Serotonin 5-HT2A receptor activity mediates adipocyte differentiation through control of adipogenic gene expression

Serotonin 5-HT2 receptors are expressed in many tissues and play important roles in biological processes. Although the 5-HT2A receptor is primarily known for its role in central nervous system, it is also expressed in peripheral tissues. We have found that 5-HT2A receptor antagonists inhibit human subcutaneous primary adipocyte differentiation. We also show that siRNA knockdown of the 5-HT2A receptor blocks differentiation. Using gene expression analysis in combination with receptor antagonists we found that activity of 5-HT2A receptors is necessary very early in the differentiation process to mediate expression of adipogenic genes, including peroxisome proliferator-activated receptor gamma (ppar-γ), adipocyte protein 2 (aP2), adiponectin, and serine/threonine-protein kinase 1 (sgk1). We show here for the first time that 5-HT2A receptor activity is necessary for differentiation of human primary subcutaneous preadipocytes to adipocytes, and that 5-HT2A receptor activity mediates key genes related to adipogenesis during this process. Importantly, this work contributes to a greater understanding of the adipocyte differentiation process, as well as to the role of 5-HT2A receptors in peripheral tissues, and may be relevant to the development of novel therapeutic strategies targeting this receptor for the treatment of obesity related diseases.


Results
5HT 2 agonist/antagonist effects on day 2 of human adipocyte differentiation. In order to determine the effects of 5-HT 2 receptor activity in the early stages of adipocyte differentiation, 5-HT 2 receptor agonist or antagonist was added to the basal medium 24 h prior to the induction medium, and added fresh with each medium change at the same concentrations. As shown in Fig. 1A on day 2, the negative control cells that were in basal medium for the entire procedure did not demonstrate any morphological changes indicative of adipocyte differentiation, whereas positive control cells in the induction media had begun to show morphological changes. As shown in Fig. 1, negative cells were cells with a long narrow phenotype, indicating that they were pre-adipocyte and had not undergone differentiation; whereas positive cells were rounded up and had a changed morphological appearance similar to differentiated adipocytes. At this time point, induced cells treated with the 5-HT 2A receptor antagonist ketanserin (1.0 µM) are mostly still elongated and not rounded up (Fig. 1C), indicating inhibition of differentiation. Induced cells pretreated with the 5-HT 2 receptor agonist (R)-DOI (1.0 µM) were mostly rounded in appearance similar to the positive control cells (Fig. 1D).

5-HT 2A antagonist blocks terminal human adipocyte differentiation.
To examine the effect of 5-HT 2 receptor activity on terminal adipocyte differentiation, primary human subcutaneous pre-adipocytes were allowed to differentiate for the full 14 days and examined with oil red O staining to assess lipid accumulation. As shown in Fig. 2A, the negative undifferentiated control cells that were maintained in basal medium for 14 days retained an overall elongated morphology and demonstrated no oil red O staining, indicating the absence of lipid accumulation. Positive control cells that were treated with the differentiation media assumed a more rounded shape and accumulated lipids, as observed by staining with oil red O, indicative of successful differentiation into adipocytes (Fig. 2B). Cells treated with the differentiation protocol, plus the addition of ketanserin (1.0 µM) throughout the experiment, were similar in appearance to the undifferentiated controls, with an overall elongated shape, and no visible lipid accumulation (Fig. 2C). Induced cells treated with (R)-DOI throughout the process were identical in morphology to the positive control, and stained with oil red O equal to positive control levels (not shown), indicating that 5-HT 2A receptor activation above levels induced by serotonin normally present in the media did not alter the differentiation process. Time course experiments demonstrated that addition of ketanserin was necessary at Day 0 to inhibit differentiation completely (data not shown).
Ketanserin dose-dependently blocks adipocyte differentiation. To examine the potency of ketanserin's effects on adipocyte differentiation, a dose-response experiment was performed. Increasing concentrations of ketanserin were added to the cells 24 h prior to, and were maintained throughout the differentiation protocol, with terminal differentiation quantified by oil red O staining. The IC 50 value of ketanserin to inhibit human adipocyte differentiation and lipid accumulation was determined to be 86 nM (Fig. 3). Although ketanserin is not selective for 5-HT 2A receptors in rodent tissues, in humans it is considered a selective 5-HT 2A receptor antagonist with more than 100× fold selectivity for 5-HT 2A over 5-HT 2C and 5-HT 2B . Further, we did not detect 5-HT 2C receptor mRNA expression in these cells, as discussed below.  (Fig. 3). As anticipated, based upon negative QPCR expression data, the 5-HT 2C receptor antagonist RS102221 had no effect (data not shown). The high IC 50 for SB204741 (956 nM)   www.nature.com/scientificreports/ suggests that this may be an off target effect at 5-HT 2A receptors, but we can not rule out a contribution of 5-HT 2B receptors to the differentiation process (Fig. 3). These results together, as discussed later, demonstrate no involvement of 5-HT 2C receptors, and that activity at 5-HT 2A receptors likely mediates the differentiation process.    Expression levels of adipocyte related genes in human primary subcutaneous preadipocytes. Q-RT-PCR experiments were performed on RNA isolated from undifferentiated human primary subcutaneous preadipocytes to determine the presence of and initial levels of each of the genes we planned to examine in the differentiation process. In undifferentiated cells, we found that 5-HT 2A receptor mRNA was expressed at a higher level than 5-HT 2B receptor mRNA. Interestingly, mRNA for the serotonin transporter gene was detected, although it was quite low compared to 5-HT 2A receptor mRNA expression (Fig. 5). The presence of c/ebp-β, ppar-γ, and sgk1, was also detected at Day 0 ( Fig. 5). Although aP2 or adiponectin mRNA was not detected at Day 0, they were detected at later stages in the process, as noted below.

Expression of adipogenic genes is dynamic throughout the differentiation process, and is influenced by 5-HT 2A receptor blockade.
To elucidate molecular mechanisms mediating the influence of 5-HT 2A receptors, expression of different adipocyte markers was profiled during each major step of the differentiation process. As shown in Fig. 6, adipocyte related transcripts show dynamic patterns of expression throughout the differentiation process. C/EBP-β and PPAR-γ are key transcription factors that control the adipocyte differentiation process 32,35 . These factors up-regulate downstream genes such as those for adipocyte lipid binding protein (aP2), and adiponectin, which lead to triglyceride elevation, lipid droplet accumulation, and a fully differentiated adipocyte phenotype. The aP2 protein is a commonly used molecular marker of adipocyte differentiation 28 . We found a dramatic increase in aP2 expression in the positive induced control cells at day 2 that persisted through day 4, and ketanserin treatment blocked this expression increase (Fig. 5A).
For c/ebp-β, we observed an initial trend for increased expression both in the positive control as well as the ketanserin treated cells, compared with the negative control (Fig. 5B). Whereas c/ebp-β expression was maintained in the positive control cells throughout the process, both the negative control and the ketanserin treated cells showed parallel decreases to day 14.
For ppar-γ, there was an increase in expression in the positive control starting from day 2 that reached its peak at day 14 ( Fig. 5C). There was very low expression of ppar-γ in the negative control. Significantly, expression increases of ppar-γ were blocked in induced cells by ketanserin treatment (Fig. 5C). Adiponectin is an adipokine secreted by differentiated adipocytes that clinical studies suggest correlates with levels of oxidative stress 38 . Further, high levels of adiponectin are observed in patients with metabolic syndrome or cardiovascular disease 38,39 . Significantly, transcription of adiponectin can be directly activated by c/ebp-β 40 . We found increased expression Figure 5. Expression of adipogenic genes are dynamic throughout the differentiation process, and are influenced by 5-HT 2A receptor blockade. "Positive" represents expression from controls that were induced to differentiate. "Negative" represents controls that were not induced, and were maintained in basal medium for 14 days. "KET" represents induced cells that were treated with 300 nM of Ketanserin at each step of the adipocyte differentiation process. (*p < 0.01 vs. negative control; #p < 0.01 vs. ketanserin treatment group; n = 3 biological replicates per treatment, with 3 technical replicates per biological replicate per gene expression data point; analysis of variance (ANOVA) with Tukey's post hoc test). Ketanserin significantly blocks differentiationinduced expression of ap2 (A), ppar-γ (C), adiponectin (D), and sgk-1 (E), but does not significantly affect c/ ebp-β expression (B). (F) aP2, c/ebp-β, ppar-γ, adiponectin, HTR2A, HTR2B, SLC6A4 (serotonin transporter; SERT), and sgk-1 expression was determined in human primary subcutaneous preadipocytes by QPCR. Relative expression of these genes was analyzed using the HTR2A expression level as 100%. All expression experiments were performed in human primary subcutaneous preadipocyte cells at passage 5. www.nature.com/scientificreports/ of adiponectin in the positive control cells beginning on day 2, increasing through day 14, and ketanserin treatment blocked adiponectin expression (Fig. 5D). Sgk-1 activity is a key component of the adipocyte differentiation process. Inactivation of sgk-1 by siRNAmediated knockdown of expression in the 3T3-L1 model blocks adipocyte differentiation 41 . Interestingly, we previously found that the powerful psychedelic drug, LSD, up-regulates sgk-1 expression in rat brain cortex through activation of 5-HT 2A receptors 42 . In adipocyte differentiation, sgk-1 activity may be downstream of PPAR-γ activation 43 . Here, we have found a large increase in sgk-1 expression in the positive control cells starting at day 2 that reached its peak at day 14 (Fig. 5E). The increases of sgk-1 expression at day 14 also were blocked by ketanserin (Fig. 5E). Expression in the negative control was increased to a much greater extent than the positive control at day 2, and decreased somewhat through day 14 (Fig. 5E).
We examined levels of 5-HT 2A receptor protein over the differentiation process by western blot. We found that there was a low level of expression in undifferentiated preadipocytes, but that through the differentiation process, expression dramatically increased (Fig. 6A). These data were in agreement with the examination of 5-HT 2A receptor mRNA (HTR2A), where mRNA expression increased dramatically throughout the differentiation process (Fig. 6B). In the positive uninduced controls, HTR2A was not significantly changed until day 14, where it was increased only twofold. HTR2A in the ketanserin treated cells mirrored expression of the positive uninduced control until day 14, where the even twofold increase was blocked (Fig. 6B). In contrast to HTR2A, for 5-HT 2B (HTR2B) mRNA there was a small decrease in the negative control starting from day 2 to day 4, that returned to baseline levels at day 14 (Fig. 6C). In the positive differentiating control, there was a dramatic down-regulation of HTR2B expression (Fig. 6C). Ketanserin treatment had little effect on HTR2B expression when compared to the positive control (Fig. 6C). For the serotonin transporter SERT, there was dramatic downregulation under all conditions at day 2, which remained lower at days 4 and 14 for both the positive control and www.nature.com/scientificreports/ ketanserin treated cells (Fig. 6D). Interestingly, the negative control had a significant upregulation of expression at day 4 that returned to positive control levels by day 14 (Fig. 6D).

Discussion
Obesity is a leading worldwide public health concern that significantly affects not only individuals, but also their families 44 . It is now widely accepted that obesity contributes to a variety of diseases, including metabolic syndrome, diabetes, and atherosclerosis 45,46 . Increases in adipose tissue mass, as well as dysfunction of adipocyte differentiation, are tightly linked to obesity. Adipocyte differentiation is a highly regulated process, and understanding this process at a greater level of detail will lead to an enhanced understanding of adipocyte biology, and potentially to new therapeutic strategies to treat obesity and its related pathologies. Transcription factors such as C/EBP-β and PPAR-γ play very important roles in the regulation of adipocyte differentiation. During the differentiation of adipocytes, these transcription factors are activated, leading to increased expression of downstream factors that include aP2 and adiponectin 47 . Although a previous report 26 has investigated 5-HT 2A receptor activity in the differentiation of mouse 3T3-L1 fibroblast cells to an adipocyte-like phenotype, they did not address potential mechanisms, and found that 5-HT 2A receptor activation decreased adipogenic biomarkers. Another report also purported to investigate the role of 5-HT 2A and 5-HT 2C receptors in 3T3-L1 differentiation and concluded that the 5-HT 2C receptor was involved 36 . Interestingly, they effectively demonstrated a role for the micro RNA, miR-488, in the differentiation process of 3T3-L1 cells through the use of siRNA, and proposed its regulation of 5-HT 2C receptor expression as a primary contributing mechanism to differentiation 36 . Because we can not detect the presence of 5-HT 2C receptor mRNA in human preadipocytes, or block the differentiation process with 5-HT 2C selective antagonists at relevant levels, the role of serotonin in the differentiation process of mouse 3T3-L1 cells from a fibroblast phenotype to an adipose-like phenotype is different from human preadipocyte to adipocyte differentiation such that 5-HT 2A receptor activity is required in humans but not in mice for proper adipocyte differentiation. A more recent study using RNA-Seq identified a potential role for 5-HT 2A receptors in the differentiation of bovine adipocytes, and implicated the PI3K-AKT pathway in this process 48 .
To establish that 5-HT 2A receptor activity is essential for differentiation of primary human preadipocytes, we first used two 5-HT 2 receptor ligands at relevant concentrations to determine whether activation or blockade affected differentiation. Whereas the high affinity 5-HT 2 receptor selective agonist (R)-DOI had no effect, the 5-HT 2A receptor antagonist ketanserin blocked adipocyte differentiation dose dependently. We validated these results using a second 5-HT 2A receptor selective antagonist, M100907. Although the 5-HT 2B receptor selective antagonist SB204741 blocked differentiation, the high concentration necessary to inhibit differentiation supports the notion that the effects of ketanserin, M100907, and SB204741 are all likely due to blockade of 5-HT 2A receptors, but we can not definitively rule out a contribution to the differentiation process by 5-HT 2B receptors. To validate our pharmacological studies, we used siRNA directed against 5-HT 2A receptor mRNA to knock down expression. Significantly, knockdown of 5-HT 2A receptor mRNA resulted in complete blockade of differentiation. Together, both pharmacological and genetic methods demonstrate that 5-HT 2A receptor activity is necessary for human adipocyte differentiation.
That serotonin receptor antagonists can block the differentiation process suggests that the process relies upon either serotonin present in the cell culture media, or in vivo, on circulating serotonin. The affinity of serotonin for the human 5-HT 2A receptor is higher than for rodent 5-HT 2A receptors (human: ~ 125 nM; rat ~ 1.1 µM; PDSP database). Typical cell culture media containing 10% FBS has a serotonin content of ~ 350 nM 49,50 . Therefore, there is likely enough serotonin present in the media to facilitate activation of the receptors and differentiation under normal culturing conditions. Interestingly, our gene expression results indicate the presence of serotonin transporter mRNA in adipose tissue, suggesting that adipocytes may themselves produce serotonin. Furthermore, adipose tissue has been shown to contain considerable amounts of serotonin 51 . It remains to be seen if putative 'endogenous' serotonin plays a significant role in adipocyte differentiation or function as it does in other peripheral cells like smooth muscle cells that synthesize their own serotonin and express SERT 14 .
How might 5-HT 2A receptor activity be required for the process of adipocyte differentiation? To answer this question we chose to examine the expression of genes previously implicated in adipocyte differentiation and function: C/EBP-β, aP2, PPARγ, adiponectin, and sgk-1 52,53 , and assess whether or not 5-HT 2A receptor activity influenced their expression. We examined gene expression at key time points throughout differentiation: Baseline (day 0), induction (day 2), differentiation (day 4), and maintenance/terminal differentiation (day 14). Expressions of all genes were dynamic throughout the process, and depending on when or if gene expression changes were affected by ketanserin, it might give clues to genes and pathways regulated by 5-HT 2A receptor activity during the differentiation process.
For expression of CEBP-β we found an initial trend for increased expression both in the positive control as well as the ketanserin treated cells, compared with the negative control. Whereas CEBP-β expression was maintained in the positive induced control throughout the differentiation process, for both the negative control and the ketanserin treated cells CEBP-β expression paralleled each other for a non-significant decrease at day 14. The absence of a ketanserin effect on c/ebp-β expression is consistent with 5-HT 2A activity likely acting downstream of c/ebp-β activation and we did find that ketanserin treatment dramatically influenced expression of genes known to be downstream of C/EBP-β activity.
Ketanserin treatment dramatically reduced ppar-γ expression, and subsequent gene expression of aP2, adiponectin, and sgk-1. Furthermore, the presence of the ppar-γ agonist rosiglitazone in the media in the induction phase abolished the inhibitory effect of ketanserin on adipocyte differentiation (data not shown). Taken together, we conclude that the role of 5-HT 2A activity on adipocyte differentiation occurs downstream of C/EBP-β but upstream of PPAR-γ. www.nature.com/scientificreports/ Interestingly, we observed that mRNA levels of HTR2A were significantly increased in the differentiated adipocytes, and that mRNA levels for HTR2B dramatically decreased early in the process. One interpretation could be that the rapid decline in 5-HT 2B receptor expression and activity is facilitating differentiation. We do not believe that to be the case. Whereas ketanserin treatment blocks the expression increase in HTR2A, it has no effect on the decrease in HTR2B. If HTR2B expression changes were involved in mediating the effects of ketanserin, one would expect that ketanserin treatment would block the observed decrease.
We also observed that for some genes, including those for the serotonin receptors and transporter, as well as sgk-1, expression increased dramatically in negative control cells that were not induced. The differentiation process is initiated by the addition of certain chemicals to the media when the plates of cells just become confluent. It may be that as the plates become confluent, and contact between cells becomes prevalent, the cells induce certain mechanisms to inhibit growth (i.e. contact inhibition, growth arrest) that may be altering expression of these genes.
In summary, our data indicate that 5-HT 2A receptor activity is necessary for adipocyte differentiation in primary subcutaneous human preadipocytes. The 5-HT 2A receptor-mediated effects on adipocyte differentiation are likely mediated through modulation of expression of key genes necessary for the differentiation process. Based upon our data, we propose a model for the involvement of 5-HT 2A receptor activity in adipocyte differentiation as shown in Fig. 7. Growth factors and second messengers lead to the production of C/EBP-β, which feeds into production of PPARγ and subsequent production of adipogenic factors including ap2, Adiponectin, and Sgk-1, whose action together leads to terminal differentiation of preadipose fibroblast cells to adipocytes. Activity of 5-HT 2A receptors is necessary for production and/or activity of PPARγ, downstream of C/EBP-β, to induce production of proadipogenic factors and terminal differentiation. One possibility for the precise role of 5-HT 2A receptors is that the receptor activates a particular effector pathway that recruits or activates, or relieves inhibition of, components of transcriptional processes activated by c/ebp-β necessary for subsequent transcription of ppar-γ. Importantly, our results contribute to a greater understanding of the adipocyte differentiation process, and particularly the role of 5-HT 2A receptors in peripheral tissues, and this information may be relevant to the development of novel therapeutic strategies targeting this receptor for the treatment of obesity related diseases. Cell culture. Primary human subcutaneous pre-adipocytes were purchased from Zen-bio, Inc (Research Triangle Park, NC, USA), and from two separate lots (#L120116E and #L050906). Individual experiments were performed with cells from only one lot, and not mixed. Each lot consists of preadipocytes from a single consenting Caucasian human female donor (non-smoker) with BMIs of 26.5 and 27.4, and ages of 52 and 46 years, respectively. As the source of human preadipocytes used in this study were from a narrow demographic, the possibility remains that preadipocytes isolated from different ages, sexes, BMIs, and races may have differential requirements for serotonin and 5-HT 2A receptor activity in their differentiation.

Reagents
Primary human preadipocytes utilized for our studies were at passage 3. All cells were grown in tissue culture treated 6-well culture dishes (Corning, Lowell MA). The following cell culture media were used for the differentiation assay: Basal Medium (BM) = DMEM + 10% FBS; Induction Medium (IM) = DMEM + 10% FBS + 0.5 mM 3-isobutyl-1-methylxanthine (IBMX) + 10 μg/mL insulin + 1 μM dexamethasone; Maintenance Medium (MM) = DMEM + 10% FBS + 10 μg/mL insulin. Cells were grown to 100% confluence in BM. Two days after full confluency, cells were transferred to IM for 2 days, followed by MM for 2 days. The cells were then Figure 7. Model of adipocyte differentiation. Growth factors and second messengers lead to the production of C/EBP-β, which feeds into production of PPARγ and subsequent production of adipogenic factors including ap2, Adiponectin, and Sgk-1, whose action together leads to terminal differentiation of preadipose fibroblast cells to adipocytes. Activity of 5-HT 2A receptors is necessary for production and/or activity of PPARγ, downstream of C/EBP-β, to induce production of proadipogenic factors and terminal differentiation.

Oil-red-O staining and microscopy.
To determine the status of adipocyte differentiation, cells were washed twice with phosphate-buffered saline (PBS, pH 7.4), then fixed with 10% formalin at room temperature for 1 h, and stained with 0.5% oil-red-O solution for 1 h. After staining, the cultures were rinsed several times with 70% ethanol and images were taken using a Motic Images Plus CCD microscope camera and software (Richmond, BC, Ca). In order to determine the extent of adipocyte differentiation, the dye-triglyceride complex was extracted with 100% isopropanol and absorbance measured at 510 nM in a spectrophotometer [54][55][56] .
RNA isolation and quantitative real-time polymerase chain reaction. RNA was extracted from cells using the Illustra RNAspin Mini kit from GE Healthcare Life Sciences (Piscataway, NJ, USA) following protocols supplied by the manufacturer. First-strand cDNA was generated using the ImProm-II cDNA synthesis kit (Promega, Madison, WI, USA) following the manufacturer's protocols using 300 ng total RNA. Quantitative real-time polymerase chain reaction (PCR) was performed using the ProbeLibrary system from Roche Diagnostics (Indianapolis, IN, USA) in combination with the HotStart-IT Probe qPCR Master Mix from USB (Cleveland, OH, USA) following the manufacturer's protocols. Primer sequences for each probe-based qPCR assay designed using Universal Probe Library Probe Finder software (Roche). Primer oligonucleotides were ordered from IDT (Coralville, IA). Corresponding probes were utilized from the Universal Probe Library (Roche). The sequences of primers and probes used for quantitative PCR are shown in Table 1. For the siRNA knockdown expression experiment, the IDT PrimeTime QPCR assay directed against the human HTR2A gene was used. For all quantitative determination of gene expression levels, reactions were performed using a two-step cycling protocol on a MyIQ-5 Cycler (Bio-Rad, Hercules CA, USA). Relative gene expression levels were calculated using the 2[− ΔΔC(T)] method 57 . Levels of all targets from the test samples were normalized to human GUSB expression.
Western blot. Cell lysates collected in triplicate at the indicated time points were normalized for protein content by bicinchoninic acid (BCA) assay, separated on 4-12% continuous gradient Bolt Bis-Tris Plus polyacrylamide gels, and transferred to nitrocellulose membranes. Blots were serially probed with primary antibodies to human 5-HT 2A receptor (Santa Cruz, sc166775; 1:500, overnight at 4C) and beta-actin (Sigma, A2228;  -GGA TGA TAA ACT GGT GGT GGA-3′  U85  Reverse  5′-CAC AGA ATG TTG TAG AGT TCA ATG C-3′   c/ebp-β  NM_005194  Forward 5′-CGC TTA CCT CGG CTA CCA -3′  U74  Reverse  5′-ACG AGG AGG ACG TGG AGA G -3′   ppar-γ  NM_138711  Forward 5′-GAC AGG AAA GAC AAC AGA CAA ATC -3′  U70  Reverse  5′-GGG GTG ATG TGT TTG AAC TTG- www.nature.com/scientificreports/ 1:5,000, 1 h at RT) followed by HRP conjugated secondary antibodies (1:50,000, 1 h at RT). Protein bands were visualized by chemiluminescence (WesternBright Sirius, Advansta) and captured on X-ray film. Densitometric analysis was performed using ImageJ. Individual bands, or an area of representative background, were encompassed within a defined area using the selection tool and mean gray values for each were determined. The background mean grey value was subtracted from all band measurements, and values between samples were normalized by ratioing the background adjusted 5-HT 2A receptor value relative to corresponding actin values. Normalized levels of 5-HT 2A receptor at each time point post-differentiation were expressed relative to undifferentiated controls. See Supplementary Figure S1 for scans of the representative western blot experiment whose data is shown in Fig. 6A.