INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

Functional cholinergic neurons make and use acetylcholine (ACh)¹ as a neurotransmitter and must coordinately express genes encoding choline acetyltransferase (ChAT, EC 2.3.1.6), for biosynthesis of ACh as well as a vesicular ACh transporter (VAChT), for accumulation of transmitter into synaptic vesicles. The VAChT gene has been cloned from Caenorhabditis elegans (1), Torpedo (2), rat (3-5), and human (3), and was found to encode a protein with 12 putative transmembrane domains which shows broad homology to the functionally characterized vesicular monoamine transporters (6, 7).

In addition to the sequence homology with vesicular monoamine transporters, genetic and biochemical studies have established the functionality of VAChT protein. Non-lethal mutations in the C. elegans VAChT gene (unc17) lead to uncoordinated movements and resistance to normally toxic levels of ACh esterase inhibitors (1, 8, 9). Modest levels of ATP-dependent vesicular transport of ACh have been shown in fibroblasts transfected with rat cDNA clones (3) and in homogenates of PC-12 cells transfected with the human VAChT cDNA (10). Furthermore, microinjection of rat VAChT cDNA into Xenopus embryos has resulted in increased quantal transmitter packaging in motor neurons (11).

In all species examined so far, the VAChT and ChAT genes form a phylogenetically conserved "cholinergic locus" in which the former gene is nested within the first intron of the latter. This unusual genomic arrangement of these two separate but related genetic functions may contribute to the coordinated regulation of both genes in some types of cholinergic neurons and may represent an example of a eukaryotic "operon" (12). In fact, several studies using cell culture systems have indicated coordinate regulatory properties for VAChT and ChAT expression (13-15). In C. elegans, alternative splicing of a common precursor RNA is proposed to be responsible for producing two distinct mRNAs from the cholinergic locus (16). In vertebrates, however, it is likely that ChAT and VAChT specific mRNAs result from a combination of differential promoter usage and/or alternative RNA processing (3, 5, 17, 18).

We have been using Drosophila melanogaster as a model organism to understand the mechanisms and physiological significance of cholinergic neuron-specific gene regulation. Our previous studies have identified the transcriptional regulatory DNA responsible for the spatial and temporal expression pattern of the Drosophila ChAT gene (19-22). We have also noted that the Drosophila ChAT gene locus produced an unidentified 7-kb transcript throughout development in addition to the 4-kb ChAT-specific mRNA (23, 24). The 7-kb transcript could be detected with a cRNA probe specific to the first exon of ChAT, but not with a probe specific to the eighth exon (24), indicating that it is transcribed in the same direction as the ChAT gene, shares at least part of the first exon with authentic ChAT mRNA, and lacks other ChAT specific exons. In this study we have isolated cDNA clones for the 7-kb transcript and demonstrate that it encodes Drosophila VAChT. In contrast to the vertebrate cholinergic locus, we also demonstrate that all Drosophila VAChT transcripts contain a shared first exon with ChAT-specific transcripts, indicating that both genetic functions are under common transcriptional control. The levels of ChAT- and VAChT-specific transcripts also appear to be independently controlled by post-transcriptional mechanisms and the structure of the VAChT 5-untranslated region (5-UTR) suggests that Drosophila VAChT may not use a conventional ribosome scanning mechanism for initiation of protein translation.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Northern and Southern Blot Analyses-- Northern and Southern blot analyses were carried out according to Sambrook et al. (25). For Northern blots, total RNA was extracted from adult Drosophila heads by the CsCl/guanidinium isothiocyanate method (26). For Southern blot analysis, P1 DNA containing the ChAT and VAChT genes was purified as described in Ref. 27, and digested with either BamHI or SalI. ³²P-Labeled probes were prepared from cDNA or genomic DNA clones using the Prime-It random primer labeling kit (Stratagene).

Cloning of VAChT cDNA and Genomic DNA-- A Drosophila adult head cDNA library constructed in the ZAP vector (Stratagene) was kindly provided by Dr. C. Hama (National Institute of Neuroscience, Tokyo). Approximately 3.6 × 10⁵ plaques were screened with either ³²P-labeled probe d (ChAT and VAChT first exon, see Fig. 1) or e (ChAT first intron). Positive clones for probe d were further screened with probe b (ChAT exon III-VIII) and negative clones were analyzed further as candidates for VAChT cDNA. The Bluescript phagemids containing candidate VAChT sequences were excised from the ZAP vector and recircularized in vivo using the ExAssist/SOLR system (Stratagene) according to the manufacturer's protocol. DNA sequencing was carried out by the didoxynucleotide chain termination method using the Sequenase 2.0 kit (U. S. Biochemical Corp.).

Polymerase Chain Reaction (PCR)-- Reverse transcription-coupled with PCR (RT-PCR) was carried out as described in Ref. 28 with adult head total RNA. Primer pairs used to amplify a gap between two partial VAChT cDNA clones (G7 and H3, see below) and the 3-end of the cDNA were: 5-TGTCATTGTGGCGGACTT-3, 5-GGTACACACATTTATGCT-3 and 5-CGAATGTCCATTTCGGTAGAA-3, 5-GAGAGAGAGAGAGAGAGAGAACTAGTCTCGAGTTTTTTTTTTTTTTTTTT-3.

Primer Extension-- Primer extension analysis was performed according to a standard protocol (29), using an oligonucleotide primer 5-CTATTCCTGGCTATACAATGTAGCT-3. One pmol of the ³²P-labeled primer was mixed with 45 µg of adult head total RNA, incubated at 65 °C for 90 min and 22 °C for 10 min. The extension reaction was performed at 42 °C for 60 min using Superscript II (Life Technologies, Inc.), and the reaction product was analyzed on a 6% polyacrylamide, 7 M urea gel.

RNase Protection Assay-- RNase protection assay was performed according to Gilman (30). Drosophila ChAT (XhoI-SacI fragment), VAChT (PstI fragment), and -tubulin (DraII-EcoRV fragment) sequences were subcloned into Bluescript (Stratagene). After restricting the plasmids with PstI, BamHI, or RsaI, respectively, [³²P]UTP-labeled cRNA probes were prepared using the MAXIscript in vitro Transcription Kit (Ambion). A probe used to detect both VAChT and ChAT mRNAs was prepared by using the BamHI-HindIII fragment from VAChT cDNA cloned into Bluescript II (Stratagene). The plasmid was digested with BamHI before transcription. Total RNA was extracted using the RNAzol method (TEL-TEST, Inc.). Quantitative analysis of transcripts was performed using ³²P-labeled ChAT and VAChT antisense probes (2 × 10⁵ cpm/each probe/tube) and a ³²P-labeled -tubulin antisense probe (2 × 10⁴ cpm/tube) mixed with unlabeled (30 ng/tube) -tubulin probe. The hybridization mixture was ethanol precipitated, dissolved in 20 µl of hybridization buffer (80% formamide in 40 mM Pipes, pH 6.4, 0.4 M NaCl, 1 mM EDTA) and incubated overnight at 50 °C. The mixture was then digested with ribonuclease T₁ (400 units/200 µl; Life Technologies, Inc.) for 30 min at 37 °C. The ribonuclease digest was electrophoresed in a denaturing gel, which was dried and exposed to a PhosphorImager screen overnight. The levels of ChAT, VAChT, and -tubulin mRNA were determined by densitometry using a Molecular Dynamics PhosphorImager and ImageQuant software (Molecular Dynamics, Sunnyvale, CA). Results are shown as the ratios of the density of the bands protected by either ChAT or VAChT mRNA to that of bands protected by tubulin mRNA.

Anti-Drosophila VAChT Polyclonal Antiserum and Western Blot Analysis-- A cDNA fragment corresponding to a portion of the predicted hydrophilic VAChT C-terminal domain (Arg-441 to Asn-546) was amplified by PCR using two oligonucleotide primers (5-CGGGATCCGAAAGCTGCGCAACATCT-3, 5-TGGAATTCATTGGCCACACGGGGCTG-3), and the amplified fragment was inserted into the BamHI/EcoRI site of the pGEX-3X fusion protein expression vector (Phamacia). The fusion protein was produced and purified according to Smith and Johnson (31). The VAChT protein fragment was separated from glutathione S-transferase by incubation with factor Xa (Boehringer Mannheim) overnight at 30 °C, followed by preparative SDS-polyacrylamide gel electrophoresis. Polyclonal antiserum was generated according to a standard procedure described in Harlow and Lane (32).

Drosophila Stocks and Temperature Shift Experiments-- For temperature shift experiments, wild type Canton-S and two temperature-sensitive mutants of the ChAT gene, Cha^ts1 and Cha^ts2 (33) were reared on a standard diet of cornmeal and agar at 18 °C, 65% humidity, on a 12-h light-dark cycle. Cha^ts1 and Cha^ts2 are point mutations in the ChAT structural gene, which result in thermolabile forms of the enzyme (34). Flies of the same age (3-4 days old) were transferred to a 30 °C incubator for the appropriate period of time. For Western blot analysis, Canton-S and Df(3R)Cha⁵/TM6 were reared at 25 °C. Df(3R)Cha⁵ is a deficiency lacking the chromosome region 91B3-91D1 where the ChAT and VAChT genes reside (35).

Prediction of RNA Secondary Structure-- Secondary structure of Drosophila VAChT mRNA 5-UTR was predicted by energy minimization using mfold RNA folding prediction program (36).

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

Northern Blot Analysis of the Cholinergic Locus-- Northern blot analysis of fly head RNA using a full-length ChAT cDNA probe reveals 2 transcripts (Fig. 1, probe a). The 4-kb transcript hybridizes with probes from ChAT exons III-VIII (Fig. 1, probe b) or a ChAT exon I probe (probe d). The longer 7-kb transcript can be detected with either probe a or d but not with probe b. Probes e and f (prepared from 2 different non-overlapping restriction fragments within the first intron of ChAT) only detect the 7-kb transcript. A probe prepared from the 5-flanking genomic DNA immediately upstream of the ChAT transcription start site (probe c) does not detect either transcript. These results indicate that the 7-kb transcript shares an exon with ChAT and that the remainder of its sequence is derived form the ChAT first intron. Based on evolutionary conservation of the cholinergic locus in other species, the nested 7-kb transcript was expected to encode Drosophila VAChT.

View larger version (23K): [in this window] [in a new window]   Fig. 1.   Northern blot analysis of the cholinergic locus. A, a schematic representation of the complete organization of the Drosophila ChAT gene. The gene is composed of eight exons (filled boxes) with a long first intron. The shaded boxes indicate DNA fragments used as probes for Northern blot analysis (probes a-f). B, Northern blot analysis of total RNA extracted from adult heads (15 µg/lane) hybridized with indicated probes.

Identification of Partial cDNA Clones for the VAChT Gene-- A Drosophila adult head cDNA library was screened with probe d which hybridized to both the 7- and 4-kb transcripts (Fig. 1). Forty-two positive clones were isolated and seven of them were shown to be negative for hybridization with probe b (a ChAT specific probe). Restriction analysis indicated that all seven cDNA clones were derived from the same transcript and partial sequencing of the 5-ends revealed that all clones contained the 5-UTR sequence of the ChAT gene. The longest clone, G7 (3.6 kb), was further analyzed and completely sequenced.

View larger version (97K): [in this window] [in a new window]   Fig. 2.   The predicted amino acid sequence of Drosophila VAChT and alignment with VAChT sequences of other species. A predicted amino acid sequence of Drosophila VAChT was aligned with VAChT sequences of rat, human, Torpedo, and C. elegans using the ClustalW algorithm. Identical amino acid residues and putative transmembrane domains (TM) are indicated with black boxes and underlines, respectively. Two potential N-linked glycosylation sites in Drosophila VAChT are indicated with arrowheads (Asn-78 and Asn-293).

Detection of the VAChT Protein by Western Blot Analysis-- Western blot analysis of adult head extracts using anti-Drosophila VAChT antiserum (see "Experimental Procedures") detected a rather broad band with an M_r of 65,000 (Fig. 3), which agreed well with the predicted molecular mass of the longest open reading frame from the cDNA sequence. The identity of this band was further confirmed when extracts from the wild type (Canton-S) and flies carrying a VAChT gene-deficient chromosome (Df(3R)Cha⁵/TM6), were compared. Although overall protein patterns and intensities were indistinguishable between Canton-S and Df(3R)Cha⁵/TM6 (data not shown), the intensity of the antibody-positive band was reduced approximately to one-half when the extract from Df(3R)Cha⁵/TM6 was used (Fig. 3).

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Western blot of VAChT protein. Extracts of Drosophila heads (CS, Canton-S; Cha5: Df(3R)Cha5/TM6) were electrophoresed on 10% SDS-polyacrylamide gel, transferred to nitrocellulose membrane, and probed with rabbit anti-Drosophila VAChT antiserum. Molecular size markers (Bio-Rad) are shown on the right.

Characterization of Full-length VAChT cDNA-- The G7 cDNA clone lacks both the 5- and 3-regions of the 7-kb transcript. In a previous study Sugihara et al. (37) tentatively determined the 5-end of ChAT exon I by sequencing a primer extension product obtained using a synthetic primer complimentary to nucleotides at position 47-70 (relative to the ChAT translation start site). We have re-examined the 5-end of this shared exon by carrying out primer extension analysis using primers that corresponds to the 5-ends of VAChT cDNA clones G1 and G12 (see Fig. 4A). Two major bands and one minor band were observed following primer extension (Fig. 4B) and the corresponding nucleotides are indicated in Fig. 4A.

View larger version (54K): [in this window] [in a new window]   Fig. 4.   The full-length cDNA. A, nucleotide sequence of the Drosophila VAChT 5-UTR. The long horizontal arrow under the sequence denotes the position of the oligonucleotide used for primer extension assay. Asterisks indicate the 5-ends of isolated VAChT cDNAs. The short vertical arrows represent putative transcription start sites determined by primer extension assay. The initiation codons for ChAT and VAChT translation, and three upstream ATG codons are indicated by black and shaded boxes, respectively. The junction between the first and second exons is indicated by an arrowhead, and the polypurine tract is double-underlined. Nucleotides that differ from those in Sugihara et al. (Ref. 37, shown in parentheses) are underlined. B, primer extension reactions (see "Experimental Procedures") are carried out with (+) or without () Drosophila head RNA. Positions of primer extension reaction products are indicated by arrowheads on the right. DNA sequencing was carried out with the primer used for primer extension and genomic DNA as a template (four left lanes). C, a schematic representation of the full-length VAChT cDNA which contains a 1734-base pair coding region (black box). The long 5- and 3-untranslated regions are indicated by the thin lines underneath the diagram. The gray box indicates the shared first exon of VAChT and ChAT. The thick lines at the bottom of the figure indicate overlapping partial cDNA clones obtained from adult head cDNA libraries (G7 and H3) or by amplifying adult head mRNA using RT-PCR (RT-PCR1 and RT-PCR2).

Genomic Organization of the Drosophila Cholinergic Locus-- Sugihara et al. (24) have previously isolated and characterized three independent phage clones (gB517, gB7, and gB1) which contained all 8 exons of Drosophila ChAT and were missing only a part of the first ChAT intron. We have now isolated this missing genomic region as part of a P1 clone (DS07917) representing polytene chromosome region 91C-D (the cytogenetic position of the ChAT/VAChT locus). This P1 clone hybridized with probe b, probe c, G7, and H3 (data not shown), indicating that it contains the entire ChAT gene including the long first intron containing the complete VAChT cDNA. Southern blot analysis of DS07917 digested with either BamHI or SalI showed four or two bands when G7 was used as a probe (Fig. 5A). The 5.5-kb band (band 3 in Fig. 5A) corresponds to the BamHI fragment located in the gap between gB517 and gB7. The size of the first intron of the VAChT and ChAT genes are thus 6.1 and 17.4 kb, respectively. This estimation was further confirmed by PCR amplification using primers corresponding to the 3-end of genomic clone gB517 and the 5-end of the VAChT second exon (data not shown). A 1.8-kb (SalI/BamHI) genomic fragment containing the boundary of the VAChT first intron/exon II was subcloned from DS07917, and the intron/exon boundary was determined. The junction showed a conventional intron/exon sequence boundary, TTAAAG/GTTGTT.

View larger version (28K): [in this window] [in a new window]   Fig. 5.   Genomic organization of the ChAT and VAChT genes. A, Southern blot of P1 clone DS07917 digested with BamHI (Bam) or SalI (Sal), and probed with VAChT cDNA (G7). Estimated sizes of positive fragments are: 1, 27 kb; 2, 7.4 kb; 3, 5.5 kb; 4, 2 kb; 5, 9.8 kb; 6, 6.5 kb. B, PCR amplification of VAChT genomic DNA or cDNA. Bent arrows represent the positions of oligonucleotide primers used to amplify cDNA and genomic DNA fragments and the black bars indicate the identical size fragments which were amplified (a-g). C, The organization of the Drosophila ChAT/VAChT locus. The extent and relationship of P1 clone DS07917 and the EMBL3 clones (gB517, gB7, and gB1) are indicated by horizontal lines and the positions of SalI and BamHI sites indicated by vertical lines. Identification of the restriction fragments detected in the Southern blot (A) are indicated. The ChAT/VAChT shared exon, VAChT and ChAT specific exons are shown as shaded, open, and black boxes, respectively.

VAChT mRNA Is Present Only in a Form Which Includes the Shared Exon I-- A class of VAChT mRNAs in rat has recently been identified that does not contain sequences shared with ChAT mRNA (18). To examine the possibility that such mRNAs may also be present in Drosophila an RNase protection assay was carried out with a probe that included the boundary sequence of the shared first exon and the unique VAChT second exon (Fig. 6A). This cRNA probe yielded only two protected fragments which correspond to ChAT and VAChT mRNA (Fig. 6B). The absence of a 238-nucleotide fragment which could correspond to VAChT second exon sequence attached to some other sequence indicates that all detectable VAChT mRNA is present in a form which includes the shared first exon with ChAT.

View larger version (35K): [in this window] [in a new window]   Fig. 6.   RNase protection assay. A, a cRNA probe (482 nucleotide, nt) containing the junction of the shared first exon and part of the VAChT specific second exon. Shaded boxes indicate possible protected fragments. An open box indicates unprotected bases derived from a vector sequence. B, protected RNA fragments. Total RNA was isolated from whole flies and an RNase protection assay was carried out using a tubulin cRNA probe and a ChAT/VAChT cRNA probe shown in A. Undigested probes (ChAT/VAChT and tubulin) are shown in the two left lanes, protected fragments are on the right and their positions indicated by arrows.

Differential Regulation of ChAT and VAChT mRNA Levels-- To examine coordinate regulation of ChAT and VAChT mRNA expression, we determined ChAT and VAChT mRNA levels on the same samples using a quantitative RNase protection assay. cRNA probes were designed to protect a part of the specific coding region for each mRNA (Fig. 7A). Transcript levels were measured in whole flies or in dissected heads or bodies. As shown in Fig. 7B the ratio of VAChT to ChAT transcript is approximately equal in heads while whole flies or bodies show an excess of VAChT transcript relative to ChAT. This result implies that the levels of VAChT and ChAT specific mRNA can be differentially regulated in different parts of the fly.

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Differential regulation of ChAT and VAChT mRNA levels. A, cRNA probes (black boxes) designed to protect a part of the coding region (gray boxes) for the ChAT- and VAChT-specific transcripts. Also indicated are positions of restriction sites used to prepare cRNA probes (P, PstI; S, SacI; B, BamHI; X, XhoI). B, ChAT and VAChT mRNA levels in the whole fly, head, and body. Total RNA was isolated from 10 whole flies, 30 bodies, or 15 heads. Data are shown as the mean ± S.E. for four RNase protection assays following four independent RNA preparations. Values correspond to one whole fly, body, or head. C, ChAT and VAChT mRNA levels in the wild type fly (CS, Canton-S) and two temperature-sensitive mutants (ts1, Chats1; and ts2, Chats2) before and after incubation at 30 °C for 1 day. Data are shown as the mean ± S.E. for seven RNase protection assays using RNA samples independently prepared for three times. nt, nucleotide.

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

The Drosophila cholinergic locus, like its counterpart in nematodes and vertebrates, has a unique genomic organization: the VAChT gene is nested within the first intron of the ChAT gene and both transcripts share a common first exon. All Drosophila VAChT transcripts appear to share the complete first exon with ChAT implying that both genetic functions are under common transcriptional control. Our previous studies identified the 5 cis-regulatory DNA responsible for the spatial and temporal expression pattern of the Drosophila ChAT gene using transgenic animals (19-22). The overall strategy used to regulate ChAT transcription is to activate or repress several regulatory sites which are distinct in different types of cholinergic neurons (19-21). Furthermore, the Drosophila ChAT gene regulatory information essential for survival has been mapped to a region of the 5-flanking DNA 300 bases in length (22). In this sequence the most strongly footprinted sequence was a 22-base pair region that contains an octamer-like motif in its center. A Drosophila transcription factor in the POU family, pdm-1 was shown to bind to the 22-base pair sequence and suggested to regulate ChAT gene expression (22). The regulatory strategies and mechanisms identified for Drosophila ChAT expression should be also applicable to expression of another cholinergic marker, VAChT. In contrast to this simple situation in Drosophila, some mammalian VAChT transcripts have been reported to share sequence with ChAT while others do not (18). These latter VAChT transcripts apparently use alternative promoters. The nematode cholinergic locus is thought to give rise to specific ChAT and VAChT transcripts through alternative splicing, although the situation is complicated by additional trans-splicing reactions to the first shared exon (16).

The organization of the Drosophila ChAT and VAChT genes suggests that two distinct transcripts are produced by alternative RNA processing of a common primary transcript. Usually, the alternatively processed transcripts code for genetic isoforms or occur in different cell types or tissues (42). Production of specific VAChT and ChAT transcripts is unique, however, since two transcripts encode metabolically related but completely distinct genetic functions. Furthermore, both transcripts must be produced in the same cells at the same times to have proper cholinergic neuron function. RNA processing to generate specific transcripts for each genetic function may involve one or both of two possible mechanisms. Alternative 3 splice site choice between VAChT and ChAT second exons would be responsible for producing the respective specific transcripts with a shared first exon from a common pre-mRNA. Alternatively, selection of poly(A) addition site could be a major determinant for which mRNA is to be made. It is not yet known which of these potential mechanisms operate in Drosophila.

Since ChAT and VAChT are under common transcriptional regulatory control in Drosophila, the levels of each transcript are likely to be fixed in different parts of the nervous system. It was somewhat surprising, therefore, to see that the ratio of VAChT to ChAT transcripts varied between heads and bodies as we have observed. Apparently, the levels of each transcript can be regulated independently of each other by post-transcriptional mechanisms in the process of either mRNA production or degradation. As we showed before (40, 41), the ChAT mRNA levels decreased in Cha^ts1 and Cha^ts2 at the non-permissive temperature probably because normal levels of ACh and/or cholinergic neuronal transmission are necessary for keeping the normal ChAT mRNA levels (41). Interestingly, the levels of ChAT and VAChT specific mRNAs can be differentially regulated in the Cha^ts mutant flies. Again a post-transcriptional regulatory phenomenon is likely to account for decreasing levels of ChAT mRNA while the VAChT mRNA levels are held constant. The most appealing rationale for having VAChT and ChAT genetic functions organized as a single genetic locus under common transcriptional control would be to ensure coordinate regulation of the expression of these two genes. The uncoupling of the levels of specific transcripts we have observed in space or in response to a temperature-sensitive structural gene mutant in the ChAT function would perhaps tend to obviate this potential regulatory advantage. Cholinergic neurons could have different physiological demands for biosynthesis and packaging of neurotransmitter in different parts of the nervous system. There may also be differential rates of turnover for the specific transcripts or proteins, thus making it attractive to fine tune the levels of each using post-transcriptional regulatory strategies.

In cultured mammalian cells ChAT and VAChT mRNA levels respond in a coordinated fashion when soluble factors, such as retinoic acid, cAMP, and the leukemia inhibitory factor are added (13-15). In addition, distribution of ChAT and VAChT mRNAs in the adult rat brain is virtually identical (3), suggesting coordinate regulation of these two genes even though specific promoters for ChAT and VAChT may exist (18). ChAT and VAChT mRNA levels, however, can be differentially regulated during rat brain development. VAChT mRNA levels are already present at up to 60% of adult levels during late embryonic stages while ChAT mRNA levels increase primarily during postnatal stages (43). It is not known if the uncoordinated changes in ChAT and VAChT mRNA levels are due to differences in transcription or RNA processing.

The 7.2-kb size of the Drosophila VAChT transcript is significantly larger than the 2-3-kb transcripts seen in other species (1-5). This difference in mRNA size is attributed primarily to the unusually long 5- and 3-UTR in Drosophila VAChT mRNA. It should be mentioned, however, that a relatively large transcript (~6 kb) has also been detected in a human neuroblastoma cell line (44), using a ChAT cDNA probe. This larger human transcript has not yet been characterized.

In nematodes and mammals, the exon shared by VAChT and ChAT does not contain any protein coding sequence. In Drosophila, however, the 3-half of the shared first exon corresponds to the ChAT protein coding region and includes an unusual ChAT GTG translation initiation codon. In addition, the VAChT 5-UTR contains three potential translation initiation codons (ATG) which are all followed by multiple termination codons and thus not likely to mark the beginning of an open reading frame (Fig. 4A). Most eukaryotic mRNAs are thought to initiate protein translation using a ribosome scanning mechanism. Ribosomes bind to a 5-cap structure and progressively scan the mRNA until encountering the first AUG codon which serves as the site of translation initiation (45). The structural features of the Drosophila VAChT 5-UTR indicate that efficient translation initiation cannot be accomplished by the conventional 5-cap dependent ribosome scanning mechanism.

One of the alternative mechanisms for Drosophila VAChT translation initiation is the use of an internal ribosome entry site (IRES). In IRES-mediated translation initiation, ribosomes bypass the 5-CAP by directly binding to an internal sequence and initiate protein translation at the first AUG downstream of the IRES site. IRES-mediated translation has been identified in a number of viral as well as cellular mRNAs (46). Although functional IRES motifs are rather divergent in sequence (47), Piconoviral IRES elements exhibit highly conserved secondary structures and contain an polypyrimidine tract followed by an ATG triplet (48). As shown in Fig. 8 computer analysis of the Drosophila VAChT 5-UTR using mfold (36) also predicts formation of secondary structures. Coupled with the presence of an extensive polypurine tract between the last upstream ATG and the initiation codon (Fig. 4A), these features may suggest that Drosophila VAChT also initiates translation via an IRES-dependent mechanism. Drosophila ChAT and VAChT may thus use different translation initiation mechanisms to achieve differential regulation. In rats, a comparison of the developmental expression patterns of VAChT mRNA and protein revealed a large excess of mRNA from late embryonic stages up to early postnatal ages (43), suggesting possible regulation of VAChT translation in mammalian nervous system during development.

View larger version (12K): [in this window] [in a new window]   Fig. 8.   Predicted secondary structure of VAChT 5-UTR. Secondary structure of the VAChT 5-UTR (the first 1037 nucleotides) was predicted by energy minimization using mfold version 2.3 with a folding temperature of 25 °C. A computed free energy is 364.7 kcal/mol. Arrows indicate 5- and 3-ends of VAChT mRNA 5-UTR. The ChAT translation initiation site and an exon junction are also shown by arrows.