig. 1). pRSV dI3E4(251-443) was constructed by using fragments made by PCR with the exon 4 primer E4(251),
5'TAATCGATGCATTGTCCTGCTTCTGAATGTGC3', and the intron 4 primer, 5'TGTCCAGCcTAGCCTAGGGT3'. The underlined NsiI and NheI sites were used for cloning the PCR fragment into the NsiI and NheI sites of pRSV dI3. The CALC-I minigenes with the branch acceptor nucleotide mutation pRSV dI3E4(50-251)U-A and pRSV dI3E4(152251)U-A were made by PCR on pRSV dI3(U-A). pRSV dE4(50-251) and pRSV dE4(152-251) were made by cloning the NcoI fragments of pRSV dI3E4(50-251) and pRSV dI3E4(152-251) into the NcoI sites of pRSV WT. pCTS9 was made by cloning the SalI-NsiI fragment (the SalI restriction site is located in the multiple cloning site in front of exon 3) of pRSV WT in the BssHII-NsiI sites of pG4hCT9, which contains the complete CALC-I gene, including its own promoter (1). The SalI and BssHII sites were made blunt ended by using the Klenow enzyme. Because of cloning procedures, pCTS9 contains 24 additional nucleotides (nt) in exon 3. pCTS9 dE4(50-251) and pCTS9 dE4(152-251) were made by inserting the SalI-NsiI fragments of pRSV dE4(50251) and pRSV dE4(152-251) into the BssHII-NsiI sites of pG4hCT9. The positions of the other exon 4 deletions in pRSV dI3 are indicated in Fig. 1. pRSV dI3E4(117-146) was constructed using fragments made by PCR with the primers HindIII; E4(117), 5'TAAGATCTjCGAfGiCTCCAACCCCAA TTGCAGT3'; E4(146), 5'TAAGATcCTCGAGACTTGGAG AGAGACCATCG3'; and NsiI, 5'AAACCACATGzCATCA AGTTA3'. One PCR fragment was made by using primers HindIII and E4(117) in a PCR on pRSV dI3 as a template and subsequent digestion with HindIII and XhoI, which cleave the underlined sites. The second PCR fragment was made by using E4(146) and the NsiI primer and subsequent digestion with XhoI and NsiI. Both fragments were inserted simultaneously into the HindIII and NsiI sites of pRSVdI3. Because of the cloning procedures, a XhoI site was introduced at the position of the deletion. The other deletion mutants were made by analogy to pRSV dI3(117-146) by using the appropriate primers. Insertion mutants were made by inserting the double-stranded oligonucleotides A, 5'TCGAGACTTCAA CAAGTTTCACAC3'; B, 5'TCGAGCACCTGGAAAGAAA AGGGATATGTCCAGC3'; or AB, 5'TCGAGACTTCAA CAAGTTTCACACGAGCACCTGGAAAGAAAAGGGA TATGTCCAGC3' in the XhoI site of a deletion mutant. The nucleotide sequences of all constructs made by PCR were confirmed by DNA sequence analysis. Transfection and RNA analysis. Cell culture, DNA transfection of 293 cells, and RNA isolation have been described
4 -
E4d
A
4b
pA
MC
1
E4(50-251)
F-5H0
251]
I
1l46
11
87
11 4
E4(63-101)
1101
E4(63-88)
88
35
E4(1 3-37)
B 137
CGRP
l
CT CT+CGRP
1 -
1
68
E4(35-68)
CT
1
1251
1521
E4(87-1 14)
F43
1251
E4(1 52-251)
114
E4(63-1 14) 87
E4(87-146) E4(63-1 46)
E4(35-146) E4(13-146)
-PA CGRP
E4 E4(251 -443)
E4(1 17-146)
56
U65
.-
-
EX4
RSV
m
5
B
146
1146
1146 1146
1
CT CT+CGRP CT+CGRP CT CT CT+CGRP CT+CGRP CGRP+(CT) CGRP+(CT) CGRP+(CT)
Oligo A: 5'TCGAGACTTCAACAAGTTTCACAC3' Oligo B: 5'TCGAGCACCTGGAAAGAAAAGGGATATGTCCAGC3' Oligo AB: 5'TCGAGACTTCAACAAGTTTCACAC GAGCACCTGGAAAGAAAAGGGATATGTCCAGC3'
FIG. 1. (A) Schematic representation of CALC-I minigenes. The CALC-I sequences are inserted in between the RSV promoter and the rabbit ,-globin polyadenylation site of pRSV328A. The sizes of the introns and exons are indicated in nucleotides (not drawn to scale). The numbers of the last nucleotide before a deletion in exon 4 and the first nucleotide after the deletion are indicated. The polyadenylation sites as well as relevant restriction enzyme sites are indicated. The primers used in the RT-PCR analysis (MC, RSV, E4a, E4b, E4d, and E5) are indicated. The 5' ends of the MC and RSV primers are located 107 and 139 nt, respectively, from the 3' end of exon 3. The 5' ends of the exon 4 antisense primers E4a, E4b, E4c, and E4d are located 206, 445, 38, and 310 nt, respectively, from the start of exon 4. The 5' end of the exon 5 antisense primer is located 113 nt from the beginning of exon 5. The products formed after transient transfection of the plasmids to 293 cells (CT and CGRP) are indicated. (B) Sequences of the oligonucleotides used for construction of insertion mutants. Underlined in oligonucleotide A is a sequence which shows similarity to the regulatory sequence present in the female-specific exon 4 of Drosophila doublesex pre-mRNA. In oligonucleotide B, a stretch of 14 consecutive purine residues is underlined.
before (4). For the reverse transcriptase (RT)-PCR analysis, 2.5 ,ug of RNA sample was reverse transcribed for 60 min at 45°C in a reaction mixture containing lx PCR mixture (10 mM Tris-HCl [pH 8.3], 50 mM KCl, 1.5 mM MgCl2, 0.001% gelatine), 0.5 mM deoxynucleoside triphosphates, 20 U of RNasin, 15 U of avian myeloblastosis virus RT, 100 ng of exon 4 primer, and 100 ng of exon 5 primer (5'CATTGGTG GGCACAAAGTTG3'). As exon 4 primers, we used E4a, 5'GAGGAGTFlTAGTTGGCATTC3'; E4b, 5'ACAGAGGA GCTCTGATGAC3'; E4c, 5'AGAAGCAGGACAACCGCT TAGATC3'; and E4d, 5'CAAGGAAAGCCACCAATA3', as indicated for each figure. Subsequently, cDNAs were amplified by adding 1.5 U of Taq polymerase, PCR mixture (final concentration, lx), and the MC primer 5'TCGACTCT AGAGGATCCC3' or the Rous sarcoma virus (RSV) primer
VOL. 14, 1994
5'ATITGGTGTGCACCTCCAAG3'. The 5' ends of the MC and RSV primers are located 107 and 139 nt, respectively, from the 3' end of exon 3. The 5' ends of the exon 4 antisense primers, E4a, E4b, E4c, and E4d, are located 206, 445, 38, and 310 nt, respectively, from the start of exon 4. The 5' end of the exon 5 antisense primer is located 113 nt from the beginning of exon 5. DNA samples were taken after different numbers of PCR cycles in the linear range of amplification, as indicated in the figures. The reaction products were treated with 50 ,ug of RNase per ml run on an agarose gel, and analyzed by Southern blotting and subsequent hybridization with different oligonucleotide probes (4) as described before for Northern (RNA) blots (9). Different numbers of PCR cycles were analyzed to ensure that the RT-PCR assays were performed in the linear range of amplification, and different exposure times of the X-ray films were obtained to ensure a linear response from the bands derived from the spliced products CT and CGRP RNA. To verify the semiquantitative nature of our RT-PCR analysis, the ratios of CT to CGRP spliced products obtained after amplification of several different RNAs isolated after transfection of cell lines with different CALC-I (mini)gene constructs were compared with the ratios of CT to CGRP RNA determined by Northern blotting. RNA isolated from a cell line (TT) derived from a human medullary thyroid carcinoma, which expresses endogenously the CALC-I gene, resulting in approximately equal amounts of CT and CGRP-I mRNA was also investigated by both methods. Similar CT/CGRP ratios were obtained with both methods. In addition, mixing experiments were performed with RNA obtained after transfection of 293 cells with pRSV dI3E4(35-146) (CGRP producer) and PRSV dI3E4(35-146) AB (CT producer) (see Fig. 5). The observed CT/CGRP ratio after RT-PCR amplification was in agreement with the input ratio of the RNAs. The ratio of CT to CGRP spliced products did not change significantly by varying the number of PCR cycles, and only small variations were observed with different primer combinations (1).
RESULTS The presence or absence of specific exon sequences can determine the RNA processing choice of alternatively processed pre-mRNAs (15, 16, 21, 27, 31, 32, 36, 38, 43, 46, 55, 58, 60, 62). We have investigated the influence of exon 4 sequences on the processing choice of the CALC-I premRNA. Transient transfection of a minigene (pRSV WT) containing the exon 3 to exon 5 region of the human CALC-I gene to 293 cells results in the predominant production of CT-specific RNA (CT RNA) (4) (Fig. 2, lane 1). After transfection, RNA was isolated and analyzed in an RT-PCR analysis with an exon 4 and an exon 5 primer for the RT reaction and addition of the RSV primer for the subsequent PCR reaction. The PCR products were analyzed on a Southern blot by hybridization to an exon 3-specific probe. Introduction of a large deletion (+860 nt) in intron 3 (pRSV dI3 [Fig. 1]) does not change the processing pattern (Fig. 2, lane 2). CT RNA is the predominant product after transient transfection of pRSV dI3 to 293 cells. However, introduction of an additional 416-nt deletion in exon 4 changes the processing pattern dramatically. Transient transfection of pRSV dI3E4 (Fig. 1) results in the exclusive production of CGRP-I-specific RNA (CGRP RNA) instead of CT RNA (Fig. 2, lane 3). This result suggests that exon 4 sequences are involved in alternative CALC-I pre-mRNA processing. Exon 4 sequences required for CT-specific processing are
REGULATION OF CALC-I PRE-mRNA PROCESSING
953
CV')
Co
LO
0 LO
)
CY)u > z CD) Er
CLJ
f cn co CV}
> > i-z>> a: CD) (i) Ct) cCI) Ircrc ca CL Qa a
a
PRE- -
PRE4rn a
~~-603
-
CT__-
*
CGRP-0.1
6
m~-Q~PRE _271
234 -194
2 3 4 5
FIG. 2. Effects of deletion of exon 4 sequences on the processing choice. After transient transfection to 293 cells, RNA was isolated and analyzed by RT-PCR. The RT-PCR analysis was performed for 18 cycles by using an exon 4 primer, the exon 5 primer (E5), and the RSV primer. As exon 4 primers, we used E4a (lanes 1, 2, and 4), E4c (lane 3), or E4b (lane 5). E4c, which is complementary to the sequence generated by the large (416-nt) deletion in exon 4 (pRSV dI3E4), is not indicated in Fig. 1. The reaction products were analyzed on a Southern blot and hybridized to an exon 3-specific oligonucleotide probe. The positions of the precursors (PRE), the CT and CGRP products, and the relevant 4X174 HaeIII restriction DNA fragments (in base pairs) are indicated. The length of the precursor band (PRE) in lanes 1 to 5 is + 1,350, 487, 319, 487, and 526 bp, respectively. The length of the band derived from CT cDNA in lanes 1 to 5 is 345, 345, 177, 345, and 384 bp, respectively. The length of the band derived from CGRP cDNA is in all lanes 252 bp. The band present just below the precursor represents heteroduplex DNA molecules. Heteroduplex DNA consists of DNA derived from spliced CT RNA hybridized to DNA derived from precursor RNA. The nature of the products was confirmed by hybridization to a CT-specific oligonucleotide probe (E3-4) and a CGRP-specific oligonucleotide probe (E3-5). These probes detect products in which exon 3 is spliced to exon 4 or exon 5, respectively.
located in the 5' end of exon 4. To locate cis-acting sequences in exon 4, we have introduced several deletions in exon 4 of the minigene with a deletion in intron 3 (pRSV dI3). The largest exon 4 deletion removes 416 nt and leaves 26 nt at the most 5' end and 56 nt at the most 3' end of exon 4 intact, including the exon 4 poly(A) site (pRSV dI3E4). Two new deletion mutants in which either the 5' or the 3' end of exon 4 was deleted were made. pRSV dI3E4(50-251) contains a 200-nt deletion in the 5' part of exon 4 and leaves the first 50 nt of the exon intact (Fig. 1). pRSV dI3E4(251-443) contains a 186-nt deletion in the 3' part of exon 4 and leaves the most 3' 56 nt of the exon intact (Fig. 1). The processing patterns observed after transient transfection of both plasmids to 293 cells are dramatically different. The deletion of the 3' part of exon 4 results in the predominant production of CT RNA (Fig. 2, lane 4), as was observed after transfection with a plasmid containing the complete exon 4 (pRSV d13). However, after deletion of the 5' part of exon 4, CGRP RNA was the only observed processing product (Fig. 2, lane 5), as after transfection with a plasmid containing the 416-nt deletion in exon 4. No CT RNA could be detected. The effect of the exon 4 deletions on the RNA processing choice could be
cu ~ ~ ~CD
VAN OERS ET AL.
954
_
-
-
1'
cN
m
t
-
m
? Y c? 7
0
7
7
r-s
C')
MOL. CELL. BIOL.
~
c's
N 00
C)
LU
LL LL C' CO1
UD
UD
:5:5:'a
S2
-603
PRE[
I
r
va
CP: RE as
-310 ~~~ CT E[:' 404p CT: _*_ _-w~~-271 -234 CGRP _ _
:
1
2
-194
3
4
5
6
7
-310
-271 CGRP-P-
_ -194
8
9
10
FIG. 3. Effects of small deletions in the 5' part of exon 4 on the processing choice. The RT-PCR analysis of exon 4 deletion mutants indicated above each lane is shown. The RT-PCR analysis was performed for 14 cycles by using primers E4a (lanes 2 to 9) or E4d (lanes 1 and 10) and E5 together with the MC primer. The length of the precursor band (PRE) in lanes 1 to 10 is 461, 436, 427, 437, 433, 422, 431, 401, 409, and 481 bp, respectively. The length of the band derived from ClT cDNA in lanes 1 to 10 is 319, 294, 285, 295, 291, 280, 289, 259, 267, and 339 bp, respectively. The length of the band derived from CGRP cDNA is in all lanes 220 bp. The blots were hybridized with an exon 3-specific oligonucleotide probe. Dots indicate the positions of splicing reaction c. In this RNA, a 5' donor site in exon 4 is spliced to the exon 5 acceptor site in addition to splicing of exon 3 to exon 4 (2). Symbols are as in Fig. 2.
due to removal of specific exon 4 sequences or to the decreased size of exon 4. However, the length of both the 5' end and the 3' end deletions are very similar. Therefore, we conclude that specific sequences in the 5' part of exon 4 are required for the CT-specific processing after transient transfection of the CALC-I minigene to 293 cells. Two regions in exon 4 are involved in CT-specific processing of the CALC-I pre-mRNA. We have further characterized the sequences involved in CT-specific processing of the human CALC-I pre-mRNA by construction of small deletions in the 5' part of exon 4 as listed in Fig. 1 and described in Materials and Methods. They are located between nt 13 and 251. The deletions each span approximately 25 to 40 nt, except for pRSV dI3E4(152-251), which contains a 98-nt deletion spanning from nt 152 to 251. All deletion mutants were transiently transfected to 293 cells, and RNA was analyzed by the RT-PCR analysis. Transfection of pRSV dI3E4(152-251) results in the production of CT RNA (Fig. 3, lane 1). We therefore concluded that the cis-acting sequences are located in the 5' part of exon 4, upstream of nt 152. Our further exon 4 deletions are located within this region. Transfection of plasmids pRSV dI3E4(13-37) or pRSV dI3E4(35-68), of which the deletions are located most 5' in exon 4, results in CT-specific processing (Fig. 3, lanes 2 and 3). In contrast, after transfection of plasmids pRSV dI3E4(63-101) or pRSV dI3E4(117-146), both CT RNA and CGRP RNA could be detected (Fig. 3, lanes 6 and 7). cis-acting sequences are located in both removed regions. The deleted sequences could be part of a single cis-acting element or consist of separate elements. Deletion of a sequence element which resides between these two deletions as in pRSV dI3E4(87114), however, does not result in an increase in the CGRP/CT ratio. Transfection of this construct results in the production of only CT RNA (Fig. 3, lane 5). Since the deletion in pRSV dI3E4(87-114) partially overlaps the deletion of pRSV dI3E4(63-101), these data imply that the actual
cis-acting region is located between nt 63 and 88, as was shown with plasmid pRSV dI3E4(63-88). Transfection of this plasmid also results in both CT RNA and CGRP RNA (Fig. 3, lane 4). The deletion in pRSV dI3E4(63-101) also partially overlaps the deletion in pRSV dI3E4(35-68), which produced only CT RNA after transient transfection to 293 cells. Therefore, we concluded that two separate regions located between nt 67 and 88 and nt 117 and 146 function as cis-acting elements in the tissue-specific alternative RNA processing of the human CALC-I pre-mRNA. Deletion of nt 50 to 251 results in the exclusive production CGRP RNA (Fig. 2, lane 5), whereas the smaller deletions from nt 63 to 88, 63 to 101, or 117 to 146 result in the production of both CGRP RNA and CT RNA (Fig. 3, lanes 4, 6, and 7). The effect of both deletions on the RNA processing choice may therefore be additive. Larger deletions were made to investigate the additive effect of both deletions. In pRSV dI3E4(63-114) and pRSV dI3E4(87-146), the size of the deleted fragment was increased compared with the former deletions, but they still contain only one of the cis-acting elements. Transfection of these plasmids results in the formation of both CT RNA and CGRP RNA, as after transfection of their corresponding smaller deletions (Fig. 3, lanes 8 and 9). A slight increase in the CGRP/CT ratio was observed after transfection of pRSV dI3E4(87-146) compared with pRSV dI3E4(117-146). Slight variations of the CGRP/CT ratio were also observed in different transfection experiments. In pRSV dI3E4(63-146), the exon 4 deletion included both regions. The CGRP/CT ratio observed after transfection of this plasmid to 293 cells was increased compared with the ratio after transfection of plasmids containing deletions of one of the individual elements. Only minor amounts of CT RNA could be observed (Fig. 3, lane 10). We conclude that both regions cooperate to promote CT-specific processing in 293 cells. The effect of deletion of these regions on the RNA processing choice is additive. The two regions are functionally equivalent. Deletion of specific regions of exon 4 results in an increase in CGRP-Ispecific splicing. This could be due to a change in secondary structure of the RNA or to the removal of specific sequences required for binding of factors which stimulate CT-specific processing. We have excluded the possibility that the effect is simply due to the decrease in length of exon 4. To investigate whether reinsertion of the deleted sequences could restore the exclusive production of CT RNA, we made several insertion mutants. Oligonucleotides were inserted into the XhoI sites which were created in exon 4 during the construction of the deletions (Materials and Methods). The sequences of the oligonucleotides used for the insertion mutants are shown in Fig. 1B. Oligonucleotide A contains the sequence from nt 67 to 88, flanked by XhoI sites. Oligonucleotide B contains the sequence from nt 117 to 146, flanked by XhoI sites. Oligonucleotide AB contains the
sequence of oligonucleotide A directly followed by the sequence of oligonucleotide B, and the total is surrounded by XhoI sites. Insertion of oligonucleotide B into pRSV dI3E4(117-146) results in a construct which has almost a wild-type exon 4 except for the presence of two additional XhoI sites [pRSV dI3E4 (117-146)B]. Insertion of oligonucleotide A into pRSV dI3E4(63-101) does not exactly result in a wild-type exon 4, since oligonucleotide A contains only 20 nt where the deletions spans 37 nt. Also in this construct, two XhoI sites were present. The resulting insertion mutants were transiently transfected to 293 cells, and the results are shown in Fig. 4 (lanes 2 and 5). Both insertion mutants produced predominantly CT RNA, which was expected
REGULATION OF CALC-I PRE-mRNA PROCESSING
VOL. 14, 1994 m
)
C\3
6 6 0l) Lo
LU
>
L.L LU
-
U) CC
-Q-o : U})U)
U) U?
aN a
aQ-
CNJ LLJ 1:
1
D1:}
LO
c
O If)
Lo
w
Lw
0)
0D)
C\
C)
a)) u)
& -
cU u) C) Ca C-) a a
>
CO CE
a CL PRE CC _-
a
-603 -
PRE
PRE -
__ mm
-603
603
CT
L:
0
Om me
CT -__. 2710
310
-31 0 -271 -234 -194
CGRP -Ol-
-
-234
CGRP-0-
2
3 4
FIG. 6. Effects of branchpoint mutation on the processing choice of exon 4 deletion mutants. The RT-PCR analysis after transfection of the plasmids indicated above each lane is shown. RT-PCR was performed for 18 cycles in the presence of exon 4 primer E4b and the exon 5 primer together with the MC primer. The length of the precursor band (PRE) in lanes 1 to 4 is 494, 494, 596, and 596 bp, respectively. The length of the band derived from CT cDNA in lanes 1 to 4 is 352, 352, 454, and 454 bp, respectively. The length of the band derived from CGRP cDNA in all lanes is 220 bp. Southern blots were hybridized to an exon 3-specific oligonucleotide probe. Symbols are as in Fig. 2.
residue is used as the major branch acceptor nucleotide in the splicing of exon 3 to exon 4 (3). Mutation of this unusual branchpoint into a more commonly preferred adenosine residue results in an increase of CT RNA production after transient transfection to 293 cells (4). We tested whether deletion of part of exon 4 still results in the production of CGRP RNA when the branchpoint was mutated to an adenosine residue. In the construct with the deletion from nt 50 to 251 [pRSV dI3E4(50-251)] the uridine branch acceptor nucleotide was mutated to an adenosine, which resulted in plasmid pRSV dI3E4(50-251)U-A. pRSV dI3E4(50-251) was chosen because after transfection of this construct, no CT RNA could be observed (Fig. 6, lane 1). Therefore, the putative effect of a branchpoint mutation may be most clearly observed with this deletion. As a control, we used plasmid pRSV dI3E4(152-251), which contains a 98-nt deletion but results in the production of CT RNA exclusively (Fig. 6, lane 3). In this plasmid, we also introduced the U-A branchpoint mutation. The resulting plasmids were transfected to 293 cells, and the results are shown in Fig. 6. Mutation of the branchpoint in pRSV dI3E4(50-251) causes a switch from CGRP RNA to CT RNA production (Fig. 6, lanes 1 and 2). Therefore, the effect of the exon 4 deletion on splice site selection is dependent on the presence of a weak splice acceptor site in front of exon 4. Mutation of the branchpoint in pRSV dI3E4(152-251) also results in an increase in CT-specific processing (Fig. 6, lane 4). All deletions in exon 4 we have tested so far were introduced into a minigene which also contained a deletion in intron 3. The intron deletion might have an effect on the processing choice of the exon 4 deletion mutants. To investigate this possibility, the deletion from nt 50 to 251 was introduced into a minigene without intron 3 deletion. As a
CGRP _
-271
1 34
_
-9
-194
1
1
271
2
3
4
5
FIG. 7. Effects of exon 4 deletions in the minigene without intron 3 deletion and in the complete CALC-I gene on the processing choice. RT-PCR analysis after transfection of the plasmids indicated above each lane is shown after 18 cycles. RT-PCR analysis was performed in the presence of exon 4 primer E4b and the exon 5 primer together with the MC primer. The length of the precursor band (PRE) in lanes 1 to 5 is ±1,350, ±1,450, ±1,550, ±1,350, and ±1,450 bp, respectively. The length of the band derived from CT cDNA in lanes 1 to 5 is 352, 454, 552, 352, and 454 bp, respectively. The length of the band derived from CGRP cDNA in all lanes is 220 bp. Southern blots were hybridized to an exon 3-specific oligonucleotide probe. Dot indicates the product of splicing reaction c (2). The deletions in exon 4 remove the donor splice site used in reaction c. Symbols are as in Fig. 2.
control, again the deletion from nt 152 to 251 was used. The resulting plasmids [pRSV dE4(50-251) and pRSV dE4(152251)] were transfected to 293 cells. The results show the same effect of the deletions on the processing choice as observed after transfection of the plasmids with the intron 3 deletion. The presence of the deletion in exon 4 from nt 50 to 251 in the minigene without intron 3 deletion results in the exclusive production of CGRP RNA (Fig. 7, lane 1), whereas transfection with pRSV dI3E4(152-251) results in the predominant production of CT RNA (Fig. 7, lane 2). Similar results were obtained with other exon 4 deletion mutants (data not shown). To exclude the possibility that the effect of the exon 4 deletions on the processing choice is limited to the minigene, the same deletions were introduced in the complete CALC-I gene. After transient transfection of the complete CALC-I gene (pCTS9) to 293 cells, both CT RNA and CGRP RNA could be observed (Fig. 7, lane 3). Introduction of the deletion in the 5' part of exon 4 [pCTS9dE4(50-251)], however, resulted in the exclusive production of CGRP RNA (Fig. 7, lane 4). Transfection with pCTS9dE4(152-251) results in a similar CT/CGRP ratio, as observed after transfection with pCTS9 (Fig. 7, lane 5). Therefore, we conclude that the effect of the deletions in exon 4 on the processing choice is not restricted to the minigene but also occurs with the complete CALC-I gene. DISCUSSION We have investigated the role of exon 4 sequences in the alternative processing of the human CALC-I gene. Transient transfection to 293 cells of a minigene, containing the exon 3
VOL. 14, 1994
REGULATION OF CALC-I PRE-mRNA PROCESSING
to exon 5 region of the human CALC-I gene, results predominantly in the production of CT-specific RNA. Introduction of a large deletion in the 5' end of exon 4 [pRSV dI3E4(50-251)] changes the processing pattern from CT RNA to CGRP RNA. The effect of this deletion is not dependent on its presence in the minigene. When the same deletion was introduced in the complete CALC-I gene, this also resulted in the production of only CGRP RNA. Transfection of the complete CALC-I gene to 293 cells results in contrast to the minigene in the production of CT and CGRP-I mRNA. The difference between the pattern of splicing detected upon transfection of the minigene and the entire CALC-I gene is currently under investigation. CGRP-specific processing in 293 cells requires the presence of sequences in the 3' end of exon 5 and/or exon 6. Deletion analysis of the 5' region of exon 4 revealed that at least two regions are involved in the alternative processing of the human CALC-I gene. The first region is located between nt 67 and 88, and the second region is located between nt 117 and 146. We have termed these regions A and B, respectively. Deletion of one of these regions did not result in a complete switch from CT RNA to CGRP RNA. Both these processing products were observed. Deletion of both regions, however [pRSV dI3E4(13-146), pRSV dI3E4(35-146), and pRSV dI3E4(63-146)], results in a further shift to the production of CGRP RNA. Only minor amounts of CT RNA
exon stimulates splicing of the preceding intron. However, the effect on the RNA processing of our largest exon 4 deletion (pRSV dI3E4) cannot be suppressed by replacement of the exon 4 poly(A) site with a strong poly(A) site (59). Replacement of the exon 4 poly(A) site with a strong splice donor site also does not result in a large increase in CTspecific processing (data not shown). In a model in which the alternative processing of the CALC-I gene is positively regulated, a trans-acting factor would be required to activate the weak exon 4 splice acceptor site. Deletion of exon 4 sequences required for binding of this putative factor results in CGRP-I-specific processing. Reinsertion of the target sequences in different contexts should reverse the splicing pattern. Oligonucleotide AB was inserted into three different deletion mutants in which both sequence elements A and B had been deleted [pRSV dI3E4(63-146), pRSV dI3E4(35-146), and pRSV dI3E4(13-146)]. All three resulting insertion mutants restored CT-specific processing. These results indicate that indeed both elements A and B positively activate CT-specific processing and that their position relative to the exon 4 acceptor site can be varied. Insertion of oligonucleotide A in the negative orientation leads to a small stimulation of CT-specific processing. Insertion of oligonucleotides AB and B in the negative orientation creates a new 3' splice acceptor site within exon 4 because of the presence of 14 consecutive pyrimidine residues. After transfection of these constructs, only splicing of exon 3 to the newly introduced 3' splice site in exon 4 and to exon 5 could be observed. Also, insertion of 4)X DNA fragments does not stimulate CT-specific splicing. CGRP-I-specific splicing is the predominant reaction, except when cryptic splice sites are present within the inserted (X sequence. CT-specific splicing is only possible when a cryptic donor splice site in the 4)X fragment is already spliced to exon 5 in the same RNA molecule (unpublished data). The involvement of exon sequences in alternative RNA processing has been shown in many genes, when they act positively or negatively on the processing of that particular pre-mRNA (15, 16, 21, 27, 31, 32, 36, 38, 43, 46, 55, 58, 60, 62). An example of a positively regulated gene is the Drosophila doublesex (dsx) pre-mRNA. The structure of this RNA resembles the CALC-I pre-mRNA (10). In the dsx pre-mRNA, the male-specific splicing of exon 3 to exon 5 is the default choice. Female-specific splicing of exon 3 to exon 4 and polyadenylation at exon 4 requires the presence of the transformer and transforner-2 gene products (10, 33, 34, 42, 44, 50). The female-specific acceptor site of dsx has a suboptimal polypyrimidine stretch that results in poor recognition of the acceptor site in males (10, 34). Regulation by tra and tra-2 is the result of activation of the female-specific 3' splice site and is dependent upon a 13-nt repeat, TC(T/ A)(T/A)C(A/G)ATCAACA, which is present six times (10, 33, 34, 43, 50). The tra and tra-2 proteins are able to bind to this cis regulatory element in exon 4 (33, 35, 57). Hoshijima et al. (34) examined deletions of different subsets of the six repeats and found that the number of repeats correlated with the efficiency of activation. The presence of only one repeat results in only low amounts of female-specific processing (34, 57). The presence of two or more copies results predominantly in female-specific processing (33-35, 50, 57). In the CALC-I exon 4, we have identified two different sequence elements which cooperate in the activation of CT-specific processing. There is no sequence similarity between these two elements. The presence of only one of these elements results in the production of both CT RNA and CGRP RNA, whereas the presence of both or a double
were detected. We therefore conclude that the effect of deletion of both regions is additive. A complete shift to the production of CGRP RNA was observed only when nt 146 to 251 in addition to nt 67 to 146 were also deleted. Although deletion of nt 152 to 251 alone resulted in the production of CT RNA, this region may have some influence on the
processing choice. However, the major cis-acting regions are located between nt 67 and 146. Cote et al. (17, 19) have reported that sequences present at positions 18 to 45 relative to the 3' splice site of exon 4 are involved in CT-specific splicing. In our constructs, deletion of sequences between nt 13 and 68 has no influence on the processing choice. Differences in the constructs, substitutions versus deletions of exon 4 sequences, and use of different cell lines may all account for these conflicting results. The effect of deletion of exon 4 sequences is dependent on
the presence of the authentic uridine branch acceptor nucleotide in front of exon 4. Previously we have shown that mutation of the unusual CT-specific branch acceptor into a commonly preferred adenosine residue results in a switch of the processing choice from CGRP RNA to CT RNA in F9 cells (1). The U-A branchpoint mutation alters the weak authentic CT-specific splice acceptor site into a relatively strong site which results in exon recognition in F9 cells. Similar results were obtained with the rat CALC-I premRNA after replacement of the CT-specific acceptor site with the exon 3 acceptor site (23). The branchpoint mutation also completely abolished the effect of the exon 4 deletion on the processing choice in pRSV dI3E4(50-251). The production of CT RNA after transfection of this mutant suggests that the exon 4 deletions exhibit their effect on 3' splice site selection and not on RNA stability. Mutation of the branchpoint in other constructs (for example, pRSV dI3E4) results in a strong increase in CT-specific processing but not in a complete reversal of the processing pattern (not shown). At the end of exon 4, a weak polyadenylation site is present (59). According to the exon definition model proposed by Robberson et al. (47), the presence of a functional poly(A) site or a consensus 5' splice site at the end of an
957
958
VAN OERS ET AL.
copy of one of them results in the production of predominantly CT RNA. Therefore, both sequence elements enhance CT-specific splicing in a similar and additive manner. We have looked for sequence similarity between the CALC-I exon 4 sequences and the 13-nt repeat sequence present in the Drosophila doublesex pre-mRNA. There is some sequence similarity between our element A and the dsx repeat (A, ACTTCAACAAGTT, and dsx, TCT'TCAATCAA CA). We have also looked for sequence similarity between our regions and the regulatory exon sequences identified in other pre-mRNAs. There does not seem to be a clear sequence similarity between the different pre-mRNAs. However, the CALC-I exon 4 element B contains a GAA-rich sequence element of 14 consecutive purine residues. In other pre-mRNAs, similar GAA-rich elements which are involved in exon recognition have been identified (16, 27, 32, 38, 62). If a consensus sequence is involved in the alternative processing of several genes, this might indicate the involvement of the same trans-acting factor. Attractive candidates for this role are the heterogeneous nuclear RNP (hnRNP) proteins. hnRNP proteins have been shown to be associated with the splicing machinery. Immunoprecipitation studies suggest a role for hnRNP C proteins in splicing (13). hnRNP Al is involved in the selection of 5' splice sites (39). Binding studies showed that the hnRNP proteins A, C, and D (56) and hnRNP I/PTB (28, 29, 45) bind to the 3' splice site. hnRNPs A, C, D, and I/PTB are part of a larger hnRNP complex, and many of the other proteins in this complex also interact with the RNA in a sequence-dependent manner (6). Bennett et al. (6) suggested that hnRNP proteins play a role in splice site selection, possibly by the selective binding to exon sequences. Binding of these proteins to the pre-mRNA may form a platform for the interaction of other factors. The sequence difference between element A and B suggests that two different factors are involved in the enhancement of CT-specific splicing. Variations in the concentration of these factors in different cells may be responsible for the processing choice, as shown for the ratio of the essential splicing factor ASF/SF2 to hnRNPA1 (39). Very recently, two papers which showed that the presence of exonic purine-rich elements similar to our element B enhance spliceosome formation and splicing and thereby promote exon inclusion have appeared (63, 64). Alternatively, secondary RNA structures present or artificially introduced in pre-mRNAs have been shown to affect splicing in vitro and in vivo (12, 20, 24-26, 52, 53, 61). For the ,B-tropomyosin pre-mRNA, it has been shown that sequences in the skeletal muscle-specific exon act as negative cis elements which prevent splicing of the skeletal musclespecific exon in myoblasts. At least part of this negative control depends on the existence of a stem-loop structure involving exon sequences (14, 36, 37). In the alternative processing of the CALC-I gene, local secondary structure might also be involved in the effect we observe after deletion of exon 4 sequences. Deletion of exon 4 sequences might induce a secondary structure which inhibits the splicing of exon 3 to exon 4. We have analyzed the secondary structure of exon 4 and its flanking sequences in the different deletion mutants by using the program of Zuker and Stiegler (65). This did not reveal a correlation between secondary structure and CT-specific processing. Also, our insertion mutants cover a variety of different constructs, which all induce CT-specific processing. Therefore, the effect of deletion of exon 4 sequences on the RNA processing is predominantly determined by the deletion of specific sequences. The small fluctuation in the CT/CGRP ratio observed after transfection
MOL. CELL. BIOL.
of plasmids containing only element A or B pRSVdI3E4 (117-146) and pRSVdI3E4(87-146) or pRSVdI3E4(63-88), pRSVdI3E4(63-101), and pRSVdI3E4(63-114), respectively, may be caused by subtle changes in the secondary structure of the pre-mRNA. The identification of two sequence elements A and B within exon 4 which activate CT-specific splicing in nonneural cells and the observation that the U-A branchpoint mutation can overcome the effect of deletion of these sequences on the RNA processing choice suggest a possible way for regulation of the tissue-specific processing of the CALC-I pre-mRNA. Inclusion of the CT-encoding exon 4 in mature RNA in nonneural cells requires, because of the presence of the weak splice acceptor site, initial binding of factors to specific exon 4 sequences to promote CT-specific splicing. Exon skipping leading to CGRP-I mRNA processing in neural cells may be accomplished either by a low concentration of these factors in neural cells or by the presence of a specific dominant neural trans-acting factor which inhibits CT-specific splicing (18, 48). The presence of a strong acceptor site in front of exon 4 converts the optional exon 4 in a constitutive one, even in neural cells (1, 23). The production of both CT and CGRP-I mRNA after transfection of the complete CALC-I gene to 293 cells in contrast to predominantly CT RNA synthesis with the minigene indicates that the balance between the two alternative processing pathways is very delicate. As shown in this study, this balance can easily be disturbed by different cis- and probably trans-acting factors. The sequences within the complete CALC-I gene responsible for the stimulation of CGRPspecific splicing in 293 cells may also play a role in the tissue-specific regulation of the alternative splicing. In conclusion, this study indicates the involvement two separate regions in the 5' end of exon 4 in the selection of the exon 4 3' splice site. It remains to be determined whether these sequences act as target for factors involved in CTspecific processing and whether they are involved in the inhibition of exon 4 splicing in neuronal cells. Further RNA binding studies and mutational analysis of the putative target sequences are needed to answer these questions. ACKNOWLEDGMENTS This work was supported in part by the Netherlands Organisation for Chemical Research (SON) with financial aid from the Netherlands Organisation for Scientific Research (NWO). REFERENCES 1. Adema, G. J., and P. D. Baas. 1991. Deregulation of alternative processing of calcitonin/CGRP-I pre-mRNA by a single point mutation. Biochem. Biophys. Res. Commun. 178:985-992. 2. Adema, G. J., and P. D. Baas. 1992. A novel calcitonin-encoding mRNA is produced by alternative processing of calcitonin/ calcitonin gene-related peptide-I pre-mRNA. J. Biol. Chem. 267:7943-7948. 3. Adema, G. J., R. A. L. Bovenberg, and P. D. Baas. 1988. Unusual branch point selection involved in splicing of the alternatively processed calcitonin/CGRP-I pre-mRNA. Nucleic Acids Res. 16:9513-9526. 4. Adema, G. J., K. L. van Hulst, and P. D. Baas. 1990. Uridine branch acceptor is a cis-acting element involved in regulation of the alternative processing of calcitonin/CGRP-I pre-mRNA. Nucleic Acids Res. 18:5365-5373. 5. Amara, S. G., R. M. Evans, and M. G. Rosenfeld. 1984. Calcitonin/calcitonin gene-related peptide transcription unit: tissue-specific expression involves selective use of alternative polyadenylation sites. Mol. Cell. Biol. 4:2151-2160. 6. Bennett, M., S. Pifiol-Roma, D. Staknis, G. Dreyfuss, and R. Reed. 1992. Differential binding of heterogeneous nuclear ribo-
VOL. 14, 1994
7. 8.
9.
10.
11.
12.
13. 14.
15. 16. 17.
18.
19.
20.
21.
22.
23.
24. 25.
26.
nucleoproteins to mRNA precursors prior to spliceosome assembly in vitro. Mol. Cell. Biol. 12:3165-3175. Bovenberg, R. A. L., G. J. Adema, and P. D. Baas. 1988. Model for tissue specific calcitonin/CGRP-I RNA processing from in vitro experiments. Nucleic Acids Res. 16:7867-7883. Bovenberg, R. A. L., T. C. Moen, H. S. Jansz, and P. D. Baas. 1989. In vitro splicing analysis of mini-gene constructs of the alternatively processed human calcitonin/CGRP-I pre-mRNA. Biochim. Biophys. Acta 1008:223-233. Bovenberg, R. A. L., W. P. M. van de Meerendonk, P. D. Baas, P. H. Steenbergh, C. J. M. Lips, and H. S. Jansz. 1986. Model for alternative RNA processing in human calcitonin gene expression. Nucleic Acids Res. 14:8785-8803. Burtis, K. C., and B. S. Baker. 1989. Drosphila doublesex gene controls somatic sexual differentiation by producing alternatively spliced mRNAs encoding related sex-specific polypep-
tides. Cell 56:997-1010. Chain, A. C., S. Zoliman, J. C. Tseng, and F. A. Laski. 1991. Identification of a cis-acting sequence required for germ line specific splicing of the P element ORF2-ORF3 intron. Mol. Cell. Biol. 11:1538-1546. Chebli, K., R. Gattoni, P. Schmitt, G. Hildwein, and J. Stevenin. 1989. The 216-nucleotide intron of the ElA pre-mRNA contains a hairpin structure that permits utilization of unusually distant branch acceptors. Mol. Cell. Biol. 9:4852-4861. Choi, Y. D., P. J. Grabowski, P. A. Sharp, and G. Dreyfuss. 1986. Heterogeneous nuclear ribonucleoproteins: role in RNA splicing. Science 231:1534-1539. Clouet d'Orval, B., Y. d'Aubenton Carafa, P. Sirand-Pugnet, M. Gallego, E. Brody, and J. Marie. 1991. RNA secondary structure repression of a muscle-specific exon in HeLa cell nuclear extracts. Science 252:1823-1828. Cooper, T. A. 1992. In vitro splicing of cardiac troponin T precursors. J. Biol. Chem. 267:5330-5338. Cooper, T. A., and C. P. Ordahl. 1989. Nucleotide substitutions within the cardiac troponin T alternative exon disrupt premRNA alternative splicing. Nucleic Acids Res. 17:7905-7921. Cote, G. J., I. N. Nguyen, S. M. Berget, and R. F. Gagel. 1990. Calcitonin exon sequences influence alternative RNA processing. Mol. Endocrinol. 4:1744-1749. Cote, G. J., I. N. Nguyen, C. J. M. Lips, S. M. Berget, and R. F. Gagel. 1991. Validation of an in vitro RNA processing system for CT/CGRP precursor mRNA. Nucleic Acids Res. 19:36013606. Cote, G. J., D. T. Stolow, S. Peleg, S. M. Berget, and R. F. Gagel. 1992. Identification of exon sequences and an exon binding protein involved in alternative RNA splicing of calcitonin/CGRP. Nucleic Acids Res. 20:2361-2366. Deshler, J. O., and J. J. Rossi. 1991. Unexpected point mutations activate cryptic 3' splice site by perturbing a natural secondary structure within a yeast intron. Genes Dev. 5:12521263. Domenjoud, L., H. Gallinaro, L. Kister, S. Meyer, and M. Jacob. 1991. Identification of a specific exon sequence that is a major determinant in the selection between a natural and a cryptic 5' splice site. Mol. Cell. Biol. 11:4581-4590. Edbrooke, M. R., D. Parker, J. H. McVey, J. H. Riley, G. D. Sorenson, 0. S. Pettengil, and R. K. Craig. 1985. Expression of the human calcitonin/CGRP gene in lung and thyroid carcinoma. EMBO J. 4:715-724. Emeson, R. B., F. Hedjran, J. M. Yeakley, J. W. Guise, and M. G. Rosenfeld. 1989. Alternative production of calcitonin and CGRP mRNA is regulated at the calcitonin-specific splice acceptor. Nature (London) 341:76-80. Eng, F. J., and J. R. Warner. 1991. Structural basis for the regulation of splicing of a yeast messenger RNA. Cell 65:797804. Eperon, L. P., I. R. Graham, A. D. Griffiths, and I. C. Eperon. 1988. Effects of RNA secondary structure on alternative splicing of pre-mRNA: is folding limited to a region behind the transcribing RNA polymerase? Cell 54:393-401. Estes, P. A., N. E. Cooke, and S. A. Liebhaber. 1992. A native RNA secondary structure controls alternative splice-site selec-
REGULATION OF CALC-I PRE-mRNA PROCESSING
27.
28.
29.
30. 31.
32.
33.
34.
35.
36.
37.
38. 39. 40. 41. 42.
43.
44.
45.
46. 47.
959
tion and generates two human growth hormone isoforms. J. Biol. Chem. 267:14902-14908. Fu, X.-D., R. A. Katz, A. M. Skalka, and T. Maniatis. 1991. The role of branchpoint and 3'-exon sequences in the control of balanced splicing of avian retrovirus RNA. Genes Dev. 5:211220. Garcia-Blanco, M. A., S. F. Jamison, and P. A. Sharp. 1989. Identification and purification of a 62,000-dalton protein that binds specifically to the polypyrimidine tract of introns. Genes Dev. 3:1874-1886. Gil, A., P. A. Sharp, S. F. Jamison, and M. A. Garcia-Blanco. 1991. Characterization of cDNAs encoding the polypyrimidine tract-binding protein. Genes Dev. 5:1224-1236. Green, M. R. 1991. Biochemical mechanisms of constitutive and regulated pre-mRNA splicing. Annu. Rev. Cell Biol. 7:559-599. Guo, W., G. J. Mulligan, S. Wormsley, and D. M. Helfman. 1991. Alternative splicing of 1-tropomyosin pre-mRNA: cisacting elements and cellular factors that block the use of a skeletal muscle exon in nonmuscle cells. Genes Dev. 5:20962107. Hampson, R. K., L. LaFolette, and F. M. Rottman. 1989. Alternative processing of bovine growth hormone mRNA is influenced by downstream exon sequences. Mol. Cell. Biol. 9:1604-1610. Hedley, M. L., and T. Maniatis. 1991. Sex-specific splicing and polyadenylation of dsx pre-mRNA requires a sequence that binds specifically to tra-2 protein in vitro. Cell 65:579-586. HoshUima, K., K. Inoue, I. Higuchi, H. Sakamoto, and Y. Shimura. 1991. Control of doublesex alternative splicing by transformer and transformer-2 in Drosophila. Science 252:833836. Inoue, K., K. HoshUima, I. Higuchi, H. Sakamoto, and Y. Shimura. 1992. Binding of the Drosophila transformer and transformer-2 proteins to the regulatory elements of doublesex primary transcript for sex-specific RNA processing. Proc. Natl. Acad. Sci. USA 89:8092-8096. Libri, D., M. Goux-Pelletan, E. Brody, and M. Y. Fiszman. 1990. Exon as well as intron sequences are cis-regulating elements for the mutually exclusive alternative splicing of the P tropomyosin gene. Mol. Cell. Biol. 10:5036-5046. Libri, D., A. Piseri, and M. Y. Fiszman. 1991. Tissue-specific splicing in vivo of the 3-tropomyosin gene: dependence on an RNA secondary structure. Science 252:1842-1845. Mardon, H. J., F. E. Sebastio, and F. E. Baralle. 1987. A role for exon sequences in alternative splicing of the human fibronectin gene. Nucleic Acids Res. 15:7725-7733. Mayeda, A., and A. R. Krainer. 1992. Regulation of alternative pre-mRNA splicing by hnRNP Al and splicing factor SF2. Cell 68:365-375. McKeown, M. 1990. Regulation of alternative splicing. Genet. Eng. 12:139-181. McKeown, M. 1992. Alternative mRNA splicing. Annu. Rev. Cell Biol. 8:133-155. McKeown, M., J. M. Belote, and R. T. Boggs. 1988. Ectopic expression of the female transformer gene product leads to female differentiation of chromosomally male Drosophila. Cell 53:887-895. Nagoshi, R. N., and B. S. Baker. 1990. Regulation of sex-specific RNA splicing at the Drosophila doublesex gene: cis-acting mutations in exon sequences alter sex-specific RNA splicing patterns. Genes Dev. 4:89-97. Nagoshi, R. N., M. McKeown, K. C. Burtis, J. M. Belote, and B. S. Baker. 1988. The control of alternative splicing at genes regulating sexual differentiation in D. melanogaster. Cell 53: 229-236. Patton, J. G., S. A. Mayer, P. Tempst, and B. Nadal-Ginard. 1991. Characterization and molecular cloning of polypyrimidine tract-binding protein: a component of a complex necessary for pre-mRNA splicing. Genes Dev. 5:1237-1251. Reed, R., and T. Maniatis. 1986. A role for exon sequences and splice-site proximity in splice site selection. Cell 46:681-690. Robberson, B. L., G. J. Cote, and S. M. Berget. 1990. Exon definition may facilitate splice site selection in RNAs with
960
vAN OERS ET AL.
multiple exons. Mol. Cell. Biol. 10:84-94. 48. Roesser, J. R., K. Liittschwager, and S. E. Leff. 1993. Regulation of tissue-specific splicing of the calcitonin/calcitonin gene-related peptide by RNA-binding proteins. J. Biol. Chem. 268: 8366-8375. 49. Rosenfeld, M. G., J.-J. Mermod, S. G. Amara, L. W. Swanson, P. E. Sawchenko, J. Rivier, W. Vale, and R. M. Evans. 1983. Production of a novel neuropeptide encoded by the calcitonin gene via tissue-specific RNA processing. Nature (London) 304:129-135. 50. Ryner, L. C., and B. S. Baker. 1991. Regulation of doublesex pre-mRNA processing occurs by 3'-splice site activation. Genes Dev. 5:2071-2085. 51. Siebel, C. W., L. D. Fresco, and D. C. Rio. 1992. The mechanism of somatic inhibition of Drosophila P-element pre-mRNA splicing: multiprotein complexes at an exon pseudo-5' splice site control Ul snRNP binding. Genes Dev. 6:1386-1401. 52. Solnick, D. 1985. Alternative splicing caused by RNA secondary structure. Cell 43:667-676. 53. Solnick, D., and S. I. Lee. 1987. Amount of secondary structure required to induce an alternative splice site. Mol. Cell. Biol. 7:3194-3198. 54. Steenbergh, P. H., J. W. M. Hoppener, J. Zandberg, A. F. M. Cremers, H. S. Jansz, and C. J. M. Lips. 1985. Structure of the human calcitonin gene and its transcripts in medullary thyroid carcinoma, p. 23-31. In A. Pecile (ed.), Calcitonin. Elsevier Science Publishers BV, Amsterdam. 55. Streuli, M., and H. Saito. 1989. Regulation of tissue specific alternative splicing: exon-specific cis-elements govern the splicing of leukocyte common antigen pre-mRNA. EMBO J. 8:787796.
MOL. CELL. BIOL. 56. Swanson, M. S., and G. Dreyfuss. 1988. RNA binding specificity of hnRNP proteins: a subset bind to the 3' end of introns. EMBO J. 7:3519-3529. 57. Tian, M., and T. Maniatis. 1992. Positive control of pre-mRNA splicing in vitro. Science 256:237-240. 58. Tsai, A. Y. M., M. Streuli, and H. Saito. 1989. Integrity of the exon 6 sequence is essential for tissue-specific alternative splicing of human leukocyte common antigen pre-mRNA. Mol. Cell. Biol. 9:4550-4555. 59. van Oers, C. C. M. 1993. Ph.D. thesis. Utrecht University, Utrecht, The Netherlands. 60. Wakamatsu, N., H. Kobayashi, T. Miyatake, and S. Tsuji. 1992. A novel exon mutation in the human P-hexosaminidase p subunit gene affects 3' splice site selection. J. Biol. Chem. 267:2406-2413. 61. Watakabe, A., K. Inoue, H. Sakamoto, and Y. Shimura. 1989. A secondary structure at the 3' splice site affects the in vivo splicing reaction of mouse immunoglobulin ,u chain premRNAs. Nucleic Acids Res. 17:8159-8169. 62. Watakabe, A., H. Sakamoto, and Y. Shimura. 1991. Repositioning of an alternative exon sequence of mouse IgM pre-mRNA activates splicing of the preceding intron. Gene Exp. 1:175-184. 63. Watakabe, A., K. Tanaka, and Y. Shimura. 1993. The role of exon sequences in splice site selection. Genes Dev. 7:407418. 64. Xu, R., J. Teng, and T. A. Cooper. 1993. The cardiac troponin T alternative exon contains a novel purine-rich positive element. Mol. Cell. Biol. 13:3660-3674. 65. Zuker, M., and P. Stiegler. 1981. Optimal computer folding of large RNA sequences using thermodynamics and auxiliary information. Nucleic Acids Res. 97:133-148.
