Tryptophan Hydroxylase 2
Historically, the monoamine serotonin (5-hydroxytryptamine, 5-HT) was first discovered in the gastrointestinal (GI) tract as a contractile substance, enteramine. Subsequently, 5-HT was discovered in blood (serum) as a potent vasoconstrictive substance (enteramine was shown to be the same substance as 5-HT), in the central nervous system (CNS) as a neurotransmitter, and in the pineal gland as an intermediate precursor in the synthesis of melatonin, the neurohormone involving in the regulation of the circadian rhythm (Amireault et al. 2013).
About 95% of the body’s 5-HT resides in the GI tract, primarily (90%) in a subtype of enteroendocrine cells distributed throughout in the GI tract called enterochromaffin (EC) cells and 10% in serotonergic neurons of myenteric plexus. 5-HT originating from the GI tract acts locally or is released in the blood stream and stored in platelets. 5-HT in the GI tract has been implicated in gastroenteric diseases, such as irritable bowel syndrome (IBS). About 5% of 5-HT is synthesized in the 5-HT neurons in the brain. The central 5-HT system, which arises from the midbrain raphe nuclei and is distributed throughout the brain, modulates diverse physiological functions, including regulation of sleep, appetite, memory, and emotion. Dysfunction of the 5-HT system in the brain has been implicated in the etiology of the wide range of neurodevelopmental disorders, including depression, anxiety, obsessive-compulsive disorder, autism, and schizophrenia.
5-HT mediates its diverse physiological and pathophysiological effects in the CNS as well as in the peripheral cells and organs through at multiple receptor subtypes. To date, 14 5-HT receptors consisting of seven subfamilies (5-HT1, 5-HT2, 5-HT3, 5-HT4, 5-HT5, 5-HT6, and 5-HT7) have been identified (Bockaert et al. 2010). With the exception of 5-HT3 receptor, which is a ligand-gated ion channel, each of the 5-HT receptors is a seven transmembrane domain G-protein-coupled receptor. The 5-HT1 and 5-HT5 receptors are negatively coupled, whereas the 5-HT4, 5-HT6, and 5-HT7 receptors are positively coupled to adenylate cyclase. The 5-HT2 receptor family couples to the phosphoinositol hydrolysis signal transduction system. In the CNS, 5-HT receptors are important target sites for different classes of antidepressants and atypical antipsychotic drugs. Downregulation of the presynaptic 5-HT1A and 5-HT1B receptors and stimulation of postsynaptic 5-HT1A receptors are considered to be critical for the efficacy of antidepressants that targeted 5-HT transporter. In the GI tract, 5-HT4 receptor agonists and 5-HT3 receptor antagonists are considered to be beneficial for the treatment of IBS with constipation and diarrhea, respectively.
It has long been thought that TPH was translated from mRNA transcribed from a single gene and that the differences in physicochemical properties of neuronal (brain stem raphe nuclei) and peripheral (pineal gland or mastocytoma cells) enzymes were possibly explained by distinct posttranslational modifications (Walther and Bader 2003; Hasegawa and Nakamura 2010). The TPH cDNA was first isolated from the rabbit pineal gland in 1987. Subsequently, a diversity of the TPH mRNA species from rats and humans was revealed, which resides in length differences of the 5′ or 3′ untranslated regions (UTRs). Strikingly, it was observed that the level of TPH mRNA was 150 times more abundant in the pineal gland than in the raphe nucleus, whereas the protein levels as quantified by immunoblot experiments using anti-TPH antibody were similar (Walther and Bader 2003). The remarkable discrepancies between the amounts of mRNA and protein could result from a difference in the translation efficiency or the stabilities of TPH mRNAs in the two structures. In 2003, based on the published Tph sequence of the mouse, Tph knockout mice were generated, and they expressed almost normal levels of brain 5-HT, showed no significant differences in 5-HT-related behaviors, but had nearly no peripheral 5-HT, in particular in the gut, in the blood, and in the pineal gland (Walther and Bader 2003; Amireault et al. 2013). These unexpected findings suggested the existence of the second TPH gene not affected by the gene targeting and led to the discovery of the second TPH gene, termed as TPH2. Eventually, it was revealed that two TPH enzymes (TPH1 and TPH2) identified are actually encoded by two distinct genes: TPH1 (formerly known as TPH) is mainly expressed in the peripheral tissues, such as pineal gland and EC cells in the GI tract, whereas the newly discovered TPH2 is primarily (restrictively) expressed in neuronal serotonergic cells in the brain and enteric nervous system (Walther and Bader 2003; Amireault et al. 2013). Immediately, it has been revealed that Tph2 knockout mice have normal blood, gut, and pineal gland levels of 5-HT, but greater than 95% depletion of forebrain 5-HT. A long-lasting question regarding the molecular discrepancy between brain and peripheral TPHs has finally been answered. The genes for human TPH1 (OMIM*191060) and TPH2 (OMIM*607478) are located on chromosome 11 (at position 11p15.1) and 12 (at position 12q21), respectively. The human TPH2 gene spans 94 kb and consists of 11 exons, whereas the human TPH1 gene spans 25 kb and consists of 11 exons. Interestingly, the human TH gene (located at 11p15.5; OMIM*191290) contains 14 exons and spans 10 kb, whereas the human PAH gene (located at 12q23.2; OMIM*612349) contains 13 exons and spans 79 kb. The locations and structural similarities of these genes might suggest that sequential gene duplications giving rise to a translocation which separated the TPH1 and TH loci on chromosome 11 from the PAH and TPH2 loci on chromosome 12.
As the rate-limiting enzyme for the synthesis of 5-HT in the CNS, TPH2 definitely plays a critical role in the modulation of 5-HT neurotransmission and is thus a promising target for the neuropsychiatric drug development.
Structural and Functional Properties of Wild-Type and Missense Variants of Human TPH2
The PAH gene, a prototypal AAAHs, was first cloned in 1983. To date, more than 550 variants of the PAH gene have been deposited in the PAHdb knowledgebase (PAH mutation map; http://www.pahdb.mcgill.ca), of which about 60% are missense mutations, 13% are deletion mutations, 11% are splicing mutations, 6% are silent mutations, and 5% are nonsense mutations. The PAH gene mutations show considerable ethnic and regional variation and cause phenylketonuria (PKU)/hyperphenylalaninemia. In contrast, to date, the total number of recorded patients with TPH2 mutations is too low. Four non-synonymous TPH2 SNPs have been reported in patients with psychiatric disorders, which severely impair TPH2 enzymatic activity by causing the amino acid substitutions (Jacobsen et al. 2012) (Fig. 2). Four variants are TPH2-S41Y (nucleotide change in cDNA (RefSeq ID, NM_173353) at position 122 C > A; exon 2), TPH2-P206S (616 C > T; exon 6), TPH2-R303W (907 C > T; exon 7), and TPH2-R441H (1322 G > A; also known as 1463 G > A; exon 11). TPH2-S41Y and TPH2-P206S variants were associated with bipolar disorder; TPH2-R303W was found in two patients with attention-deficit/hyperactive disorder (ADHD) and TPH2-R441H with unipolar depression. In 2005, a representative missense mutant of the TPH2-R441H was first identified in a relatively small cohort of senile patients with unipolar major depression (Jacobsen et al. 2012). Functional analysis in PC12 cells demonstrated an about 80% reduction in the 5-HT synthesis of the TPH2-H441 enzyme compared with the wild-type TPH2-R441. Coincidently, amino acid R441 in TPH2 is analogous to amino acid R408 in PAH. An R408W mutation in PAH causes a severe loss of function of PAH and has been identified in about 10% of the PKU patients, representing the most prevalent pathogenic mutation in PAH (PAH mutation map; http://www.pahdb.mcgill.ca). However, R441H mutation in TPH2 was not replicated in subsequent extensive studies in multiple ethnically diverse cohorts. Thus R441H mutation in TPH2 is not considered to be a general genetic risk factor for depression, rather this particular mutation may be prevalent in restricted populations as consequence of a founder effect (Jacobsen et al. 2012). Since then, several additional missense mutations of the TPH2 were identified.
By using three different expression systems, it has been revealed that the properties of variants affecting the regulatory domain (TPH2-L36V, TPH2-L36P, TPH2-S41Y, and TPH2-R55C) were indistinguishable from the wild-type TPH2 (McKinney et al. 2009). Moderate loss-of-function effects observed for variants in the catalytic and tetramerization domains (TPH2-P206S, TPH2-A328V, TPH2-R441H, and TPH2-D479E) were manifested via stability and solubility effects, whereas TPH2-R303W had severely reduced solubility and was completely inactive (McKinney et al. 2009). It is worth noting that except for the TPH2-R303W variant, all the variants have some residual enzymatic activity, possibly reflecting the biological significance and indispensability of TPH2 in human health (McKinney et al. 2009).
Gene Expression and Regulation
An early pioneering study revealed that an inverted CCAAT box (−67/−63, transcription start site is designated as +1) is essential for cAMP-mediated induction of the human TPH1 gene transcription through the binding of NF-Y. The CCAAT displacement protein (CDP), homologous to the Drosophila homeobox protein, Cut, binds to the element (−270/−263) to repress the human TPH1 promoter activity. Identification of a functional TPH1 polymorphism (rs7130929) associated with bowel habit subtypes in IBS demonstrated that early growth response factor 1 (EGR-1) binds to the GC-rich sequences (−348/−330) to upregulate the human TPH1 promoter activity.
The core promoter of the human TPH2 gene was localized to the region between nt −107 and +7, which also includes regulatory elements necessary for the cell-type specific induction of the human TPH2 gene transcription by calcium mobilization (Lenicov et al. 2007; Chen et al. 2008). Close inspection of the sequences conserved between the human and rat TPH2 genes indicates that AP-2 (−17/−9) and AP-1 (−41/−35) elements are relevant to the Ca2+-induced human TPH2 gene regulation (Lenicov et al. 2007). It was previously reported that a rapid Ca2+-dependent activation of cAMP-response element-binding protein (CREB) regulates AP-1 activity, whereas a Ca2+-/diacylglycerol-mediated protein kinase C activation regulates AP-2 activity (Lenicov et al. 2007). The human TPH2 gene also contains a potential C/EBP (−53/−46) site as a Ca2+-sensitive element. However, the C/EBP site is not conserved in the rat TPH2 promoter. In the human TPH2 promoter, an inverted cAMP response element (CRE) is found at −243/−235 (5′-TAACGTCA-3′) (Lenicov et al. 2007). Interestingly, potential CREs were also found at corresponding positions of the rat (−203/−196, 5′-TGACGCAT-3′) and mouse (−374/−367, 5′-TAACGTCA-3′) TPH2 genes. It was reported that a CREB-mediated activation of target gene transcription requires an activating cofactor, CREB-regulated transcriptional activator (CRTC). CRTC is thought to be constitutively phosphorylated by the upstream CRTC kinase(s), such as salt-induced kinase (SIK) (Altarejos and Montminy 2011). Elevated intracellular calcium activates Ca2+-/calmodulin-dependent protein phosphatase, calcineurin which actively dephosphorylates CRTC. Coordinately, elevation of cAMP activates protein kinase A (PKA) to liberate catalytic subunits (PKAα), which in turn phosphorylates and deactivates SIK, leading to an increase in dephosphorylated state of CRTC. When actively dephosphorylated, CRTC translocates to the nucleus. Once in the nucleus, CRTC form a multi-complex with CRE-bound CREB on the target gene promoter and enhance its transcriptional activity (Altarejos and Montminy 2011). These findings may imply that the CRTC dictates changes in both cAMP-dependent and Ca2+-sensitive signal transductions and then transmits integrated information to CRE-bound CREB to eventually regulate TPH2 gene transcription. At present, the precise role of CREB and CRTC for TPH2 gene regulation remains to be determined.
Examination of the function of the TPH2 gene polymorphism rs11178997 in 5’UTR revealed that POU3F2 (POU Class 3 Homeobox 2) would be a potential candidate regulating the human TPH2 promoter activity through binding to the element (−491/−461) containing the relevant SNP (Scheuch et al. 2007). It was demonstrated that overexpression of POU3F2 caused about 2.7-fold increase in the human TPH2 promoter, which was negated by co-expression with PQBP1 (polyglutamine-binding protein 1) that binds to the polyglutamine tract of POU3F2, resulting in suppression of the POU3F2-mediated gene transactivation. Mammalian POU3F2 has three homopolymeric amino acid repeats (glycine, glutamine, and proline), whereas nonmammalian POU3F2 orthologs lack most or all of these repeats. It was reported that the Xenopus Pou3f2 knock-in female mice showed the significant decrease in the TPH2 levels in raphe nuclei and the remarkable deterioration of characteristic mammalian maternal behaviors (Nasu et al. 2014). These results imply that mammalian-specific sequences in Pou3f2 may contribute to maternal behaviors through regulation of the TPH2 gene expression and consequently modulation of the brain 5-HT levels. A functional analysis using deletion and site-directed mutations identified the estrogen response element (ERE) half site located within the human TPH2 promoter (−792/−787) that may be important for the estrogen receptor β-induced human TPH2 gene transcriptional activity (Hiroi and Handa 2013). 5-HT and vitamin D have been proposed to play a role in autism; however, no causal mechanism has been established. A scanning in silico the proximal 5′ 10 kb of the human TPH2 gene identified two potential distal activating vitamin D response elements (VDREs) (−7059/−7045 and −9771/−9757) that are presumably associated with transcriptional activation. Thus, the human TPH2 gene is likely to be activated by vitamin D (Patrick and Ames 2014). At present, the role of VDREs for TPH2 gene regulation remains to be determined.
RE-1/NRSE and TPH2 Gene Regulation
Among human AAAH genes, besides TPH2, TH has a functional canonical NRSE (5′-TTAGATTCCACGGACGAGCCC-3′) within its promoter region located at −204/−184 that is responsible for repression of the TH gene transcription by recruiting HDAC complexes. TH catalyzes the initial and rate-limiting step in biosynthesis of catecholamines (dopamine, epinephrine, and norepinephrine) in central and peripheral nervous system. It is evolutionarily interesting that TPH2 and TH genes acquired NRSE, whereas PAH and TPH1 appear to be devoid of NRSE. The presence of the NRSEs on their promoter region further explains the neuron specificity of the TPH2 and TH genes.
NRSF, a member of the Kruppel class C2H2 zinc-finger transcription factors, contains a DNA-binding domain that is localized within the cluster of eight zinc fingers (I to VIII) and recognizes NRSE in the regulatory region of target genes. Once NRSF binds to the NRSE in target genes, it acts as a molecular scaffold to recruit several chromatin modifiers to its N- and C-terminal repressor domains (RD1 and RD2, respectively). The RD1 interacts with Sin3 and subsequently recruits the histone deacetylase complex (HDAC1 and HDAC2), whereas the RD2 associates with CoREST (REST corepressor 1), which also interacts with Sin3, HDAC1, HDAC2, and methyl-CpG-binding protein 2 (MeCP2). Thus NRSF can repress target gene expression through the generation of dynamic macromolecular complexes (Ballas and Mandel 2005; Ooi and Wood 2009) (Fig. 4).
A1 mes-c-myc (A1) cell line, obtained from mesencephalic primary cultures generated from 11-day-old mouse embryos, has been widely used as a model to study neural differentiation (Gentile et al. 2012). A1 cells have been reported to show some properties of the serotonergic phenotype and thus have proven to be useful as a central serotonergic neuron model.
It was observed that NRSF is significantly more expressed in undifferentiated A1 cells as compared with differentiated A1 cells; namely, the NRSF mRNA and protein expression levels were decreased as A1 cells differentiated (Gentile et al. 2012). Curiously, the NRSE-bearing endogenous mouse TPH2 gene transcript and protein are upregulated by neuronal differentiation in A1 cells. Transient transfection analyses revealed that upon mutation of the NRSE motif, the promoter activity driven by the human TPH2 5′-upstream region strongly increases with A1 cell differentiation (Gentile et al. 2012). It has been reported that NRSF undergoes proteasomal degradation by ubiquitination during neuronal differentiation (Ballas and Mandel 2005). These results suggest that the expression of the NRSE-bearing human TPH2 gene is negatively controlled depending on the NRSF activity. Once the NRSF levels decrease to below a certain threshold that is enough to repress the human TPH2 gene transcription, the human TPH2 promoter would be relieved from the NRSF-mediated negative regulation. It was reported that NRSF mRNA expression in dorsal raphe neurons was significantly increased in female subjects affected by major depressive disorder compared with matched controls (Goswami et al. 2010), suggesting that the NRSE-NRSF system and its dysregulation might be related to the dysregulation of the hTPH2 gene expression, TPH2-dependent 5-HT synthesis, and the subsequent 5-HT neural system functions in the brain. Future studies will further address a critical role of the NRSE-NRSF interaction, the posttranslational modulation of NRSF stability, and the NRSF-centered epigenetic mechanisms in regulating the hTPH2 gene expression (Ballas and Mandel 2005; Ooi and Wood 2009).
The TPH1 and TPH2 gene have nonoverlapping expression patterns and different functions and are independently regulated (no compensatory activation of TPH1 or TPH2 expression). TPH2, the rate-limiting enzyme in 5-HT biosynthesis in serotonergic neurons in the brain and enteric nervous system, is strictly controlled at several distinct and overlapping steps, including gene transcription, pre-mRNA alternative splicing, translation, posttranslational modification, and protein degradation.
Extensive efforts have been devoted to examine the gene for functional variants (SNPs) and to determine whether these variants can be correlated with clinical aspects of neuropsychiatric disorders. Although extremely rare, four coding SNPs associated with psychiatric disorders are reported. Functional analysis of naturally occurring mutations should yield valuable information on the structure-function relationships of TPH2. In contrast, numerous non-coding SNPs deposited into database, the precise functional consequences remain largely unknown. The phosphorylation of TPH2 at Ser19 by CaMKII/PKA leads to the interaction with 14-3-3 proteins, and their interactions are important for the activity and stability of TPH2. The TPH2 promoter contains multiple transcription factor binding sites that regulate the gene transcription by neuronal activity, hormone stimulations, and a variety of stresses. Extracellular stimuli acting through several protein kinases also mediate transcriptional regulation of the TPH2 gene. Regarding the epigenetic regulation of the TPH2 gene, the NRSE-bound NRSF definitely plays a critical role to repress the gene expression by recruiting chromatin modifiers including HDAC 1 and 2. Future studies will further address a critical role of the NRSE-NRSF interaction, the posttranslational modulation of NRSF stability, and the NRSF-centered epigenetic mechanisms in regulating the hTPH2 gene expression.
- Goswami DB, May WL, Stockmeier CA, Austin MC. Transcriptional expression of serotonergic regulators in laser-captured microdissected dorsal raphe neurons of subjects with major depressive disorder: sex-specific differences. J Neurochem. 2010;112:397–409. doi:10.1111/j.1471- 4159.2009. 06462.x.PubMedCrossRefPubMedCentralGoogle Scholar
- Patel PD, Bochar DA, Turner DL, Meng F, Mueller HM, Pontrello CG. Regulation of tryptophan hydroxylase-2 gene expression by a bipartite RE-1 silencer of transcription/neuron restrictive silencing factor (REST/NRSF) binding motif. J Biol Chem. 2007;282:26717–24. doi: 10.1074/jbc.M705120200.PubMedCrossRefPubMedCentralGoogle Scholar
- Scheuch K, Lautenschlager M, Grohmann M, Stahlberg S, Kirchheiner J, Zill P, Heinz A, Walther DJ, Priller J. Characterization of a functional promoter polymorphism of the human tryptophan hydroxylase 2 gene in serotonergic raphe neurons. Biol Psychiatry. 2007;62:1288–94. doi: 10.1016/j.biopsych.2007.01.015.PubMedCrossRefPubMedCentralGoogle Scholar
- Winge I, McKinney J, Haavik J. Tryptophan hydroxylase. In: JFP D’M, editor. Amino acids in human nutrition and health. Wallingford: CAB International; 2012. p. 150–72.Google Scholar