Journal of Neuroscience Research 2k457-465 (1991)

Isolation and Structural Characterization of the Murine Tryptophan Hydroxylase Gene J. Stoll and D. Goldman Laboratory of Clinical Studies, National Institute on Alcohol Abuse and Alcoholism, Bethesda, MD 1987; Darmon et al., 1988; Dumas et al., 1989; Stoll et al., 1990). Although serotonin is found in many sites throughout the body, TrpOHase has a more limited, cell-specific distribution of expression. In the brain, serotonergic neurons are located primarily in the raphe nuclei of the brainstem and send projections throughout the rest of the CNS (Ishimura et al., 1988). In the periphery, TrpOHase is found in the pineal gland, in mast cells, and in the enteric neurons of the gut (Hakanson et al., 1967; Lovenberg et al., 1967, 1968; Gershon, 1981). An mRNA for TrpOHase was detected in these tissues and in several mastocytoma cell lines (Grenett et al., 1987; Darmon et al., 1988; Stoll et al., 1990). TrpOHase is a member of the family of aromatic amino acid hydroxylases which also includes tyrosine hydroxylase (TyrOHase) and phenylalanine hydroxylase (PheOHase) (Kaufman and Fisher, 1974). cDNA clones have been isolated for these enzymes and analysis of their sequences revealed that the degree of similarity varied along the length of the sequence; with the greatest homology found at the carboxy-terminal two-thirds of Key words: tryptophan hydroxylase, gene structure, the amino acid sequence (Farquet et al., 1988). The abilpromoter sequence, evolutionary relationship ity of TrpOHase constructs containing only the conserved regions to catalyse the hydroxylation of tryptophan suggests the C-terminal regions encode the catalytic sites (Stoll et al., 1990). The non-similar NINTRODUCTION terminal regions are thought to encode tissue-specific Tryptophanhydroxylase(Trp0HaseE.C.1.14.16.4) and enzyme-specific regulatory domains. is the rate-limiting enzyme in the biosynthesis of serotoThe structures of the genes encoding tyrosine and nin (Grahame-Smith, 1964; Jequier et al., 1967) which phenylalanine hydroxylase have been determined makes this enzyme an important target for mechanisms (DiLella et al., 1986; Brown et al., 1987; O’Malley et that regulate serotonin synthesis. TrpOHase activity is al., 1987) and although large differences exist in the altered by glucocorticoids (Sze et al., I976), methylenesizes of the introns, the majority of the positions of the dioxymethamphetamine (Stone et al., 1989), stimulation intronlexon boundaries are identical. The gene encoding of the serotonin autoreceptor (Sawada and Nagatsu, TrpOHase is located on mouse chromosome 7 as is the 1986); and neuronal activity (Boadle-Biber et al., 1983). gene for TyrOHase, while the gene for PheOHase is on The enzyme is regulated post-translationally by phosphorylation (Ehret et al., 1989) and pre-translationally by reserpine treatment (Zivkovic et al., 1974). Transcrip- Received June 12, 1990; revised August 22, 1990; accepted August tional regulation of TrpOHase has not been demonstrated 31, 1990. directly, primarily since a cDNA clone has only recently Address reprint requests to James Stoll at his current address, Labobeen reported and because of the fact that the mRNA is ratory of Neuroscience, National Institute on Aging, Bldg. 10, Rm present in such low levels in the brain (Grenett et al., 6C103, Bethesda, MD 20892.

The mouse tryptophan hydroxylase gene was isolated and its introdexon boundaries and putative regulatory sequences identified. To isolate the gene a mouse mastocytoma cDNA clone encoding tryptophan hydroxylase was used to identify and isolate ten overlapping DNA fragments from a mouse genomic library. Restriction mapping and sequence analysis of the clones revealed that the gene contains 11 exons and covers a region of DNA of approximately 21 kb. The transcription initiation site was mapped and the major site of initiation yields an untranslated leader sequence of 124 nucleotides. A minor initiation site is located 9 nucleotides 3’ of the major site. The 5‘ untranslated sequence is interrupted by the first intron. Analysis of the sequence upstream of the initiation site showed the presence of several putative promoter and regulatory sequences. Nine of the ten introdexon boundaries of tryptophan hydroxylase are conserved with tyrosine hydroxylase and phenylalanine hydroxylase, further delineating the evolutionary relationship of these three genes.

0 1991 Wiley-Liss, Inc.

458

Stoll and Goldman

chromosome 11. These two chromosomes have been proposed to arise by duplication of an ancestral chromosome (Ledley et a]., 1987; Stoll et al., 1990). Analysis of the gene encoding TrpOHase would provide additional information on the evolution of the aromatic amino acid hydroxylase family. Further, characterization of the gene will allow the identification of cis-acting elements involved in the cell-specific expression and regulation of TrpOHase. In this paper we report the isolation and sequence analysis of murine genomic clones encoding TrpOHase.

MATERIALS AND METHODS Isolation of Genomic Clones The entire 1729 bp cDNA clone of TrpOHase obtained from mouse P815 cells was used as a probe to screen a commercial library of BALB/c mouse DNA in AFixII (Stratagene) using the hybridization conditions previously described (Stoll et al., 1990). Following plaque purification and restriction mapping, restriction fragments were subcloned into pCEM3Zf( + ) and pGEMSZf( + ) (ProMega Biotec, Madison, WI). The plasmid subclones were sequenced as single-stranded templates prepared according to the protocol supplied by the manufacturer (ProMega Biotec) or as doublestranded templates using the protocol supplied with the sequencing kit (Sequenase 2.0, U.S. Biochemicals, Cleveland, OH). Alternatively, the sequence was determined using single-stranded templates prepared by asymmetric PCR, (primer ratio 20:l) (Ausubel et al., 1989). The entire exons and the adjacent intron/exon boundaries were sequenced on both strands. Sequencing primers were synthesized according to the available sequence (Gene Assembler, Pharmacia, Piscataway , NJ). Mapping of Transcription Initiation Site The transcription initiation site was mapped by primer extension or RNAse protection using standard procedures (Ausubel et al., 1989). For primer extension a 25 nucleotide oligomer, 5’:TGCTGGGAGTCTTCTGATCCGATGT 3’ (Trh27) corresponding to the complement of nucleotides 63 to 87 of exon 1, was labelled with y-[32P]ATP and T4 polynucleotide kinase to a specific activity of 5 x lo8 d p d p g and hybridized 16 hr at 30°C to 10 pg of total RNA from P815 mastocytoma cells. To generate RNA, which served as a positive control, the TrpOHase cDNA was subcloned into pGEM4Z and the resulting plasmid was transcribed in vitro using Sp6 RNA polymerase to yield a sense TrpOHase RNA. For a negative control, Trh 27 was hybridized to 25 p g yeast tRNA. The products of Mo-MuLV reverse transcriptase (Life Technologies, Gaithersberg, MD) were analysed on a 7% sequencing gel and compared to a

sequencing ladder prepared from the same oligonucleotide. For RNAse protection, the Not1 insert of phage clone 6 (see Figs. 1,4) was subcloned into pGEMSZ( + ) (ProMega Biotec). This plasmid, p6N-2, contains 12 kb of DNA upstream of the gene and terminates at nucleotide 94 of exon 1 . The plasmid was digested with NarI and EcoRI to yield a 5.8 kb fragment containing the T7 RNA polymerase promoter adjacent to 155 bp of genomic sequence extending 5’ from nucleotide 94. Transcription with T7 RNA polymerase in the presence of a-[”PICTP yields an antisense probe of 257 nucleotides and was used without gel purification. The probe was hybridized to RNA as indicated in Figure 3 for 16 hr at 55°C. The RNAse-protected fragments were resolved on a 6% sequencing gel and compared to a sequencing ladder.

RESULTS AND DISCUSSION Isolation and Characterization of Genomic Clones Genomic DNA encoding tryptophan hydroxylase was identified by hybridizing the entire cDNA obtained from P815 mouse mastocytoma cells (Stoll et al., 1990) to a BALB/c murine genomic library prepared in AFIXII. A total of 1 x lo6 plaques were screened, from which ten positive clones were identified and plaque purified. The clones were mapped by restriction enzyme digestion and found to represent eight individual overlapping fragments (clones 4 and 5 were identical as were clones I and 8) encompassing a total of approximately 40 kb (Fig. 1). Southern blot hybridization of the digested clones to TrpOHase oligonucleotides showed the cDNA sequence is contained in a region of DNA encompassing approximately 21 kb. Clone 6 extends 12 kb upstream from the gene and clone 3 continues downstream for an additional 14 kb. This would mean the size of TrpOHase is intermediate between the 8-1 0 kb reported for tyrosine hydroxylase and the 100 kb reported for phenylalanine hydroxylase (DiLella et al., 1986; Brown et al., 1987; O’Malley et al., 1987). TrpOHase cDNA clones also detect a second, larger mRNA by Northern blot analysis (Grenett et al., 1987: Darmon et al., 1988; Stoll et a]., 1990). This larger mRNA was shown to be an extension of the 3’ untranslated region (Darmon et at., 1988). A second mRNA of larger size can also be detected in mastocytoma cells, although it is only a minor component, typically 5 % . We have not characterized the genomic sequences giving rise to this larger mRNA. This sequence could conceivably result from an additional exon or might represent heterogeneous processing at the 3‘ terminus. The sequence at the 3‘ end of the gene contains a weak poly A addition signal and lacks a CACTG se-

Mouse Tryptophan Hydroxylase Gene

exon

B R

HBH

S

1

2

R

3

BBSS

R S

S

4

56

L

H HSBR

7 8

9 10

459

H

11

H

1kb

Fig. 1. Structure of the mouse tryptophan hydroxylase gene. A schematic representation of the gene is provided. Middle: The linear DNA sequence is indicated by the bold horizontal line. The vertical bars indicate the exons and each is numbered directly below. Top: A restriction map of the gene is presented

corresponding to the linear DNA molecule. S: Sst I; H: Hind 111; B: Bam HI; R: Eco RI; L: Sal I. Bottom: The ten different phage clones are shown relative to the position they map in the gene. The size corresponding to Ikb is indicated.

quence. Both of these elements have been proposed to be important in the processing of the 3' terminus of mRNA (Birnstiel et al.. 1985).

sequence (amino acids 7-12) which is found in mouse but not in rat or rabbit TrpOHase cDNA clones (Grenett et al., 1987; Dannon et al., 1988; Stoll et al., 1990). This repeat occurs in the middle of exon 2. Approximately 8 kb of the total of 18 kb of intron sequence has been determined. Comparison of these sequences with the GenBank database revealed no homology to other unique sequences. However, a number of repetitive elements were found. The B 1 element (Krayev et al., 1980) was found in introns I , 6, and 9. The B2 element (Krayev et al., 1982) was found in intron 4. Portions of the LlMd family (Tolberg et al., 1987) were detected in introns 2 and 10. Simple repeat sequences were also present. Just upstream of exon 8 is a stretch of (GT)16.Following exon 8 is a 60 nucleotide polypurine stretch immediately followed by a (CA), dinucloetide stretch. Immediately downstream of exon 10 is a (AAAC), repeat. Immediately following the end of the final exon a polypyrimidine stretch is encountered (see Table I). Since only half of the intron sequences have been determined, further examples of repetitive elements

IntrodExon Structure Each exon was sequenced on both strands along with the immediately adjacent portion of the intron to determine the introdexon boundaries. Table I gives the sequences at these boundaries and also lists the sizes of the exons and introns. There are a total of 1 1 exons. The obligate GT . . . AG dinucleotides are present at the ends of the introns. The first ten exons average approximately 100 bp in size, varying from 68 to 197 bp. The last exon is 470 bp long and contains the entire 3' noncoding sequence along with the final 189 bp of coding sequence. The first exon contains only 5' noncoding sequences and terminates 7 bp 5' to the predicted initiation codon. No discrepancies were found between the exon sequences of the genomic clones and the sequence of the cDNA from P815 cells determined previously. This includes the 9 nucleotide/3 amino acid repeat at the 5' end of the coding

460

Stoll and Goldman

TABLE I. Exonhtron Structure of the Mouse Tryptophan Hydroxylase Gene Acceptor splice site

Exon

1 Met cacacttcag/ATTCACCATG 43 G l u A s n H i s ctttctacag/GAGAATCATG 104 a l M e t G l u tattttacag/TTATGGAGAC 138 GlyPheLys caattctcag/GGCTTCAAAG 160 sGlyAspPro atctcctcag/TGGGGACCCC

Size

Donor splice site

Intron size

I

117

-2200

I 1

133

TAGGAACCAG/gtgaagtcct I l e P h e G l n 42 AATCTTCCAG/gtatggattt

-2700

111

184

LysGluAspV 104 AAGGAAGACG/gtacgtgtga

-1 5 0 0

IV

100

AspHisPro 137 CGACCACCCT/gtaaatatgt

-3200

V

68

T y r L y s H i 160 ACTACAAACA/gtaagtatgt

197

PheLeuLysG 226 TTTTTAAAAG/gtaatacgcc

-2800

136

P r o G l u P r 271 CTCCAGAGCC/gtaagtactg

-800

127

L e u A l a T h r 31 3 ACTGGCAACG/gtatagacat

-2000

96

G l u L e u L y s 345 TGAACTCAAA/gtaagagtca

134

L y s M e t A r 390 AGAAGATGAG/gtaaaaaaca

470

SerValStop CAGTGTGTGA

VI

226 l u A r g T h r tgctttgcag/AACGCACTGG

VII

271 o A s p T h r C y s tgtgttgcag/AGACACCTGC

VIII

31 4 C y s T y r P h e CgttttgcagITGCTACTTTT

I X

346 H i s A l a L e u ttgtctccag/CATGCACTTT

X

390 g G l u P h e A l a tcttaaatag/AGAATTTGCC

X I

. . . . . 3'

noncoding

151

905 -2400

.....

i

ATTAAAATGTAATTGAATCATtctctcccttccctttcttccctccaaacccctccctctcaaattcatggccttt

ctaccaaatgcccgtctttgatgggtgagttgtctgacactggccagagctggactcatggtgatgccat The nucleotide sequence of each exoniintron junction is shown. Exon sequences are given in upper case letters: intron sequences are in lower case. The corresponding amino acids at the junctions are given and numbered as before (Stoll et al., 1990). The polyadenylation signal is underlined. The site of polyadenylation is indicated by the asterisk. Note the polypyrimidine stretch immediately following the polyadenylation site.

are expected. The function of these elements is unknown, however certain polypurine stretches have been postulated to be termination sites for DNA polymerase (Lapidot et al., 1989). Sequences 3' of the polyA addition site are presumed to be important in the correct processing of the 3' terminus, however the sequence determined shows no similarity to other described sequences (Birnstiel et al., 1985). Comparison of the introdexon structure with that of the other members of the aromatic amino acid hydroxylase family extends the pattern of conservation of intervening sequence locations first developed by comparing TyrOHase and PheOHase (Brown et al., 1987). The majority of the introns occur in homologous positions in TrpOHase when compared to the other two hydroxylases (Fig. 2). This is consistent with a common evolutionary origin for members of this family. Previously, cDNA clones for the three hydroxylases were shown to have significant sequence identity in the carboxy two-thirds of the predicted amino acid sequence, suggesting a functional similarity. Experiments in which truncated portions of the cDNA were expressed have shown that the portion of the cDNA containing the final nine exons is

necessary for enzymatic activity (Stoll, J., unpublished observations). The final eight introns of TrpOHase occur in a homologous position to introns in PheOHase and TyrOHase. Interestingly, TrpOHase has two fewer introns than either of the other two enzymes. PheOHase has an intron near the C-terminus which is missing in both TrpOHase and TyrOHase. TyrOHase has an intron at Trp(233) which is missing in TrpOHase and PheOHase. Finally, PheOHase and TyrOHase share an intron that is not present in the homologous position of TrpOHase. The presence of an intron in a conserved location in TyrOHase and PheOHase but not TrpOHase is interesting in light of the proposition that PheOHase and TrpOHase are more closely related (Grenett et al., 1987). In Drosophila, only one gene encodes Trp/PheOHase (Neckameyer, 1989). Thus, the absence of this intron would suggest the perfect excision of an intron during the evolution of TrpOHase. The other additional introns in TyrOHase and PheOHase do not occur in conserved locations and may have been introduced after the divergence of the family. The first two exons encoding TrpOHase mRNA show only limited similarity (