Conserved Alternative Splicing in the 5'-Untranslated Region of the Muscle-Specific Enolase Gene....

Eur. J. Biochem. 232, 141-149 (1995) 0 FEBS 1995

Conserved alternative splicing in the 5’-untranslated region - of the muscle-specific enolase gene Primary structure of mRNAs, expression and influence of secondary structure on the translation efficiency

Daniele OLIVA’, Silvana VENTURELLA’, Rosa PASSANTINO’, Salvatore FEO’ and Agata GIALLONG02 ‘ Dipartimento di Biologia Cellulare e dello Sviluppo, Universiti di Palermo, Italy ’ Istituto di Biologia dello Sviluppo del Consiglio Nazionale delle Ricerche, Palermo, Italy

(Received 10 April 1995) - EJB 95 0568/2

We report here the isolation and characterization of cDNAs covering the 5’-end region of mouse and rat mRNAs that encode the p or muscle-specific isoform of the glycolytic enzyme enolase. As previously determined for humans, two classes of p-enolase transcripts with distinct sequences in their 5’-untranslated regions are present in both mouse and rat muscles. A mechanism of alternative splicing, conserved from mouse to man, generates the two forms of mRNA. Secondary-structure predictions indicated that, in all cases, a more stable secondary structure could exist in the 5’ end of the message with the longer leader. In vitro transcripts containing defined human or mouse 5’-untranslated sequences were obtained by fusion of the different cDNA clones and tested for their relative translational efficiencies in rabbit reticulocyte lysates. Transcripts containing the human long and short leader sequences showed differences in the translational rate, suggesting a role for the 5’-untranslated region in the regulation of translation. No detectable difference was found between transcripts with the two distinct mouse leader sequences. In addition, both transcripts are bound to polysomes and are equally distributed along differently sized polysomes in C2C12 myogenic cells. The relative expression of the two spliced forms in developing and adult muscle tissues by means of reverse transcription and polymerase chain reaction did not show a stage-specific or a tissue-type-specific pattern. A putative functional role for the 5’-untranslated sequences of p-enolase transcripts is discussed.

Keywords: muscle-specific enolase ; alternative splicing ; untranslated region ; translation efficiency.

The glycolytic enzyme enolase catalyzes the interconversion of 2-phosphoglycerate to phosphoenolpyruvate and is present in three different isoforms in mammals and birds [l]. The muscle- specific isoform, referred to as p-enolase, is encoded by a gene whose primary structure has been elucidated in different species [2-71. Studies on the expression of the p-enolase gene in the mouse revealed an interesting pattern of expression in the course of cardiac and skeletal muscle development, when the p isoform progressively replaces the nearly ubiquitous a isoform [8]. Early in embryogenesis, transcripts are first detectable, by in situ hybridization, in the cardiac tube and in the myotome [9] ; expression remains extremely low in skeletal primary fibers, while a striking increase occurs in the second generation of fibers at the fetal stage of development [9, lo]. A further increase is observed after birth and, in mouse, rat and human adult muscles, higher levels of expression of both Benolase protein and transcripts are detected in fast-twitch fibers than in slow-twitch fibers [9, 11 - 131. Expression of p-enolase increases with terminal differentia-

Correspondence to A. Giallongo, Istituto di Biologia dello Sviluppo

Fax: +39 91 6165665. Abbreviations. RT-PCR, reverse transcription PCR; RACE, rapid

Enzyme. Enolase, 2-phospho-~-glycerate hydrolase (EC 4.2.1.1 1). Note. The sequence reported in this paper is deposited in the EMBLI

GenBankDDBJ Nucleotide sequence Database and is available under accession number X70182.

C. N. R., Via Archirafi 20, 1-90123 Palermo, Italy

amplification of cDNA ends.

tion, as established by studies on myogenic cell lines and primary myoblasts induced to differentiate in vitro; expression al- ready occurs in proliferating myoblasts [2, 131, in contrast with what is observed for the majority of the muscle-specific genes. Furthermore, it has been reported that, in mutated myogenic cells which do not differentiate into myotubes, and do not ex- press muscle-specific gene products as well as the entire helix- loop-helix family of transcription factors involved in muscle differentiation, p-enolase transcripts are present at levels comparable to those detected in wild-type cells [13]. These data suggest that transcription of the gene might be under the control of different muscle-specific regulators.

In addition, the studies mentioned above show that changes in the level of p-enolase protein or in the level of its enzymic activity are similar and temporally correlate with those of the corresponding mRNA, suggesting that this enolase isoform is under transcriptional control during muscle development and fi- ber differentiation. It cannot be excluded that variation in the mRNA stability may result in different transcript levels. We previously reported the existence, in human muscle, of two forms of p-enolase mRNA that differ in their 5’-untranslated sequences and are generated by alternative splicing [7]. On the basis of secondary-structure predictions, indicating that the leader sequences of the spliced forms could form hairpin structures with different free energies, we proposed a potential functional role of the heterogeneous 5’-untranslated region in translational control. Keller et al. [9] observed that, in the central part of the tongue

142 Oliva et al. (Em J. Biochem. 232)

of 18-day mouse fetuses, p-enolase transcripts are present at a high level but no protein could be detected, suggesting the existence of a translational or post-translational mechanism of control. At present, no other cases of discrepancy between accumulation of p-enolase transcripts and of the protein have been reported. In an attempt to understand the functional role of the two mRNA species encoding the same protein but bearing different leader sequences, we first established the presence of two classes of p-enolase mRNAs in murine and rat muscles that differ in the 5’-untranslated region and are generated by alternative splicing. The finding that this mechanism of alternative splicing has been conserved in the course of evolution from mouse to man prompted us to further search for functional significance. We have examined the expression of the two transcripts during muscle development, in different muscle fibers and in the myogenic cell line C2C12. We have also explored the possibility that the diversity in the leader sequence of p-enolase mRNA might affect translation by generating in vitro transcripts, by comparing their in vitro translational efficiencies and, in addition, by ana- lyzing the polysomal distribution of the two forms of mRNA in myogenic cells.

MATERIALS AND METHODS Isolation of RNA. Total RNA was isolated from embryonic

and adult tissues and cultured C2C12 cells [14] by the method described by Chomczynski and Sacchi [15]. Murine tissues (limbs and hearts) were obtained from CD1 outbred mouse embryos and adult mice, as previously reported [lo]. Fast-twitch and slow-twitch skeletal muscles were dissected from rat hind limbs and the identity of the fibers was confirmed by hybridization of the RNA with a cDNA probe, plslow MyHC [16], kindly donated by S. Schiaffino (Dept. Biomedical sciences, Padova, Italy), which is specific for the p isoform of the slow-type myosin heavy chain (data not shown). Proliferating C2C12 myoblasts were plated in a growth-factor-rich medium consisting of Dulbecco’s Modified Eagles Medium (Gibco) supplemented with 20% fetal calf serum, differentiation into myotubes was induced by exposure of confluent cultures to a medium containing 5% horse serum and 5 pg/ml insulin (Sigma) for 48-70 hours.

cDNA cloning. Mouse p-enolase cDNA clones containing 5’ ends were isolated by the rapid amplification of cDNA ends (RACE) procedure [17] using RNA extracted from murine skeletal muscle (whole legs). The strategy and the oligomers used have been previously described for the isolation of the human cDNAs [7]. Rat p-enolase cDNAs and additional mouse cDNAs were isolated by direct subcloning of PCR products obtained by amplification of reverse-trancribed total muscle RNA. PCR reactions were treated with 1 unit Klenow polymerase (Boeh- ringer) for 30 min at room temperature. DNA was digested with BclI, which is present in the mouse and rat p enolase sequences in the 5‘ portion of the coding region [2, 31; fragments were separated on a 1.5% agarose gel and specific products were isolated and ligated into BamHI-HincII-digested Bluescript vector (S tratagene).

DNA analysis and secondary-structure prediction. The Bluescript clones were used for nucleotide sequencing according to the method of Sanger [18] using T7 DNA polymerase (Sequenase, USB). All sequences were determined on both strands. Sequences alignments and maximum similarity were obtained using the HIBIO DNAsis program of the Hitachi Soft- ware Engineering Co. The RNA secondary structure was predicted using the FOLD program [19] included in the Sequence Analysis Software package of the Genetic Computer Group, University of Wisconsin.

Reverse transcription PCR (RT-PCR) of total RNA. cDNA was synthesized from 10 pg total RNA by extension of primer BEN 3 (5’-CTTGAAGGAGCTGGCTCCCA-3‘), complementary to bases 545-564 in the coding region of the human p-enolase cDNA [4]. The human primer BEN 3 was used to synthetize mouse and rat cDNAs, since it contained a single internal mismatch which did not affect annealing. Briefly, RNA was annealed with 5 pmol complementary primer than was incubated at 42°C for 2 hours in 50 p150 mM Tris/HCl, pH 8.3, containing 75 mM KC1, 10 mM MgC12, 10 mM dithiothreitol, 50 pg/ml BSA, 1 mM each dNTP, 40 units placental ribo- nuclease inhibitor (Rnasin, Promega) and 50 pnits avian myelo- blastosis virus reverse transcriptase (Boehringer). Following al- kaline hydrolysis of RNA and ethanol precipitation, the cDNA was amplified using oligomer BEN 3 and oligomer BEN 4 (5’- GACACTGTCCCAGCTGCCACCTAG-3’), corresponding to the extreme portion of the 5’-untranslated sequence of human and rat p-enolase mRNAs (the first nucleotide represents the transcription start site previously determined for the human gene).

For mouse cDNAs, oligomer TOP 4 (5’-GACACTGTCC- CAGCTGCTACCTAGA-3’), corresponding to the analogous portion of the leader sequence of the mouse mRNA, was used as forward primer instead of BEN 4 (Fig. 1). PCR reactions were performed with 1 unit Taq polymerase (Perkin Elmer Cetus), 1 pM each primer, 0.2 mM each dNTP in 50 pl 10 mM Trid HC1, pH 8.3, 50 mM KC1, 1.5 mM MgC1, and 10% (by vol.) dimethylsulfoxide. After incubation at 92°C for 10 min, 20 cycles of PCR were performed (denaturation, 2 min at 92°C; annealing, 2 min at 51°C; extension, 2 min at 72°C). Aliquots of the reactions were electrophoresed on a 3% NuSieve agarose gel (FMC Bioproducts) and transferred to a nylon membrane (Hybond, Amersham). Hybridizations were carried out as previously described [20] with a 3ZP-labeled 430-bp SmaI-XbaI fragment (XbaI is a polylinker site) derived from the mouse p- enolase cDNA and representing the 5’ portion of the coding region corresponding to amino acids 9-145 (this work).

Construction of fusion cDNAs. Truncated human p-enolase cDNAs, isolated by the RACE procedure and containing the long and short leader sequences [7], were fused to a previously isolated cDNA clone containing the 3’ half of the coding sequence, the complete 3’hntranslated sequence and a 20-bp poly(A) tail [4]. Briefly, the PvuII site at nucleotide position 398 was linked to the RsaI site at nucleotide position 797, generating two human p-enolase cDNAs both lacking a 399-bp fragment encoding amino acids 124-256 [4] and containing the short and long leaders, respectively. These Bluescript plasmids were designated HS and HL with respect to the human short and long leader sequences contained in the cDNA. One of these Bluescript plasmids, HL, was then double digested with restriction endonucleases SmaI and XhoI in order to remove the DNA fragment containing the 5’-untranslated sequence. The SmaI site is located 20 bp downstream from the ATG initiation codon both in the human and mouse sequences and XhoI is the first available site upstream of the cloning insertion. Similarly, the 206- bp and 94-bp SmaI-XhoI fragments containing the long and short leaders of the mouse p-enolase cDNA, respectively, were isolated from the PCR clones (see above) and used to replace the analogous fragment present in the starting plasmid HL. These replacements resulted in the construction of chimeric cDNAs containing either the short or the long leaders and the first 25 codons of the mouse p-enolase cDNA fused to the coding region of human cDNA present in the plasmids HL and HS. These new constructs were designated ML and MS, respectively, on the basis of which mouse leader, long or short, was present in the chimeric cDNA (Fig. 4).

Oliva et al. (Eua J. Biochem. 232) 143

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MoL - GACACTGTCCCAGCTGCTACCTAGAGGAGACACTCCACTCARAGTTCC~GGAAG~TTTCCAG~ttggtaaa MoS - GACACTGTCCCAGCTGCTACCTAGAGGAGACACTCCACTCARAGTTCCAAGG~GGCTTTCCAG~ttggtaaa

~ - . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . RaL - tiACACTGTCCCAGCTGCTACCTAG-G----CACTCTACTCARAGTTCCAG~GACATTACAGGTTTGG~ RaS - GACACTGTCCCAGCTGCTACCTAG-G----CACTCTACTCAGTTCCARAG~GACATTACAG~ttggaaaa -

HuL . T-C----TTCCA-GCCCTG-G--GG-GTGGAG-GTAGTAAGG~gagcatg ... lntron ... tatttttagc HuS . t-c----ttcca-gccctg-g--gg-gtggag-gtagtaaagg~gagcatg ... intron . . . tatttttagc MoL . a-g---gttacatgg--tgtgcaggagtgg-g-gtggt---gc-tg-gcttg ... intron ... tattttt~c MoS . a-g---gttacatgg--tgtgcaggagtgg-g-gtggt ---g c-tg-gcttg . . . intron ... tattttt~c RaL . ACGTCAGTTACGCGG--TG-GCAG~gcgg-gcggggt---gg-tg .. cttg . . . intron ... tatttttagc RaS . acgtcagttacgcgg--tg-gcag~gcgg-gcggggt---gg-tg--cttg ... intron . . . tatttttagc

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..........

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..........

HuL - tccagcacctgccttcttggagtggggaagaatctt-aaagggcaagggatttctggttccttaagagatcaac HuS - tccagcacctgccttcttggagtggggaagaatctt-aaagggcaagggatttctggttccttaagagatcaac

MoL - TCCAGCACCTGCCTTCTTCGGGTGGAGAAGACTCTCTT~GGGCAAGGGATTTCTAGTTCCTTAAGGGATCAAC MoS - tccagcacctgccttcttcgggtggagaagactcttaaaagggcaagggatttctagttccttaagggatcaac

RaL - tccagcacctgccttcttcgggtggagaagactcttaaaagggcaagggatttctagttccttaagggatcaac RaS - tccagcacctgccttcttcgggtggagaagactcttaaaagggcaagggatttctagttccttaagggatcaac

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.......................................................................... ..........................................................................

HuL - tgtctacac--------tcactcacacctcctgtcctgc~C~T~CATGCAG~ HuS - tgtctacac--------tcactcacacctcctgtcctgc~C~~CAT~AG~ - . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..................................... MOL - TGTCCACTCTTG----CTCACTCACATCTCCTGTGGTGCAGCA~GCCATGCAAAAA MoS - tgtccactcttg----ctcactcacatctcctgtggtgc~C~TGGCCATGC~

RaL - tgtccacacttgctcactcactcacatctcttgtggtgc~C~GGCCATGC~ RaS - t g t c c a c a c t t g c t c a c t c actcacatctcttgtggtgc~C~T~CAT~~

- . . . . . . . . . . . ......................................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

-

Fig. 1. Comparison of nucleotide sequences of cDNAs encoding the 5’-untranslated regions of human, mouse and rat b-enolase mRNAs in relation to the genomic sequences. Nucleotides present in the transcripts are shown in capital letters while intron sequences are in lower case letters, data for the human gene are from Giallongo et al. [7], for the rat gene from Sakimura et al. [5] and for the mouse gene from the genomic sequence deposited in the EMBL Databank (Lamande, N., Brosset, S., Keller, A., Lucas, M. and Lozar, M., accession no X61600). The six sequences, Hu (human), Mo (mouse), Ra (rat) are aligned at the initiation codon ATG, indicated in bold-face letters, and gaps (dashes) have been introduced to obtain the maximum match. L and S indicate the alternatively spliced mRNA species with the long and short leader, respectively. Dinucleotides gt (donor site, 5’ end of intron) and ag (acceptor site, 3’ end of intron) are double underlined. Sequences of forward primers, BEN 4 and TOP 4, used in the PCR reactions are underlined.

In vitro transcription and quantification of mRNA. In order to generate in vitro transcripts that would most faithfully represent the primary structure of the two forms of p-enolase mRNA, HS, HL, MS and ML cDNAs were transferred into the PGEM3Z vector (Promega), utilizing the EcoRI site that is located five nucleotides downstream from the T7 RNA polymerase transcription initiation site. Briefly, the pGEM3Z vector was digested with EcoRI, then the EcoRI site was converted into a blunt-ended site by treatment with mung bean nuclease (Phar- macia) and the plasmid was further digested with XbaI (polylinker site). Similarly, the Bluescript plasmids containing the human and mouse fusion cDNAs were cut with XhoI, treated with mung bean nuclease, then digested with XbaI. Finally, the fragments were isolated and inserted in the pGEM3Z vector. The sequences of the pGEM constructs were confirmed by sequencing (data not shown). Plasmids containing the HS, HL, MS and ML cDNAs were linearized by XbaI digestion and transcribed by T7 RNA polymerase using a M A Capping Kit (Strata- gene). In vitro transcription of capped mRNA was performed in a 50-pl reaction mixture following the manufacturer’s instructions and in the presence of 2 $3 [3H]UTP (10 Ci/mmol, Amer- sham). The amount of mRNA synthesized was accurately esti- mated by measuring the incorporation of [3H]UTP into the mRNA.

In vitro translation and mRNA decay assay. In vitro translation was performed using rabbit reticulocyte lysate (Pro- mega). Equal amounts (10-40 ng) of each p-enolase transcript generated from the pGEM constructs were incubated in the pres-

ence of [35S]methionine (1200 Ci /mmol, Amersham) for 60 min at 30°C. As a control, 20ng unrelated in vitro transcribed mRNA was added to each reaction mixture; H1 mRNA coding for the 38-kDa Surf 2 protein [21] was used in all the translation experiments. Translation products were resolved by SDSPAGE. Gels were fixed, treated with Amplify TM (Amersham) according to the manufacturer’s instructions, then dried and exposed to X-ray film for autoradiography. The amount of synthesized protein was determined by densitometric scanning using a BioRadGs-670 densitometer and the Molecular Analyst version 1.1 program. The degradation course of the in vitro transcribed mRNA containing the different leaders as a function of the incubation time in the rabbit reticulocyte lysate was determined as described by Ryabova et al. [22]. The integrity of the mRNA was assessed by electrophoresis on a 5% polyacrylamide gel in the presence of 8 M urea and subsequent autoradiography. None of the mRNA species analyzed showed a preferential degradation during the 60-min incubation (data not shown).

Isolation of polysomes and RNA analysis. Polysomal RNA was isolated from proliferating (myoblasts) and differentiated (myotubes) C2C12 cells as described by Buckingham et al. [23]. Sucrose gradient fractions corresponding to polysomes, mono- somes and ribosomal subunits were pooled (Fig. 6A) and phenol extracted after adding 0.2% (mass/vol.) SDS. The RNA from each fraction was ethanol precipitated at -20°C overnight, washed with 70% ethanol, dried and dissolved in an equal vol- ume of water. The presence of alternatively transcribed p-eno-

144 Oliva et al. ( E m J. Biochem. 232)

lase mRNAs was analyzed by RT-PCR using cell equivalents of RNA for each fraction.

RESULTS

Alternatively spliced /I-enolase mRNAs are conserved in humans, mice and rats. To investigate the presence in murine skeletal muscle of distinct p-enolase transcripts differing in the length of their 5'-untranslated sequences as those identified in the human tissue, we first performed a primer-extension analysis. An oligomer complementary to the first 36 nucleotides of the coding region of the human enolase sequence [4] was used as primer, relying on the high similarity shared by the cDNAs isolated from the two species. Two extended fragments differing in length by about 100 bp were obtained in the presence of murine muscle poly(A)-rich RNA (data not shown), suggesting the existence of mouse a-enolase mRNAs with different leaders. TO confirm these data, the RACE protocol was used to generate murine p-enolase cDNA clones containing 5' ends. Nucleotide sequence analysis of several cDNA clones obtained from a first screening identified a p-enolase transcript bearing a 66-nucleotide leader and corresponding to the shorter product obtained by primer extension. Sequence comparison between human and mouse leader sequences showed identity of 24 out of the first 25 nucleotides, therefore an oligomer corresponding to this region and an oligomer complementary to a sequence in the coding region were used to isolate additional mouse p-enolase cDNAs by direct cloning of reverse PCR products, as described in Materials and Methods. Nucleotide sequence analysis of these clones showed that they belonged to two groups representing two transcripts differing from one another by the presence or absence of a 112-nucleotide insert in the 5' untranslated sequences, that are 66 nucleotides and 178 nucleotides, respectively, in size (Fig. 1). Alignment of the cDNAs and a genomic sequence deposited in the EMBL Databank (Lamande, N., Brosset, S., Keller, A., Lucas, M. and Lazar, M., accession number X61600) indicated that the two species of p-enolase mRNA present in murine skeletal muscle are generate by use of a common splice-donor site at the 3' end of exon I, and alternate acceptor sites at the 5' end of exon 11. This utilization of splice sites is different from what was previously observed in humans where alternate donor sites and a common acceptor site distin- guish the two classes of transcripts (Fig. 1) [7].

In order to explore whether alternative splicing of the ,@ enolase primary transcript occurs in rat, PCR amplification of reverse-transcribed RNA isolated from rat skeletal muscle was performed using the same pair of primers previously described, the sequences of which are conserved across species [3-5, 71. Two distinct spliced forms of p-enolase mRNA are present in rat muscle (Fig. 2A) and the relative abundance of the two classes of transcripts, the shorter transcript being more abundant than the longer transcript, is comparable in mouse, rat and human tissues. Sequence analysis of the subcloned PCR fragments obtained from rat muscle RNA demonstrated that the two amplification products correspond to transcripts bearing different leaders, 92 nucleotides and 61 nucleotides in length, respectively, which arose via differential utilization of a 31-nucleotide sequence. In this case, the alternative splicing in the 5' untranslated region involves different donor sites and a common acceptor site (Fig. l), as established by comparison of our cDNA sequences with the previously reported rat p-enolase genomic sequence [5]. Sequences alternatively spliced in human, mouse and rat p-enolase transcripts are different in length, being 42, 112 and 31 nucleotides, respectively, and, as far as concerns the longer mouse message, the portion of the intron differentially

A B

AUG MoS

AUG RaS

HuS 4 Fig. 2. Detection by RT-PCR of j7-enolase alternatively spliced transcripts in mouse, rat and human tissues. (A) Specific hybridization of /I-enolase cDNA fragments amplified from RNAs extracted from skeletal muscles. PCR and hybridization conditions are described in Materials and Methods. Molecular size markers (bp) from a 123-bp ladder are shown at right. (B) Schematic representation of the alternative splicing events in mouse (Mo), rat (Ra) and human (Hu) /I-enolase mRNAs. L and S indicate the long and short leader, respectively. The coding regions are represented by open boxes and the alternatively spliced 5'-untranslated regions by solid bars. Position of the translation initiation site, AUG, is shown.

spliced is different (Fig. 1). However, a high degree of sequence conservation is observed among the analogous regions of the three genes involved in the alternative splicing (Fig. 1). Struc- tural features of the alternatively spliced transcripts of p-enolase present in human, mouse and rat muscle tissues are depicted (Fig. 2 B).

Our findings are consistent with two mouse p enolase cDNA sequences deposited in the EMBL Databank, (Peterson, C. A., Cho, M., Rastinejad, F. and Blau, H. M., accession no. X62667), whose partial 5' untranslated sequence corresponds to the long leader we identified, the other (Lazar, M., Lamande, N., Brosset, S., Lucas, M. and Keller, A., accession no. X57747) that has a partial 5' untranslated region matching the short leader of the most abundant message.

Secondary-structure predictions. We previously reported that the predicted secondary structure of the two splice forms of the human p-enolase mRNA have very different free energies of formation; the shorter leader has the least stable predicted secondary structure (AG = -115 kJ/mol) while the longer form forms a potentially more stable structure (AG = -270.7 kJ/mol) [7]. Similarly, folding of the 5' end of the p-enolase mRNAs identified in mouse and rat skeletal muscle was analyzed using sequences from the transcription start site to 36-nucleotides downstream from the AUG start codon (Fig. 3). Introduction of the 112-nucleotide fragment in the leader sequence of the mouse mRNA had a strong impact on the secondary structure, leading to the formation of a very stable hairpin structure with AG = -353.9 kJ/mol, while the shorter form showed a potentially less stable secondary structure with AG = -119.6 kJ/mol. The AG values of the predicted most stable secondary structures in the 5' end of rat p-enolase transcripts were -146.8 kJ/mol and -84.9 kJ/mol for the long and short leaders respectively, (Fig. 3).

Translational activity of various /I-enolase mRNAs with different leaders. It has been demonstrated that the translational efficiency can be affected when secondary structure is artificially introduced into the 5'-untranslated region of an mRNA [24, 251

Oliva et al. ( E m J. Biochem. 232) 145

MoL AG = -353.9 kJ/rnol

\ , G . \ I .* , I 1..

RaL AG = -146.8 kJ/mol

5'c . c "< .< 1 c c-,4

c 1 , \

" 1

c 1

\ I " A \ ,

A MoS AG = -1 19.6 kJ/rnol

RaS AG = -84.9 kJimol

Fig. 3. Predicted secondary structure of the 5'-untranslated region of mouse and rat /3-enolase mRNA. The program FOLD [19] was used to predict the most stable stem-loop structure in the 5'-untranslated regions of the mouse p-enolase mRNAs from positions -178 to +36 for the long (MoL) and from positions -66 to +36 for the short form (MoS). Similarly, the rat sequences were analyzed from positions -94 to +36 for the long transcript (RaL) and from position -63 to position +36 for the short transcript (RaS); since no other structures involved nucleotides from positions +18 to +36, sequences up to position +17 are shown. The numbering indicates the nucleotide position relative to the initiation codon AUG, indicated by an asterisk in the figure and designated +l. For the rat p-enolase mRNA, the transcription start site determined by Sakimura et al. [ 5 ] , positioned two nucleotides upstream with respect to the human and mouse 5' ends, has been considered. The predicted secondary structure of the 5'-untranslated region of the human p-enolase mRNA has been previously reported [7].

and leader sequences have been associated with control of translation (for review, see [26, 271). To examine whether the difference in the leader between the alternatively spliced j3-enolase transcripts also reflected differences in the translational efficiency, fusion mRNAs were constructed following the strategy described in Materials and Methods. Human j3-enolase cDNAs containing the short and long leaders respectively, were fused in-frame to a cDNA encompassing the 3'-coding region and the 3'-untranslated sequence of the human transcript [4]. The result- ing cDNAs (HS and HL; Fig. 4), do not contain the 399-bp fragment encoding amino acids 124-256 of the human j3-enolase [4]. Human leader sequences in HS and HL plasmids were then replaced by the short and long leaders of the mouse j3-enolase mRNAs, respectively, to give chimeric cDNAs, MS and ML, containing the alternative spliced 5'-untranslated regions of the murine transcripts linked to most of the human coding region (Fig. 4). In vitro transcripts produced by T7 RNA polymerase from the four pGEM plasmids, HS, HL, MS and ML, all contained an open reading frame of identical length and should be translated into a polypeptide of about 33 kDa. The relative translational efficiencies of the four fusion mRNAs were determined testing different amounts of the transcripts in rabbit reticulocyte lysate (Fig. 5). All the mRNAs produced a polypeptide of the expected molecular mass, indicating that the initial AUG codon was properly used. The translational rate for the mouse mRNA with the longer leader was the same as that for mRNA containing the shorter leader (Fig. 5A). In contrast, the human transcript containing the longer 5'-untranslated region was found to be

HS

ATG Sma I A 3 9 9 bD

HL

ATG Sma I ~ 3 9 9 bp

MS

ATG Sma I A 399 bo

ML

100 bp H

Fig. 4. Schematic representation of the pGEM3Z constructs used to generate in viho transcripts. Benolase fusion cDNAs were obtained as described in Materials and Methods. HS and HL plasmids contain the human short and long 5'-untranslated sequences (cross-hatched boxes), respectively. MS and ML contain the mouse short and long 5'-untranslated sequences (hatched boxes), respectively. The open boxes represent the human p-enolase coding region with a deletion of 399 bp as indicated. The black boxes represent the 25-bp fragment of mouse coding sequence present in the MS and ML plasmids. The 3'-untranslated region of the human cDNA, contained in all the constructs, is represented by vertical striped boxes. The T7 polymerase transcription start site, the ATG initiation codon and relevant restriction sites discussed in the text are shown.

146 Oliva et al. ( E M J. Biochem. 232)

I A B

WMHHHHH ( A B C D E h A B C D E F G

TOP BOTTOM

Fig. 5. In vitro translation of synthetic /I-enolase transcripts containing either mouse or human 5’-untranslated sequences. Increasing amounts of in vitro transcripts, generated from the MS, ML, HS and HL constructs (see Fig. 4), were translated in rabbit reticulocyte lysates and the 35S-labeled peptides were separated by SDSRAGE. 20% each reaction mixture was loaded onto the gel. Lane 1, reaction containing only 20 ng H1 mRNA as a control. Lanes 2,4 and 6 contain 10, 20 and 40 ng MS or HS transcripts, respectively, in addition to 20ng H1 mRNA. Lanes 3, 5 and 7 contain the same corresponding amounts of ML or HL transcripts. Relative positions of H1 mRNA translation product and ,5- enolase polypeptide (ENO) are indicated. The migration of molecular mass standards is shown (ma).

translated three times less efficiently than was the human mRNA with the shorter leader (Fig. 5B). The same results were obtained with three different preparations of synthetic transcripts in inde- pendent experiments. Furthermore, analysis of the stability of the mRNAs in the translation mixture, as described in Materials and Methods, demonstrated that the difference in translational efficiency observed between the two forms of the human transcripts was not due to a preferential degradation of the mRNA with the longer 5’-untranslated sequence.

Polysomal localization of B-enolase mRNA in C2C12 cells. To study the translational status of the two alternatively spliced p-enolase-encoding mRNAs, their association with ribosomes was examined in the mouse myogenic cell line C2C12. Lysates from proliferating (myoblasts) and differentiated (myotubes) C2C12 cells were fractionated on a 10-50% sucrose gradient in order to separate ribosomes and the different size classes of polysomes. To avoid polysome run-off, the cultures were treated with cycloheximide prior to fractionation, and the drug was added to all buffers. The gradient fractions were pooled (Fig. 6A) and the distribution of the two spliced forms of p- enolase mRNA was determined in each pool by PCR amplification of reverse-transcribed RNA (Fig. 6B). Sample A (Fig. 6) contained the material sedimenting more slowly than ribosomal subunits; sample B contained ribosomal subunits and mono- somes; samples C-G contained the smallest to the largest polysomes, respectively. The results indicated that essentially all the p-enolase mRNA was associated with ribosomes and that the two alternatively spliced transcripts were equally distributed along differently sized polysomes (Fig. 6B). A much longer exposure revealed trace amounts of both transcripts in fraction B (data not shown). Similar results were obtained with cell lysates from proliferating myoblasts (data not shown), where both p- enolase mRNA species are present at a much lower level than in myotubes (see insert in Fig. 6A).

Expression patterns of alternatively spliced B-enolase mRNAs in developing and adult muscle tissues. In order to obtain insight into the physiological significance of heterogeneity in p-enolase mRNA species, the expression patterns of the two classes of transcripts were examined using PCR amplifi-

Fig. 6. Polysomal distribution of /I-enolase mRNAs. (A) Sedimentation profile of polyribosomes from C2C12 myotubes. A cytoplasmic extract from approximately 5 X lo7 C2C12 cells, exposed to differentiation medium for 70 hours, was subjected to centrifugation on a 10-50% sucrose gradient as described in Material and Methods. Fractions were collected and the absorbance at 254 nm (Az5.+) determined. The gradient fractions were pooled as shown, yielding samples A-G. (B) Analysis by RT-PCR of enolase mRNAs in gradient samples. Cell equivalents of RNA from each fraction (A-G) were reverse transcribed and PCR amplification products detected by hybridization with a /3-enolase-specific probe. Pat- terns of expression of alternatively transcribed mRNAs in proliferating (Mb) versus differentiated (Mt) C2C12 cells are shown in the insert (A).

cation of reverse-transcribed total RNAs isolated from limbs and hearts of mouse embryos at different days of gestation and from adult tissues. The two alternatively spliced mRNAs are expressed in the embryonic (12 d) and fetal (16 d) stages of development both in developing skeletal muscle and heart (Fig. 7A). The relative amount of the smaller transcript compared to the larger transcript does not vary during development and is equiv- alent to that observed in adult tissues (Fig. 7A). Both transcripts are bound to polysomes in embryonic and fetal skeletal muscle, as established by analysis of the mRNA present in a crude polysomal pellet (data not shown).

In addition, we explored whether p-enolase mRNA splicing displays different patterns of expression in different fibers. Total RNA was isolated from individual fibers of the rat fast-twitch extensor digitorum longus, EDL, and slow-twitch soleus muscle, and PCR amplification was performed using the primers that identify the two spliced mRNA forms. The two alternatively spliced transcripts (Fig. 7B) are present in both types of muscle fibers in the same relative amounts.

Discussion We have isolated and characterized cDNA clones encoding

mouse and rat muscle-specific enolase that show heterogeneity in the 5’-untranslated region. In both cases, transcripts with leaders of different lengths are generated by alternative RNA splicing via differential utilization of two donor sites and one acceptor splice site for the rat mRNAs and, vice versa, by the use of alternate acceptor sites and a common donor site for the murine mRNAs. These data, in conjunction with our previous work on the presence of alternatively spliced p-enolase mRNAs in human muscle, show that this mechanism that generates two mRNA species encoding the same protein but bearing different leader sequences, has been conserved in species which diverged 80 million years ago, such as mouse and human, suggesting the existence of a strong evolutionary constraint to preserve this difference in the 5‘-untranslated sequences. In the last few years, an increasing number of examples have provided evidence that gene expression can be regulated in the cytoplasm of eukaryotic cells at the level of mRNA stability, subcellular localization of mRNAs and translation 127 - 291. Translation regulatory element


Fig. 7. Analysis by RT-PCR of alternative splicing of j3-enolase mRNA in developing and adult muscle tissues. Southern blots were performed on RT-PCR amplified products (20 cycles of amplification) of enolase transcripts and hybridized with a specific probe (see details in Materials and Methods). (A) Total RNA was isolated from limbs and hearts of 12-day-old mouse embryos (12d), 16-day-old fetuses (16d) and adult mice (a). For the adult skeletal muscle sample, ten times less cDNA synthetized by reverse transcription was used in the PCR reaction relative to that used in the other samples. (B) Total RNA was from fast- twitch extensor digitorum longus, EDL, and slow-twitch Soleus muscles isolated from adult rats. Molecular size markers (bp) from a 123-bp ladder are shown.

have been identified in the 5‘-untranslated and 3’-untranslated regions of mature transcripts [27, 301. It has been demonstrated that an important role is played by specific interaction between cis-acting sequences contained in the leader and regulatory RNA-binding proteins [31] ; furthermore, the presence of secondary structures at the 5’ end of mRNA molecules can influence translation by interfering with the migration of the ribosomal initiation complex [24]. It was demonstrated by Kozak [25] that a moderately stable hairpin (AG = -125.5 kJ/mol) can drastically inhibit translation when positioned within the first 12 nucleotides of the leader sequence, but does not affects translation when it occurs around the AUG initiation codon, while a very stable stem-and-loop structure (AG = -209.2 kJlmol or -251 kJ/mol) inhibits, irrespective of the position in the 5’-untranslated region [25].

Secondary-structure predictions of the two splice forms of human [7], mouse and rat p-enolase mRNA show a common feature; the leader of the short transcript can form a less stable secondary structure than the leader of the long form that, in all the species, can assume a very stable stem-and-loop structure. The values of the calculated Gibbs energy of formation are < -125.5 M/mol for all the short forms and > -209.2 kJ/mol for the human and mouse long transcripts; the rat mRNA with the longer leader has a predicted secondary structure very close to the 5‘ end with AG = -146.8 kJ/mol. We have constructed synthetic transcripts encoding the same polypeptide and only differing in their 5’-untranslated regions, which faithfully represent the authentic 5‘ ends of mouse and human p-enolase mRNAs. These transcripts have allowed us to test the effect that these regions have on the in vitro translational efficiency. The results presented in this paper demonstrate that the p-enolase mRNA with the human short leader is translated at a higher efficiency than the transcript with the long leader. Unexpectedly, the transcripts bearing the mouse short and long leader sequences show no differences in translational efficiency. Since the same strategy was used to construct the plasmids containing the human and mouse leader sequences and because of the similar structural features involving the 5’-untranslated regions of the mRNA from the two species, we expected that the transcript with the shorter leader should translate more efficiently than the

other form in both cases. The results obtained suggest that the secondary structure alone cannot account for our observations. It has been proposed that the cytosolic trans-acting factor involved in the translational regulation of ferritin mRNA, the IRE- binding protein, carries out its negative effect not only by binding to specific sequences located at the 5‘ end of the transcript, but also by stabilizing the 5’ secondary structure with conse- quent inhibition of ribosome entry [26]. If this is the case for the /3-enolase mFWA, putative cytosolic trans-acting factors could be present in low amounts or absent in the rabbit reticulocyte lysate translation system used; the contrasting results obtained with the synthetic transcripts bearing the human or mouse leader sequences may be due to the limitations of the in vitro system. However, we cannot exclude the possibility that the results of in vitro translation reflect a functional diversity between mouse and human mRNAs. These theoretical considerations can be addressed experimentally by transient transfection, in myogenic cells, of a reporter gene under the control of the alternatively spliced 5‘-untranslated sequences.

In the past few years, the number of eukaryotic genes that produce two mRNAs with different 5‘-untranslated regions but identical coding and 3’-untranslated regions has increasing (for review, see [27]); for most of them on one version of mRNA, the leader is shorter and less loaded with secondary structures than on the other, and for a few of them, it has been demonstrated that the short form supports translation more efficiently in vitro andlor in vivo [32-361. In the case of the insulin-like growth factor I1 mRNAs, a short transcript is actively engaged in protein synthesis and is found on polysomes, while a longer message is sequestered in non-polysomal ribonucleoprotein par- ticles [36]. Our data on the translational status of the endogenous p-enolase mRNAs in the murine myogenic cell line C2C12, show that the distribution of the two spliced forms on polysomes is identical and does not change with differentiation. Almost all the p-enolase mRNA is engaged on polysomes and levels of the two alternatively spliced messages are maximal in the larger polysomes. Therefore, translation initiation and elongation seem to proceed efficiently for both transcripts, irrespective of the leader they carry. However, these results do not exclude the possibility of regulation at a post-initiation level through a mechanism not affecting the polysomal distribution of the mRNA, which is not common but has been well documented in a few cases [37, 381.

For some RNAs which are alternatively spliced, a different pattern of splicing has been shown not only in different tissues, but also at separate stages of development. Usually the splicing events occur in the coding sequences and cause the synthesis of functionally distinct protein isoforms. Consequently, it has been suggested that different trans-acting splicing factors exist in different tissues and during development [39,40]. In this study, we show that alternatively spliced p-enolase transcripts are present during cardiac and skeletal muscle development and that their relative amounts are almost constant in embryonic, fetal and adult tissues and in different fibers type. Furthermore, prelimi- nary experiments on myogenic cells in culture seem to exclude the possibility of a diverse stability of the two mRNA forms.

At present, evidence for the existence in vivo of a post-transcriptional regulation of p-enolase expression has not been documented and the observations made in this study lead to the con- clusion that a putative functional role for the alternatively spliced transcripts encoding p-enolase might be linked to physiological or metabolic changes. Recently, it has been reported that heat shock induces an alternative splice-site selection which generates transcripts with different leader sequences that are efficiently translated under heat-shock conditions [41]. Muscles are highly specialized tissues which undergo physiological changes

148 Oliva et al. (Eur J. Biochem. 232)

according to the contractile activity and in response to influences of the cellular environment. Furthermore, enolase is a key energy-generating enzyme whose presence might be required at different levels a n d o r in discrete cell compartments under par- ticular conditions. Variations in the a-enolase protein levels, re- lated to the functional status of the muscle and to the innerva- tion, have been reported [ 12, 421. Data presented here provide a basis for further investigations to determine whether the 5‘- untranslated regions of a-enolase mRNAs, whose differences have been conserved over the 160 million years of divergent evolution that separate mouse and human [43, 441, might be involved in functional regulation in response to extracellular or intracellular stimuli.

This work was supported by grants from Telethon no 416 and CNR Bilateral project to A. G., and from CNR target project Zngegneria Ge- netica and the Italian Association for Cancer Research (AIRC) to S. F. We thank S. Alemi for the gift of the C2C12 cell line, G. Cossu for providing the mouse embryonic tissues and S. Schiaffino for the gift of the /l/slow MyHC probe. We also thank M. Boucht for help and sugges- tions in the preparation of polysomes and G. Cossu for helpful discus- sions and critical reading of the manuscript.

REFERENCES 1 . Zomzely-Neurath, C. E. (1983) Enolase, in Handbook of neuro-

chemistry (Lajtha, A,, ed) vol. 4, 2nd edn, pp. 403-433, Plenum Press, New York.

2. LamandB, N., Mazo, A. M., Lucas, M., Montarras, D., Pinset, C., Gros, F, Legault-Demare, L. & Lazar, M. (1989) Murine muscle- specific enolase: cDNA cloning, sequence, and developmental expression, Proc. Natl Acad. Sci. USA 86, 4445-4449.

3. Oshima, Y., Mitsui, H., Takayama, Y., Kushiya, E., Sakimura, K. & Takahashi, Y. (1989) cDNA cloning and nucleotide sequence of rat muscle-specific enolase (beta-beta enolase), FEBS Lett. 242, 425-430.

4. Cali, L., Feo, S., Oliva, D. & Giallongo, A. (1990) Nucleotide sequence of a cDNA encoding the human muscle-specific enolase (MSE), Nucfeic Acids Res. 18, 1893.

5. Sakimura, K., Kushiya, E., Ohshima-Ichimura, Y., Mitsui, H. & Takahashi, Y. (1990) Structure and expression of rat muscle- specific enolase gene, FEBS Lett. 277, 78-82.

6. Peshavaria, M. & Day, I. N. M. (1991) Molecular structure of the human muscle-specific enolase gene (EN03), Biochem. J. 275,

7. Giallongo, A., Venturella, S., Oliva, D., Barbieri, G., Rubino, P. & Feo, S. (1993) Structural features of the human gene for muscle- specific enolase. Differential splicing in the 5’-untraslated sequence generates two forms of mRNA, Eur J. Biochem. 214,

8. Fletcher, L., Rider, C. C., Adamson, E. D., Luke, B. M. & Greaham, C. F. (1978) Enolase isoenzymes as markers of differentiation in teratocarcinoma cells and normal tissues of mouse, Dev. Biol. 65,

9. Keller, A,, Ott, M. O., Lamand6, N., Lucas, M., Gros, F., Bucking- ham, M. & Lazar, M. (1992) Activation of the gene encoding the glycolytic enzyme beta-enolase during early myogenesis precedes an increased expression during fetal muscle development, Mech. Develop. 38, 41 -54.

10. Barbieri, G., De Angelis, L., Feo, S., Cossu, G. & Giallongo, A. (1990) Differential expression of muscle-specific enolase in embryonic and fetal myogenic cells during mouse development, Dq- ferentiation 45, 179- 184.

11. Ibi, T., Sahashi, K., Kato, K., Takahashi, A. & Sobue, I. (1983) Immunoistochemical demonstration of beta enolase in human skeletal muscle, Muscle & Nerve 6, 661-663.

12. Kato, K., Shimizu, A., Semba, R. & Satoh, T. (1985) Tissue distribution, developmental profiles and effect of denervation of enolase isozymes in rat muscles, Biochim. Biophys. Acta 841, 50-58.

13. Peterson, C. A., Cho, M., Rastinejad, F. & Blau, H. M. (1992) Beta- enolase is a marker of human myoblast heterogeneity prior to differentiation, Dev. Bid. 151, 626-629.

427 -433.

367-374.

462-475.

14. Blau, H. M., Chiu, C. P. & Webster, C. (1983) Cytoplasmic activation of human nuclear genes in stable heterocaryons, Cell 32,

15. Chomczynski, P. & Sacchi, N. (1987) Single step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform ex- traction, Anal. Biochem. 162, 156-159.

16. Schiaffino, S., Samuel, J. L., Sassoon, D., Lompr6, A. M., Gamer, I., Marotte, F., Buckingham, M., Rappaport, L. & Schwartz, K. (1989) Nonsynchronous accumulation of alpha-skeletal actin and beta-myosin heavy chain mRNAs during early stages of pressure- overload-induced cardiac hypertrophy demonstrated by in situ hybridization, Circ. Res. 64, 937-948.

17. Frohman, M. A,, Dush, M. K. & Martin, R. G. (1988) Rapid production of full-length cDNAs from rare transcripts : amplification using a single gene-specific oligonucleotide primer, Proc. Nut1 Acad. Sci. USA 85, 8998-9002.

18. Sanger, F., Nicklen, S . & Coulson, A. R. (1978) DNA sequencing with the chain-terminating inhibitors, Proc. Natl Acad. Sci. USA 74, 5463-5467.

19. Zuker, M. & Stiegler, P. (1981) Optimal computer folding of large RNA sequences using thermodynamics and auxiliary information, Nucleic Acids Res. 9, 133-148.

20. Giallongo, A,, Oliva, D., Cali, L., Barba, G., Barbieri, G. & Feo, S. (1990) Structure of the human gene for alpha enolase, Eur J. Biochem. 190, 567-573.

21. Williams, T. J. & Fried, M. (1986) The MES-1 murine enhancer element is closely associated with the heterogeneous 5’ ends of two divergent transcription units, Mol. Cell. Biol. 6, 4558-4569.

22. Ryabova, L. A,, Torgashov, A. F., Kumasov, 0. V., Bubunenko, M. G. & Spirin, A. S. (1993) The 3’-terminal untraslated region of alfalfa mosaic virus RNA 4 facilitates the RNA entry into translation in a cell-free system, FEBS Lett. 326, 264-266.

23. Buckingham, M., Cohen, A. & Gros, F. (1976) Cytoplasmic distribution of pulse-labelled poly(A)-containing RNA, particularly 26 S RNA, during myoblast growth and differentiation, J . Mol. Biol. 103, 611-626.

24. Kozak, M. (1986) Influences of mRNA secondary structure on initiation by eukaryotic ribosomes, Proc. Natl Acad. Sci. USA 83, 2850-2854.

25. Kozak, M. (1989) Circumstances and mechanisms of inhibition of translation by secondary structure in eucaryotic mRNAs, Mol. Cell. Biol. 9, 5134-5142.

26. Kozak, M. (1991) Structural features in eukaryotic mRNAs that modulate the initiation of translation, J. Biol. Chem. 266, 19867- 19 870.

27. Kozak, M. (1991) An analysis of vertebrate mRNAs sequences: in- timations of translational control, J. Cell Biol. 115, 887-903.

28. Cleveland, D. W. (1989) Gene regulation through messenger RNA stability, Cum Opin. Cell I , Biol. 1, 1148-1153.

29. Ding, D. & Lipshitz, H. D. (1993) Localized RNAs and their func- tions, BioEssays 15, 651 -658.

30. Jackson, R. J. (1993) Cytoplasmic regulation of mRNA function: the importance of the 3’ untraslated region, Cell 74, 9-14.

31. Melefors, 0. & Hentze, M. W. (1993) Translational regulation by mRNNprotein interactions in eukaryotic cells: ferritin and be- yond, BioEssays 15, 85-90.

32. Guan, K. L. & Weiner, H. (1989) Influence of 5’-end region of alde- hyde dehydrogenase mRNA on translational efficiency, J. Biol. Chem. 264, 17764-17769.

33. Manzella, J. M. & Blackshear, P. J. (1990) Regulation of rat omith- ine decarboxylase mRNA translation by its 5’-untraslated region, J. Biol. Chem. 265, 11817-11822.

34. Lopes-Casillas, F. & Kim, K-H. (1991) The 5’ untraslated regions of acetyl-coenzyme A carboxylase mRNA provide specific translational control in vitro, EUK J. Biochem. 201, 119-127.

35. Aasheim, H-C, Aas-Eng, D. A, Blomhoff, H. K., Funderud, S. & Smeland, E. B. (1 993) Cell-specific expression of human p-galac- toside a2,6-sialyltransferase transcripts differing in the 5’ untraslated region, Eur J. Biochem. 213, 467-475.

36. Nielsen, F. C., Gammeltoft, S. & Christiansen, J. (1990) Transla- tional discrimination of mRNAs coding for human insulin-like growth factor 11, J. Biol. Chem. 265, 13431-13434.

37. Gross, M. K. & Merrill, G. F. (1989) Thimidine kinase synthesis is repressed in nonreplicating muscle cells by a translational mecha-

1171 -1180.


nism that does not affect the polysomal distribution of thymidine kinase mRNA, Proc. Natl Acad. Sci. USA 86, 4981-4991.

38. Rousseau, D., Khochbin, S., Gorka, C. & Lawrence, J. (1991) Regu- lation of histone H1 accumulation during induced differentiation of murine erythroleukemia cells, J. Mol. Biol. 21 7, 85 -92.

39. Bies, R. D., Phelps, S. F., Cortez, M. D., Roberts, R., Caskey, C. T. & Chamberlain, J. S. (1992) Human and murine dystrophin mRNA transcripts are differentially expressed during skeletal muscle, heart, and brain development, Nucleic Acids Res. 20,

40. Su, M. S., Suzuki, H. R., Bieker, J. J., Solursh, M. & Ramirez, F.(1991) Expression of two nonallelic type I1 procollagen genes during Xenopus laevis embryogenesis is characterized by stage-

1725-1731.

specific production of alternatively spliced transcripts, I . Cell Biol. 115, 565 -515.

41. Takechi, H., Hosokawa, N., Hirayoshi, K. & Nagata, K. (1994) Al- ternative 5’ splice site selection induced by heat shock, Mol. Cell. Biol. 14, 561-575.

42. Satoh, T., Matsushita, H., Kato, K., Yarnada, S. & Adachi, M.(1991) Muscle p-enolase under tetrodoxin-paralysis or chronic stimula- tion of the sciatic nerve, Biomed. Res.12, 165-168.

43, Benton, M. J. (1990) Phylogeny of the major tetrapod groups: mor- phological data and divergence dates, J. Mol. Evol. 30, 409-424.

44. Goodman, M., Czelusniack, J., Koop, B. F., Tagle, D. A. & Slightom, J. L.11987) Globins: a case study in molecular phylogeny, Cold Spring Harbor Symp. Quant. Biol. 52, 815-890.

Conserved Alternative Splicing in the 5'-Untranslated Region of the Muscle-Specific Enolase Gene....

Documents

Transcript of Conserved Alternative Splicing in the 5'-Untranslated Region of the Muscle-Specific Enolase Gene....