The generation of Tph1-floxed mice, Htr2a-floxed mice, and adiponectin (Adipoq)-Cre mice have previously been reported [11-13]. C57BL/6 J mice were purchased from Charles River Japan (Yokohama, Japan). Tph1 FKO mice were crossed with uncoupled protein 1 (Ucp1)-luciferase transgenic mice. The mice were housed in climate-controlled, specific pathogen-free barrier facilities, under a 12-hour light-dark cycle, with chow and water provided ad libitum. Male mice (aged 8 weeks) were fed either a standard chow diet (SCD; 12% fat calories, Research Diets D10001) or an HFD (60% fat calories, Research Diets D12492). In case of the transgenic mice, we compared the data between KO mice and their wild type (WT) littermates. The experimental protocols for this study were approved by the Institutional Animal Care and Use Committee (LML 15-535) at the Korea Advanced Institute of Science and Technology. These experiments were performed unblinded.

The mice were housed individually in a 12-chamber, open-circuit Oxymax/CLAMS (Columbus Instruments Comprehensive Lab Animal Monitoring system) system and the metabolic rate was measured as previously described [11]. After one day of acclimation, each mouse was assessed for 72 hours in the fed state to assess the metabolic rates. The respiratory exchange ratio (RER=VCO₂/VO₂) and heat production (HP=3.185×VO₂+ 1.232×VCO₂) were calculated. Fat mass and lean body mass were measured using Minispec time-domain nuclear magnetic resonance analyzer (Bruker Optics, Billerica, MA, USA).

To perform the glucose tolerance test, the mice were fasted overnight. Then, 2 g/kg D-glucose in phosphate-buffered saline (PBS) was intraperitoneally injected into the mice. For the insulin tolerance test (ITT), 0.75 U/kg human insulin (Humulin R, Lilly, Indianapolis, IN, USA) was intraperitoneally injected into 6 hours fasted mice. Blood samples were collected from the tail vein and glucose concentrations were measured using a Gluco DR plus glucometer (Allmedicus, Anyang, Korea) as previously described [10].

Total RNA extractions from the mice adipose tissues were performed using TRIzol as previously described [14]. After TURBO DNase (Invitrogen, Waltham, MA, USA) treatment, 2 μg of total RNA was used to generate complementary DNA with Superscript III reverse transcriptase (Invitrogen) according to the manufacturer’s protocol. Quantitative real-time polymerase chain reaction (RT-PCR) was performed with Fast SYBR Green Master Mix (Applied Biosystems, Foster City, CA, USA) and a ViiA 7 Real-time PCR system (Applied Biosystems). Gene expressions were analyzed using the delta-delta Ct method as previously described [11]. The primer sequences are presented in Table 1.

Murine 3T3-L1 cells (American Type Culture Collection) were cultured in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal calf serum and 100 μg/mL penicillin/streptomycin in a humidified atmosphere of 5% CO₂ at 37°C. Cells were differentiated and analyzed as previously described [10]. Immunostaining was performed on differentiated day 8.

Inguinal, epididymal, and interscapular adipose tissues were harvested, fixed in 4% (w/v) paraformaldehyde in PBS and embedded in paraffin. Next, 5-μm-thick tissue sections were deparaffinized, rehydrated, and used for hematoxylin and eosin staining, immunohistochemistry and immunofluorescence [10].

We obtained Ucp1-luciferase mice from the laboratory of Dr. Shingo Kajimura (University of California, San Francisco) [15]. Briefly, 150 mg/kg of D-Luciferin potassium salt (GoldBio, St. Louis, MO, USA) was injected into WT and Tph1 FKO with Ucp1-luciferase fed SCD or HFD. Fifteen minutes post-injection, we detected luciferase activity using IVIS Lumina S5, using the Living Image software (PerkinElmer, Waltham, MA, USA) to setup (autoexposure), analyze, and organize data.

The correlations of BXD white adipose tissue (WAT; subcutaneous fat) Tph1 mRNA and phenotypes were analyzed by using database (http://www.genenetwork.org/). The comparison of Tph1 mRNA levels and total 2000 phenotypes correlated with Tph1 in the WAT expressed by the volcano plot. The X axis represents Spearman’s Rho, with positive values representing positive correlations and negative values representing negative correlations, and the y-axis representing inversed P values. The gray horizontal solid line on the y-axis are statistically significant with a P value of less than 0.5, while the vertical solid line on the x-axis has a blue or red color with only Rho values greater than 0.5 (right) or less than −0.5 (left). Expressed in color can be viewed as having a positive or negative correlation. Marked in red is the metabolic phenotype. The result was obtained without bias using the whole without excluding one value from the entire BXD data.

All values are expressed as the mean±standard error of the mean. Statistical significance was determined by Student’s t test or two-way analysis of variance (ANOVA) with Bonferroni’s post hoc test. Two-way ANOVA was performed in body weight (Figs. 1B, 4B), glucose tolerance test (Fig. 1F, Supplemental Fig. S3B), ITT (Supplemental Figs. S2A, S3C), and energy expenditure assays (Fig. 2A, Supplemental Figs. S1E, S2B). P<0.05 was considered statistically significant.

To investigate whether serotonin is associated with body weight and composition, we analyzed RNA expression in WAT of the BXD family from GeneNetwork. Tph1 mRNA expression has shown a positive correlation with fat mass and glucose levels and a negative correlation with lean mass (Fig. 1A). We hypothesized that if serotonin increase by TPH1 in WAT has a positive correlation with increased fat mass, serotonin depletion in WAT may reduce fat mass and body weight gain. Previously, we generated inducible Tph1 knockout (adipocyte protein [aP2]-CreERT2^+/−/Tph1 ^flox/flox) mice and reported that serotonin depletion at the adult stage prevents HFD induced obesity [10]. In this study, we investigated the role of peripheral serotonin in mature adipocytes in the basal status as well as overnutrition status. For this, we generated adipocyte-specific Tph1 knockout (Adipoq-Cre^+/−/Tph1 ^flox/flox, Tph1 FKO) mice and analyzed the phenotypes of these mice at young and mature adult stages (Supplemental Fig. S1A, B). The Ap2 gene can be expressed in adipocyte progenitors and other tissues [16,17]. Notably, Adipoq gene expression is more specific in mature adipocytes than the Ap2 gene expression [18]. In young adults (8 weeks of age), Tph1 FKO mice did not show significant difference in body weight compared to WT mice (Fig. 1B). However, inn mature adults (20 weeks of age), Tph1 FKO mice demonstrated lower body weight and fat mass compared to the WT group when fed a HFD (Fig. 1B, Supplemental Fig. S1C, D). These results are consistent with the correlation analysis of the BXD reference group (Fig. 1A).

To explain the result, we measured the energy expenditure in these mice groups. As expected, Tph1 FKO mice with SCD feeding showed increased energy expenditure (increase VO₂, VCO₂, and HP) compared to WT mice fed SCD (Supplemental Fig. S1E). Under HFD conditions, after 12 weeks on HFD, these differences observed between Tph1 FKO and WT mice are greater than those observed under SCD conditions. Tph1 FKO fed the HFD showed reduced weight gain and lower body fat mass compared to WT mice (Fig. 1B-D). Notably, Tph1 FKO mice fed the HFD reported an improvement in glucose tolerance and insulin resistance when compared to WT mice fed the HFD (Fig. 1F, Supplemental Fig. S2A).

Tph1 FKO mice fed the HFD demonstrated increased energy expenditure (Fig. 2A, Supplemental Fig. S2B) compared to WT. To elucidate this phenotype, we generated in vivo reporter system for brown adipose tissue (BAT) activity using Ucp1-luciferase reporter mice [15]. Fig. 2B shows increased luciferase activity in the BAT area of Tph1 FKO and WT mice. Tph1 FKO mice fed the SCD and HFD showed significantly increased luciferase activity compared to WT mice fed the SCD and HFD (Fig. 2C). Then, we analyzed the histologic changes in BAT by aging. At 8 weeks of age, brown adipocytes in Tph1 FKO mice and WT mice appeared identical (Fig. 2D). This is consistent with their body weight at 8 weeks of age (Fig. 1B). As the mice grew older, from 8 to 20 weeks of age, the size of adipose cells and lipid droplets increased. However, Tph1 FKO BAT maintained a similar size of adipose cells at 20 weeks of age, even after 12 weeks of HFD (Fig. 2D).

Previously, we reported that serotonin depletion in the subcutaneous adipose tissue induced beige adipocyte formation in mice fed the HFD [10]. We observed similar changes after depleting serotonin in mature subcutaneous adipose tissue after HFD (Fig. 3A). Interestingly, serotonin depletion also increased beige adipocyte formation in the subcutaneous adipose tissue of mice fed the SCD (Fig. 3B). In visceral adipose tissue, Tph1 FKO mice fed HFD demonstrated a reduced adipose cell size and lipid droplets compared to WT mice fed on the HFD (Fig. 3C, D). Under the SCD conditions, Tph1 FKO mice did not indicate significant differences in the visceral adipose tissue compared to WT mice (Fig. 3C).

Previously, we identified the role of HTR2A in 3T3-L1 adipocytes [10]. In 3T3-L1 cell lines, HTR2A agonist increased lipogenesis gene expression in differentiated 3T3-L1 adipocytes [10]. Using these in vitro data, we suggested that HTR2A may regulate lipogenesis in WAT. To evaluate this hypothesis, we performed both in vitro and in vivo studies. We observed the lipid accumulation process in the in vitro system. We stained differentiated 3T3-L1 adipocytes with lipid (green, BODIPY 493/503) and serotonin (red, 5-HT). Fig. 4A shows the presence of serotonin and lipid droplets in differentiated 3T3-L1 adipocytes. Interestingly, we observed 5-HT and BODIPY copositive adipocytes (white arrowhead, Fig. 4A). The yellow staining implied that serotonin is related to lipid accumulation. Accordingly, the lipogenic gene and Htr2a gene expressions, in 3T3-L1 adipocytes, were increased during differentiation (Supplemental Fig. S3A).

Next, we generated adipocyte-specific Htr2a knockout (Adipoq-Cre^+/−/Htr2a ^flox/flox, Htr2a FKO) mice. Htr2a FKO mice fed the HFD demonstrated a lower body weight gain and improved glucose tolerance compared to WT mice fed the HFD (Fig. 4B, Supplemental Fig. S3B, C). As expected, the adipose cell size and lipid droplets were decreased not only in the visceral adipose tissue but also in subcutaneous and BAT of Htr2a FKO mice (Fig. 4C, D, Supplemental Fig. S3D, E). The gene expression analysis also showed reduced lipogenic gene expression in Htr2a FKO mice (Fig. 4E). These in vivo data strongly suggested that HTR2A in WAT regulates lipogenesis.