2016 Volume 39 Issue 4 Pages 570-577
The mechanism of serotonin 5-HT2 receptor subtype-stimulated DNA synthesis and proliferation was investigated in primary cultures of adult rat hepatocytes to elucidate the intracellular signal transduction pathways. DNA synthesis and proliferation were detected in hepatocyte parenchymal cells grown in serum-free, defined medium containing 5-HT (10−6 M) or the selective 5-HT2B receptor agonist BW723C86 (10−6 M). In addition, exogenous transforming growth factor (TGF)-α (1.0 ng/mL) significantly increased hepatocyte DNA synthesis and proliferation, which reached plateau after 4 h of culture. Use of blocking monoclonal antibodies demonstrated that TGF-α, but not insulin-like growth factor-I, was involved in hepatocyte proliferation mediated by 5-HT or BW723C86. TGF-α levels in the culture medium increased significantly versus baseline within 5 min in response to 5-HT (10−6 M) or BW723C86 (10−6 M), and the maximum TGF-α level (30 pg/mL) was reached 10 min after 5-HT or BW723C86 stimulation. Secretion of TGF-α into the culture medium was inhibited by addition of the selective phospholipase C (PLC) inhibitor, U-73122 (10−6 M), or somatostatin (10−7 M). These results indicate that the proliferative mechanism of action of 5-HT is mediated mainly through a 5-HT2B receptor/Gq/PLC-stimulated increase in autocrine secretion of TGF-α from primary cultured hepatocytes.
Under normal physiological conditions, the mature rat liver is quiescent and does not proliferate. However, the liver has a remarkable capacity to regenerate. For example, after 70% partial hepatectomy, the remaining hepatocytes proliferate to restore the liver to its original mass within 2 weeks.1) Among the many growth-promoting factors important for liver regeneration, platelet-derived serotonin (5-hydroxytryptamine; 5-HT) is directly involved in liver regeneration.2–5)
In addition to its role as a neurotransmitter, 5-HT also functions as a hormone with various peripheral effects. 5-HT mediates many types of biological responses such as platelet aggregation, vascular constriction, and cell proliferation and mediates a diverse array of responses by interacting with 5-HT receptor subtypes in mammals.6) Seven receptor classes including 14 subtypes of 5-HT receptors have been identified, reflecting the diversity of 5-HT actions. With the exception of the 5-HT3 receptors, which are ligand-gated ion channels, all known 5-HT receptors are G-protein-coupled receptors that are positively linked to phosphatidylinositol turnover or cAMP production and further linked to a variety of downstream pathways.7,8)
In a previous report, we demonstrated that 5-HT-induced hepatocyte DNA synthesis and proliferation are mediated through 5-HT2B receptors.9) Moreover, 5-HT2B receptor-mediated hepatocyte mitogenesis involves activation of the Gq/phospholipase C (PLC) pathway and the epidermal growth factor (EGF)/transforming growth factor (TGF)-α receptor tyrosine kinase (RTK)/phosphatidylinositol 3-kinase (PI3K)/extracellular signal-regulated kinase (ERK)2/mammalian target of rapamycin (mTOR) pathway. However, how these two signaling pathways are associated with each other remains unknown. We hypothesized that 5-HT2B receptor stimulation induces secretion of a putative primary mitogen via the Gq/PLC pathway in hepatocytes, followed by induction of DNA synthesis and proliferation that is mediated by the EGF/TGF-α RTK/PI3K/ERK2/mTOR pathway.9)
To test this model, demonstrating that treatment of primary cultured hepatocytes with 5-HT or the selective 5-HT2B receptor agonist BW723C8610) leads to an increase in secretion of a putative primary mitogen is important. If so, determining how autocrine secretion of this mitogen via 5-HT2B receptor activation is regulated is also important. Here we show that in primary cultured hepatocytes, 5-HT or BW723C86 induces autocrine secretion of TGF-α, which then activates EGF/TGF-α RTK/PI3K/ERK2/mTOR signaling to induce DNA synthesis and proliferation.
Male Wistar rats weighing 200–220 g were obtained from Tokyo Experimental Animal Co. (Tokyo, Japan). Adaptation to a light-, humidity-, and temperature-controlled room occurred over a minimum 3-d period prior to beginning the experiments. Rats were fed a standard diet and given tap water ad libitum. All animals used in this study were given humane care in compliance with the Guiding Principles for the Care and Use of Laboratory Animals approved by The Japanese Pharmacological Society and Josai University.
Hepatocyte Isolation and CultureRats were anesthetized by intraperitoneal injection of sodium pentobarbital (45 mg/kg). Hepatocytes were isolated from normal livers by the two-step in situ collagenase perfusion technique devised by Seglen to facilitate disaggregation of the adult rat liver.11) Viability as determined with Trypan blue exclusion was >97%.
Freshly isolated hepatocytes (3.3×104 cells/cm2) were plated onto 35-mm plastic culture dishes (Iwaki Glass Co., Tokyo, Japan) coated with collagen and incubated for 3 h in Williams’ medium E supplemented with 5% newborn calf serum, 0.1 nM dexamethasone, 100 U/mL penicillin, 100 µg/mL streptomycin, and 0.10 µg/mL aprotinin in 5% CO2 in air at 37°C to allow attachment.12) The medium was changed to Williams’ medium E without serum or dexamethasone, and 5-HT or BW723C86 with or without test substances was added.
Determination of DNA Synthesis and Cell ProliferationHepatocyte DNA synthesis was assessed by measuring [3H]-thymidine incorporation into acid-precipitable materials.12,13) Briefly, following 3-h attachment and washing with serum-free Williams’ medium E, cells were grown for another 4 h in medium containing 5-HT or selective 5-HT receptor subtype agonists. [3H]Thymidine (1.0 µCi/well) was added to cells for 2 h after addition of 5-HT or agonists. Data are expressed as dpm/h·mg cellular protein. Cellular protein was determined using a modified Lowry procedure with bovine serum albumin as a standard.14)
Nuclei rather than cells were counted to assess proliferation.12)
Analysis of Cell Cycle ProgressionAnalysis of cell cycle progression induced by 5-HT, BW723C86, and TGF-α was performed according to the method described previously.15) The basic method for quantitative measurement of the cell cycle phase has been described elsewhere.16)
Determination of TGF-α Concentrations Secreted into the MediumSome components of Williams’ medium E inhibit enzyme-linked immunosorbent assays (ELISA). Therefore, we measured TGF-α levels that were secreted into phosphate-buffered saline (PBS; pH 7.4; conditioned medium) in the presence of 5-HT (10−6 M) or BW723C86 (10−6 M) with or without test substances.17) Cells were allowed to attach for 3 h, washed three times with PBS, and preincubated for 5 min in PBS containing 1.0 mM CaCl2, 5.5 mM glucose, and 0.10 µg/mL aprotinin (pH 7.4). Hepatocytes were then stimulated with 5-HT (10−6 M) or BW723C86 (10−6 M) with or without test substances. Conditioned medium (50 µL) was collected at various times, and TGF-α levels were determined with an ELISA kit (Calbiochem, San Diego, CA, U.S.A.). Absorbance was measured at a wavelength of 490 nm using a micro-plate reader (Tecan Japan, Kanagawa, Japan). A standard curve was obtained in the linear range of 15.6–800 pmol/mL, with a minimum detectable limit of about 7.8 pmol/mL.
MaterialsThe following reagents were obtained from Sigma Chemical Co. (St. Louis, MO, U.S.A.): serotonin hydrochloride (5-HT), aphidicolin, dexamethasone, somatostatin, verapamil hydrochloride, aprotinin, and BW723C86. Selective antagonists for 5-HT2 receptor subtypes were obtained from the following chemical companies: ketanserin tartrate (Enzo Life Sciences, Farmingdale, NY, U.S.A.), LY272015 hydrochloride (Santa Cruz Biotechnology, Dallas, TX, U.S.A.), and SB242084 (Tocris Bioscience, Bristol, U.K.). Recombinant human TGF-α was obtained from Pepro Teck, Inc. (Rocky Hill, NJ, U.S.A.). Monoclonal antibodies against insulin-like growth factor-I (IGF-I) were obtained from Oncogene Research Products (San Diego, CA, U.S.A.). Monoclonal antibodies against TGF-α and BAPTA/AM [1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid tetrakis(acetoxymethyl ester)] were obtained from Santa Cruz Biotechnology. U-73122 (1-[6-[17β-3-methoxyestra-1,3,5(10)-trien-17-yl]amino]hexyl]-1H-pyrrol-2,5-dione), U-73343 (1-[6-[17β-3-methoxyestra-1,3,5(10)-trien-17-yl]amino]hexyl]-2,5-pyrrolidine-dione), GF109203X (2-[1-(3-dimethylaminopropyl)-1H-indol-3-yl]-3-(1H-indol-3-yl)maleimide), 2,4-dideoxyadenosine, H-89 (N-[2-(p-bromocinnamylamino)ethyl]-5-isoquinolinesulfonamide dihydrochloride), AG1478 (2-[4-morpholinyl]-8-phenyl-1(4H)-benzopyran-4-one), LY294002 (N-[3-chlorophenyl]-6,7-dimethoxy-4-quinazolinamine), and rapamycin were obtained from Enzo Life Sciences, Inc. PD98059 (2′-amino-3′-methoxyflavone) was obtained from Calbiochem-Behring (La Jolla, CA, U.S.A.). Williams’ medium E and newborn calf serum were purchased from Flow Laboratories (Irvine, Scotland). Collagenase (type II) was obtained from Worthington Biochemical Co. (Freehold, NJ, U.S.A.). [Methyl-3H]thymidine (20 Ci/mmol) was purchased from DuPont-New England Nuclear (Boston, MA, U.S.A.). All other reagents were of analytical grade.
Data Analysis and StatisticsData are expressed as the mean±standard error of the mean (S.E.M.). Group comparisons were made by ANOVA for unpaired data followed by post-hoc analysis using Dunnett’s multiple comparison test. Differences of p<0.05 were considered to be statistically significant.
At the time that the culture medium of adult rat hepatocytes was changed to serum-free medium (3 h after plating), 5-HT (10−6 M), a 5-HT2B receptor agonist BW723C86 (10−6 M),10) or TGF-α (1.0 ng/mL) was added to examine the effects of these compounds on DNA synthesis and proliferation (Fig. 1). DNA synthesis in cultured hepatocytes was observed in the presence of 5-HT or BW723C86 beginning around 2.5 h and peaking at about 3.5 h after addition of the compounds (Fig. 1A). Significantly higher numbers of nuclei were first seen at 3 h; numbers peaked at about 4 h and were maintained for an additional 17 h (Fig. 1B). These results were in agreement with our previous report.9,15) In addition, exogenously added TGF-α (1.0 ng/mL) also significantly increased time-dependent DNA synthesis and proliferation in primary cultures of adult rat hepatocytes (Figs. 1A, B).
Hepatocytes at a cell density of 3.3×104 cells/cm2 were plated and allowed to attach for 3 h (time 0). The medium was then quickly replaced with serum-free Williams’ medium E, and the cells were grown in the presence of 5-HT (10−6 M), BW723C86 (10−6 M), or TGF-α (1.0 ng/mL). DNA synthesis (A) and cell proliferation (B) were assessed over time. Results are expressed as the mean±S.E.M. of three separate experiments. * p<0.05, ** p<0.01 compared with control.
As shown in Fig. 2A, the percent of total hepatocyte nuclei in the G0/G1 phase of the cell cycle decreased gradually during the 3-h attachment period. The percent of total hepatocyte nuclei in the G0/G1 phase of the cell cycle decreased gradually during an additional 4 h of culture of control hepatocytes (medium alone). On the other hand, percentages decreased significantly when 5-HT (10−6 M), BW723C86 (10−6 M), or TGF-α (1.0 ng/mL) was added to the culture compared with the respective control (medium alone) at 2, 3 and 4 h after addition of 5-HT, BW723C86, or TGF-α. As shown in Fig. 2B, the percent of total hepatocyte nuclei in the S phase of the cell cycle did not change significantly during the 3-h attachment period. On the other hand, we found that the percent of total hepatocyte nuclei in the S phase of the cell cycle increased significantly when 5-HT (10−6 M), BW723C86 (10−6 M), or TGF-α (1.0 ng/mL) was added to the culture compared with the respective control at 1, 2, 3, and 4 h after addition of 5-HT, BW723C86, or TGF-α. Percentages were not affected during an additional 4 h of culture in the medium-only control.
Hepatocytes at a cell density of 3.3×104 cells/cm2 were plated (time −3) and allowed to attach for 3 h (time 0). After a 3-h attachment period, the medium was rapidly replaced with serum-free Williams’ medium E, and hepatocytes were cultured with 5-HT (10−6 M), BW723C86 (10−6 M), or TGF-α (1.0 ng/mL) for various durations. At each time point, isolated hepatocyte nuclei (105/mL) were subjected to cell cycle analysis, and the percentages of total hepatocyte nuclei in the G0/G1 phase (A) and the S phase (B) of the cell cycle were assessed. Arrows indicate the time of agonist addition. Results are expressed as the mean±S.E.M. of 3–4 separate experiments. * p<0.05, ** p<0.01 compared with hepatocyte nuclei before culture; # p<0.05, ## p<0.01 compared with the respective control.
To examine the dose-dependent effects of selective 5-HT2 receptor subtype antagonists on DNA synthesis and proliferation in primary cultured hepatocytes induced by 5-HT (10−6 M) or BW723C86 (10−6 M), we used the 5-HT2B receptor antagonist LY272015 (3×10−8 to 3×10−7 M).18) Mitogenesis in hepatocytes induced by 5-HT or BW723C86 was blocked by LY272015 in a dose-dependent manner after culturing for 4 h (Figs. 3A, B). In contrast, 5-HT- or BW723C86-induced hepatocyte DNA synthesis and proliferation were not affected by a 5-HT2A receptor antagonist ketanserin (10−10 to 3×10−6 M)19) or a 5-HT2C receptor antagonist SB242084 (10−10 to 3×10−6 M).20)
Hepatocytes at a cell density of 3.3×104 cells/cm2 were plated and cultured for 3 h. After changing the medium, the cultured hepatocytes were treated with increasing concentrations of ketanserin (10−10 M to 3×10−6 M), LY272015 (10−10 M to 3×10−6 M), or SB242084 (10−10 M to 3×10−6 M) in the presence of 5-HT (10−6 M) or BW723C86 (10−6 M) for 4 h. DNA synthesis (A) and cell proliferation (B) were assessed. Results are expressed as the mean±S.E.M. of three separate experiments. * p<0.05, ** p<0.01 compared with control.
To determine how the 5-HT2B receptor/Gq/PLC pathway and the EGF-RTK/PI3K/ERK2/mTOR pathways are associated with each other, we hypothesized that 5-HT2B receptor stimulation may induce secretion of putative primary mitogens (e.g., IGF-I, TGF-α) through the Gq/PLC pathway in cultured hepatocytes, followed by induction of hepatocyte mitogenesis through RTK/PI3K/ERK2/mTOR pathways. We tested our working hypothesis by using a monoclonal antibody against IGF-I (1–100 ng/mL) or TGF-α (1–100 ng/mL) and examined the effects of these antibodies on hepatocyte mitogenesis induced by 5-HT (10−6 M) or BW723C86 (10−6 M). The TGF-α monoclonal antibody (1–100 ng/mL) blocked the induction of DNA synthesis and proliferation by 5-HT (10−6 M) or BW723C86 (10−6 M) in hepatocytes (Figs. 4A, B), with IC50 values of 29 ng/mL (DNA synthesis) and 25 ng/mL (proliferation) after 4 h of culture. On the other hand, the monoclonal antibody against IGF-I (1–100 ng/mL) did not block these effects, suggesting that secretion of endogenous TGF-α, but not IGF-I, mediates DNA synthesis and proliferation. When used alone, the monoclonal antibodies did not significantly affect DNA synthesis or proliferation of hepatocytes after a 4-h culture (data not shown).
Hepatocytes at a cell density of 3.3×104 cells/cm2 were plated and cultured for 3 h. After attachment for 3 h (time 0), the medium was rapidly replaced with serum-free Williams’ medium E, and the hepatocytes were cultured with various concentrations of monoclonal antibodies against TGF-α (1–100 ng/mL) or IGF-I (1–100 ng/mL) in the presence of 5-HT (10−6 M) or BW723C86 (10−6 M) for an additional 4 h. Hepatocyte DNA synthesis (A) and proliferation (B) were determined as described in Materials and Methods. Data are expressed as the mean±S.E.M. of three separate experiments. * p<0.05, ** p<0.01 compared with the respective control.
To further test our hypothesis, we needed to show that 5-HT or BW723C86 induces secretion of TGF-α by primary hepatocytes. We used ELISA for TGF-α to determine the time course of TGF-α secretion into the culture medium after addition of 5-HT (10−6 M) or BW723C86 (10−6 M) for 1–30 min (Fig. 5). The TGF-α concentration (30 pg/mL) peaked 10 min after addition of 5-HT or BW723C86, and levels remained significantly increased for up to 30 min. A 5-HT2B receptor antagonist LY272015 (10−7 M) completely blocked the secretion of TGF-α that was induced by 5-HT or BW723C86. LY272015 (10−7 M) by itself did not significantly affect TGF-α concentrations in the culture medium (0–30 min) (data not shown).
Hepatocytes were isolated and cultured and then incubated for various lengths of time (0–30 min) in the presence of 5-HT (10−6 M) or BW723C86 (10−6 M) with or without LY272015 (10−7 M). TGF-α concentrations in 50 µL culture medium were assessed with ELISA at various times. The results are expressed as the mean±S.E.M. of three separate experiments. * p<0.05, ** p<0.01 compared with the respective control.
5-HT (ED50=7.2×10−8 M; Fig. 6A) or BW723C86 (ED50=1.1×10−7 M; Fig. 6B) induced increased levels of TGF-α in a dose-dependent manner for 10 min. TGF-α secretion induced by either compound was almost completely blocked by a PLC inhibitor U-73122 (10−6 M)21) or somatostatin (10−7 M), which is known to inhibit the secretion of growth hormone.22)
Hepatocytes were isolated, cultured, and allowed to attach for 3 h (time 0). Then, the medium was quickly replaced with phosphate-buffered conditioned medium, and 5-HT (10−6 M) or BW723C86 (10−6 M) with or without U-73122 (10−6 M) or somatostatin (10−7 M) was added for 10 min. TGF-α concentrations in 50 µL culture medium were assessed with ELISA. The results are expressed as the mean±S.E.M. of three separate experiments. * p<0.05, ** p<0.01 compared with the respective control.
Addition of a 5-HT2B receptor antagonist LY272015 indicated that 5-HT2B receptor signaling is involved in 5-HT- and BW723C86-induced secretion of TGF-α (Fig. 7). We further investigated the signaling pathway involved using the PLC inhibitor, U-73122,21) the membrane-permeable Ca2+ chelator, BAPTA/AM,23) and somatostatin. In addition, the effects of the protein kinase C (PKC) inhibitor, GF109203X,24) and the L-type Ca2+ channel blocker, verapamil, were also tested.
LY272015 (10−7 M) or U-73122 (10−6 M) blocked the increase in the TGF-α concentration mediated by 5-HT (10−6 M) or BW723C86 (10−6 M) (Figs. 7A, B). These results are consistent with the hypothesis that the 5-HT2B receptor/Gq/PLC pathway plays an important role in induction of TGF-α secretion by 5-HT or BW723C86 in cultured hepatocytes. As a control, the inactive structural analogue of U-73122, U-73343 (10−6 M), had no effect on TGF-α levels induced by 5-HT or BW723C86. Somatostatin is known to block the secretion of some gastrointestinal and pancreatic hormones by inhibiting adenylate cyclase as well as blocking of cytosolic Ca2+ increase.25) Treatment with somatostatin (10−7 M) for 10 min inhibited TGF-α release by hepatocytes that was induced by 5-HT or BW723C86. In addition, the increase in TGF-α levels induced by these compounds was also blocked by verapamil (10−6 M) and BAPTA/AM (10−7 M), indicating that intracellular Ca2+ plays a role in TGF-α secretion.
Hepatocytes were isolated, cultured, and incubated with 5-HT (10−6 M) or BW723C86 (10−6 M) for 10 min in the presence or absence of various inhibitors. Then, TGF-α concentrations in 50 µL culture medium were assessed with ELISA. The concentrations of the various inhibitors of signal transducers were: LY272015 (10−7 M), U-73122 (10−6 M), U-73343 (10−6 M), GF109203X (10−6 M), dideoxyadenosine (10−6 M), H-89 (10−6 M), BAPTA/AM (10−7 M), verapamil (10−6 M), somatostatin (10−7 M), AG1478 (10−6 M), LY294002 (3×10−7 M), PD98059 (10−6 M), and rapamycin (10 ng/mL). The results are expressed as the mean±S.E.M. of three separate experiments. ** p<0.01 compared with the respective control.
In contrast, the increase in TGF-α levels induced by 5-HT or BW723C86 was not affected by a PKC inhibitor GF109203X (10−6 M), an adenylate cyclase inhibitor dideoxyadenosine (10−6 M),26) or a PKA inhibitor H-89 (10−6 M)27) (Figs. 7A, B), indicating that PKC and adenylate cyclase/PKA signaling are unlikely to play a role in TGF-α secretion mediated by 5-HT or BW723C86.
We next tested the effects of inhibitors of signal transducers that modulate growth, including a RTK inhibitor AG1478 (10−6 M),28) a PI3K inhibitor LY294002 (3×10−7 M),29) a MEK inhibitor PD98059 (10−6 M),30) and a mTOR inhibitor rapamycin (10 ng/mL).31) In the presence of 5-HT (10−6 M) or BW723C86 (10−6 M), none of these compounds significantly changed the increase in TGF-α secretion mediated by 5-HT or BW723C86. When used alone, these signal transducer inhibitors did not affect basal concentrations of TGF-α (data not shown).
We showed that 5-HT and a 5-HT2B receptor agonist BW723C86 as well as TGF-α increased DNA synthesis and proliferation in a time-dependent manner in primary cultures of adult rat hepatocytes (Figs. 1A, B). In addition, in primary adult rat hepatocytes not treated with EGF and hepatocyte growth factor, 5-HT acts via the 5-HT2B receptor, but not the 5-HT2A or 5-HT2C receptors, to induce increased DNA synthesis and proliferation, as shown by treatment with 5-HT2 receptor subtype-specific agonists and antagonists (Figs. 1, 3).
On the other hand, the doubling time of primary cultured hepatocytes (approximately 4 h) seems to be very fast. Thus, we performed cell cycle analysis of primary cultures. As shown in Fig. 2, under our culture conditions (i.e., low cell density and low dexamethasone level), hepatocytes were able to progress from the G0 to the G1 phase in the absence of exogenous primary mitogens during the 3-h attachment period. These hepatocytes were arrested in the G1 phase, and further progression to the S phase was dependent on the presence of 5-HT, BW723C86, or TGF-α. In support of our results, a previous study showed that collagenase perfusion of the liver triggers the G0/G1 transition of quiescent normal rat hepatocytes.32) Progression through the G1 phase is essential for making hepatocytes competent to respond to these primary growth factors.33)
We previously showed that DNA synthesis and proliferation are induced by 5-HT in hepatocytes via 5-HT2B receptors.9) Moreover, 5-HT2B receptor-mediated hepatocyte mitogenesis involves activation of both the Gq/PLC pathway and the EGF/TGF-α RTK/PI3K/ERK2/mTOR pathway. However, how these signaling pathways are associated with each other remains unknown. Here, we tested the hypothesis that stimulation of autocrine secretion of a putative mitogen occurs via the 5-HT2B receptor/Gq/PLC/Ca2+ pathway in primary cultured hepatocytes. We also tested the idea that DNA synthesis and cell proliferation occur by stimulation of the downstream RTK/PI3K/ERK2/mTOR pathway. We considered TGF-α and IGF-I as possible primary growth factors because hepatocytes express both TGF-α and IGF-I mRNA, and these cells synthesize and store both factors.1) In support of this hypothesis, we previously showed in hepatocytes that DNA synthesis and proliferation are stimulated by prostaglandin EP1 and a prostacyclin IP receptor agonist via autocrine secretion of TGF-α, which then stimulates hepatocyte mitogenesis via the EGF/TGF-α RTK/ERK2/mTOR pathway.17,34)
Consistent with our hypothesis, DNA synthesis and proliferation in hepatocytes induced by 5-HT or BW723C86 were blocked by monoclonal antibodies recognizing TGF-α, but not antibodies recognizing IGF-I (Fig. 4). Thus, the cytokine TGF-α, which is stored in parenchymal hepatocytes, is secreted into the culture medium following 5-HT2B receptor stimulation. Although IGF-I is also stored in hepatocytes, IGF-I secretion appears not to be induced by 5-HT. We provided further evidence for this mechanism by demonstrating that a significant time- and dose-dependent increase in TGF-α was induced by 5-HT or BW723C86 (Figs. 5, 6). Consistent with our previous study, TGF-α induced growth primarily via RTK/PI3K/ERK/mTOR signaling.35) Thus, 5-HT or BW723C86 stimulates secretion of TGF-α, which then induces DNA synthesis and proliferation in primary adult rat hepatocyte cultures.
We next examined the mechanisms responsible for regulating the rapid increase in TGF-α concentrations after stimulation of 5-HT2B receptors. Secretion of TGF-α that was stimulated by 5-HT or BW723C86 was completely blocked by the PLC inhibitor, U-73122,21) but not by the inactive analog, U-73343 (Fig. 7). Our data suggest that the primary mechanism of hepatocyte DNA synthesis and proliferation induced by 5-HT2B receptor agonists involves functional 5-HT2B receptors expressed on hepatocytes and stimulation of PLC and mobilization of Ca2+ to induce TGF-α secretion (Figs. 1–3). The important role of intracellular Ca2+ was shown by the observation that via inhibition of TGF-α secretion, a L-type Ca2+ channel blocker verapamil, a cell-permeable Ca2+ chelator BAPTA/AM, and somatostatin each blocked DNA synthesis and proliferation in hepatocytes that were induced by 5-HT or BW723C869) (Fig. 7). These results suggest that somatostatin also blocks the release of TGF-α by inhibiting cytosolic Ca2+ increase.35) Conversely, PKC and adenylate cyclase/PKA signaling are unlikely to play a role in these processes because GF109203X (10−6 M), dideoxyadenosine (10−6 M), and H-89 (10−6 M) failed to block secretion of TGF-α in the presence of 5-HT or BW723C86 (Fig. 7).
In contrast, the growth-related signal transducer inhibitors, a RTK inhibitor AG1478, a PI3K inhibitor LY294002, a MEK inhibitor PD98059, and a mTOR inhibitor rapamycin, which block signal transducers that modulate cell growth, did not significantly affect the increase in TGF-α concentrations that were stimulated by 5-HT or BW723C86 (Fig. 7). Therefore, these inhibitors only block growth downstream of TGF-α (Figs. 4A, B). Taken together, our data suggest that 5-HT or a 5-HT2B receptor agonist BW723C86 promotes hepatocyte mitogenesis indirectly via stimulation of autocrine secretion of TGF-α, which occurs via the 5-HT2B receptor/Gq/PLC pathway.
We propose that stimulation of the serotonin 5-HT2B receptor subtype induces PLC activation, possibly via Gq protein, which increases intracellular Ca2+ levels, resulting in autocrine TGF-α secretion. The secreted TGF-α then induces DNA synthesis and proliferation in hepatocytes by activating a second pathway, the EGF/TGF-α RTK (p-175 kDa)/PI3K/ERK2/mTOR pathway. A schematic of the proposed mechanism for increasing hepatocyte DNA synthesis and proliferation via 5-HT2B receptors is shown in Fig. 8.
RTK: receptor tyrosine kinase, PLC: phospholipase C, PIP2: phosphatidylinositol 4,5-bisphosphate, DG: diacylglycerol, IP3: inositol 1,4,5-trisphosphate, PI3K: phosphoinositide 3-kinase, ERK2: extracellular signal-regulated kinase 2, mTOR: mammalian target of rapamycin, TGF-α: transforming growth factor-α.
In conclusion, our data indicate that 5-HT stimulates DNA synthesis and proliferation in primary cultures of adult rat hepatocytes by acting via autocrine secretion of TGF-α induced by the serotonin 5-HT2B receptor/Gq/PLC/Ca2+ pathway. 5-HT is one of the main molecules released by platelets. Therefore, liver regeneration in vivo may be due to hepatocyte mitogenesis that is triggered by 5-HT from platelets in response to liver damage or injury.
The authors declare no conflict of interest.
