SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

In addition to its function as a neurotransmitter and vascular active molecule, serotonin is also a mitogen for hepatocytes and promotes liver regeneration. A possible role in hepatocellular cancer has not yet been investigated. Human hepatocellular cancer cell lines Huh7 and HepG2 were used to assess the function of serotonin in these cell lines. Characteristics of autophagy were detected with transmission electron microscopy, immunoblots of microtubule-associated protein light chain 3(LC3) and p62 (sequestosome 1). Immunoblots of the mammalian target of rapamycin (mTOR) and its downstream targets p70S6K and 4E-BP1 were used to investigate signaling pathways of serotonin. Two different animal models served as principle of proof of in vitro findings. Clinical relevance of the experimental findings was evaluated with a tissue microarray from 168 patients with hepatocellular carcinoma. Serotonin promotes tumor growth and survival in starved hepatocellular carcinoma cells. During starvation hepatocellular carcinoma cells exhibited characteristics of autophagy, which disappeared in serotonin-treated cells. Rapamycin, an inhibitor of mTOR, is known to induce autophagy. Serotonin could override rapamycin by an mTOR-independent pathway and activate common downstream signals such as p70S6K and 4E-BP1. In two tumor models of the mouse, inhibition of serotonin signaling consistently impaired tumor growth. Human biopsies revealed expression of the serotonin receptor HTR2B, correlating with downstream signals, e.g., phosphorylated p70S6K and proliferation. Conclusion: This study provides evidence that serotonin is involved in tumor growth of hepatocellular cancer by activating downstream targets of mTOR, and therefore serotonin-related pathways might represent a new treatment strategy. (HEPATOLOGY 2010.)

Serotonin (5HT), a well-known neurotransmitter and vasoactive substance, also regulates a wide range of physiological actions in the gastrointestinal tract.1 5HT is a potent mitogen for many different cell types,2 including hepatocytes,3 and it is crucial for liver regeneration.4 On a cellular level 5HT acts predominately by way of G-protein-coupled receptors (GPCRs). Seven receptor classes including 14 subtypes of 5HT receptors (HTR) reflect the diversity of serotonergic actions.5 Among the numerous 5HT receptors, particularly the 5HT2B-receptor has been frequently described to mediate proliferation3, 4, 6–8 and to exhibit antiapoptotic properties.7, 9 Within the liver 5HT is emerging as a mediator of different pathological conditions. It contributes to liver fibrosis,7 mediates oxidative stress in nonalcoholic steatotic hepatitis,10 and aggravates viral hepatitis.11 All these conditions are involved in the tumorgenesis of hepatocellular carcinoma (HCC),12 the third cause of cancer-related death worldwide.13 Our group has shown that 5HT promotes tumor growth in a mouse model of subcutaneous colon cancer allografts. 5HT deficiency led to decreased vascularity and increased necrosis reflecting cell death of the tumor.14 Because of the rising role of 5HT in liver disease and tumor growth, on the one hand, and the role in liver regeneration on the other, we asked whether 5HT contributes to the biology of HCC.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Experimental Setup.

Human hepatocellular cell lines, Huh7 and HepG2, were seeded into 24-well plates at a density of ≈25% corresponding to 2.5 × 104 cells per well and allowed to adhere overnight before the medium was changed to the specified conditions, containing different concentrations of 5HT creatinine complex (Sigma Aldrich), 5HT agonists, or antagonists. Experiments with different concentrations of 5HT agonists and/or antagonists were performed after serum withdrawal for 48 hours, followed by 24-hour stimulation. Time-dependent experiments were performed with subsequent stimulation after serum withdrawal. In agonist/antagonist experiments cells were incubated with the antagonist for 20 minutes before addition of the corresponding agonist.

Cell Culture, Agonist, and Antagonists.

Cell lines were purchased from the American Type Culture Collection (Rockville, MD) and Huh7 and HepG2 cultured in Dulbecco's Minimal Essential Medium with 4.5 g/L glucose, sodium pyruvate, GlutaMAX (Invitrogen), and 10% fetal calf serum (PAA Laboratories), with the addition of 100 units/mL of penicillin and 100 μg/mL of streptomycin (Invitrogen). Cells were maintained at 37°C in a 5% CO2 atmosphere.

NAN-190 hydrobromide (NAN-190), SB204741 (SB204), SB269970, α-methyl-5-hydroxytryptamine maleate (α-Me-HTP), 5-carboxyamidotryptamine maleate (5-CT), rapamycin, and phorbol 12-myristate 13-acetate were purchased from Tocris. 2,5-Dimethoxy-4,5-iodoamphetamine hydrochloride (DOI), ketanserin, ritanserin and demecolcine were purchased from Sigma Aldrich. Stock solutions were prepared in H2O or dimethyl sulfoxide (DMSO) as appropriate and stored at −20°C. Depending on the used agonist/antagonist, 0.1% DMSO was added to the cell culture media in control groups.

Proliferation, Viability, and Apoptosis Assays.

After 48-hour serum deprivation, cells were pulsed with 0.1 μCi/well methyl-3H thymidine (100 Ci/mmol, Amersham Pharmacia Biotech), and incubated for 48 hours. Plates were harvested by rinsing in cold phosphate-buffered saline (PBS) followed by a 60-minute incubation in cold 10% trichloroacetic acid (TCA). Wells were washed with 10% TCA and 500 μL of 1N NaOH was added. Then 250 μL of this suspension was neutralized with 250 μL 1N HCl and added to 5 mL scintillation fluid. Measurement was performed in a liquid scintillation counter (Kontron Instruments).

5-Bromo-2′-deoxy-uridine (BrdU) was added after 48-hour serum withdrawal to a final concentration of 10 μM. Labeling and detection of the cell were performed with BrdU-immunofluorescence following the manufacturer's instructions (Roche Applied Science).

Viable cells were distinguished with the fluorescent dyes Calcein AM and Ethidium homodimer 1 (Live/Dead, Viability/Cytotoxicity Kit, Molecular Probes, L-3224). Stainings were performed according to the manufacturer's instructions.

The number of viable cells was quantified by the addition of 25 μL of a 0.5% tetrazolium salt solution [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT); Sigma Aldrich]. After 45 minutes of incubation, the formation of the formazan product was monitored by measuring absorbance at 570 nm after solubilization in acidic isopropranol (5% formic acid in isopropranol). Values were calculated as the percentage of untreated controls.

Cytotoxicity was assessed using CytoTox-Fluor-Assay (Promega, G9260) following the manufacturer's instruction. The test measures the relative number of dead cells in cell populations and uses a fluorogenic peptide substrate (bis-alanyl-alanyl-phenylalanyl-rhodamine 110; bis-AAF-R110) to measure “dead-cell protease activity,” which has been released from cells that have lost membrane integrity.

DNA fragmentation was visualized with fluorescein-dUTP [TUNEL (TdT-mediated dUTP-X nick end labeling)] using a Cell Death Detection Kit (Roche Applied Science).

Caspase-3 activity was measured using Caspase-Glo 3/7 Assay (Promega) as well as with a pan-caspase inhibitor (Z-DEVD-FMK) and caspase-3 substrate (Ac-DEVD-AFC) according to the manufacturer's instruction.

Western Blotting.

Preparation of cell extracts, sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) blotting, and secondary staining were performed according to standard protocols. Primary antibodies were phospho-mTOR(Ser2448) (#2971, Cell Signaling), phospho-p70S6K(Thr421/Ser424) (#2971, Cell Signaling), phospho-4E-BP1(Thr37/Ser46) (#2855, Cell Signaling), mTOR (#2983, Cell Signaling), p70S6K (#9202, Cell Signaling), 4E-BP1(#9644, Cell Signaling), LC3B (#2775, Cell Signaling), p62/SQSTM1 (M162-3; MBL), HTR2B (ab32944; Abcam), β-tubulin (ab6046; Abcam), GAPDH (ab9485; Abcam), and Ki67 (M7240, Dako). All results were measured by densitometry and displayed as relation of phosphorylated/nonphosphorylated protein and fold induction to untreated controls.

Patients.

A total of 168 carcinomas of the liver were retrieved from the surgical pathology files of the Department of Pathology, University Hospital Zürich, Switzerland, and the Institute of Pathology, University of Regensburg, Germany. Tumors were classified according to the World Health Organization (WHO) Classification of the Digestive System (2000). The study was approved by the local ethics committees of Zurich (Kantonale Ethikkommission Zurich, StV 26–2005).

Tissue Microarray.

Formalin-fixed paraffin-embedded tumor tissues were used to construct tissue microarrays (TMA) with cores of 168 HCC and 20 normal liver tissues. The TMAs were constructed as described.15 Two tissue cores per tumor with a diameter of 0.6 mm were punched out of the donor block and transferred to the recipient block.

Histology, Immunohistochemistry.

Three-micron-thick sections of TMA blocks and formalin-fixed, paraffin-embedded tissues were mounted on glass slides (SuperFrost Plus; Menzel), deparaffinized, rehydrated, and stained with hematoxylin-eosin using standard histological techniques. For immunohistochemical staining the Ventana Benchmark automated staining system (Ventana Medical Systems) and Ventana reagents were used. After deparaffinization in xylene, slides were rehydrated in decreasing concentrations of ethanol. Endogenous peroxidase was blocked using the Ventana endogenous peroxidase blocking kit after a rinse with distilled water. For antigen retrieval slides were heated with cell conditioning solution (CC1, Ventana) according to the manufacturers instructions. The same primary antibodies as for western blotting were used against and adjusted to the Ventana Benchmark system after performing titrations. iVIEW-DAB was used as chromogen. Ki-67 proliferation index was defined as percentage of positive nuclei per 100 tumor cells and was determined on TMA tumor tissue cores.

Animal Experiments.

All animal experiments were in accordance with Swiss federal animal regulations and approved by the cantonal veterinary office of Zurich. Athymic female NU/NU mice ages 6 to 8 weeks were kept on a 12-hour day/night cycle with free access to food and water. Huh7 cells were trypsinized (Invitrogen), washed in PBS, and resuspended in serum-free DMEM. Three × 106 Huh7 cells in 200 μL DMEM were injected subcutaneously in the lower back. After 3 weeks 33% of the mice bore visible tumors. When tumors reached a mean volume of 150 mm3 mice were randomized into two groups (six animals per group): Animals were treated with 0.16% DMSO in 500 μL H20 subcutaneously or 10 mg/kg BW SB204741 in 500 μL H20 twice a day. Tumor growth was monitored daily with a sliding caliper and the volume was calculated according to the formula (lxw2)/2, where l is the length and w the width.

For the second animal model,14 aging (1-year-old) C57Bl/6 and Tph (−/−) male mice on a background of C57Bl/64, 10 were fed carbon tetrachloride (4 mL/kg CCl4 mixed with an equal volume of soybean oil) three times per week for 6 weeks by gavage (10 animals per group).

Statistical Analysis.

Data are given as mean and standard deviation (SD). Differences between the groups were assessed by 1-way or 2-way analysis of variance (ANOVA) using an appropriate post-test, including Dunnett's and Bonferroni post-hoc tests. Differences of tumor weight were assessed with a two-tailed t test. The level of statistical significance was set at P < 0.05. Statistical analyses were performed with Prism 4.0 (GraphPad) The TMA was analyzed with an SPSS databank (SPSS 12.0). Association between categorical data was tested with the two-tailed χ2 test. Correlation between categorical and continuous data was measured with Kendall's τ (two-tailed).

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Does 5HT Promote Proliferation in HCC Cells?

To test whether 5HT is a mitogen for hepatocellular cancer cells we measured thymidine incorporation in two human hepatocellular cancer cell lines, Huh7 and HepG2. In the presence of 10% fetal calf serum (FCS) thymidine-incorporation was tripled compared to serum-free media (SFM) alone (Supporting Fig. 1A). A dose response over six log scales revealed a maximal incorporation of thymidine at 100 μM 5HT (in the absence of FCS), similar to the activity in the presence of 10% FCS. Qualitative assessment of proliferating cells was performed with staining of incorporated BrdU and Ki67 (Supporting Fig. 1B). The total number of cells, determined with nuclear Hoechst staining, was lower in SFM compared to media containing 10% FCS or 100 μM 5HT in the absence of serum (Supporting Fig. 1C). Interestingly, the percentage of BrdU- or Ki67-positive cells in relation to the total number of cells was the same in FCS, SFM, or 5HT. We concluded that the assays reported the number of viable cells, and not whether there was proliferation.16 Therefore, we tested whether 5HT treatment improved cell survival. After 72 hours of serum deprivation almost all cells (Huh7) underwent complete necrotic cell death as demonstrated by light microscopy, calcein/ethidium staining (green cells were alive, red cells were dead), and Hoechst/TUNEL staining (blue cells were alive, green cells were dead) (Fig. 1A). However, upon treatment with 100 μM 5HT, cell death could be prevented and viability was maintained to a similar degree as with standard culture conditions in the presence of 10% FCS. Thus, we concluded that 5HT promotes survival, which was also supported by two different viability assays with both cells lines, Huh7 and HepG2. MTT (Fig. 1B) and CytoTox-Fluor (Fig. 1C) exhibited a dose-dependent increase of vital cells or a decrease of dead cells after stimulation with 5HT. To clarify whether 5HT promotes predominantly survival or acts as a mitogen, MTT activity of Huh7- and HepG2 cells was measured for 120 hours in the presence of 0.1 μg/mL demecolcine, a potent mitotic poison, in combination with 100 μM 5HT. After 5 days 5HT in SFM caused 7- to 8-fold higher values of MTT activity in Huh7 and HepG2, whereas serum deprivation led to complete cell death (Fig. 1D,E). This clearly could not only be attributed to increased cell viability but also to an enhanced rate of proliferation. Initial proliferation in SFM was prevented by demecolcine. The combination of 5HT and demecolcine abolished cell proliferation, but importantly, MTT activity did not drop below the baseline of 100% MTT activity. Thus, we concluded that 5HT predominantly promotes cell survival of Huh7 and HepG2 and this prerequisite facilitates cell proliferation.

thumbnail image

Figure 1. Effect of 5HT on survival of hepatocellular cancer cells. (A) Phase contrast microscopy of Huh7 48 hours after serum deprivation followed by stimulation for 24 hours either with 10% FCS or with 100 μM 5HT. Note bullous change of Huh7 under serum deprivation. Combined calcein/ethidium-stainings with living cells (green) and dead cells (read). Hoechst/TUNEL-stainings demonstrated death of Huh7 after serum deprivation and survival of 5HT-treated cells (Hoechst: blue, TUNEL: green). The same experimental conditions as in the qualitative assessment were used to quantify survival of Huh7 and HepG2. Survival increases 2-fold in dose dependence of 5HT as shown concordantly with MTT assay (B) and CytoTox-Fluor-Assay (C) (n = 16, **P < 0.001, *P < 0.01). (D,E) Quantification of survival and proliferation with 100 μM 5HT in the presence of 0.1 μg/mL demecolcine, a mitotic spindle poison. Cells were incubated in SFM as control and subsequently stimulated with 5HT for 120 hours. MTT assays were used to assess viability and proliferation. All experiments were performed in the absence of FCS. 5HT caused a 5- to 8-fold increase of cell mass indicating proliferation (SFM versus 5HT: P < 0.001). Demecolcine (De) prevented proliferation and lead to cell death, whereas 5HT prevented cell death even in the presence of De (Huh7: 5HT/De versus De: P < 0.001, HepG2: 5HT/De versus De: P < 0.01, n = 6 for each treatment).

Download figure to PowerPoint

Which 5HT Receptor Is Involved in Survival of HCC Cells?

To identify a potential target receptor that is involved in 5HT-mediated survival of HCC cells we tested different agonists and antagonists. A scheme of the experimental setup is shown in Supporting Fig. 2A. 5HT2 receptors belong to the Gq/11 family of G-proteins. Their stimulation results in the activation of phospholipase C (PLC) and further downstream in the activation of protein kinase C (PKC).5 The use of a classical PKC activator, phorbol 12-myristate 13-acetate (PMA), in Huh7 cells could mimic the effect of 5HT in serum-free culture conditions, indicating that class 2 receptors are potentially involved in 5HT-mediated cell survival (Supporting Fig. 2B). After administration of specific agonists for the receptors 1/7 (5-CT), 2A (DOI), and 2B (α-Me-HTP), we concluded that the 2B receptor is responsible receptor for 5HT-mediated cytoprotection. This finding was further supported with ritanserin, a general 5HT2-receptor antagonist. Ritanserin was able to reverse the cytoprotection conferred by 5HT, whereas antagonists targeted for the 5HT-1A, -2A, and -7 receptor failed to provide an effect on cell viability (Supporting Fig. 2C). The presence of HTR2B was demonstrated by western blotting (Supporting Fig. 2D). In time-dependent experiments 5HT caused significantly higher values of MTT activity in Huh7 and HepG2 cells after 120 hours of serum deprivation compared to untreated cells (Fig. 2A,C). The same effect was also found with α-Me-HTP, a selective 5HT2B agonist17, 18 (Fig. 3B,D). When SB204741 (SB204), a selective 5HT2B antagonist,17, 18 was added during incubation with 5HT or α-Me-HTP the effect was abolished in a dose-dependent manner (Fig. 2A-D). Taken together, these experiments identified (1) the 5HT2B receptor as a mediator of survival, and (2) demonstrated that 5HT mediates proliferation of HCC cells as a result of cytoprotection.

thumbnail image

Figure 2. Viability and proliferation in the presence of 100 μM 5HT (A,C) or 100 μM α-Me-HTP (HTR2B selective agonist) (B,D) and in combination with the HTR2B selective antagonist SB204741 (SB204). Three different concentrations of SB204 were used in the presence of 5HT or α-ME-HTP. α-ME-HTP had the same effect on survival as 5HT (SFM versus α-ME-HTP: P < 0.001). SB204 abolished or reduced the 5HT and α-ME-HTP mediated proliferation in Huh7 and HepG2 (5HT versus 50 μM SB204: P < 0.001; α-ME-HTP versus 50 μM SB204: P < 0.001). Data represent six experiments for each treatment.

Download figure to PowerPoint

thumbnail image

Figure 3. Evaluation of apoptosis and autophagy of Huh7 under serum deprivation. (A) Activity of caspase-3 as an indicator of apoptosis. Serum deprivation did not cause elevated caspase-3 activity. As positive control, cells were treated with actinomycin-D and TNF-α (Actino/TNF-α) (n = 4 for each treatment). (B) Extracellular protease activity, as a sign of necrotic cell death, was elevated in serum-deprived and apoptotic cells, but not after treatment with 5HT or FCS (SFM versus SFM/5HT: P < 0.001) (n = 4 for each treatment). (C) TEM. Treatment with actinomycin-D/TNF-α exhibited chromatin condensation (1). Cell culture with 10% FCS (2) or 5HT only (3) showed confluent intact cells after 72 hours. Serum deprivation of 48 hours (SFM) induced embedded cell bodies consistent with autophagosomes (black arrows) (4). Picture 5 shows a 10-fold magnification of an autophagosome. After 72 hours of serum deprivation cells revealed an extended vacuolization (6).

Download figure to PowerPoint

How Does 5HT Mediate Cytoprotection of HCC Cells?

The combined Hoechst/TUNEL and calcein/ethidium stainings (Fig. 1A) suggested either necrotic or apoptotic cell death after serum deprivation. Therefore, we measured caspase-3 activity in Huh7 under different conditions to elucidate the cytoprotective mechanism of 5HT. Interestingly, serum deprivation was not associated with any detectable caspase-3 activity (Fig. 3A). By determination of proteases (CytoTox-Fluor-Assay) released by necrotic cells we observed a strong increase after serum deprivation. Addition of 5HT prevented the cell death and subsequently necrosis (Fig. 3B). The specificity of both assays was confirmed by the use of tumor necrosis factor alpha (TNF-α) in combination with actinomycin-D, a classical inducer of apoptosis. To specify cell death that is distinct from apoptosis, we performed an ultrastructural analysis with transmission electron microscopy (TEM) (Fig. 3C). TEM failed to show pyknosis or karyorrhexis, both morphological criteria for apoptosis, but revealed lysosomal organelles in serum-deprived cells consistent with autophagosomes, which typically appear during macroautophagy. After 72 hours of serum deprivation cells were markedly vacuolized, whereas the nucleus was intact. This has been considered a distinct morphological sign of autophagy.16 Under 5HT treatment neither autophagosomes nor vacuolization were apparent. Macroautophagy (herein referred to as autophagy) is a catabolic process whereby cells undergo a self-digestion of intracellular organelles. It has been realized as a mechanism of cell survival as well as cell death. In response to cellular stress like starvation, growth factor withdrawal or high bioenergetic demands the degradation of cytoplasmatic material enters the tricarboxylic acid cycle to generate ATP.19 Excessive autophagy leads to cell death and has been described as type II cell death that is morphologically and mechanistically distinct from apoptosis.20 From the findings of the TEM we hypothesize that serum deprivation leads to autophagy, which may be inhibited by 5HT.

Does 5HT Inhibit Autophagy in HCC Cells?

To explore the role of 5HT in autophagy different characteristics of autophagy were investigated. First, the microtubule-associated protein light chain 3 (LC3B) is essential for the assembly of autophagosomes and serves as a marker for autophagy.19 We found 10-fold elevated expression of LC3B in Huh7 cells after 72 hours of serum deprivation (Fig. 4A,B). In the presence of 5HT the increase in LC3B was significantly blunted in serum-deprived Huh7. Second, p62, also called sequestosome 1 (SQSTM1), can be used as an additional marker of autophagy. An interaction of p62 with LC3 causes a specific degradation by autophagy. Because its degradation is dependent on autophagy, the level of p62 increases in response to inhibition of autophagy.21 We found a 7-fold elevated expression of p62 after 24 hours of 5HT treatment. The expression levels remained elevated after 72 hours. Under serum deprivation the expression of p62 increased during the first 48 hours and decreased afterwards. Third, the mammalian target of rapamycin (mTOR) is a key regulator of autophagy and an essential controller of cell growth. When growth conditions are favorable mTOR is active and maintains ribosome biogenesis, translation initiation, and nutrient import.22 We investigated the activation of mTOR and its downstream targets p70S6K and 4E-BP1 with immunoblots. During serum deprivation we could not detect significant changes of phosphorylated mTOR (Fig. 4A,B), but the amount of phosphorylated p70S6K and 4E-BP1 after 72 hours differed significantly. 5HT treatment sustained activation of p70S6K and 4E-BP1, whereas serum deprivation caused a continuous decrease in the phosphorylation of these proteins (Fig. 4A,B). These findings indicate (1) that serum withdrawal activates autophagy and leads to cell death and that (2) 5HT inhibits autophagy and modulates cellular downstream targets of mTOR.

thumbnail image

Figure 4. (A) Representative immunoblots of autophagy related proteins LC3B, p62, phosphorylated and nonphosphorylated mTOR, p70S6K, and 4E-BP1. (B) Quantification of immunoblots: Serum deprivation caused a 10-fold increased expression of LC3B that was attenuated with 100 μM 5HT (n = 3, **P < 0.001, *P < 0.01). 5HT lead to a sustained accumulation of p62 after 24 hours and 72 hours compared to serum deprivation (n = 3, *P < 0.001). Phosphorylation of mTOR did not change after 5HT treatment (n = 3). 5HT caused a sustained phosphorylation of p70S6K (n = 3, *P < 0.001) and 4E-BP1 (n = 3, **P < 0.01, *P < 0.05), direct downstream targets of mTOR.

Download figure to PowerPoint

Does 5HT Bypass Signaling Pathways of mTOR?

Although 5HT did not affect the phosphorylation of mTOR in serum-deprived Huh7 cells, we detected sustained activation of direct downstream targets of mTOR. P70S6K and 4E-BP1 are important regulators of protein synthesis and translation initiation. Inhibition of these proteins by targeting mTOR with rapamycin leads to cell cycle arrest and induces autophagy.22, 23 Therefore, we assumed that 5HT could promote cell survival by bypassing mTOR activation, i.e., in the presence of rapamycin. To test this hypothesis we performed viability assay and immunoblots with rapamycin in the presence or absence of 5HT. Under serum deprivation Huh7 and HepG2 disclosed a strong reduction in viability after an initial phase of cell growth within the first 48 hours. This initial cell growth was abolished with rapamycin. Strikingly, the cytotoxic effect of rapamycin was strongly attenuated by 5HT in both Huh7 and HepG2 cells within 120 hours (Fig. 5A). Inhibition of the 5HT-2B receptor by SB204 in the presence of 10% FCS led to a marked decrease in cell growth, even beyond the effect of rapamycin administration alone (Fig. 5B), whereas a combined treatment with rapamycin and SB204 had no further effect. This suggests (1) that serum-derived 5HT promotes cell growth and survival and (2) at least in part by an mTOR-independent pathway. To substantiate these findings we tested the activation of mTOR, p70S6K, and 4E-BP1 in the presence of rapamycin (Fig. 5C,D). In support of our hypothesis rapamycin reduced the phosphorylation of all three proteins. 5HT increased the activation of p70S6K and 4E-BP1 also in the presence of rapamycin, whereas activation of mTOR remained unchanged. Additionally, with 5HT the expression of LC3B was decreased (Fig. 5C,D). In conclusion, 5HT bypassed mTOR and activated p70S6K and 4E-BP1 to facilitate proliferation. The findings are summarized in Supporting Fig. 3.

thumbnail image

Figure 5. Viability and proliferation of Huh7 and HepG2 in the presence of 100 μM 5HT and 100 nM rapamycin (Rapa) without FCS (A), and in the presence of 100 μM SB204 with FCS (B). 5HT attenuated the cytotoxic effect of rapamycin in SFM in Huh7 (Rapa versus 5HT/Rapa: P < 0.001) In the presence of 10% FCS growth inhibition of Huh7 was mediated by SB204 (FCS versus SB204: P < 0.001). Rapa had no additive effect (SB204 versus SB204/Rapa: ns). In SFM HepG2 responded like Huh7 to rapamycin (Rapa versus 5HT/Rapa: P < 0.001). In the presence of FCS, Rapa did not inhibit proliferation of HepG2 and SB204 contributed to growth inhibition (FCS versus SB204: P < 0.001). Data represent six experiments for each treatment. Representative immunoblots (C) and quantification (D) of immunoblots of autophagy-related proteins treated with 100 μM 5HT and 100 nM rapamycin after 48 hours. 5HT suppressed expression of LC3B and phosphorylated p70S6K and 4E-BP1 even in combination with rapamycin (n = 3, *P < 0.05, **P < 0.01). Rapa decreased the phosphorylation of mTOR, but no differences were found with 5HT in the presence of rapa (n = 3, *P < 0.05).

Download figure to PowerPoint

Does 5HT Play a Role in Animal Models of HCC?

As the 5HT2B receptor mediated cell survival and growth in vitro we tested the 5HT2B receptor antagonist SB204741 in vivo in a subcutaneous xenograft model with Huh7 cells. The growth and weight of tumors in athymic mice treated with SB204741 was significantly decreased compared to the control group (Fig. 6A,B). Further support for a role of 5HT in tumor formation was gained in a second animal model in which CCl4 was chronically fed to 1-year-old mice. B/6-mice showed a 33% liver tumor incidence (4/12) (Fig. 6C,D). In contrast, mice lacking peripheral 5HT due to a deletion of tryptophan hydroxylase (Tph) had no liver tumor (0/11), suggesting that the presence of 5HT was required for tumor formation. These data from both animal models, as a proof of principle, suggest the 5HT2B receptor as a molecular target for HCC and 5HT as a deleterious factor in tumor formation.

thumbnail image

Figure 6. Growth of subcutaneous HCC xenografts in vivo. Huh7 cells were injected in athymic mice and after tumor lodgment mice were treated with the 5HTR2B antagonist SB204 for 10 days. Tumor xenograft growth (A) and weight (B) was decreased with SB204 group compared to control (n = 6 per group, A: *P = 0.0055, B: *P = 0.0174). (C) Immunohistochemical staining of Collagen IV demonstrates irregular hepatocytic cords indicating tumorous tissue in old CCl4-treated wildtype mice. Black arrows indicate tumor margin. (D) Incidence of macroscopically detectable tumors (gray bar) in CCl4-treated young (10 weeks) and old (60 weeks) wildtype and tph−/− mice (*P = 0.04, n = 10 per group). (E) Expression of HTR2B in HCC. Representative immunohistochemistry of HTR2B and p-p70S6K in normal human liver and HCC. (F) Correlation of HTR2B and p-p70S6K with Ki67 (HTR2B: n = 168, r = 0.160, P < 0.014; p-p70S6K: n = 168, r = 0.370, P < 0.0001).

Download figure to PowerPoint

Are the Experimental Findings Relevant for Human HCC?

To answer whether our in vitro findings may be useful in a clinical setting we prepared a TMA from 168 patients who underwent resection or transplantation due to HCC. Immunohistochemistry revealed that 48/168 (28.6%) of these tumors were positive for HTR2B and 46/168 (27.4%) were positive for p-p70S6K (Fig. 6E). A chi-squared test revealed that HTR2B and p-p70S6K were significantly associated (P = 0.001) (Supporting Table 1). Immunohistochemistry of HTR2B and p-p70S6K correlated with the proliferation index as assessed by Ki67 staining (HTR2B: n = 168, r = 0.160, P < 0.014; p-p70S6K: n = 168, r = 0.370, P < 0.0001) (Fig. 6F). These data strongly support our in vitro and in vivo findings that 5HT promotes cell survival and growth of hepatocellular cancer cells by activation of the 5HT2B receptor.

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The study reveals a novel function of 5HT as a survival factor of HCC cells. We demonstrated that activation of HTR2B leads to sustained phosphorylation of two downstream targets of mTOR, p70S6K and 4E-BP1, thereby facilitating survival and inhibiting autophagy. Targeting the HTR2B receptor reduced cancer cell growth in vitro and in vivo. Furthermore, an analysis of a TMA of 168 patients with HCCs points toward a contribution of the HTR2B in the biology of HCC.

In our previous study we demonstrated that 5HT mediates angiogenesis and growth of colon cancer allografts in vivo.14 In contrast to the current study, that report suggested a receptor-independent mechanism. However, both studies demonstrate a harmful role of 5HT in cancer. In line with our work of liver regeneration, the proliferation of hepatocytes was attributed to an increased expression of HTR2B in the liver4 and the effect of 5HT on cancer cells may depend on the cell type. A specific effect of 5HT on hepatocellular cancer cells was also supported by initial experiments excluding a general survival effect in different cell types (Supporting Fig. 4).

Our results from the cell culture suggest an involvement of 5HT in autophagic pathways. The decreased maturation of autophagosomes reflected by the expression of LC3B together with the accumulation of p62 in 5HT-treated cells indicates that 5HT inhibits autophagy. But these findings alone do not distinguish whether autophagy leads to cell death or autophagy occurs together with cell death.19 Experiments detecting DNA-fragmentation with TUNEL staining suggested that 5HT suppresses apoptosis. Because TUNEL staining may be positive also in necrotic cells,24 we investigated caspase activity in serum-deprived cells. Serum deprivation did not lead to apoptosis, as shown by caspase activity and in TEM. Our data show necrotic processes, e.g., membrane permeabilization and DNA-fragmentation together with autophagy. Therefore, either excessive self-digestion or other caspase-independent death mechanisms may lead to cell death, which is inhibited by 5HT. Our data suggest that autophagy represents a survival mechanism where growth conditions are not favorable and that autophagy becomes unnecessary in the presence of 5HT. Thus, 5HT may directly or indirectly inhibit autophagy by, e.g., facilitating glucose uptake25 or autocrine mechanisms.26

In terms of cancer biology, autophagy acts in tumor suppression as well as in cytoprotection of cancer cells.27, 28 The inhibition of mTOR is a potential target to treat HCC.29 In accordance with other groups, HCC cells did not or only weakly respond to mTOR-inhibition in the presence of FCS,30, 31 presumably due to mTOR-independent activation of its classical mTOR downstream targets as observed in our case by the action of 5HT. Therefore, inhibition of 5HT may be suppressive by preventing the phosphorylation of p70S6K and 4E-BP1 and by restoring autophagic mechanisms. The question whether autophagy enhances or suppresses tumor progression in patients with HCC remains open. However, the results of the animal experiments and the human biopsies indicate clearly a deleterious role of 5HT in HCC.

In conclusion, the presence of HTR2B in HCC and the activation of autophagy-related mechanisms demonstrate novel insights of 5HT in cancer biology and propose 5HT-mediated signaling as a therapeutic target.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

We thank Udo Ungethüm for technical assistance and Ursula Lüthi (Center for Microscopy and Image Analysis, University Zurich) for preparing and analyzing electron microscopy samples.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
HEP_23441_sm_SupFig1.tif24136KSupporting Fig. 1: Effect of 5HT on proliferation of hepatocellular cancer cells. (A) 3H-thymidin-incorporation was determined in Huh7 and HepG2. After 48h serum deprivation cells were labelled and stimulated for 24h with FCS 10%, SFM or 5HT (in the absence of FCS). Dose response curve of 5HT compared to SFM revealed a maximum at 100 μM 5HT (n=6, *p<0.01). (B) Same experimental conditions as in (A) were used for BrdU- and Ki67-staining. Overlay with Hoechst illustrated less cells in SFM and a similiar number of cells in media containing FCS or 5HT. (C) Quantification of BrdU-staining showed significantly less cells in SFM (n=3, *p<0.05). A similar observation was made with Ki67 staining. The ratio of BrdU or Ki67 positive cells to the total number of cells (Hoechst) was equal under the three different conditions.
HEP_23441_sm_SupFig2.tif3517KSupporting Fig. 2: (A) Experimental setup for concentration dependent experiments. (B) Agonists to 5HT receptors HTR1, HTR2A, HTR7 do not increase the survival of Huh7, but phorbol-myristate-acetate (PMA) and α-Me-HTP (selective for HTR2B) increase survival two- to three-fold (n=4, *p<0.01). (C) Accordingly, antagonists to 5HT receptor HTR1, HTR2A and HTR7 have no effect on survival of Huh7 (n=3). Ritanserin, a general 5HT antagonist of HTR2 abolished the survival effect of 5HT (n=8, *p<0.001). (D) Identification of the HT2B receptor in Huh7 and HepG2 by western blot analysis.
HEP_23441_sm_SupFig3.tif8201KSupporting Fig. 3: Illustration of the suggested mechanism. Black lines indicate activation of p70S6K and 4E-BP1, red lines illustrate inhibition.
HEP_23441_sm_SupFig4.tif5215KSupporting Fig. 4: HeLa cells (human cervical cancer) (A) and MCF7 cells (human breast cancer) (B) were treated with four concentrations of 5HT for 24h after an initial serum deprivation of 48h. MTT assay was used to quantify survival. 5HT did not increase survival in these cells compared to serum free media (SFM) and media containing 10% fetal calf serum (FCS) (n=6 per group). HEF5 cells (human fibroblasts) (C) and PC3 (human prostate cancer) (D) were subsequently treated with 100 μM 5HT and survival was compared with standard culture conditions (FCS) and SFM using MTT assay. 5HT did not increase survival (n=6 per group).
HEP_23441_sm_SupTab.doc27KSupporting Table. 1: Cross tabulation for HTR2B and p-p70S6K of 168 HCC biopsies

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.