Identification of New Non-BBB Permeable Tryptophan Hydroxylase Inhibitors for Treating Obesity and Fatty Liver Disease

Serotonin (5-hydroxytryptophan) is a hormone that regulates emotions in the central nervous system. However, serotonin in the peripheral system is associated with obesity and fatty liver disease. Because serotonin cannot cross the blood-brain barrier (BBB), we focused on identifying new tryptophan hydroxylase type I (TPH1) inhibitors that act only in peripheral tissues for treating obesity and fatty liver disease without affecting the central nervous system. Structural optimization inspired by para-chlorophenylalanine (pCPA) resulted in the identification of a series of oxyphenylalanine and heterocyclic phenylalanine derivatives as TPH1 inhibitors. Among these compounds, compound 18i with an IC50 value of 37 nM was the most active in vitro. Additionally, compound 18i showed good liver microsomal stability and did not significantly inhibit CYP and Herg. Furthermore, this TPH1 inhibitor was able to actively interact with the peripheral system without penetrating the BBB. Compound 18i and its prodrug reduced body weight gain in mammals and decreased in vivo fat accumulation.


Introduction
Serotonin (5-hydroxytryptamine ) is an ancient biochemical that acts in the central nervous system and peripheral nervous system [1]. In the central nervous system, 5-HT is a neurotransmitter, which affects mood, appetite, sleep, and memory [2][3][4][5]. An imbalance in the serotonin system in the central nervous system has been implicated in a multitude of neuropsychiatric diseases. Recently, the new function regarding peripheral serotonin in energy homeostasis ranging from the endocrine regulation by gut-derived serotonin to the autocrine/paracrine regulation by adipocyte-derived serotonin was reported [6]. Serotonin plays an important role in metabolic regulation in peripheral tissues and is emerging as a possible target for anti-obesity and fatty liver disease treatment. Serotonin is synthesized from tryptophan by the sequential action of two enzymes, tryptophan hydroxylase (TPH) and aromatic amino acid decarboxylase with TPH catalyzing the rate-limiting step ( Figure 1).
Molecules 2022, 27, x FOR PEER REVIEW 2 of 22 5-HT is a neurotransmitter, which affects mood, appetite, sleep, and memory [2][3][4][5]. An imbalance in the serotonin system in the central nervous system has been implicated in a multitude of neuropsychiatric diseases. Recently, the new function regarding peripheral serotonin in energy homeostasis ranging from the endocrine regulation by gut-derived serotonin to the autocrine/paracrine regulation by adipocyte-derived serotonin was reported [6]. Serotonin plays an important role in metabolic regulation in peripheral tissues and is emerging as a possible target for anti-obesity and fatty liver disease treatment. Serotonin is synthesized from tryptophan by the sequential action of two enzymes, tryptophan hydroxylase (TPH) and aromatic amino acid decarboxylase with TPH catalyzing the rate-limiting step ( Figure 1). Oh et al. [7] reported that serotonin regulates white and brown adipose tissue function. Mice with inducible TPH1 KO in adipose tissues exhibit inhibition of lipogenesis in epididymal white adipose tissue (WAT), induction of browning in inguinal WAT, and activation of adaptive thermogenesis in brown adipose tissue (BAT).
The inability of serotonin to cross the Blood Brain Barriers (BBB) enforces the dualistic character of the serotonin system by creating two physiologically separated serotonin pools in the body. These characteristics of serotonin and KO study results prompted us to develop a TPH1 inhibitor that only acts in peripheral tissue for treating obesity and fatty liver disease.
Several tryptophan hydroxylase type I (TPH1) inhibitors have been reported in the literature and patents. For example, para-chlorophenylalanine (pCPA), [8] also known as fenclonine, is a classic TPH1 inhibitor that has been developed to treat carcinoid syndrome. However, the in vitro binding activity of pCPA to TPH1 is very weak (IC50 > 50 µ M). Additionally, pCPA causes side effects such as depression because it passes through the central nervous system. Recently, telotristat ethyl was marketed as a TPH1 inhibitor for carcinoid syndrome [9]. Scientists at Karos Pharmaceuticals have reported that spirocyclic TPH1 inhibitors such as telotristat ethyl can be used to treat Pulmonary Arterial Hypertension (PAH) and Inflammatory Bowel Disease (IBD) [10].
Herein, we attempt to identify new TPH1 inhibitors starting from pCPA by introducing a non-BBB permeable moiety. The design, synthesis, and biological evaluation of a diverse suite of phenylalanine derivatives for preventing obesity and fatty liver disease are presented.

Chemistry
Tyrosine, which is similar to pCPA, was selected as the starting material because it is suitable for derivatization. The general methods for the synthesis of oxy-phenylalanine derivatives are outlined in Scheme 1. Commercially available L-tyrosine (1) was first esterified, and Boc-anhydride was then introduced to provide Boc protection, which yielded compound 2. Compound 2 was subsequently coupled with 7-bromo-4-chlorothieno [3,2d] pyrimidine (3) to yield compound 4. Compound 4 was treated with NaOH and subjected to Boc deprotection with HCl (4 M) in 1,4-dioxane to give compound 6. Compound 4 also underwent Suzuki coupling with 3-hydroxyphenylboronic acid (5a) and 4-hydroxyphenylboronic acid (5b) in the presence of a palladium catalyst to afford compounds 7a Oh et al. [7] reported that serotonin regulates white and brown adipose tissue function. Mice with inducible TPH1 KO in adipose tissues exhibit inhibition of lipogenesis in epididymal white adipose tissue (WAT), induction of browning in inguinal WAT, and activation of adaptive thermogenesis in brown adipose tissue (BAT).
The inability of serotonin to cross the Blood Brain Barriers (BBB) enforces the dualistic character of the serotonin system by creating two physiologically separated serotonin pools in the body. These characteristics of serotonin and KO study results prompted us to develop a TPH1 inhibitor that only acts in peripheral tissue for treating obesity and fatty liver disease.
Several tryptophan hydroxylase type I (TPH1) inhibitors have been reported in the literature and patents. For example, para-chlorophenylalanine (pCPA), [8] also known as fenclonine, is a classic TPH1 inhibitor that has been developed to treat carcinoid syndrome. However, the in vitro binding activity of pCPA to TPH1 is very weak (IC 50 > 50 µM). Additionally, pCPA causes side effects such as depression because it passes through the central nervous system. Recently, telotristat ethyl was marketed as a TPH1 inhibitor for carcinoid syndrome [9]. Scientists at Karos Pharmaceuticals have reported that spirocyclic TPH1 inhibitors such as telotristat ethyl can be used to treat Pulmonary Arterial Hypertension (PAH) and Inflammatory Bowel Disease (IBD) [10].
Herein, we attempt to identify new TPH1 inhibitors starting from pCPA by introducing a non-BBB permeable moiety. The design, synthesis, and biological evaluation of a diverse suite of phenylalanine derivatives for preventing obesity and fatty liver disease are presented.

Chemistry
Tyrosine, which is similar to pCPA, was selected as the starting material because it is suitable for derivatization. The general methods for the synthesis of oxy-phenylalanine derivatives are outlined in Scheme 1. Commercially available L-tyrosine (1) was first esterified, and Boc-anhydride was then introduced to provide Boc protection, which yielded compound 2. Compound 2 was subsequently coupled with 7-bromo-4-chlorothieno [3,2-d] pyrimidine (3) to yield compound 4. Compound 4 was treated with NaOH and subjected to Boc deprotection with HCl (4 M) in 1,4-dioxane to give compound 6. Compound 4 also underwent Suzuki coupling with 3-hydroxyphenylboronic acid (5a) and 4-hydroxyphenylboronic acid (5b) in the presence of a palladium catalyst to afford compounds 7a and 7b; these compounds were hydrolyzed and underwent Boc deprotection to give compounds 9a and 9b. Compounds 7a and 7b also underwent the Mitsunobu reaction with benzyl alcohol to afford compounds 10a and 10b, which were treated with NaOH and underwent Boc-deprotection with HCl (4 M) in 1,4-dioxane to give compounds 11a and 11b. Compounds 15a and 15b were also synthesized, as shown in Scheme 1. Thienopyrimidine (3) was ammonified to afford compound 13a, and thienopyrimidine (3) was treated with aqueous NaOH to afford compound 13b. Compounds 12 and 13b were coupled via Suzuki coupling to afford compounds 14a and 14b. Triflic anhydride and bis(pinacolato) diboron were used to prepare compound 12 from compound 2. Compounds 14a and 14b were then hydrolyzed, and subsequent acid deprotection afforded compounds 15a and 15b.
The synthesis of compound 30 is shown in Scheme 3. Compound 25 was cyclized with urea and then brominated to afford compound 26 which was subsequently chlorinated with POCl3 to obtain compound 27. Compound 27 was further treated with compound 16i and p-methoxybenzylamine to afford compound 28. TFA deprotection afforded compound 29, which was then hydrolyzed; acidic deprotection of 29 afforded compound 30.

Biological Evaluations
The in-house screening revealed that thienopyrimidine tyrosine (derivative 6) was active in the inhibition of TPH1, with an inhibition of 64% at 100 µ M. Consequently, the core structure of thienopyrimidine tyrosine (derivative 6) was selected as the starting point for the synthesis, evaluating the structure-activity relationships (SAR), and optimization of TPH1 inhibitors. The ability of the synthesized compounds to inhibit TPH1 at 100 µ M and 1 µ M, respectively, was evaluated. In the evaluation, pCPA and LP533401 (4-
The optimized parameters for the oxy-heteroaryl phenylalanine derivatives of compound 6 are summarized in Table 1. Compound 6 exhibited an inhibition that was greater than 60%. When substituents were introduced at the para position (compounds 9a and 11a), little loss of the inhibitory activity occurred (45% and 58% inhibition at 100 µM, respectively). Compounds 9b and 11b (82% and 94% inhibition at 100 µM, respectively) showed better inhibitory activity than compounds 9a and 11a, respectively.  Based on these results, we explored a wide range of substituents on posi thienopyrimidine core to enhance the inhibitory activity; the results are sum Table 2. Naphthalene ethane amine (18b) showed less inhibition than compo tempts were made to introduce benzyloxy derivatives at position-4 of thieno However, these derivatives were unstable under the acidic conditions used fo tection, as evidenced by the cleavage of the benzyl moiety to form compo improve the stability of these derivatives, attempts were made to synthesize 1 18e, which have a CF3-substituted benzyloxy moiety. The molecular docking that the CF3 group interacted with Leu236 and Phe241 of TPH1. Contrastingly 18c without a substituent and compound 18d with a phenyl substituent at th tion on the phenyl ring completely lost their activity. However, when the sub in the ortho position (compound 18e), the percentage inhibition dramatically a concentration of 1 µ M. Although the interactions among compound 18e and and key residues of TPH1 were enhanced, the hydrophobic tail of compound 90% Based on these results, computational molecular docking studies were conducted to improve the molecular level understanding of effective binding to the active site; these studies are fundamental for facilitating further structural modification. For the docking studies, the high-resolution structure of human TPH1 (Protein Data Bank access code 3HF8; chain A) was used in AutoDock4.2 [11,12]. The standard protocols and parameters described in our previous work were used to pre-process the structures [13]. The molecular docking results showed that compound 11b interacted with the catalytic residues of TPH1 (Arg257 and Thr265). In addition, the hydrophobic tail of compound 11b was bound to the exterior of the active site of TPH1 with weak interactions and unable to block the entrance of the TPH1 active site. Such binding of the hydrophobic tail of compound 11b is leaving the possibility of substrate binding with the TPH1 in a competitive manner ( Figure 2). We hypothesized that the interaction of the hydrophobic tail of synthesized compounds at the entrance of the active site of TPH1 might prevent the substrate access the active site of TPH1 which might lead to better inhibition. Based on these observations and our hypothesis, we first replaced oxy-phenylalanine to produce compounds 15a, 15b, and 18a. Although compounds 15a and 15b showed weak activity, biphenyl methyl amine (18a) was a promising TPH1 inhibitor with an inhibition of 90% at 100 µM.
Based on these results, we explored a wide range of substituents on position-4 of the thienopyrimidine core to enhance the inhibitory activity; the results are summarized in Table 2. Naphthalene ethane amine (18b) showed less inhibition than compound 18a. Attempts were made to introduce benzyloxy derivatives at position-4 of thienopyrimidine. However, these derivatives were unstable under the acidic conditions used for Boc deprotection, as evidenced by the cleavage of the benzyl moiety to form compound 15b. To improve the stability of these derivatives, attempts were made to synthesize 18c, 18d, and 18e, which have a CF 3 -substituted benzyloxy moiety. The molecular docking study shows that the CF 3 group interacted with Leu236 and Phe241 of TPH1. Contrastingly, compound 18c without a substituent and compound 18d with a phenyl substituent at the para position on the phenyl ring completely lost their activity. However, when the substituent was in the ortho position (compound 18e), the percentage inhibition dramatically improved at a concentration of 1 µM. Although the interactions among compound 18e and the catalytic and key residues of TPH1 were enhanced, the hydrophobic tail of compound 18e partially occupied the entrance of the active site of TPH1. This observation suggests that further modifications are required. Halogens are widely used in drug design to improve the drug-target binding affinity and slow drug metabolism; therefore, we introduced chlorine at the para position of the benzyloxy moiety. However, the introduction of chlorine in 18e yielded compound 18f, which showed a loss in the percentage inhibition. Based on the molecular docking simulation, the R isomer of the trifluoroethoxy linker was also introduced as a substituent. This approach resulted in the design and synthesis of compounds 18g, 18h, and 18i with furyl, dihydropyran, and methyl pyrazole substituents. Compared with compound 18g, compounds 18h and 18i showed a higher percentage inhibition, with IC 50 values of 263 nM and 37 nM, respectively. Based on these results, we explored a wide range of substituents on position-4 of the thienopyrimidine core to enhance the inhibitory activity; the results are summarized in Table 2. Naphthalene ethane amine (18b) showed less inhibition than compound 18a. Attempts were made to introduce benzyloxy derivatives at position-4 of thienopyrimidine. However, these derivatives were unstable under the acidic conditions used for Boc deprotection, as evidenced by the cleavage of the benzyl moiety to form compound 15b. To improve the stability of these derivatives, attempts were made to synthesize 18c, 18d, and 18e, which have a CF3-substituted benzyloxy moiety. The molecular docking study shows that the CF3 group interacted with Leu236 and Phe241 of TPH1. Contrastingly, compound 18c without a substituent and compound 18d with a phenyl substituent at the para position on the phenyl ring completely lost their activity. However, when the substituent was in the ortho position (compound 18e), the percentage inhibition dramatically improved at a concentration of 1 µ M. Although the interactions among compound 18e and the catalytic and key residues of TPH1 were enhanced, the hydrophobic tail of compound 18e partially occupied the entrance of the active site of TPH1. This observation suggests that further modifications are required. Halogens are widely used in drug design to improve the drugtarget binding affinity and slow drug metabolism; therefore, we introduced chlorine at the para position of the benzyloxy moiety. However, the introduction of chlorine in 18e yielded compound 18f, which showed a loss in the percentage inhibition. Based on the molecular docking simulation, the R isomer of the trifluoroethoxy linker was also introduced as a substituent. This approach resulted in the design and synthesis of compounds 18g, 18h, and 18i with furyl, dihydropyran, and methyl pyrazole substituents. Compared with compound 18g, compounds 18h and 18i showed a higher percentage inhibition, with IC50 values of 263 nM and 37 nM, respectively.
To optimize the binding of the CF3 group and thienopyrimidine ring with TPH1, compound 18i was synthesized as an R-enantiomer with a modified hydrophobic tail,     Therefore, compound 18i is a more effective inhibitor than compounds 1 because of its increased potency, as evidenced by its IC50 value of 37 nM. It assumed that adding an amine group to the thienopyrimidine ring would lik different potency profile to compound 18i. However, compound 30 dispro sumption because it showed reduced potency, as evidenced by the IC50 valu (Table 3). During this evaluation, LP533401 was used as a reference [14]. The compound 18i were synthesized to improve its oral absorption ( Table 3). The e pound 18i, compound 19, exhibited nanomolar inhibitory activity, with an I 994 nM. In addition, the hippurate salt (21) was synthesized as an oral drug study. To optimize the binding of the CF 3 group and thienopyrimidine ring with TPH1, compound 18i was synthesized as an R-enantiomer with a modified hydrophobic tail, which was capable of effectively inhibiting TPH1. The optimal interactions were observed among the thienopyrimidine ring and phenylalanine moiety of compound 18i and the aforementioned key and catalytic residues of TPH1.
Therefore, compound 18i is a more effective inhibitor than compounds 18e and 18h because of its increased potency, as evidenced by its IC 50 value of 37 nM. It was further assumed that adding an amine group to the thienopyrimidine ring would likely confer a different potency profile to compound 18i. However, compound 30 disproved this assumption because it showed reduced potency, as evidenced by the IC 50 value of 208 nM (Table 3). During this evaluation, LP533401 was used as a reference [14]. The prodrugs of compound 18i were synthesized to improve its oral absorption ( Table 3). The ester of compound 18i, compound 19, exhibited nanomolar inhibitory activity, with an IC 50 value of 994 nM. In addition, the hippurate salt (21) was synthesized as an oral drug for in vivo study. Therefore, compound 18i is a more effective inhibitor than compounds 18e because of its increased potency, as evidenced by its IC50 value of 37 nM. It wa assumed that adding an amine group to the thienopyrimidine ring would likely different potency profile to compound 18i. However, compound 30 disproved sumption because it showed reduced potency, as evidenced by the IC50 value of ( Table 3). During this evaluation, LP533401 was used as a reference [14]. The pro compound 18i were synthesized to improve its oral absorption ( Table 3). The este pound 18i, compound 19, exhibited nanomolar inhibitory activity, with an IC50 994 nM. In addition, the hippurate salt (21) was synthesized as an oral drug fo study. Therefore, compound 18i is a more effective inhibitor than compounds 18e a because of its increased potency, as evidenced by its IC50 value of 37 nM. It was assumed that adding an amine group to the thienopyrimidine ring would likely c different potency profile to compound 18i. However, compound 30 disproved sumption because it showed reduced potency, as evidenced by the IC50 value of ( Table 3). During this evaluation, LP533401 was used as a reference [14]. The prod compound 18i were synthesized to improve its oral absorption ( Table 3). The ester pound 18i, compound 19, exhibited nanomolar inhibitory activity, with an IC50 v 994 nM. In addition, the hippurate salt (21) was synthesized as an oral drug for study. Therefore, compound 18i is a more effective inhibitor than compounds 18e because of its increased potency, as evidenced by its IC50 value of 37 nM. It was assumed that adding an amine group to the thienopyrimidine ring would likely c different potency profile to compound 18i. However, compound 30 disproved sumption because it showed reduced potency, as evidenced by the IC50 value of ( Table 3). During this evaluation, LP533401 was used as a reference [14]. The prod compound 18i were synthesized to improve its oral absorption ( Table 3). The ester pound 18i, compound 19, exhibited nanomolar inhibitory activity, with an IC50 v 994 nM. In addition, the hippurate salt (21) was synthesized as an oral drug for study. Therefore, compound 18i is a more effective inhibitor than compounds 18e because of its increased potency, as evidenced by its IC50 value of 37 nM. It wa assumed that adding an amine group to the thienopyrimidine ring would likely different potency profile to compound 18i. However, compound 30 disproved sumption because it showed reduced potency, as evidenced by the IC50 value o (Table 3). During this evaluation, LP533401 was used as a reference [14]. The pro compound 18i were synthesized to improve its oral absorption ( Table 3). The este pound 18i, compound 19, exhibited nanomolar inhibitory activity, with an IC50 994 nM. In addition, the hippurate salt (21) was synthesized as an oral drug fo study. Based on the in vitro data, compound 18i was selected for inhibiting the in v permeability. Several in vitro methods and computational models have been used 79% 21 because of its increased potency, as evidenced by its IC50 value of 37 nM. It was assumed that adding an amine group to the thienopyrimidine ring would likely different potency profile to compound 18i. However, compound 30 disproved sumption because it showed reduced potency, as evidenced by the IC50 value of ( Table 3). During this evaluation, LP533401 was used as a reference [14]. The pro compound 18i were synthesized to improve its oral absorption ( Table 3). The ester pound 18i, compound 19, exhibited nanomolar inhibitory activity, with an IC50 994 nM. In addition, the hippurate salt (21) was synthesized as an oral drug for study. Based on the in vitro data, compound 18i was selected for inhibiting the in v permeability. Several in vitro methods and computational models have been used 92% Based on the in vitro data, compound 18i was selected for inhibiting the in vitro BBB permeability. Several in vitro methods and computational models have been used in drug discovery to predict the potential of a drug to penetrate the BBB [15][16][17][18][19][20][21]. The parallel artificial membrane permeation assay (PAMPA) was used to evaluate the BBB permeability of the selected compounds. The PAMPA-BBB assay is advantageous because it predicts passive penetration at the blood-brain barrier with high success, high throughput, low cost, and high reproducibility [21]. Table 4 shows that compound 18i has non-BBB permeable characteristics, with Pe≈ 0.00 × 10 −6 cm/s (i.e., Pe < 2.0), whereas pCPA exhibited a higher Pe of 1.85 × 10 −6 cm/s. Therefore, compound 18i was chosen as a prototype for further investigation. Compound 18i showed good liver microsomal stability (99% of the parental microsomes remained after incubation for 30 min). Table 5 shows that compound 18i did not significantly inhibit any CYP isoforms (1A2, 2C19, 2D6, and 3A4). cost, and high reproducibility [21]. Table 4 shows that compound 18i has non-BBB permeable characteristics, w 0.00 × 10 −6 cm/s (i.e., Pe < 2.0), whereas pCPA exhibited a higher Pe of 1.85 × 10 Therefore, compound 18i was chosen as a prototype for further investigation. Com 18i showed good liver microsomal stability (99% of the parental microsomes re after incubation for 30 min). Table 5 shows that compound 18i did not significantly any CYP isoforms (1A2, 2C19, 2D6, and 3A4).  Compound 18i was evaluated for its BBB ratio by iv injection and PK profil as shown in Table 6. The levels of compound 18i in the plasma (2266 and 1439 indicate that it did not penetrate the BBB. Furthermore, the level of compound 18 plasma was below the quantification limit (7 ng/mL). The PK profiles of 18i w evaluated which revealed that compound 18i had a reasonable AUC of 1.2 μg h/m ever, 18i itself could not be absorbed into the body due to its low permeability. Th 19 and 21 were introduced as the ethyl ester and hippurate salt prodrug, respe affording improved oral exposure, with 16% oral bioavailability. cost, and high reproducibility [21]. Table 4 shows that compound 18i has non-BBB permeable characteris 0.00 × 10 −6 cm/s (i.e., Pe < 2.0), whereas pCPA exhibited a higher Pe of 1.8 Therefore, compound 18i was chosen as a prototype for further investigatio 18i showed good liver microsomal stability (99% of the parental microso after incubation for 30 min). Table 5 shows that compound 18i did not signif any CYP isoforms (1A2, 2C19, 2D6, and 3A4).  Compound 18i was evaluated for its BBB ratio by iv injection and PK as shown in Table 6. The levels of compound 18i in the plasma (2266 and indicate that it did not penetrate the BBB. Furthermore, the level of compo plasma was below the quantification limit (7 ng/mL). The PK profiles of evaluated which revealed that compound 18i had a reasonable AUC of 1.2 μ ever, 18i itself could not be absorbed into the body due to its low permeabil 19 and 21 were introduced as the ethyl ester and hippurate salt prodrug affording improved oral exposure, with 16% oral bioavailability.  Compound 18i was evaluated for its BBB ratio by iv injection and PK profile in rats as shown in Table 6. The levels of compound 18i in the plasma (2266 and 1439 ng/mL) indicate that it did not penetrate the BBB. Furthermore, the level of compound 18i in the plasma was below the quantification limit (7 ng/mL). The PK profiles of 18i were also evaluated which revealed that compound 18i had a reasonable AUC of 1.2 µg h/mL; however, 18i itself could not be absorbed into the body due to its low permeability. Therefore, 19 and 21 were introduced as the ethyl ester and hippurate salt prodrug, respectively, affording improved oral exposure, with 16% oral bioavailability. Table 6. BBB ratio and pharmacokinetic parameters of 18i in male rats. Recent reports have shown that TPH inhibitors, such as pCPA and LP533401, protect against diet-induced obesity in vivo by reducing body weight gain and lipogenesis in adipose tissue in high-fat diet (HFD) fed mice [7,22]. Based on the pharmacological profiles of newly developed TPH1 inhibitors, we selected compound 18i and its prodrug compound 19 and 21 for further efficacy tests in HFD-induced obesity mouse models. First, we tested the effects of compound 18i in vitro by treating 3T3-L1 adipocyte differentiation and found decreased expression of key lipogenic genes, Fasn and Srebp1c, along with the adipogenic transcription factors, Pparg and Cebpa ( Figure 3A), indicating an anti-adipogenic effect for compound 18i during adipocyte differentiation. To assess in vivo efficacy, we performed daily intraperitoneal injections of compound 18i to HFD-fed C57BL6 mice. We found that compound 18i treatment, as compared to vehicle, resulted in blunted weight gain accompanied by a significant reduction in epididymal white adipose tissue (eWAT) and brown adipose tissue (BAT) adipocyte size as assessed by histological examination ( Figure 3B). Similar to the effects of compound 18i, compound 19 treatment attenuated body weight gain in HFD-fed mice compared to vehicle ( Figure 4A). In addition to decreased adiposity, HFD-fed mice treated with compound 18i had lower fasting blood glucose levels compared to vehicle-treated mice ( Figure 4B). Histological analysis of adipose tissues from compound 19-treated mice showed decreased adipocyte size in both eWAT and inguinal white adipose tissue (iWAT) ( Figure 4C). iWAT from compound 19 treated mice had a higher content of mitochondrial uncoupling protein 1 (UCP1), a marker for thermogenic beige adipose tissue, as assessed by immunohistochemistry ( Figure 4D). Notably, oral administration of compound 21 to HFD-fed mice dramatically decreased lipid accumulation in liver tissue compared to vehicle-treated mice ( Figure 4E), as assessed by histological examination and oil red O staining, without significant changes in body and liver weights ( Figure 4F,G). This effect by compound 21 treatment is consistent with our finding that gut tissue-specific deletion of TPH1 in mice leads to a reduction in liver steatosis in HFD-fed mice without affecting body and liver weights (unpublished data). Taken together, these results demonstrate the in vivo efficacy of our newly developed TPH1 inhibitors for ameliorating obesity and fatty liver diseases.    Docking studies revealed that compound 18i had optimal interactions with the catalytic and other important residues of TPH1, with predicted binding free energy of −14.33 Docking studies revealed that compound 18i had optimal interactions with the catalytic and other important residues of TPH1, with predicted binding free energy of −14.33 kcal/mol, which effectively inhibited TPH1. The amino group in the phenylalanine moiety of compound 18i interacted with the catalytic residues, Arg257 and Thr265, and the TPH1 residues, Gly333 and Ser335 ( Figure 5). The phenyl ring of the phenylalanine moiety interacted with His272 and Phe318 of TPH1. The phenyl ring of the phenylalanine moiety and the thienopyrimidine ring were optimally oriented because of Pro268 of TPH1, which facilitated the interaction of the thienopyrimidine ring with the key residues, Phe241 and Glu317. Additional interactions were observed between the sulfur of the thienopyrimidine ring and the Phe313 residue of TPH1. Modification of the hydrophobic tail of compound 18i improved its interactions, thereby blocking the entrance of the active site of TPH1. To summarize, compound 18i adopts a compact conformation that is expected to inhibit the binding of substrate cofactor and interferes with the binding of iron with TPH1, thereby leading to the effective inhibition of TPH1. kcal/mol, which effectively inhibited TPH1. The amino group in the phenylalanine moiety of compound 18i interacted with the catalytic residues, Arg257 and Thr265, and the TPH1 residues, Gly333 and Ser335 ( Figure 5). The phenyl ring of the phenylalanine moiety interacted with His272 and Phe318 of TPH1. The phenyl ring of the phenylalanine moiety and the thienopyrimidine ring were optimally oriented because of Pro268 of TPH1, which facilitated the interaction of the thienopyrimidine ring with the key residues, Phe241 and Glu317. Additional interactions were observed between the sulfur of the thienopyrimidine ring and the Phe313 residue of TPH1. Modification of the hydrophobic tail of compound 18i improved its interactions, thereby blocking the entrance of the active site of TPH1. To summarize, compound 18i adopts a compact conformation that is expected to inhibit the binding of substrate cofactor and interferes with the binding of iron with TPH1, thereby leading to the effective inhibition of TPH1.

Chemistry
All reported yields are isolated yields after column chromatography or crystallization. All solvents and chemicals were used as purchased without further purification. 1 H NMR spectra and 13 C spectra were recorded on a JEOL JNM-ECS400 spectrometer at 400 MHz for 1 H NMR and 100 MHz for 13 C NMR, respectively. The chemical shift (δ) is expressed in ppm relative to tetramethylsilane (TMS) as an internal standard, and CDCl3, DMSO-d6, and CD3ODwere used as solvents. The multiplicity of peaks is expressed as s (singlet), d (doublet), t (triplet), q (quartet), dd (doublet of doublets), td (triplet of doublets), qd (quartet of doublets), dt (doublet of triplets), and m (multiplet). Melting points were determined on a Melting Point M-560, purchased from Buchi. Fast atom bombardment-high-resolution mass spectrometry (FAB-HRMS) data were obtained by a JMS 700 (JEOL, Japan). The purity of all tested compounds was ≥ 95%, as estimated by high performance liquid chromatography (HPLC) analysis. Samples were analyzed on a Waters Agilent HPLC system equipped with a PDA detector and a Waters SB-C18 column (1.8 μm, 2.1 × 50 mm 2 ). The mobile phase was used with buffer A (ultrapure H2O containing 0.1% TFA) and buffer B (chromatographic-grade CH3CN). The flow rate was 0.5 mL/min. The inset (left) shows the optimal binding of compound 18i at the active site of TPH-1 marked with the red box and the zoomed-in view (right) highlights the key interactions of hydrogen bonds, pi-sulfur, fluorine interactions, and hydrophobic interactions, which were depicted with green, yellow, cyan, and pink, respectively, according to their interaction type. The alkyl interactions were not shown for ease of presentation and clarity, which are explained in the main text. Figures were prepared with Discovery Studio Visualizer, Dassault Systèmes.

Chemistry
All reported yields are isolated yields after column chromatography or crystallization. All solvents and chemicals were used as purchased without further purification. 1 H NMR spectra and 13 C spectra were recorded on a JEOL JNM-ECS400 spectrometer at 400 MHz for 1 H NMR and 100 MHz for 13 C NMR, respectively. The chemical shift (δ) is expressed in ppm relative to tetramethylsilane (TMS) as an internal standard, and CDCl3, DMSO-d6, and CD3ODwere used as solvents. The multiplicity of peaks is expressed as s (singlet), d (doublet), t (triplet), q (quartet), dd (doublet of doublets), td (triplet of doublets), qd (quartet of doublets), dt (doublet of triplets), and m (multiplet). Melting points were determined on a Melting Point M-560, purchased from Buchi. Fast atom bombardment-high-resolution mass spectrometry (FAB-HRMS) data were obtained by a JMS 700 (JEOL, Japan). The purity of all tested compounds was ≥ 95%, as estimated by high performance liquid chromatography (HPLC) analysis. Samples were analyzed on a Waters Agilent HPLC system equipped with a PDA detector and a Waters SB-C18 column (1.8 µm, 2.1 × 50 mm 2 ). The mobile phase was used with buffer A (ultrapure H2O containing 0.1% TFA) and buffer B (chromatographic-grade CH3CN). The flow rate was 0.5 mL/min.

CYP Inhibition Assay
Five major screening systems: P450-Glo CYP1A2 Screening System (Catalog# V9770), P450-Glo CYP2C9 Screening System (Catalog# V9790), P450-Glo CYP2C19 (Catalog# V9880), P450-Glo CYP 2D6 (Catalog# V9890), and P450-Glo CYP3A4 (Catalog# V9920) were attained from Promega Corp. (Madison, WI). α-naphthoflavone, sulfaphenazole, amitriptyline, quinidine, and ketoconazole were identified as positive control inhibitors for CYP1A2, CYP2C9, CYP2C19, CYP2D6, and CYP3A4, respectively. The CYP enzyme and substrate were combined in a potassium phosphate (KPO4) buffer and the reaction is initiated by adding an NADPH regenerating system. The volume of the mixture (12.5 µL in a 96well plate) was combined with an equal volume of test compound solution and positive control inhibitor (12.5 µL added to make the volume 25 µL). The reaction of the NADPH regeneration system was initiated by adding a 2× concentration of NADPH solution. Then an equal volume of luciferin detection reagent was added. The plates were incubated at room temperature for 20 min and the luminescence signal was measured by a microplate reader device.

The hERG Inhibition Activity
The inhibition of binding affinity by small molecules was measured using the Predic-tor™ hERG FP kit (Thermo Fisher Scientific, Inc., Rockford, IL, USA). The procedure was carried out in accordance with the manufacturer's instructions. Briefly, tracer (Predictor™ hERG tracer red) was prepared with dilution in the binding buffer. The binding assay was conducted in 384 well black flat-bottom microplates (Corning Life Sciences, Lowell, MA, USA) after the serial addition of 5 µL of a 4 nM tracer, 5 µL of test compounds, and a 10 µL membrane fraction containing hERG channel protein. The mixtures were incubated at room temperature for 2 h. The fluorescence polarization (FP) was measured with an excitation filter of 530 nm and an emission filter of 585 nm using a microplate reader (Infinite M1000PRO; Tecan, Mannedorf, Switzerland).

Blood Brain Barriers PK
The biodistribution study in plasma and brain tissue was performed using eight-weekold male SD rats (Orient Bio Inc., Seongnam, Korea). Dosing vehicles were composed of DMSO, PEG400, and deionized water (5:40:55, v/v/v). 18i compound was administrated at a dose of 5 mg/kg via intravenous (I.V.). Plasma from the inferior vena cava and whole brain tissue were collected at 0.5 and 3 h after drug administration. Brain tissues were rinsed with saline and four volumes of acetronitrile containing disopyramide were added as an internal standard for LC-MS/MS analysis. The brain samples were homogenized using a sonicator (IKA Labortechnik T10 basic ULTRA-TURRAX, Staufen, Germany) for 10 s on the ice. The plasma samples were added to nine volumes of acetonitrile and vortexed for 5 min. The samples were centrifuged (15, The pharmacokinetic study of 18i was performed using eight-week-old male SD rats (Orient Bio Inc., Seongnam, Korea). 18i and its prodrug 21 (5 mg/kg, 2 mL/kg) were administered via intravenous and oral routes. Blood samples were collected at 0.033, 0.16, 0.5, 1, 2, 4, 6, 8, and 24 h for intravenous injection at 0.15, 0.5, 1, 2, 4, 6, 8, and 24 h for oral dosing after drug administration, and then immediately centrifuged at 10,000× g for 3 min. The plasma concentrations of 18i were determined by LC-MS/MS. The plasma concentration-time profiles and pharmacokientic parameters were analyzed by a noncompartmental method using the nonlinear least-squares regression program WinNonlin 5.3 (Pharsight, Mountain View, CA, USA).

LC-MS/MS Analysis
The LC-MS/MS analysis was performed on an Agilent 1200 series (Agilent Technologies, Santa Clara, CA, USA) with an API 4000 linear ion trap triple quadruple mass spectrometer (AB Sciex, CA, USA). The chromatographic separation of the drug was carried out using a kinetex column (100 × 2.1 mm, 3 µm, Phenomenex, CA USA) with a SecurityGuard C18 guard column (4 mm × 2.0 mm i.d., Phenomenex). The flow rate was 300 µL/min with a mobile phase consisting of 0.1% formic acid in water (eluent A) and 0.1% formic acid in acetonitrile (eluent B) by linear gradient condition as follows: starting with 0 to 0.3 min, B: 5%; 0.3 to 2.5 min, B: 5 → 95%; 2.5 to 2.9 min, B: 95%; 2.9 to 3.0 min, B: 95 → 5%, and then re-equilibrated to initial conditions until 3 min. The total run time was 6 min and the injection volume of 5 µL.

In Vitro Screening Test
In vitro screening was performed as described previously [PMID: 33417443]. Briefly, TPH1 enzyme activity was measured with a commercially available kit (BPS Bioscience). Compounds were dissolved in DMSO (Sigma). Then, 10 µL of the compound was dispensed into a 96-well microplate, and 40 µL of the TPH1 enzyme solution (320 ng enzyme/reaction) was added. Then, 50 µL of the TPH1 reaction solution was added, and the microplate was sealed with aluminum foil. The microplate was immediately cooled to 4 • C, gently shaken, and incubated for 4 h. After a defined time, 10 µL of the TPH1 quench solution was added. Then, the fluorescence was measured with a Flexstation3 microplate reader at 300 nm for excitation and 360 nm for emission. All experiments were performed in triplicate.

In Vivo Efficacy Test
All animal study protocols were approved by the Institutional Animal Care and Use Committee at the Korea Advanced Institute of Science and Technology. Male C57BL/6J mice were purchased from SLC (Shizuoka, Japan) and acclimatized for a week. Mice were housed in a humidity and temperature-controlled environment under a 12 h light-dark cycle with ad libitum access to water and food. After acclimatization, male mice were fed a high-fat diet (HFD, 60% fat calories, Research diet; New Brunswick, NJ) and received daily treatment with TPH1 inhibitors. Intraperitoneal administration of PBS or 300 mg kg −1 pCPA (Sigma, St. Louis, MO, USA) was performed concomitant to the start of HFD feeding. Compound 18i was dissolved in 5% dimethyl sulfoxide (DMSO) (Sigma, St. Louis, MO, USA) and 25% β-cyclodextrin (Sigma, St. Louis, MO, USA). Vehicle or compound 18i (100 mg kg −1 ) was injected intraperitoneally in mice for 10 days after two weeks of HFD feeding. Both compounds 19 and 21 were prepared in 10% DMSO and 10% Kolliphor EL (Sigma, St. Louis, MO, USA). Vehicle and 100 mg kg −1 of compound 19 or compound 21 were administered daily by intraperitoneal injection or oral feeding, respectively, concomitant to the initiation of HFD feeding. For analyzing blood glucose concentration, glucose in blood collected from the tail vein was measured after fasting mice for 16 h using a Gluco Dr. Top glucometer (Allemedicus, Anyang, Korea). Body weight was measured daily and mice were sacrificed for harvesting adipose tissue and liver.

Histological Analysis and Immunohistochemistry
Tissue samples were fixed in 10% formalin, embedded in paraffin, and cut into 5 µm thick sections, followed by deparaffinization and rehydration for H&E staining and UCP1 immunohistochemistry. After heat-induced antigen retrieval with a citrate buffer (pH 6.0) for 10 min at 95 • C, tissue sections were incubated with BLOXALL Blocking solution (Vector Laboratories, Burlingame, CA, USA) to block endogenous peroxidase, followed by incubation with 2% normal goat serum in phosphate-buffered saline (PBS) for 30 min at room temperature to block nonspecific binding. Sections were incubated overnight with the anti-UCP1 primary antibody diluted 1:200 in 2% normal goat serum in PBS (Abcam, Cambridge, UK) at 4 • C, followed by 30 min incubation with biotinylated anti-rabbit IgG secondary antibody diluted 1:400 in PBS for 30 min at room temperature. After incubation with Vectastain ABC-AP reagent (Vector Laboratories) for 30 min, peroxidase activity was visualized with 3,3 -diaminobenzidine substrates (Vector Laboratories). For measuring average adipocyte size, we analyzed three random fields of H&E stained micrographs per adipose tissue sample from two representative mice of each group using AdipoCount software [23].

Liver Oil-Red O Staining
Fresh liver tissues were flash frozen in O.C.T. compound (Sakura Finetek, Torrance, CA, USA) and 10 µm cryostat sections were collected on glass slides. Sections were allowed to air-dry for 1 h, rinsed with PBS for 5 min, and washed with distilled water for 5 min. After washing, sections were placed in 30% isopropanol for 5 min, 60% isopropanol for 5 min twice, and then stained in Oil Red O solution (0.5% Oil red O (Sigma, St. Louis, MO, USA) dissolved in 60% isopropanol) for 15 min. The stained sections were rinsed twice with distilled water, mounted, coverslipped, and documented using bright field microscopy.
3.6.13. RNA Isolation and Real-Time Quantitative PCR (RT-qPCR) Total RNA was extracted from 3T3-L1 cells using Trizol reagent (Invitrogen, Carlsbad, CA, USA), purified using a RNeasy Lipid Tissue Mini kit (Qiagen, Hilden, Germany), and an aliquot (2 µg) of total RNA was reversed transcribed using the RevertAidTM First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer's instructions. Real-time PCR was carried out using SYBR Green (Power SYBR Green PCR Master Mix, Thermo Fisher Scientific) with a ViiA ™ 7 Real-Time PCR system (Applied Biosystems, Waltham, MA, USA). Mouse Rplp0 gene expression was used as the internal control to normalize gene expression values. Primer sequence information for RT-qPCR is shown in Supplementary Materials.

Docking Studies
The experimentally determined structure of TPH1 was downloaded from Protein Data Bank. The choice of three-dimensional structure [PDB access code 3HF8] is based on its high resolution, extensive literature, and our experience with the previous studies. Chain A was retrieved and preprocessed in a Discovery Studio Visualizer with a standard protocol to add missing atoms and residues. The structures of the synthesized compounds were manually drawn in ChemDraw and their 3D conformations were obtained from the Chem3D with the energy minimization protocol. The 3D conformations of TPH1 and synthesized compounds were preprocessed into 'pdbqt' files with AutoDockTools [11]. The docking studies were carried out with Lamarckian genetic algorithm implemented in AutoDock4.2 for 200 runs and 270,000 grid points of 0.325 grid spacing that covered the entire active site of TPH1 [11]. The detailed parameterization was described in our previous work [13]. The intermolecular interactions between the synthesized compounds and TPH1 were investigated with Discovery Studio Visualizer.

Conclusions
A series of oxyphenylalanine and heterocyclic phenylalanine derivatives were identified as TPH1 inhibitors. Among these derivatives, fused heterocyclic derivatives were identified as potent TPH1 inhibitors. Compound 18i with an IC 50 value of 37 nM was the most active compound in vitro. Compound 18i showed good liver microsomal stability and did not significantly inhibit CYP and Herg. As a TPH1 inhibitor, this compound was able to interact with the peripheral system without penetrating the BBB. Compound 18i and its prodrugs, compounds 19 and 21, reduced body weight gain and decreased in vivo fat accumulation in mammals. Consequently, this compound as a therapeutic agent is a useful non-BBB permeable TPH1 inhibitor that can act on the peripheral system while preventing obesity and fatty liver disease.