Tryptophan Hydroxylase-2 Controls Brain Serotonin Synthesis

See allHide authors and affiliations

Science  09 Jul 2004:
Vol. 305, Issue 5681, pp. 217
DOI: 10.1126/science.1097540


Dysregulation of brain serotonin contributes to many psychiatric disorders. Tryptophan hydroxylase-2 (Tph2), rather than Tph1, is preferentially expressed in the brain. We report a functional (C1473G) single-nucleotide polymorphism in mouse Tph2 that results in the substitution of Pro447 with Arg447 and leads to decreased serotonin levels in PC12 cells. Moreover, in BALB/cJ and DBA/2 mice that are homozygous for the 1473G allele, brain serotonin tissue content and synthesis are reduced in comparison to C57Bl/6 and 129X1/SvJ mice that are homozygous for the 1473C allele. Our data provide direct evidence for a fundamental role of Tph2 in brain serotonin synthesis.

Dysregulation of brain serotonin (5-hydroxytryptamine, or 5-HT) is an important contributing factor in many psychiatric disorders. Central 5-HT neurons are the primary targets for tricyclic antidepressants and selective serotonin reuptake inhibitors (SSRIs), as well as psychostimulants and hallucinogenic drugs (13). Tryptophan hydroxylase (Tph1) has long been considered as the sole rate-limiting enzyme for the synthesis of 5-HT. However, the recently identified Tph2 is preferentially expressed in the brain, whereas Tph1 is mainly expressed and controls 5-HT synthesis in the periphery (4).

To study the biological function of Tph2, we performed reverse transcription–polymerase chain reaction (PCR) to obtain cDNA of Tph2 from brainstem tissue combined from mice of different genotypes (5). We obtained two different clones of mTph2, each encoding either Arg447 (CGC, mTph2R) or Pro447 (CCC, mTph2P) (Fig. 1A). Comparison with other members of this family of enzymes, tyrosine hydroxylase, phenylalanine hydroxylase, and Tph1 in humans, mice, and rats, as well as with the mouse genome sequence, revealed that Pro447 was highly conserved (fig. S1). To explore the possible consequences of this (C1473G) single-nucleotide polymorphism, we used PC12 cells that endogenously synthesize dopamine and norepinephrine (6). We generated stable cell lines that expressed similar levels of hemagglutinin (HA)–tagged mTph2P and mTph2R, respectively (Fig. 1B). 5-HT levels in PC12 cells expressing HA-mTph2R were reduced by ∼55% as compared to those in cells expressing HA-mTph2P (Fig. 1B), whereas no 5-HT was detected in mock-transfected cells.

Fig. 1.

(A) (C1473G) polymorphism in mTph2. The C or G polymorphism is highlighted and indicated with arrowheads. Nucleotide numbers are shown for the start and stopcodons of mTph2 as well as the site of polymorphism. (B) 5-HT levels in PC12 cells expressing HA-mTph2R were lower than those in PC12 cells expressing HA-mTph2P (n = 6 experiments), whereas no differences in dopamine levels were observed (9). Similar levels of HA-mTph2P and HA-mTph2R were expressed in respective PC12 stable cell lines (n = 3 experiments). (C) BALB/cJ mice exhibited lower 5-HT synthesis rates and tissue contents as compared to 129X1/SvJ mice (n = 7 mice). (D) Genotyping of 129X1/SvJ and BALB/cJ mice. Positive-control [523 base pairs (bp)] and allele-specific PCR products (307 bp) are indicated with arrowheads. All data are presented as means ± SEMs. The statistical significance of all data presented was analyzed by Student's t test: **, P < 0.01; ***, P < 0.001.

We then applied a modified tetra-primer amplification refractory mutation system–PCR (7), with C- or G-allele–specific primers to identify the mouse strain(s) that harbored the (C1473G) polymorphism (fig. S2). We first tested 129X1/SvJ and BALB/cJ mice that display marked 5-HT–related behavioral differences (8) and identified that 129X1/SvJ mice were homozygous for the 1473C allele, whereas BALB/cJ mice were homozygous for the 1473G allele (Fig. 1D and table S1).

Next, we determined in vivo 5-HT synthesis in 129X1/SvJ and BALB/cJ mice (5). We treated mice with m-hydroxybenzylhydrazine to determine the accumulation of the 5-HT precursor, 5-hydroxytryptophan (5-HTP), and found that BALB/cJ mice showed ∼50% and ∼70% reductions in 5-HTP synthesis in the frontal cortex and striatum, respectively, as compared to 129X1/SvJ mice (Fig. 1C). Similarly, tissue contents of 5-HT in untreated mice were measured and revealed a substantial ∼40% decrease in the frontal cortex and striatum of BALB/cJ mice as compared to those in 129X1/SvJ mice (Fig. 1C). No significant differences of mTph2 mRNA levels in brainstem tissue between 129X1/SvJ and BALB/cJ strains were detected (9). Further genotyping revealed the existence of a homozygous 1473C allele in C57Bl/6 mice and a homozygous 1473G allele in DBA/2 mice (table S1), which also showed an ∼45% difference in striatal 5-HTP synthesis (fig. S3). Because mice of these strains are widely used in biomedical research, identification of this functional polymorphism has immediate practical impact.

Variations in brain 5-HT levels can contribute to behavioral differences in mice (8, 10) and psychiatric disorders in humans (13). Our data provide direct evidence for the fundamental role of Tph2 in 5-HT synthesis in the central nervous system and may set the stage for a better understanding of 5-HT–related behaviors.

Supporting Online Material

Materials and Methods

Figs. S1 to S3

Table S1

References and Notes

References and Notes

View Abstract

Navigate This Article