Phenylalanine hydroxylase contributes to serotonin synthesis in mice

Abstract

Serotonin is an important signaling molecule in the periphery and in the brain. The hydroxylation of tryptophan is the first and rate-limiting step of its synthesis. In most vertebrates, two enzymes have been described to catalyze this step, tryptophan hydroxylase (TPH) 1 and 2, with expression localized to peripheral and neuronal cells, respectively. However, animals lacking both TPH isoforms still exhibit about 10% of normal serotonin levels in the blood demanding an additional source of the monoamine. In this study, we provide evidence by the gain and loss of function approaches in in vitro and in vivo systems, including stable-isotope tracing in mice, that phenylalanine hydroxylase (PAH) is a third TPH in mammals. PAH contributes to serotonin levels in the blood, and may be important as a local source of serotonin in organs in which no other TPHs are expressed, such as liver and kidney.

Keywords: phenylalanine hydroxylase; serotonin; tryptophan hydroxylase.

FIGURE 1

PAH is a candidate for serotonin synthesis in mammals. A, 5‐HT levels in blood, spleen, brain, and duodenum of wild‐type (WT), TPH1‐KO, and TPH1/TPH2‐double KO mice. **P < .01, ***P < .001, ****P < .0001, one‐way ANOVA with Bonferroni post hoc test. B, Multiple sequence alignment and derived phylogenetic tree showing stronger homology between TPH‐ and PAH families than to THs or either TPHs or PAHs

FIGURE 2

PAH is a tryptophan hydroxylating enzyme in vitro. A, In vitro 5‐HTP generation in lysates from HEK293 cells either untransfected or transfected with empty vector (pcDNA3.1) or vector expressing PAH. B, In vitro 5‐HTP production at baseline and following PAH‐inhibition by α‐MPA in liver lysates harvested from wild‐type (WT) animals. C, In vitro 5‐HTP generation in liver lysates harvested from WT, TPH1, TPH1/TPH2, and PAH‐deficient animals under standard chow or phenylalanine‐free (Phe‐free) diet. D, 5‐HTP accumulation in primary hepatocytes from WT, TPH1, and TPH1/PAH animals incubated with Trp for 2 hours. Micrographs represent cells immunofluorescently labeled with DAPI (blue) and anti‐5‐HTP antibodies (red). 5‐HTP can be detected in hepatocytes harvested from WT and TPH1 animals but not in hepatocytes harvested from TPH1/PAH animals. ***P < .001, ****P < .0001, one‐way ANOVA with Bonferroni post hoc test

FIGURE 3

Pharmacological inhibition or genetic ablation of PAH activity reduces blood 5‐HT levels in vivo. A, Blood 5‐HT levels in TPH1‐KO mice treated either with saline (vehicle) or PAH‐inhibitor α‐MPA. B, Blood 5‐HT levels in TPH1‐KO, PAH‐KO, and TPH1/PAH‐double KO animals either on a standard chow or a phenylalanine‐free (Phe‐free) diet. Data are presented as % of WT blood 5‐HT levels. **P < .01, ***P < .001, ****P < .0001, ns, not significant, one‐way ANOVA with Bonferroni post hoc test

FIGURE 4

Animals lacking TPH1 and PAH show no 5‐HT neo‐synthesis as detectable via stable‐isotope resolved mass spectrometry. A, Experimental setup. Mice were treated with ¹³C‐labeled Trp and their blood was evaluated for fully ¹³C‐labeled 5‐HT via LC‐MS/MS. B, ¹³C‐labeled Trp and c ¹³C‐labeled 5‐HT levels in the blood of WT, TPH1‐KO, TPH1/TPH2‐double KO, PAH‐KO, and TPH1/PAH‐double KO mice 4 hrs after ¹³C‐Trp treatment. PAH‐KO and TPH1/PAH‐double KO animals were kept on a Phe‐free diet. *P < .05, **P < .01, one‐way ANOVA with Bonferroni post hoc test

FIGURE 5

Human PAH crystal structure analysis. A superimposition of the crystal structures of the catalytic domains of human PAH (PDB ID 1mmk, ⁵⁷ ) and chicken TPH1 (PDB ID 3e2t, ⁵⁸ ) is shown. The cartoon model of the overall fold is displayed in light orange (PAH) and grey (TPH1), respectively. The PAH crystal structure is complexed with iron (Fe^II, orange sphere), the cofactor tetrahydrobiopterin (BH4), and the substrate analog 3‐(2‐thienyl)‐L‐alanine (THA) (orange stick models), TPH1 is bound to tryptophan (Trp, green stick model). The interior surface (light blue) of the active‐site cavity of PAH was calculated with the program Hollow and displayed with PyMOL Molecular Graphics System (Version 1.3, Schrödinger, LLC). The superimposition illustrates that the PAH active site is generally capable of Trp binding. However, in PAH, the efficiency of Trp conversion is diminished by the presence of the bulky side chain of a Trp residue (W326 in human PAH, orange stick model) that is positioned only 3 Ångstrom away from the substrate‐binding site and thus influences substrate binding both sterically and by hydrophobic interactions (red dashed line). In comparison, in TPH, a phenylalanine residue is conserved at this position (F314 in chicken TPH1, green stick model)