JBC Connect with Cosmo for Collagen Detection

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Boularand, S.
Right arrow Articles by Mallet, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Boularand, S.
Right arrow Articles by Mallet, J.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Volume 270, Number 8, Issue of February 24, 1995 pp. 3748-3756
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
The Human Tryptophan Hydroxylase Gene
AN UNUSUAL SPLICING COMPLEXITY IN THE 5`-UNTRANSLATED REGION (*)

(Received for publication, October 6, 1994)

Sylviane Boularand(§), (¶),  ,  Michèle C. Darmon ,  Jacques Mallet

From the Laboratoire de Génétique Moléculaire, de la Neurotransmission, et des Processus Neurodégénératifs, C.N.R.S., F91198 Gif-sur-Yvette Cedex, France

ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES

ABSTRACT

We report the isolation and the organization of the gene encoding human tryptophan hydroxylase (TPH) and an analysis of the corresponding mRNAs. The gene spans a region of 29 kilobases, which contains at least 11 exons and a variably spliced 5`-untranslated region (5`-UTR). The sequence of the coding region and the majority of the positions of the intron-exon boundaries of human TPH gene are very similar to those encoding human tyrosine hydroxylase and phenylalanine hydroxylase, the other members of the aromatic amino acid hydroxylase family. Phylogenetic analysis evidences the early divergence and the independent evolution of the three hydroxylase types. TPH cDNA cloning and anchored polymerase chain reaction revealed a diversity of the TPH mRNA, which is restricted to the 5`-UTR. Four TPH mRNA species were detected by Northern blot with pineal gland and carcinoid tumor RNAs. These messengers are transcribed from a single transcriptional initiation site, and their diversity results from differential splicing of three intron-like regions and of three exons located in the 5`-UTR. Analysis by S1 nuclease protection revealed that the intron-like regions in the 5`-UTR are mostly unspliced and that TPH mRNA species where the three intron-like regions are eliminated are present at low level in pineal gland and not detectable in carcinoid tumors.

INTRODUCTION

Tryptophan hydroxylase (TPH) (^1)is the key enzyme in the biosynthetic pathway of the neurotransmitter serotonin. In mammals, TPH is only expressed by a small number of tissues, namely the brainstem raphe nuclei, the pineal gland in the central nervous system, and the pancreatic and intestinal enterochromaffin cells in the periphery. The serotonergic neurons of the raphe nuclei form a highly divergent neuronal system controlling the basic activity of many target regions distributed throughout the forebrain, the cerebellum, and the spinal cord. This neuronal system modulates a variety of psychological and physiological processes including thirst and appetite, sleep and memory, and reproduction(1) . In the pineal gland, the concentration of serotonin, which is an intermediate in melatonin synthesis, is higher than in any other region of the brain or in any other organ analyzed. The production of melatonin is characterized by a dark-light circadian rhythm, which synchronizes the circadian and ultradian cycles involved in a variety of functions including sleep, sexual behavior, and body temperature(2, 3) . However, the mechanisms involved in regulating these diverse activities remain uncharacterized. At the periphery, TPH is also present in the autonomous nervous system, in the intestinal and pancreatic enterochromaffin cells, and in carcinoid tumors. These tumors develop from intestinal enterochromaffin cells, produce large amounts of TPH, and secrete large amounts of serotonin, accounting for most of the symptoms associated with this pathology(4, 5) .

TPH is a member of the aromatic amino acid hydroxylase family, which also includes tyrosine hydroxylase and phenylalanine hydroxylase. These three enzymes require a reduced pterin cofactor to hydroxylate their amino acid substrate and interact with Fe in their tetrameric quaternary structure to coordinate oxygen molecules(6) . The human TPH amino acid sequence is similar to those of tyrosine hydroxylase (7) and phenylalanine hydroxylase(8) , and the most conserved region comprises the 340 C-terminal amino acids of the proteins. This region shows 49% identity among the mammalian hydroxylases without any gaps and with mostly conservative substitutions (the TPH enzyme contains an extra 5 amino acids at the C-terminal end not found in the other enzymes). Biochemical (9, 10, 11) and sequence analyses (12) suggest that this domain corresponds to the catalytic part of the enzyme. The N-terminal domain differs more in size and sequence between the enzymes and has been proposed to modulate the enzyme activity. Relatively little is known about the regulation and the biochemistry of TPH because it is not abundant and is extremely unstable in vitro. Therefore, cloning the gene could lead to a much better understanding of the biochemistry and regulation of TPH.

Recently, cDNAs encoding rabbit and rat TPH have been cloned from pineal gland, and this was followed by the isolation of homologous human and mouse cDNAs from carcinoid tumor and mastocytoma cDNA libraries, respectively(13, 14, 15, 16) . The coding region of human TPH cDNA extends over 1332 bp, and the deduced amino acid sequence is very similar all along the sequence to those of rat (91.2%(18) ) rabbit (94.6%(13) ), and mouse (90.1%(16) ). The rat TPH gene is characterized by a diversity restricted to the 5`- and 3`-UTR of the TPH mRNAs, which could provide the basis for post-transcriptional regulation of TPH gene expression(18, 19, 20, 21) . We report the isolation of the human TPH gene and show by a combination of cDNA cloning, polymerase chain reaction (PCR) analysis, and S1 nuclease protection an unexpected diversity in the 5`-UTR of TPH mRNAs. These messengers are transcribed from a single promoter, and their diversity results from the conservation of one or several intron-like sequences and from the differential splicing of three exons in the 5`-UTR of the TPH mRNAs.

MATERIALS AND METHODS

Screening a Carcinoid Tumor cDNA Library and Sequence Analysis

A cDNA library was constructed from poly(A) RNA extracted from a human intestinal carcinoid tumor (kindly provided by J. M. Launay, Paris) using the method developed by Gubler and Hoffman(22) . Double-stranded cDNA was size-fractionated on a 10-30% sucrose gradient (containing 1 M NaCl) and cloned into the EcoRI-digested ZapII vector (Stratagene). 4 times 10^5 recombinant plaques of this human cDNA library were screened with a rat TPH cDNA probe at low stringency. Positive clones were isolated and analyzed by restriction mapping and Southern blotting. The EcoRI DNA fragments of the longest type 1 and type 2 TPH cDNA clones were sequenced on both strands.

The coding and noncoding sequences of the human TPH mRNA were aligned with those of the other aromatic amino acid hydroxylases from different animal species. Computer-aided sequence comparisons and analysis were performed with the GCG program(23) . After sequences were optimally aligned, phylogenetic distance trees were constructed by the Neighbor Joining Method of Saitou and Nei(24) , and bootstrap analysis was performed by using the MUST package(25) .

Screening a Human Genomic Library

A human genomic library constructed in the EMBL3 vector was obtained from Clontech and screened by standard procedures with a probe containing the human TPH coding sequence and part of the 5`-flanking region. 6 times 10^5 bacteriophages corresponding to about 3 times the equivalent of the size of the human genome were screened, and the positive clones were purified by four sequential rounds of plating and hybridization. Bacteriophage DNA was prepared from selected plaques according to the method described in Sambrook et al.(26) and analyzed by restriction mapping and Southern blotting. Genomic DNA fragments were subcloned into pBluescript vectors, and the intron-exon junctions were sequenced using TPH-specific primers.

Northern Blot Analysis

Total RNA was extracted from human pineal gland, carcinoid tumor, liver, dorsal raphe nuclei, and colon (kindly provided by F. Javoy-Agid, Paris) according to the technique described by Civelli et al.(27) . The poly(A) RNA of human pineal gland, carcinoid tumor, and liver were purified by oligo(dT)-cellulose chromatography. Northern blots were performed as described by Faucon Biguet et al.(28) .

S1 Nuclease Protection Analysis

S1 nuclease mapping was performed according to Sambrook et al.(26) . Aliquots of the labeled probes and 0.25 µg of poly(A) mRNA from human carcinoid tumor, pineal gland, and liver were hybridized for 16 h at 45 °C. The samples were digested with S1 nuclease (100-400 units/ml) at 37 °C for 75 min. The specifically protected DNA fragments were separated on a 4-5% denaturing polyacrylamide gel and visualized by autoradiography.

PCR Amplification

Single-stranded cDNAs were synthesized from 0.2 µg of poly(A) mRNA prepared from carcinoid tumor and pineal with 10 µM of pd(N)(6) primer and avian myeloblastosis virus reverse transcriptase. The amplification was carried out for 30 cycles with 0.3 µM of forward and reverse primers in 1.5 mM MgCl(2). Thermal cycling was as follows: denaturation at 93 °C for 40 s; annealing at 53 °C or 58 °C (depending on the T(m) of the primers) for 40 s; elongation at 72 °C for 120 s. The primers O(1), O, and O(2) corresponded to parts of exons E(1), E, and E(2), respectively. Primers O(b) and O overlapped exons E(1),E(b) and E(c),E(2), respectively, to amplify transcripts containing these two sets of exons. PCR amplification with the primers O(b) and O was performed with an annealing temperature of 50 °C to prevent the annealing of these primers to only one exon. The primer sequences were as follows: O(1), CGACCCAGCCTGCACCTAC; O, TACTGGCGCCCGAGGTGAG; O(2), TCCCCTTTCTAAGGAATGGTCTTTG; O(b), GCACCTACTGGCGCCCGAGTGGTA; O, TTTGGAGTAATTCTCTAAAACCATT. This technique is hereafter referred to as RT-PCR.

Cloning by anchored PCR of the 5`-Ends of the mRNA

Anchored PCR was performed as described by Dumas-Milne-Edwards et al.(29) . Briefly, the first cDNA strand was synthesized by avian myeloblastosis virus reverse transcriptase initiated from 0.5 µM of a TPH-specific primer (SBPE1), which corresponds to a sequence 168 bases downstream of the translation initiation codon and with 0.2 µg of poly(A) RNA from carcinoid tumor and pineal gland. Oligonucleotide BM-5` (29) was ligated to the 3`-end of the cDNA with T4-RNA ligase. Half of the ligation product was amplified by PCR with 0.15 µM primers (BM-5`-1 (29) and a TPH oligonucleotide (SBSLIC) overlapping the SBPE1 extension primer) in 1.5 mM MgCl(2). Two nested PCR amplifications were performed on the previously amplified products with primers BM-5`-2, BM-5`-3 (29) and with a TPH-specific primer, O(2), which corresponds to a sequence 80 nucleotides upstream of the extension oligonucleotide (SBPE1). PCR products were analyzed on 3% agarose gels, blotted onto nylon membranes, and hybridized with a primer complementary to a sequence of the first coding exon (SBCRIB). The bands hybridizing with SBCRIB were eluted from the gel, subcloned into M13 mp8/SmaI and sequenced on both strands. The sequences of the primers were as follows: SBPE1, 5`-TCTTCTTTTTGATTTTCGGGAC; SBSLIC, 5`-TTTTCGGGACTCGATATGTAACAGATTC; SBCRIB, 5`-TTTTCGGGACTCGATATGTAACAGATTC.

RESULTS

Multiple Human TPH mRNAs

Approximately 4 times 10^5 bacteriophages from a human carcinoid tumor cDNA library were screened at low stringency with a rat TPH cDNA probe. Twenty positive clones were isolated and analyzed by restriction mapping, Southern blotting, and sequencing. The coding regions of these clones and most of their 3`-untranslated regions (UTR) were identical; they differed, however, by the length of their 5`-ends. The open reading frame of 1332 bp (previously published; (15) ) was very similar to that of the rat, rabbit, and mouse TPH cDNA. The 3`-noncoding sequence, including the poly(A) tail, was about 2 kb long and also exhibited a high degree of identity with part of the corresponding rat and rabbit sequences. However, the size and sequence of the human TPH 5`-UTR were different from those of the rodent and were found similar only to the 50 bases immediately upstream of the AUG codon of rabbit TPH cDNA.

The human TPH cDNA clones fell into two categories, which differed by the organization of their 5`-leader sequences and their abundance. The majority of TPH cDNA clones, referred to hereafter as type 1 cDNAs, all had long 5`-UTR of variable size and probably corresponded to incompletely reverse-transcribed mRNAs (the longest type 1 cDNA was 6 kb long with a 5`-UTR of 2.5 kb). In contrast, the type 2 cDNA clones were about 3.6 kb long with a short 5`-UTR (the longest type 2 cDNA has a 5`-UTR of 310 bp). They differed from the type 1 cDNA clones by the absence of a 1.7-kb sequence (named I) 26 nucleotides upstream of the translation initiation codon (Fig. 1A). The 5`- and 3`-extremities of the I region showed intron splice site sequences, suggesting that I may be recognized as an intron in some TPH transcripts. Therefore, the two types of cDNA clones may result from alternative splicing in the 5`-UTR; or possibly the type 1 cDNA clones are TPH precursor RNA, whereas the type 2 is the mature transcript.

Figure 1: Schematic representation of the various 5`-cDNA clones for human TPH and organization of TPH 5`-UTR on genomic DNA. A, two types (1 and 2) of TPH cDNA clones were obtained by cDNA cloning. The two longest species of these different cDNAs were 6 kb and 3.5 kb, respectively. Open and shaded boxes indicate the 5`-noncoding and coding exons, respectively. Thick horizontal lines represent the 3`-UTR and the intron-like sequences in the 5`-leader sequence. The broken lines indicate the elimination of the intron-like region. B, two 5`-noncoding extremities of TPH cDNA clones were isolated by anchored PCR: Slic type 1 and Slic type 3. C, restriction map and organization of the TPH 5`-noncoding region in genomic DNA. The arrow shows the position of the transcription initiation site. Some of the restriction sites present in the TPH gene are shown.


To determine the size and the number of human TPH mRNAs, Northern blots were performed in denaturing conditions with RNA extracted from various tissues that do or do not produce TPH. No hybridization signal was detected with RNA purified from liver and dorsal raphe nuclei. In contrast, the cDNA probe labeled four major transcripts with apparent sizes of 5, 6, 7.5, and 9 kb in mRNA from pineal gland and carcinoid tumors (Fig. 2A). These results agreed with the known distribution of the enzyme in tissues, with the notable exception of the raphe nuclei area of the brainstem, which, as found in the rat, did not show any hybridization signal(20) . The 5-kb species appeared to be the most abundant and was the only signal detected in RNA prepared from the intestine. The high molecular weight TPH mRNA forms were more abundant in carcinoid tumor than in pineal gland RNA.

Figure 2: Northern blot analysis of TPH mRNA expression. A, tissue distribution of TPH transcripts. Two µg of poly(A) RNA from carcinoid tumor (Carcinoid T.), pineal gland (Pineal and Pineal (5d)), and 20 µg of total RNA from raphe nuclei, colon, and liver were subjected to gel electrophoresis, blotted onto a nylon filter, and hybridized to P-labeled TPH probe. Pineal and pineal are identical samples with different exposure times (pineal, 18-h exposure; pineal, 5-day exposure). The arrows show the TPH mRNAs and indicate their molecular weights. B, Northern blot analysis of TPH transcripts with intron and exon probes. Carcinoid tumor (right panel) and pineal gland RNAs (left panel) were hybridized with a 1-kb-long TPH coding sequence (lane 1), I intron-like region (lane 2), I intron-like region (lane 3), or I intron-like region probes (lane 4).


To characterize better the diversity of TPH mRNA, the whole coding sequence and the 3`-UTR were analyzed by PCR amplification from pineal gland and carcinoid tumor cDNAs. A single fragment was detected for each of the four overlapping subregions (defined by the primers) spanning these domains, in agreement with the length of the cloned TPH cDNA sequences (data not shown). In this respect, human TPH mRNA differs from rat and mouse TPH mRNAs, which possess two different 3`-untranslated regions generated by alternative polyadenylation sites. Thus, the diversity of TPH mRNAs may arise from RNA splicing in the 5`-noncoding region. The size of the cDNA clones cannot easily be reconciled with that of the TPH mRNAs detected on Northern blots. This could be because of the cDNA clones being incomplete at the 5`-end. We therefore cloned the entire human TPH gene to facilitate the analysis of the diversity of its mRNAs.

Cloning and Mapping of the Human TPH Gene and Phylogenetic Analysis

A human genomic DNA library in EMBL3 was screened with the TPH cDNA probe, and six independent positive clones were isolated. The restriction maps of the clones were determined, and three clones that span the entire coding region of the TPH gene were selected for further analysis (Fig. 3A). The coding region, parts of the introns, and the 5`- and 3`-noncoding flanking regions were sequenced. The individual coding exons were mapped, and the sequences of each exon-intron boundary were determined with TPH-specific oligonucleotides (Fig. 3B). TPH mRNA is transcribed from a minimum of 11 exons, contained within a region of genomic DNA of 29 kb. The intron locations in the three paralogous hydroxylase genes are strongly conserved (Fig. 4A). Interestingly, the number, position, and size of introns in the coding region are conserved between the human TPH gene and the mouse TPH gene whose organization has been previously reported(17) . Moreover, the nucleotide sequences in the vicinity of the splice sites of the human introns are similar to the corresponding regions in the mouse. The exon sizes are between 63 and 197 nucleotides, and all of the donor and acceptor splicing sites conform to the consensus sequences for eukaryotic genes(30) . The exon that spans the 3`-part of the open reading frame and the whole 3`-untranslated sequence is about 2 kb long and contains 172 coding nucleotides. The length of the 3`-UTR is conserved among the human, rat, and mouse TPH genes(16, 18) , and their sequences are very similar for the first 200 nucleotides downstream of the stop codon. The sequence at the 3`-end of the human gene contained a weak polyadenylation signal, AAUAGA, as compared with the canonical AAUAAA polyadenylation signal(31) .

Figure 3: Structural organization and exon-intron junctions of the human tryptophan hydroxylase gene. A, restriction map of the TPH gene. Filled and open boxes indicate the noncoding and coding exons, respectively. Open and filled circles represent HindIII and EcoRI sites, respectively. Three different phage clones (12, 13, 15) cover the whole TPH gene. B, exon-intron structure of the human TPH gene. The exon sequences are denoted in uppercase letters; intron sequences are in lowercase and are given for each junction. The amino acids and their corresponding positions on the cDNA sequence are also shown underneath each exon-intron boundary.


Figure 4: Conservation of intron-exon junction positions between the human TPH, tyrosine hydroxylase (TH), and phenylalanine hydroxylase (PAH) genes. A, the three genes are aligned to maximize amino acid identity. Boxes and thick lines indicate the exons and the introns, respectively. Numbers in the boxes indicate the percentages of identity in the exons between the hydroxylase genes. The colors white, gray, and black provide a visual representation of the degree of identity: 0-35%, 36-59%, and 60-100%, respectively. B, phylogenetic distance tree based on the protein sequences of the aromatic amino acid hydroxylases. After sequence alignment, the distance tree was constructed by the Neighbor Joining Method (24) (left; (24) ) and bootstrap analysis (right) as described under ``Materials and Methods.'' The tree was arbitrarily rooted on the tyrosine hydroxylase Drosophila sequence. Note the heterogeneity of the molecular clock between the three subfamilies of hydroxylases, which does not allow the sequence duplications to be dated with confidence.


A phylogenetic distance analysis was performed after having aligned all of the available sequences of the three aromatic amino acid hydroxylases (AAAH) isolated from the different animal species. The topologies of the trees obtained from the nucleotide and the amino acid sequences were identical. Moreover, tree branching was also unchanged by using the whole sequences or only the C-terminal two-thirds of the molecules where the sequences aligned without any gap or deletion. Thus, all the parts of the sequences have presumably evolved at roughly the same relative rates. Each of the three vertebrate AAAH, i.e. tyrosine hydroxylase, phenylalanine hydroxylase, and tryptophan hydroxylase, clearly constitutes a monophyletic group, a contention unambiguously supported by the bootstrap analysis (Fig. 4B). The phylogeny of animal species is correctly reproduced by the AAAH sequences, but the rate of sequence divergence varies significantly among the hydroxylases (note branch lengths for human, rat, and mouse in each of the three groups; Fig. 4B). Thus, there is no satisfactory molecular clock to date with confidence the divergence of the three AAAH. Similarly, it is not possible to determine which of the three AAAH diverged first, although the absence of a bona fide TPH in Drosophila(32) suggests an early divergence of tyrosine hydroxylase before that of TPH and phenylalanine hydroxylase from a common ancestral gene.

One of the genomic clones, 12, which contained both the 5`-noncoding region of TPH cDNA and upstream sequences, was analyzed by detailed restriction mapping (Fig. 1C). The sequence of the region upstream of the translation initiation site on the genomic DNA was identical to the entire 5`-UTR of the type 1 cDNA clones. It was therefore clear that the domain I characterizing the type 1 cDNAs corresponded to an unspliced, intron-like sequence.

Cloning of the 5`-UTR of the Human TPH mRNA by Anchored PCR

The TPH mRNA diversity revealed by cDNA cloning and Northern blotting was likely to be confined to the 5`-UTR. We therefore used the anchored PCR technique (SLIC, (29) ) to isolate other 5`-leader sequences for the TPH mRNA. Several 5`-directed cDNAs were synthesized and cloned from human pineal gland and carcinoid tumor, both tissues that strongly express the TPH gene. Only two classes of transcripts were obtained, as shown by restriction mapping and sequence analysis (Fig. 1B). In both tissues, most of the isolated clones were about 0.3 kb long and contained the 3`-end of the I region. They corresponded to the type 1 TPH cDNA clone. This result agreed with the cDNA library screening, where the majority of cDNA clones isolated carried the I sequence. The second type of SLIC clone was about 0.35 kb long and characterized by the absence of the I region, as were the type 2 cDNA clones. However, it differed from type 2 cDNA 5`-UTR in that the first 224 bases were replaced by a previously unidentified sequence of 210 bases (Fig. 1B). This clone was called type 3 SLIC cDNA. The comparison of the type 3 SLIC clone sequence to that of the genomic DNA showed that the 210-nucleotide stretch is located 2.4 kb upstream of the identified 5`-UTR sequence. Therefore, this 2.4-kb region could be recognized as an intron, which when spliced together with the I intron-like region, generates an additional exon of 63 nucleotides in the 5`-leader sequence (named exon Ec; Fig. 1C). The nucleotide sequence, upstream of exon E(c), found in type 1 and type 2 cDNA also corresponded to the nonspliced 3`-end of this 2.4-kb intervening sequence.

The analysis of the 5`-ends of TPH mRNA obtained by the anchored PCR technique confirmed the predominance of the TPH mRNA containing the intron-like sequence of 1.7 kb (I) and also showed the existence of a new 5`-noncoding extremity. In short, cloning TPH cDNAs revealed a diversity in the 5`-UTR of the human TPH mRNAs resulting from the conservation or the elimination of two intron-like regions, I and I.

Determination of the mRNA Cap Site by S1 Nuclease Protection

We mapped the transcriptional initiation site of the TPH gene as follows. The nucleotide sequence of the genomic DNA encompassing the 5`-end of the type 3 SLIC clone and extending 1.5 kb upstream was completely analyzed by S1 nuclease protection assays (Fig. 5). The labeled fragment of 0.57 kb (probe D) spanning the 5`-end of the type 3 SLIC clone was fully protected in pineal gland and carcinoid tumor mRNAs, indicating that the TPH mRNA extended upstream of the probe D (Fig. 5D). This result, also confirmed by PCR analysis (data not shown), showed that the cloning techniques we used did not allow us to reach the transcription start site. A more extensive S1-mapping study was undertaken with three TPH genomic probes (probes A, B, and C) upstream of the 5`-end of the type 3 SLIC clone and covering 1.05 kb of genomic DNA. The labeled fragments B and C were completely protected from S1 nuclease by RNAs extracted from both tissues. In contrast, hybridization of the RNAs with probe A yielded a major protected fragment (126 bases long) much smaller than the probe, suggesting divergence between the genomic and the mRNA sequences (Fig. 5A). The genomic sequence did not contain any acceptor splice site that coincided with the S1 nuclease cleavage. Moreover, numerous putative cis-acting DNA elements, commonly found in eukaryote gene promoters (TATA box, inverted CCAAT box, CCACCC box, GC-rich sequences, AP-2 and AP-4 binding sites), were present immediately upstream of the S1 nuclease cleavage site. This site, therefore, probably corresponded to the TPH transcription initiation site. Primer extension, using an oligonucleotide hybridizing to the 3`-end of probe A, yielded a single extension product consistent with transcription beginning at the residue predicted by S1 nuclease protection (data not shown). Thus, the position of TPH RNA start site in genomic DNA is located 5.5 kb upstream of the ATG codon. In addition, S1 nuclease protection analysis revealed that the region of 1.4 kb (named I), localized between the cap site and the 5`-end of the type 3 SLIC clone, is present at the 5`-extremity of many of the human TPH mRNAs.

Figure 5: Determination of the transcription initiation site by S1 nuclease protection. Poly(A) RNA (0.25 µg) from pineal (Pi); carcinoid tumor (CT), and liver (L) was subjected to S1 nuclease protection analysis. Pr-S1, probe without S1 nuclease; Pr+S1, probe with S1 nuclease. The arrows show the protected fragments and their corresponding sizes. A, probe A was a 0.42-kb SmaI-PstI fragment. B, probe B was a 0.347-kb HindIII-SmaI fragment. C, probe C was a 0.287-kb EcoRI-HindIII fragment. D, probe D was a 0.574-kb KpnI-EcoRI fragment. The bottom panel represents the position of probes A, B, C, and D in genomic DNA (horizontal arrows). The vertical arrows indicate the position of restriction enzyme sites, and the broken arrow indicates the transcription initiation site. Open boxes indicate the 5`-noncoding exons.


The S1-mapping experiment using probe A, which spans the cap site, also revealed an additional protected fragment of 29 bases in pineal gland RNA (Fig. 5A). The presence on the genomic sequence of a 5`-splice donor site 29 nucleotides downstream of the transcription initiation site suggested that this fragment could correspond to a small exon located at the cap site. This was confirmed by PCR experiments with a primer bordering the cap site and another primer localized inside the 5`-end of the type 3 SLIC clone. Sequence analysis of the PCR fragment confirmed the existence of the 29-base exon (called exon E(1)) and showed the splicing of the 1.4-kb region (data not shown). In this fragment, exon E(1) was joined to a 179-base sequence (named exon E(b)) within the 210-bp stretch of the type 3 SLIC clone (Fig. 1C). Therefore, type 3 SLIC clone contained a part of the region I, which, like the I and I regions, could be recognized as intron but is present in most TPH mRNA extracted from the pineal gland and carcinoid tumors (see above).

Characterization of a Novel TPH mRNA Species with Short 5`-UTR by RT-PCR Analysis

To help decipher the complete organization of the 5`-untranslated region of TPH mRNAs, both pineal gland and carcinoid tumor mRNAs were exhaustively analyzed by RT-PCR. Two forward primer sequences were selected downstream of the transcription initiation site. One was specific for the 29-bp exon E(1) (O(1)) and was chosen to reveal the possible presence of short TPH mRNA 5`-UTR, which would result from the splicing of the three intron-like regions (I, I, and I). The second was specific for the I sequence (O) and was used to confirm the presence of long 5`-extremities in TPH mRNAs. The sequence of the reverse primer (O(2)) corresponds to a region downstream of the translation initiation site. Several very short PCR products were obtained with each of the two sets of primers, demonstrating the existence of short 5`-UTR in TPH mRNA in the two studied tissues. Surprisingly, no long PCR product was amplified. However, when fragments of very different sizes are simultaneously transcribed by RT-PCR, the amplification of the shortest transcripts is strongly favored over the longer ones, whatever the initial abundance of their respective mRNAs(33, 34) . Three fragments (103, 166, and 282 bp) were amplified by PCR with primers O(1)-O(2) (Fig. 6). Their sequences were determined and showed that the 103-bp band corresponded to exon E(1) joined to the first coding exon E(2) (containing 26 noncoding nucleotides at its 5`-end). The 166- and 282-bp products resulted from the insertion between the exon E(1) and the exon E(2) of the exon E(c) and the exon E(b), respectively (Fig. 6A). Only two fragments (187 and 250 bp) were generated with primers O-O(2) and corresponded to an extension of 100 bp from the first 29-nucleotide exon E(1) linked to either exon E(2) or exon E(c) and exon E(2) (Fig. 6B). Thus, there appeared to be two 5`-splice donor sites in the first noncoding exon, both of which are used in the pineal gland and carcinoid tumor (the resulting fragments were named E(1) and E). An additional 5`-untranslated extremity containing all of the exons E(1), E(b), E(c), and E(2) joined together was also identified but only by using a set of primers covering the junction E(1)-E(b) and the junction E(2)-E(c), suggesting that this transcript is of very low abundance (Fig. 6C). Thus, in TPH mRNAs containing exon E(1), all combinations of exons were found upstream of exon E(2). However, when the exon E was present, exon E(b) was never found between E and E(2). This suggests that there are a variety of patterns of splicing, which all generate TPH mRNAs of about 3.5 kb, with short 5`-UTR.

Figure 6: PCR analysis of the 5`-UTR of human TPH mRNA. PCR was performed with specific primers after first-strand cDNA synthesis using poly(A) mRNA isolated from carcinoid tumor (CT) and pineal gland (Pi). Open and hatched boxes represent the noncoding and coding exons, respectively. The organization of the 5`-noncoding extremity for each PCR fragment is schematized. A, radiolabeled PCR with specific TPH primers (O(1) and O(2)). B, radiolabeled PCR with specific TPH primers (O and O(2)). The products of the two sets of primers were separated on a 5% denaturing polyacrylamide gel and visualized by autoradiography. C, PCR with primers overlapping exons E(1),E(b) and E(c),E(2). The products of PCR were separated on a 3% agarose gel, transferred to a membrane, and hybridized with a labeled oligonucleotide (SBCRIB).


Determination of the Abundance of the TPH mRNA with Short 5`-UTR as Compared with That of Long 5`-UTR TPH mRNAs

To study the relative abundance of the various 5`-UTR of the TPH mRNAs in pineal gland and carcinoid tumor, additional S1 nuclease protection experiments were performed. TPH probes hybridizing with sequences in either the introns or overlapping noncoding and coding exon-intron junctions were used ( Fig. 5and Fig. 7). Part of the coding region was analyzed with the cDNA probe H, which overlaps three exons (E(6), E(7), and E(8)). It was entirely protected by both pineal gland and carcinoid tumor mRNAs, indicating that intervening introns were eliminated in this domain (Fig. 7D). In contrast, probes C, E, and F located in the intron-like regions (I I, and I) of the 5`-UTR, each were fully protected by the same RNA (Fig. 5C and Fig. 7, A and B). Therefore, the three intron-like regions in the 5`-UTR (I, I, and I) were conserved in most of the TPH mRNAs when the introns interrupting the coding region were usually eliminated. The splicing of the intron-exon junctions E(1)-I and I-E(2) was investigated with probes A and G, respectively. Probe G yielded two fragments, which corresponded to the spliced and to the nonspliced intron-like sequence I, thereby showing the presence of two classes of transcripts in which the 1.7-kb region was and was not present (Fig. 7C). The TPH mRNA where unspliced intron-like regions are maintained in the 5`-UTR are much more abundant in carcinoid tumor than in pineal gland. Similar results were obtained with probe A, which covered the junction of the exon E(1) and the intron-like region I (Fig. 5A). No signal corresponding to the transcript from which intron-like region I had been eliminated was detected in the tumoral tissue. Therefore, the relative abundance of the spliced transcript was higher in pineal gland than in carcinoid tumor RNAs.

Figure 7: S1 nuclease analysis of the 5`-UTR exons (C), 5`-intron-like regions (A, B), and coding region (D) of TPH mRNA. Poly(A) RNA (0.25 µg) from pineal (Pi); carcinoid tumor (CT), and liver (L) was subjected to S1 nuclease protection. Pr-S1, probe without S1 nuclease; Pr+S1, probe with S1 nuclease. A and B, S1 nuclease mapping with the E and F probes containing 290 and 362 nucleotides from the intron-like regions. C, probe G was 250 nucleotides long and contains a part of the exon E(2) (117 nucleotides). D, probe H was 330 nucleotides long and covered three coding exons. The arrows show the protected fragments for each probe used. The bottom panel shows the positions of probes (E, F, G, H) in genomic DNA (horizontal arrows). Open and shaded boxes indicate the 5`-noncoding and coding exons, respectively.


The low abundance of the spliced TPH mRNA 5`-UTR also was confirmed with two other probes (I and J). Probe I corresponds to 187 bases of the 5`-noncoding extremity of TPH mRNA and contains the first noncoding exon linked to the exon E(2) (Fig. 8A). The entire probe was fully protected by the pineal gland RNA, showing that short mRNA 5`-extremities are present in normal tissue. Only a very weak protection was obtained in the carcinoid tumor, where splicing of intron-like regions in the TPH 5`-UTR is rare. In contrast, probe J (349 bases), complementary to the sequence of the type 3 SLIC clone, was poorly protected in the two tissues (Fig. 8B). Therefore, the type 3 SLIC clone isolated by anchored PCR is a minor form of TPH mRNA.

Figure 8: S1 nuclease analysis of the TPH 5`-UTR region. Poly(A) RNA (0.25 µg) from pineal gland (Pi), carcinoid tumor (CT), and liver (L) were subjected to S1 nuclease protection. Pr-S1, probe without S1 nuclease; Pr+S1, probe with S1 nuclease. A, S1 nuclease mapping with probe I of PCR fragment (187 bases) corresponding to one of the TPH 5`-noncoding extremities. B, probe J is 349 bases long and corresponds to the type 3 SLIC clone. The arrows show the fragments protected by each probe. The schematic organization of each protected fragment is represented. Open boxes indicate the 5`-noncoding exons and horizontal lines indicate the intron-like region.


Finally, to determine which of the three intron-like regions was retained in the 5`-UTR of the major species of TPH transcripts, probes specific to these three regions were generated by PCR amplification and hybridized to different Northern blots (Fig. 2B). One intron, the I, contained three antisense Alu sequences separated by up to a few hundred bases of non-Alu DNA. Thus, we chose an intronic probe in a region located outside of the repetitive elements. However, no simple conclusion can be drawn from these hybridizations. In both carcinoid tumor and pineal gland, the major 5-kb band was recognized by each of three probes, suggesting that it may correspond to several species of TPH mRNA with different 5`-UTR but similar size. In practice, it was very difficult to quantify the relative abundance of the various other transcripts labeled by the intronic probes from the RNA material available. Therefore, although each of the mRNA bands detected on Northern blot may correspond to a different and complex exon-intron arrangement in the 5`-UTR, it was impossible to unravel the organization of these TPH transcripts by simple hybridizations to the blots. Nevertheless, Northern analysis supported the main conclusions of the extensive and more accurate nuclease protection experiments, which were that the TPH transcripts having eliminated the three intron-like regions in the 5`-UTR generally represented a minor population as compared with the partially spliced TPH mRNAs, a situation that is much more pronounced in carcinoid tumor than in pineal gland.

DISCUSSION

The study of the human TPH mRNA by cDNA cloning, anchored PCR amplification, and S1 nuclease protection assays revealed a very unusual organization of its 5`-UTR. Human TPH mRNA exhibited a large diversity in the 5`-leader sequence, whereas the coding region was identical in all of the tissues studied. Four TPH transcripts were visualized by Northern blotting of both pineal gland and carcinoid tumor RNA. To unravel this complex organization, the corresponding gene was isolated and mapped. The sequence and the locations of the intron-exon junction of the human TPH gene revealed very strong similarity to those of the genes encoding other aromatic amino acid hydroxylases (AAAH).

The human TPH locus spans 29 kb and contains at least 11 exons, and its mRNA appears to undergo differential splicing in the 5`-UTR. The locations of intron-exon junctions of the mammalian AAAH genes are very similar, particularly in the region corresponding to the catalytic core of the enzyme (see Fig. 4). The only exceptions are one intron specific to tyrosine hydroxylase (I-6)), one specific to phenylalanine hydroxylase (I-12), and one common to tyrosine hydroxylase and phenylalanine hydroxylase, which is absent from TPH and could therefore have been lost during the course of evolution. The 5`-extremity of the fourth exon is 4 amino acids longer in the TPH gene than in the corresponding exon of the other hydroxylases. The N-terminal region is less well conserved and encoded by a number of exons that varies from one enzyme to another, and even among mammalian species, as in the case of tyrosine hydroxylase(7) . Interestingly, the junction between the regulatory and catalytic domains of the proteins corresponds to an intron-exon junction in all the genes of the family. Among the hydroxylase genes, the mouse and human TPH genes alone have introns in the 5`-UTR (one and three introns, respectively). It is also very likely that an intron is present in the 5`-UTR of the rabbit TPH mRNA. Indeed, the 5`-leader sequence of the rabbit TPH cDNA shares 74.4% identity and 85.7% identity, respectively, with the human TPH exons E and E(2), which border the 5`-UTR and are separated by introns (Fig. 9). However, this suggestion awaits experimental confirmation. In addition, the good conservation of the TPH 5`-UTR sequence between human and rabbit is not found in the other known mammalian species, indicating major sequence shuffling in this region.

Figure 9: Comparison of human and rabbit TPH gene 5`-UTR sequences. A, nucleotide sequences of human and rabbit TPH are numbered with respect to the transcription start site and to the first nucleotide of the cDNA, respectively. =, nucleotides conserved between the two species. B, schematic representation and alignment of the rabbit TPH 5`-UTR with the exons (E(1) and E(2)) and part of the I intron of human TPH 5`-UTR. Hatched boxes represent the human or rabbit exons of the TPH gene 5`-UTR. Thick lines indicate the introns of the TPH gene 5`-UTR.


Phylogenetic distance analysis, using either the amino acid or the nucleotide sequences of this gene family, only partly supports the conclusions drawn by Woo and colleagues (12, 35) about the evolution of the AAAH genes. These authors proposed that two major gene duplications have occurred; the first one separated tyrosine hydroxylase from the common ancestor, and the second one gave birth to phenylalanine hydroxylase and TPH. However, the uncertainty about the regularity of the molecular clock in this protein family (Fig. 4B) and the small number of sequences available from animal species belonging to different phyla do not allow the duplications to be dated with confidence. Nevertheless, it was recently proposed that Drosophila melanogaster possesses only two aromatic amino acid hydroxylase genes, one being tyrosine hydroxylase-homologous and the other having both phenylalanine hydroxylase and TPH activities (32) . If one could rule out the possibility that one of the hydroxylases was eliminated as redundant in the Drosophila phylum, it should be proposed that the first duplication occurred presumably before and the second one after the divergence of arthropods from the other taxa (600 million years ago).

Human TPH mRNAs are characterized by large diversity within, and restricted to, the 5`-noncoding region. This diversity results from the conservation of one or more intron-like regions in the 5`-leader sequence of the TPH mRNA and from differential splicing of three exons in the spliced TPH mRNAs 5`-UTR. Generally, mRNAs that display a 5`-UTR diversity, for example the genes encoding mouse choline acetyltransferase(36) , human insulin-like growth factor II(37) , and aldolase-A gene(38) , are transcribed from alternative promoters, followed by the splicing of intervening sequences in the 5`-UTR. An extreme example is the hydroxymethylglutaryl-CoA reductase gene, where a more complex mechanism involves the combination of multiple transcription initiation sites and various 5`-splice donor sites for one intron(39) . This diversity within the mRNA 5`-leader sequences is therefore associated with the use of alternative promoters, which could be preferentially activated in particular tissues or stages of development(37, 40, 41) . In the case of the human TPH gene, however, the multiple mRNA species are transcribed from a single promoter, and the variety of TPH messengers is the result only of the differential splicing of three intron-like regions and of the three exons located in the 5`-UTR.

It is surprising that the three intron-like sequences in the 5`-UTR of TPH mRNAs are in many cases retained when the introns of the coding region are eliminated. Northern blotting clearly identified high molecular weight TPH transcripts, which may result from differential splicing, generating unusually long 5`-noncoding sequences. These transcripts are more abundant than would be expected of processing intermediates. This led to their isolation directly from the screening of the cDNA library, a rather uncommon event. PCR experiments only allowed the cloning and the characterization of several rare, differentially spliced TPH mRNA species where the three intron-like regions are eliminated. Generally, mRNAs with long 5`-leader sequences correspond to precursors. In this latter case, the abundance of long 5`-UTR in TPH mRNAs should imply that the excision of three intron-like regions is suffi ciently slow and that these messengers are sufficiently stable to allow their accumulation. The limiting step of the TPH mRNA processing could be the splicing of the 5`-leader sequence rather than nuclear RNA degradation. Therefore, there appear to be two steps in the processing of human TPH mRNA. The first is rapid, eliminating the introns of the coding region. The second is slower, leading to a complex pattern of 5`-UTR maturation.

The mRNAs not containing region I (such as the type 2 TPH cDNA), are characterized by the presence of a supplementary in-phase AUG codon, 27 bases upstream of the presumed translation start site. The putative use of this initiator codon would generate a longer N-terminal sequence. Nevertheless, it remains to be determined whether or not this protein is produced and whether the two resulting proteins possess the same characteristics (i.e. stability or activity). The recent cloning of Xenopus laevis TPH cDNA has shown that it potentially encodes a TPH protein with 37 extra amino acids at the N terminus as compared with TPH in other species(42) . To date, this extension has no known functional consequences.

It is generally thought that the 5`-UTR contributes to the stability (43) and to the regulation of translation of the messengers(44, 45) . However, the presence of long 5`-noncoding sequences in TPH mRNAs poses many problems with regard to translation mechanisms. The initiation of translation in higher eukaryotes is modulated by several structural features in the 5`-untranslated region of mRNA. They include the m7G cap, the position of the AUG codon, the length of the leader sequence, and secondary structures(46) . mRNAs with long 5`-UTR could correspond to precursors or to otherwise nonfunctional transcripts. Indeed, translation initiation optimally requires a short 5`-noncoding region and no AUG codon upstream of that used to initiate translation(47) . The introns that are retained in the 5`-leader sequence considerably impair translation efficiency because they often contain an AUG-burdened leader sequence. Several AUG codons upstream from the translation initiator AUG are present in the type 1 and type 2 human TPH cDNA clones. All of these upstream AUG codons are followed by short open reading frames that could potentially encode peptides. According to the scanning model of translation, these AUG-burdened RNA sequences corresponding to the high molecular weight TPH transcripts are expected to be poorly translated, a characteristic that could be compensated for by the abundance of these mRNAs. In this case, these mRNAs would be translated without additional maturation. In addition, there have been several reports of abundant, incompletely spliced transcripts that enter the cytoplasm (48) and also which have been found on polysomes (49) . This observation suggests that the introns, when they are maintained in the transcripts, could play a role in the regulation of gene expression.

In contrast, it has recently been shown that precursor RNAs can be synthesized and stored for later processing(50) . In this model, the large TPH RNAs would be precursors, to be translated only after a maturation step, a mechanism that easily accounts for the abundance of TPH mRNAs bearing long 5`-UTR relative to those short 5`-UTR. Only the TPH mRNAs with a short 5`-leader sequence would be effectively translated, and the low abundance of these transcripts could result from their rapid degradation. The conversion of a stable, untranslatable precursor to a functional mRNA generates a supplementary step in the regulation of gene expression.

Finally, internal translation intiation as described for some viral and eukaryotic genes (51, 52, 53) could also explain the abundance of long TPH mRNAs. This translation initiation mechanism allows messengers with long 5`-leader sequences to be efficiently translated. Each of these three models could account for the large difference between the amounts of human TPH mRNAs with long and short 5`-UTR, and the intracellular localization of the high molecular weight TPH mRNAs may indicate whether or not they can be translated. In any case, the diversity exhibited by the 5`-UTR of the human TPH mRNAs may play a physiological role in the production of TPH enzyme. It increases the possibility of modulation of TPH gene expression at post-transcriptional and translational levels.

Another particularity of TPH mRNA expression is the discrepancy between the tissues in which the TPH enzyme and mRNA is found. Northern blot analysis detected TPH transcripts in the pineal gland, intestine, and carcinoid tumor but not in the brainstem raphe nuclei, which nevertheless contain TPH. There have been similar observations in rat, rabbit, and mouse(13, 16, 18) . The discrepancy between TPH mRNA and protein levels in the brainstem could be explained by (i) the existence of another TPH gene expressed specifically in the raphe nuclei, (ii) better translation efficiency of very small amounts of TPH mRNA, or (iii) enhanced stability of the TPH protein. Measurements of the TPH gene transcription rate have shown that the level of gene expression was similar in the pineal gland and in the brainstem, suggesting post-transcriptional or translational regulation of the TPH mRNA(21) . It is possible that no high molecular weight TPH mRNAs are transcribed in human raphe nuclei brainstem and that only TPH transcripts with spliced 5`-UTR are synthesized. These short mRNAs may be efficiently translated and then rapidly degraded. In the carcinoid tumors and pineal gland, large amounts of TPH mRNAs are produced. Surprisingly, no short 5`-leader sequences of TPH messengers are detected in the carcinoid tumors by S1 nuclease protection, although they are in the pineal gland. The abundance of these high molecular weight TPH mRNAs in carcinoid tumors could reflect a high transcription rate or RNA stability peculiar to the mitotic character of this tissue. Although these tumors synthesize and secrete very high levels of serotonin, it is not known if the pathological cells produce more active TPH than healthy enterochromaffin cells.

In conclusion, the cells expressing the TPH gene contain a large and complex variety of TPH mRNA forms differing in the 5`-UTR. Although the functional consequences of this phenomenon are only beginning to be investigated, it provides interesting clues to novel mechanisms of regulation of gene expression. An important aspect of TPH expression in the pineal gland is its rhythmicity. In rat, TPH activity and the mRNA levels have been shown to vary during the circadian rhythm(42, 54) . (^2)This type of variation could imply an integrated regulation of TPH gene expression. An attractive hypothesis is that it evolves from differential splicing events leading to mRNAs, which differ only by their 5`-leader sequence.

FOOTNOTES

*
This work was supported by grants from the Centre National de la Recherche Scientifique, the Institut National de la Santé et de la Recherche Medicale, the Association pour la Recherche contre le Cancer, the Institut de Recherche sur la Moëlle Epiniere and Rhône-Poulenc-Rorer. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked ``advertisement'' in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBank(TM)/EMBL Data Bank with accession number(s) X83213[GenBank].

§
To whom correspondence should be addressed: Laboratoire de Génétique Moléculaire, de la Neurotransmission, et des Processus Neurodégénératifs, C.N.R.S., F91198 Gif-sur-Yvette Cedex, France. Fax: 33-1-69-82-35-80.

Supported by fellowships from the Institut de Recherche sur la Moëlle Epiniere.

(^1)
The abbreviations used are: TPH, tryptophan hydroxylase; kb, kilobase(s); bp, base pair(s); PCR, polymerase chain reaction; RT-PCR, reverse transcriptase PCR.

(^2)
T. L. Green and R. P. Hart, personal communication.

ACKNOWLEDGEMENTS

We thank E. Jean-Gilles, I. Brunet, and D. Samolyk for technical assistance; P. Ravassard, N. Faucon Biguet, and J. F. Julien for critical comments; and P. Vernier for the phylogenetic analysis and critical reading of this manuscript.

REFERENCES

  1. Jacobs, B. L., and Azmitia, E. C. (1992) Physiol. Rev. 72, 165-215 [Free Full Text]
  2. Wurtman, R. J., Axelrod, J., and Chu, E. W. (1963) Science 141, 277-278 [Abstract/Free Full Text]
  3. Cardinali, D. P. (1981) Endocr. Rev. 2, 327-346 [Medline] [Order article via Infotrieve]
  4. Feldman, J. M., and O'Dorisio, T. M. (1986) Am. J. Med. 81, 41-48 [CrossRef][Medline] [Order article via Infotrieve]
  5. Feldman, J. M. (1987) Semin. Oncol. 14, 237-246 [Medline] [Order article via Infotrieve]
  6. Kuhn, D. M., Ruskin, B., and Lovenberg, W. (1980) J. Biol. Chem. 255, 4137-4143 [Abstract/Free Full Text]
  7. Grima, B., Lamouroux, A., Boni, C., Julien, J. F., Javoy-Agid, F., and Mallet, J. (1987) Nature 326, 707-711 [CrossRef][Medline] [Order article via Infotrieve]
  8. Kwok, S. C. M., Ledley, F. D., DiLella, A. G., Robson, K. J. H. and Woo, S. L. C. (1985) Biochemistry. 24, 556-561 [CrossRef][Medline] [Order article via Infotrieve]
  9. Kaufman, S., and Fisher D. B. (1974) in Molecular Mechanisms of Oxygen Activation (Hayaishi, O., ed) pp. 285-369, Academic Press, New York
  10. Vigny, A., and Henry, J. P. (1982) Biochem. Biophys. Res. Commun. 106, 1-7 [CrossRef][Medline] [Order article via Infotrieve]
  11. Yang, X. J., and Kaufman, S. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 6659-6663 [Abstract/Free Full Text]
  12. Ledley, F. D., DiLella, A. G., Kwok, S. C. M., and Woo, S. L. C. (1985) Biochemistry 24, 3389-3394 [CrossRef][Medline] [Order article via Infotrieve]
  13. Grenett, H. E., Ledley, F. D., Reed, L. I., Woo, S. L. C. (1987) Proc. Natl. Acad. Sci. U. S. A. 84, 5530-5534 [Abstract/Free Full Text]
  14. Darmon, M. C., Grima, B., Cash, C. D., Maitre, M., and Mallet, J. (1986) FEBS Lett. 206, 43-46 [CrossRef][Medline] [Order article via Infotrieve]
  15. Boularand, S., Darmon, M. C., Ganem, Y., Launay, J. M., and Mallet, J. (1990) Nucleic Acids Res. 18, 4257 [Free Full Text]
  16. Stoll, J., Kozak, C. A., and Goldman, D. (1990) Genomics 7, 88-96 [CrossRef][Medline] [Order article via Infotrieve]
  17. Stoll, J., and Goldman, D. (1991) J. Neurosci. Res. 28, 457-465 [CrossRef][Medline] [Order article via Infotrieve]
  18. Darmon, M. C., Guibert, B., Leviel, V., Ehret, M., Maitre, M., and Mallet, J. (1988) J. Neurochem. 51, 312-316 [Medline] [Order article via Infotrieve]
  19. Delort, J., Dumas, J. B., Darmon, M. C., and Mallet, J. (1989) Nucleic Acids Res. 17, 6439-6448 [Abstract/Free Full Text]
  20. Dumas, S., Darmon, M. C., Delort, J., and Mallet, J. (1989) J. Neurosci. Res. 24, 537-547 [CrossRef][Medline] [Order article via Infotrieve]
  21. Hart, R. P., Yang, R., Riley, L. A., and Green, T. L. (1991) Mol. Cell. Neurosci. 2, 71-77 [CrossRef]
  22. Gubler, U., and Hoffman, B. J. (1983) Gene (Amst.) 25, 263-269 [CrossRef][Medline] [Order article via Infotrieve]
  23. Devereux, J., Haeberli, P., and Smithies, O. (1984) Nucleic Acids Res. 12, 387-395
  24. Saitou, N., and Nei, M. (1987) Mol. Biol. Evol. 4, 406-425 [Abstract]
  25. Philippe, H. (1993) Nucleic Acids Res. 21, 5264-5272 [Abstract/Free Full Text]
  26. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, Vol. 1, pp. 2.73-2.81 and 7.63-7.70, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY
  27. Civelli, O., Birnberg, N., and Herbert, E. (1982) J. Biol. Chem. 257, 6783-6787 [Abstract/Free Full Text]
  28. Faucon Biguet, N., Buda, M., Lamouroux, A., Samolyk, D., and Mallet, J. (1986) EMBO. J. 5, 287-291 [Medline] [Order article via Infotrieve]
  29. Dumas Milne Edwards, J. B., Delort, J., and Mallet, J. (1991) Nucleic Acids Res. 19, 5227-5232 [Abstract/Free Full Text]
  30. Breathnach, R., and Chambon, P. (1981) Annu. Rev. Biochem. 50, 349-383 [CrossRef][Medline] [Order article via Infotrieve]
  31. Birnstiel, M. L., Busslinger, M., and Strub, K. (1985) Cell 41, 349-359 [CrossRef][Medline] [Order article via Infotrieve]
  32. Neckameyer, W. S., and White, K. (1992) J. Biol. Chem. 267, 4199-4206 [Abstract/Free Full Text]
  33. Jeffreys, A. J., Wilson, V., Neumann, R., and Keyte, J. (1988) Nucleic Acids Res. 16, 10953-10971 [Abstract/Free Full Text]
  34. Delort, J., Dumas, J. B., Darmon, M. C., and Mallet, J. (1989) Nucleic Acids Res. 17, 6439-6448
  35. Ledley, F. D., Grenett, H. E., Bartos, D. P., van Tuinen, P., Ledbetter, D. H., and Woo, S. L. C. (1985) Somatic Cell. Mol. Genet. 13, 575-580 [CrossRef]
  36. Misawa, H., Ishii, K., and Deguchi, T. (1992) J. Biol. Chem. 267, 20392-20399 [Abstract/Free Full Text]
  37. Sussenbach, J. S. (1989) Prog. Growth Factor Res. 1, 33-48 [CrossRef][Medline] [Order article via Infotrieve]
  38. Izzo, P., Costanzo, P., Lupo, A., Rippa, E., Paolella, G., and Salvatore, F. (1988) Eur. J. Biochem. 174, 569-578 [Medline] [Order article via Infotrieve]
  39. Reynolds, G. A., Goldstein, J. L., and Brown, M. S. (1985) J. Biol. Chem. 260, 10369-10377 [Abstract/Free Full Text]
  40. Schibler, U., Hagenbuchle, O., Wellauer, P. K., and Pittet, A. C. (1983) Cell 33, 501-508 [CrossRef][Medline] [Order article via Infotrieve]
  41. Savakis, C., Ashburner, M., and Willis, J. H. (1986) Dev. Biol. 114, 194-207 [CrossRef]
  42. Green, C. B., and Besharse, J. C. (1994) J. Neurochem. 62, 2420-2428 [Medline] [Order article via Infotrieve]
  43. Brawerman, G. (1987) Cell 48, 5-6 [CrossRef][Medline] [Order article via Infotrieve]
  44. Nielsen, F. C., Gammeltoft, S., and Christiansen, J. (1990) J. Biol. Chem. 265, 13431-13434 [Abstract/Free Full Text]
  45. Kosak, M. (1986) Proc. Natl. Acad. Sci. U. S. A. 83, 2850-2854 [Abstract/Free Full Text]
  46. Kosak, M. (1989) J. Cell Biol. 108, 229-241 [Abstract/Free Full Text]
  47. Kosak, M. (1991) J. Cell Biol. 115, 887-903 [Abstract/Free Full Text]
  48. Rorsman, F., Bywater, M., Knott, T. J., Scott, J., and Betsholtz, C. (1988) Mol. Cell. Biol. 8, 571-577 [Abstract/Free Full Text]
  49. Weil, D., Brosset, S., and Dautry, F. (1990) Mol. Cell. Biol. 10, 5865-5875 [Abstract/Free Full Text]
  50. Xie, W., Chipman, J. G., Robertson, D. L., Erikson, R. L., and Simmons, D. L. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 2692-2696 [Abstract/Free Full Text]
  51. Macejak, D. G., and Sarnow, P. (1991) Nature 353, 90-94 [Medline] [Order article via Infotrieve]
  52. Kosak, M. (1992) Crit. Rev. Biochem. Mol. Biol. 27, 385-402 [Medline] [Order article via Infotrieve]
  53. Oh, S. K., Scott, M. P., and Sarnow, P. (1992) Genes & Dev. 9, 1643-1653
  54. Shein, H. M., and Wurtman, R. J. (1971) Life. Sci. 10, 935-940 [CrossRef]
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.


Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 CirculationHome page
S. Eddahibi, C. Guignabert, A.-M. Barlier-Mur, L. Dewachter, E. Fadel, P. Dartevelle, M. Humbert, G. Simonneau, N. Hanoun, F. Saurini, et al.
Cross Talk Between Endothelial and Smooth Muscle Cells in Pulmonary Hypertension: Critical Role for Serotonin-Induced Smooth Muscle Hyperplasia
Circulation, April 18, 2006; 113(15): 1857 - 1864.
[Abstract] [Full Text] [PDF]

Home page
 FASEB J.Home page
A. Slominski, J. Wortsman, and D. J. Tobin
The cutaneous serotoninergic/melatoninergic system: securing a place under the sun
FASEB J, February 1, 2005; 19(2): 176 - 194.
[Abstract] [Full Text] [PDF]

Home page